
[Federal Register: April 29, 2008 (Volume 73, Number 83)]
[Notices]               
[Page 23222-23240]
From the Federal Register Online via GPO Access [wais.access.gpo.gov]
[DOCID:fr29ap08-56]                         

-----------------------------------------------------------------------

DEPARTMENT OF ENERGY

Federal Energy Regulatory Commission

[Docket No: PL07-2-000]

 
Composition of Proxy Groups for Determining Gas and Oil Pipeline 
Return on Equity; Policy Statement

Issued April 17, 2008.
Before Commissioners: Joseph T. Kelliher, Chairman; Suedeen G. 
Kelly, Marc Spitzer, Philip D. Moeller, and Jon Wellinghoff.

    1. On July 19, 2007, the Commission issued a proposed policy 
statement concerning the composition of the proxy groups used to 
determine gas and oil pipelines' return on equity (ROE) under the 
Discounted Cash Flow (DCF) model.\1\ Historically, in determining the 
proxy group, the Commission required that pipeline operations 
constitute a high proportion of the business of any firm included in 
the proxy group. However, in recent years, there have been fewer gas 
pipeline corporations that meet that standard, in part because of the 
greater trend toward Master Limited Partnerships (MLPs) in the gas 
pipeline industry. Additionally, there are no oil corporations 
available for use in the oil pipeline proxy group. These trends have 
made the MLP issue one of particular concern to the Commission and are 
the reason that the Commission issued the Proposed Policy Statement.\2\
---------------------------------------------------------------------------

    \1\ Composition of Proxy Groups for Determining Gas and Oil 
Pipeline Return on Equity, 120 FERC ] 61,068 (2007) (Proposed Policy 
Statement).
    \2\ After an initial round of comments and reply comments, the 
Commission concluded that it required additional comment on the 
issue of the growth rates of MLPs. After notice to this effect and 
the receipt of a round of initial and reply comments, staff held a 
technical conference involving an eight member panel on January 23, 
2008 that was transcribed for the record. Comments and reply 
comments were filed thereafter.
---------------------------------------------------------------------------

    2. After review of an extensive record developed in this 
proceeding, the Commission concludes: (1) MLPs should be included in 
the ROE proxy group for both oil and gas pipelines; (2) there should be 
no cap on the level of distributions included in the Commission's 
current DCF methodology; (3) the Institutional Brokers Estimated System 
(IBES) forecasts should remain the basis for the short-term growth 
forecast used in the DCF calculation; (4) there should be an adjustment 
to the long-term growth rate used to calculate the equity cost of 
capital for an MLP; and (5) there should be no modification to the 
current respective two-thirds and one-third weightings of the short- 
and long-term growth factors. Moreover, the Commission will not explore 
other methods for determining a pipeline's equity cost of capital at 
this time. The Commission also concludes that this Policy Statement 
should govern all gas and oil rate proceedings involving the 
establishment of ROE that are now pending before the Commission, 
whether at hearing or in a decisional phase at the Commission.

I. Background

A. The DCF Model

    3. The Supreme Court has stated that ``the return to the equity 
owner should be commensurate with the return on investments in other 
enterprises having corresponding risks. That return, moreover, should 
be sufficient to assure confidence in the financial integrity of the 
enterprise, so as to maintain its credit and to attract capital.'' \3\ 
Since the 1980s, the Commission has used the DCF model to develop a 
range of returns earned on investments in companies with corresponding 
risks for purposes of determining the ROE to be awarded natural gas and 
oil pipelines.
---------------------------------------------------------------------------

    \3\ FPC v. Hope Natural Gas Co., 320 U.S. 591 (1944). Bluefield 
Water Works & Improvement Co. v. Public Service Comm'n, 262 U.S. 679 
(1923).
---------------------------------------------------------------------------

    4. The DCF model was originally developed as a method for investors 
to estimate the value of securities, including common stocks. It is 
based on

[[Page 23223]]

the premise that ``a stock's price is equal to the present value of the 
infinite stream of expected dividends discounted at a market rate 
commensurate with the stock's risk.'' \4\ With simplifying assumptions, 
the DCF model results in the investor using the following formula to 
---------------------------------------------------------------------------
determine share price:

    \4\ CAPP v. FERC, 254 F.3d 289, 293 (2001) (CAPP).
---------------------------------------------------------------------------

    P = D/(r-g)

where P is the price of the stock at the relevant time, D is the 
current dividend, r is the discount rate or rate of return, and g is 
the expected constant growth in dividend income to be reflected in 
capital appreciation.\5\
---------------------------------------------------------------------------

    \5\ Id. National Fuel Gas Supply Corp., 51 FERC ] 61,122, at 
61,337 n.68 (1990). Ozark Gas Transmission System, 68 FERC ] 61,032, 
at 61,104 n.16. (1994).
---------------------------------------------------------------------------

    5. Unlike investors, the Commission uses the DCF model to determine 
the ROE (the ``r'' component) to be included in the pipeline's rates, 
rather than to estimate a stock's value. Therefore, the Commission 
solves the DCF formula for the discount rate, which represents the rate 
of return that an investor requires in order to invest in a firm. Under 
the resulting DCF formula, ROE equals current dividend yield (dividends 
divided by share price) plus the projected future growth rate of 
dividends:
r = D/P + g

    6. Over the years, the Commission has standardized the inputs to 
the DCF formula as applied to interstate gas and oil pipelines. The 
Commission averages short-term and long-term growth estimates in 
determining the constant growth of dividends (referred to as the two-
step procedure). Security analysts' five-year forecasts for each 
company in the proxy group (discussed below), as published by IBES, are 
used for determining growth for the short term. The long-term growth is 
based on forecasts of long-term growth of the economy as a whole,\6\ as 
reflected in the Gross Domestic Product (GDP which are drawn from three 
different sources.\7\ The short-term forecast receives a two-thirds 
weighting and the long-term forecast receives a one-third weighting in 
calculating the growth rate in the DCF model.\8\
---------------------------------------------------------------------------

    \6\ Northwest Pipeline Company, 79 FERC ] 61,309, at 62,383 
(1997) (Opinion No. 396-B). Williston Basin Interstate Pipeline 
Company, 79 FERC ] 61,311, at 62,389 (1997) (Williston I), aff'd, 
Williston Basin Interstate Pipeline Co. v. FERC, 165 F.3d 54, 57 (DC 
Cir. 1999) (Williston v. FERC).
    \7\ The three sources used by the Commission are Global Insight: 
Long-Term Macro Forecast--Baseline (U.S. Economy 30-Year Focus); 
Energy Information Agency, Annual Energy Outlook; and the Social 
Security Administration.
    \8\ Transcontinental Gas Pipe Line Corp., 84 FERC ] 61,084, at 
61,423-4 (Opinion No. 414-A), reh'g denied, 85 FERC ] 61,323, at 
62,266-70 (1998) (Opinion No. 414-B), aff'd sub nom. North Carolina 
Utilities Commission v. FERC, 203 F.3d 53 (DC Cir. 2000) 
(unpublished opinion). Northwest Pipeline Co., 88 FERC ] 61,057, 
reh'g denied, 88 FERC ] 61,298 (1999), aff'd CAPP v. FERC, 254 F.3d 
289 (DC Cir. 2001).
---------------------------------------------------------------------------

    7. Most gas pipelines are wholly-owned subsidiaries and their 
common stocks are not publicly traded. This is also true for some 
jurisdictional oil pipelines. Therefore, the Commission must use a 
proxy group of publicly traded firms with corresponding risks to set a 
range of reasonable returns for both natural gas and oil pipelines. For 
both oil and gas pipelines, after defining the zone of reasonableness 
through development of the appropriate proxy group for the pipeline, 
the Commission assigns the pipeline a rate within that range or zone, 
to reflect specific risks of that pipeline as compared to the proxy 
group companies.\9\ The Commission has historically presumed that 
existing pipelines fall within a broad range of average risk. A 
pipeline or other litigating party has to show highly unusual 
circumstances that indicate anomalously high or low risk as compared to 
other pipelines to overcome the presumption.\10\
---------------------------------------------------------------------------

    \9\ Williston v. FERC, 165 F.3d at 57 (citation omitted).
    \10\ Transcontinental Gas Pipe Line Corp., 90 FERC ] 61,279, at 
61,936 (2000).
---------------------------------------------------------------------------

    8. The Commission historically required that each company included 
in the proxy group satisfy the following three standards.\11\ First, 
the company's stock must be publicly traded. Second, the company must 
be recognized as a natural gas or oil pipeline company and its stock 
must be recognized and tracked by an investment information service 
such as Value Line. Third, pipeline operations must constitute a high 
proportion of the company's business. Until 2003, the Commission's 
policy was that the third standard could only be satisfied if a 
company's pipeline business accounted for, on average, at least 50 
percent of a company's assets or operating income over the most recent 
three-year period.\12\
---------------------------------------------------------------------------

    \11\ Id. at 61,933.
    \12\ Williston Basin Interstate Pipeline Company, 104 FERC ] 
61,036, at P 35 n.46 (2003) (Williston II).
---------------------------------------------------------------------------

    9. However, in recent years fewer corporations have satisfied the 
Commission's standards for inclusion in the gas and oil pipeline proxy 
groups. Mergers and acquisitions have reduced the number of publicly 
traded corporations with natural gas pipeline operations. Most of the 
remaining corporations are engaged in such significant non-pipeline 
business that their pipeline business accounts are significantly less 
than 50 percent of their assets or operating income. At the same time, 
there has been a trend toward MLPs owning natural gas pipelines. This 
trend has been even more pronounced in the oil pipeline industry, with 
the result that there are now no purely oil pipeline corporations 
available for inclusion in the oil pipeline proxy group and virtually 
all traded oil pipeline equity interests are owned by MLPs. Thus, for 
both oil and gas pipeline rate cases, the composition of the proxy 
group has become a significant issue, and the central question is 
whether, and how, to include MLPs in the proxy group.

B. The MLP Business Model

    10. MLPs consist of a general partner, who manages the partnership, 
and limited partners, who provide capital and receive cash 
distributions, but have no management role. The units of the limited 
partners are traded on public exchanges, just like corporate stock 
shares. In order to be treated as an MLP for Federal income tax 
purposes, an MLP must receive at least 90 percent of its income from 
certain qualifying sources, including natural resource activities. 
Natural resource activities include exploration, development, mining or 
production, processing, refining, transportation, storage and marketing 
of any mineral or natural resource, including gas and oil.\13\
---------------------------------------------------------------------------

    \13\ See Wachovia Securities, Master Limited Partnerships: A 
Primer, November 10, 2003, (Wachovia Primer 1) at 1, 3-4, reproduced 
in full in Docket No. OR96-2-012, Ex. SEP ARCO-22 and also in Kern 
River Gas Transmission Company, Docket No. RP04-274-000, Ex. No. BP-
19 filed October 25, 2005; J.P. Morgan, Industry Analysis, Energy 
MLPS, dated March 28, 2002 (J.P. Morgan 2002 Energy MLPs) at 5-6, 
reproduced in full in Docket No. OR92-8-025, Ex. No. SWST-18, filed 
October 20, 2005; Wachovia Capital Markets, LLC, Equity Research 
Department, Master Limited Partnerships: Primer 2nd Edition, A 
Framework for Investment dated August 23, 2005 (Wachovia 2nd Primer) 
at 8-9, reproduced in full in Docket No. RP06-72-000 at Ex. S-36, 
filed May 31, 2006); Coalition of Publicly Traded Partnerships, 
Publicly Traded Partnerships: What they are and how they work 
(undated) (Publicly Traded Partnerships) at 1-3, reproduced in full 
in Docket No. RP06-72-000 at Ex. S-35, filed May 31, 2006, and 
Docket No. OR96-2-012, Ex. No. BP-19, filed October 25, 2005; CAPP 
Reply Comments, Attachment A at 2-3; APGA Additional Comments dated 
December 21, 2007.
---------------------------------------------------------------------------

    11. MLPs generally distribute most available cash flow to the 
general and limited partners in the form of quarterly distributions. At 
their inception, MLPs establish agreements between the general and 
limited partners, which define cash flow available for distribution and 
how that cash flow is

[[Page 23224]]

to be divided between the general and limited partners. Most MLP 
agreements define ``available cash flow'' as (1) net income (gross 
revenues minus operating expenses) plus (2) depreciation and 
amortization, minus (3) capital investments the partnership must make 
to maintain its current asset base and cash flow stream.\14\ 
Depreciation and amortization may be considered a part of ``available 
cash flow,'' because depreciation is an accounting charge against 
current income, rather than an actual cash expense. Thus, depreciation 
does not reduce the MLP's current cash on hand. The MLP agreement may 
provide for the general partner to receive increasingly higher 
percentages of the overall distribution if it raises the quarterly 
distribution. This gives the general partner incentives to increase the 
partnership's business and cash flow.\15\
---------------------------------------------------------------------------

    \14\ The definition of available cash may also net out short 
term working capital borrowings, the repayment of capital 
expenditures, and other internal items.
    \15\ Wachovia Primer 1 at 6-7; J.P. Morgan 2002 Energy MLPs at 
5, 14; Wachovia 2nd Primer at 9, 15-19.
---------------------------------------------------------------------------

    12. The general partner has discretion not to distribute the entire 
amount of available cash flow for the proper exercise of the business, 
to create reserves for capital expenditures, for the payment of debt, 
and for future distributions. However, pipeline MLPs have typically 
distributed 90 percent or more of available cash flow. As a result, the 
MLP's cash distributions normally include not only the operating profit 
component of ``available cash flow,'' but also the depreciation 
component. This means that, in contrast to a corporation's dividends, 
an MLP's cash distributions generally exceed the MLP's reported 
earnings. The pipeline MLP's ability to distribute a high percentage of 
available cash flows reflects the stable cash flows underpinning its 
businesses.\16\
---------------------------------------------------------------------------

    \16\ J.P. Morgan 2002 Energy MLPs at 11-13; Wachovia 2nd Primer 
at 24-25; Enbridge Initial Comments Attachment A, Wachovia Capital 
Markets, LLC, MLPs: Safe to Come Back Into the Water (Wachovia MLPs) 
dated August 20, 2007, at 2-4.
---------------------------------------------------------------------------

    13. Because of their high cash distributions, MLPs have financed 
capital investments required to significantly expand operations or to 
make acquisitions through debt or by issuing additional units rather 
than through retained cash, although the general partner has the 
discretion to do so. These expansions financed through external debt 
are intended to provide a return equal to the cost of the capital plus 
some additional return for the existing unit holders, i.e., it is 
accretive. Thus, the return on any newly issued units is expected to be 
sufficiently high to avoid dilution of the current distributions to the 
existing unit holders.\17\
---------------------------------------------------------------------------

    \17\ Id.
---------------------------------------------------------------------------

    14. MLPs may also provide significant tax advantages to their unit 
holders. Some MLPs allocate depreciation, amortization, and tax credits 
to the limited partners and away from the general partner. In some 
cases, the limited partner may have no net taxable income reported on 
the income tax information document (the K-1) the limited partner 
receives from the partnership each year, a pattern that may continue 
for years. In that case, the limited partner will not pay any taxes on 
the cash received from the partnership in the year of the distribution. 
To the extent a limited partner is allocated items of depreciation, 
credit, or losses that exceed the limited partner's ownership 
percentage, income taxes will be due on the difference when the unit is 
sold. However, this may not occur for many years. Over time the real 
cost of the future taxes declines while the future return of any tax 
savings that is reinvested increases. This can significantly increase 
the return to the investor over the holding period of the limited 
partnership unit.\18\
---------------------------------------------------------------------------

    \18\ See PSCNY Initial Comments at 12-13 and Attachment 1 
thereto at 2; Wachovia Primer at 4-5; Publicly Traded Partnerships 
at 2-3; Wachovia 2nd Primer at 1, 5, 20-22; J.P. Morgan 2002 Energy 
MLPs at 18-19.
---------------------------------------------------------------------------

    15. Moreover, distributions in excess of earnings are not taxed as 
long as the limited partner has a tax basis. Rather, the limited 
partner's tax basis is reduced and again any taxes are deferred until 
the unit is sold. By this tax deferral, the cash flow distributed in 
excess of earnings can be made available for reinvestment much earlier 
than would be the case of a corporate share.\19\ This reduces the 
limited partner's risk because the limited partner's cash basis in the 
unit is reduced, but the distribution would not normally reduce the 
market price of the unit nor, if the firm has access to external 
capital, would this necessarily reduce its long term growth potential.
---------------------------------------------------------------------------

    \19\ Id.
---------------------------------------------------------------------------

C. The Recent Cases on the Shrinking Proxy Group

1. Natural Gas Pipeline Cases
    16. The Commission first addressed the problem of the shrinking 
natural gas pipeline proxy group in Williston II, 104 FERC ] 61,036 at 
P 34-43. In that NGA section 4 rate case, the Commission relaxed the 
requirement that natural gas business account for at least 50 percent 
of the corporation's assets or operating income. Instead, the 
Commission approved the pipeline's proposal to use a proxy group based 
on the corporations listed in the Value Line Investment Survey's list 
of diversified natural gas firms that own Commission-regulated natural 
gas pipelines, without regard to what portion of the company's business 
comprises pipeline operations. The proxy group approved in that case 
included four corporations that satisfied the Commission's historic 
standards \20\ and five corporations with less pipeline business and 
more local distribution business than the Commission had previously 
allowed. The Commission set Williston's ROE at the median of this proxy 
group.
---------------------------------------------------------------------------

    \20\ The Commission noted that two of those four companies were 
in the process of merging so that in the future there would be only 
three pipeline corporations that satisfied our historic proxy group 
standards. Williston II, 104 FERC ] 61,036 at P 35.
---------------------------------------------------------------------------

    17. The Commission next addressed the proxy group issue in a 2004 
order in Petal Gas Storage, LLC, 97 FERC ] 61,097 (2001), reh'g granted 
in part and denied in part, 106 FERC ] 61,325 (2004) (Petal). In that 
case, a jurisdictional storage company with market-based rates had 
applied for a certificate under NGA section 7 to construct pipeline 
facilities to transport gas from its existing storage facility to a new 
interconnection with Southern Natural Gas Co. The Commission found that 
Petal was not a new entrant in the jurisdictional gas transportation 
business, but was simply expanding its existing business and had not 
shown that it faced any unusual risks. Ordinarily in such circumstances 
the Commission would use the pipeline's own currently approved ROE for 
its existing services in determining an initial incremental rate for 
the expansion. However, because Petal had market-based rates for its 
existing services, there was no such currently approved ROE to use. 
Therefore, the Commission calculated the initial rate for Petal's 
expansion using the same median ROE which it had approved in Williston, 
which was the most recent litigated gas pipeline section 4 rate case.
    18. When the Commission next addressed the proxy group issue, in 
High Island Offshore System, LLC (HIOS),\21\ and Kern River Gas 
Transmission Company (Opinion No. 486),\22\ the Williston II proxy 
group had shrunk to six corporations. Moreover, the Commission found 
that two of those

[[Page 23225]]

corporations should be excluded from the proxy group on the ground that 
their financial difficulties had lowered their ROEs to such a low level 
as to render them unrepresentative.\23\ This left only four 
corporations eligible for the proxy group under the standards adopted 
in Williston II, three of whom derived more revenue from the 
distribution business than the pipeline business. The two pipelines 
contended that, in these circumstances, the Commission should include 
natural gas pipeline MLPs in the gas pipeline proxy group. They 
asserted that MLPs have a much higher percentage of their business 
devoted to pipeline operations than most of the corporations eligible 
for the proxy group under Williston II, and therefore are more 
representative of the risks faced by pipelines.
---------------------------------------------------------------------------

    \21\ 110 FERC ] 61,043, reh'g denied, 112 FERC ] 61,050 (2005).
    \22\ 117 FERC ] 61,077 (2006), reh'g pending.
    \23\ HIOS, 110 FERC ] 61,043 at P 118. Opinion No. 486, 117 FERC 
] 61,077 at P 140-141.
---------------------------------------------------------------------------

    19. In HIOS and Opinion No. 486, the Commission rejected the 
proposals to include MLPs in the proxy group, and approved proxy groups 
using the four corporations still available under the Williston II 
approach of basing the proxy group on the Value Line Investment 
Survey's group of diversified natural gas corporations that own 
Commission-regulated pipelines. In HIOS, the Commission set the 
pipeline's ROE at the median of the four-corporation proxy group. In 
Opinion No. 486, the Commission took the same general approach as in 
HIOS, but set the pipeline's ROE 50 basis points above the median to 
account for the fact its pipeline operations have a higher risk than 
its distribution business.\24\
---------------------------------------------------------------------------

    \24\ Id. at P 171-176.
---------------------------------------------------------------------------

    20. In rejecting the proposals to include MLPs in the proxy group 
in both cases, the Commission made clear that it was not making a 
generic finding that MLPs cannot be considered for inclusion in the 
proxy group if a proper evidentiary showing is made.\25\ However, the 
Commission pointed out that data concerning dividends paid by the proxy 
group members is a key component in any DCF analysis, and expressed 
concern that an MLP's cash distributions to its unit holders may not be 
comparable to the corporate dividends the Commission uses in its DCF 
analysis. In Opinion No. 486, the Commission explained its concern as 
follows:
---------------------------------------------------------------------------

    \25\ Id. at P 147. See also HIOS, 110 FERC ] 61,043 at P 125.

    Corporations pay dividends in order to distribute a share of 
their earnings to stockholders. As such, dividends do not include 
any return of invested capital to the stockholders. Rather, 
dividends represent solely a return on invested capital. Put another 
way, dividends represent profit that the stockholder is making on 
its investment. Moreover, corporations typically reinvest some 
earnings to provide for future growth of earnings and thus 
dividends. Since the return on equity which the Commission awards in 
a rate case is intended to permit the pipeline's investors to earn a 
profit on their investment and provides funds to finance future 
growth, the use of dividends in the DCF analysis is entirely 
consistent with the purpose for which the Commission uses that 
analysis. By contrast, as Kern River concedes, the cash 
distributions of the MLPs it seeks to add to the proxy group in this 
case include a return of invested capital through an allocation of 
the partnership's net income. While the level of an MLP's cash 
distributions may be a significant factor in the unit holder's 
decision to invest in the MLP, the Commission uses the DCF analysis 
solely to determine the pipeline's return on equity. The Commission 
provides for the return of invested capital through a separate 
depreciation allowance. For this reason, to the extent an MLP's 
distributions include a significant return of invested capital, a 
DCF analysis based on those distributions, without any adjustment, 
will tend to overstate the estimated return on equity, because the 
'dividend' would be inflated by cash flow representing return of 
equity, thereby overstating the earnings the dividend stream 
purports to reflect.\26\
---------------------------------------------------------------------------

    \26\ Opinion No. 486, 117 FERC ] 61,077 at P. 149-150.

    21. The Commission stated that it could nevertheless consider 
including MLPs in the proxy group in a future case, if the pipeline 
presented evidence addressing these concerns. The discussion in the 
order suggested that such evidence might include some method of 
adjusting the MLPs' distributions to make them comparable to dividends, 
a showing that the higher ``dividend'' yield of the MLP was offset by a 
lower long-term growth projection, or some other explanation why 
distributions in excess of earnings do not distort the DCF results for 
the MLP in question.\27\ However, the Commission concluded that Kern 
River had not presented sufficient evidence to address these issues, 
and that the record in that case did not support including MLPs in the 
proxy group.
---------------------------------------------------------------------------

    \27\ Proposed Policy Statement at P 10-11
---------------------------------------------------------------------------

    22. In addition, Opinion No. 486 pointed out that the traditional 
DCF model only incorporates growth resulting from the reinvestment of 
earnings, not growth arising from external sources of capital.\28\ 
Therefore, the Commission stated that if growth forecasted for an MLP 
comes from external capital, it is necessary either (1) to explain why 
the external sources of capital do not distort the DCF results for that 
MLP or (2) propose an adjustment to the DCF analysis to eliminate any 
distortion.
---------------------------------------------------------------------------

    \28\ Id. at P 152.
---------------------------------------------------------------------------

2. Oil Pipeline Cases
    23. In some oil pipeline rate cases decided before HIOS and Opinion 
No. 486, the Commission included MLPs in the proxy group used to 
determine oil pipeline return on equity on the ground that there were 
no corporations available for use in the oil proxy group.\29\ In those 
cases, no party raised any issue concerning the comparability of an 
MLP's cash distribution to a corporation's dividend. However, that 
issue did arise in the first oil pipeline case decided after HIOS and 
Opinion No. 486, which involved SFPP's Sepulveda Line.\30\ The 
Commission approved inclusion of MLPs in the proxy group in that case 
on the grounds that the included MLPs in question had not made 
distributions in excess of earnings. The order found these facts 
sufficient to address the concerns expressed in HIOS and Opinion No. 
486.
---------------------------------------------------------------------------

    \29\ SFPP, L.P., 86 FERC ] 61,022, at 61,099 (1999).
    \30\ SFPP, L.P., 117 FERC ] 61,285 (2006) (SFPP Sepulveda 
Order), rehearing pending.
---------------------------------------------------------------------------

D. Court Remand of Petal and HIOS

    24. Both Petal and HIOS appealed the Commission's orders in their 
cases to the United States Court of Appeals for the District of 
Columbia Circuit. The court considered the appeals together, and it 
vacated and remanded the proxy group rulings in both cases.\31\ The 
court emphasized that the Commission's ``proxy group arrangements must 
be risk-appropriate.'' \32\ The court explained that this means that 
firms included in the proxy group should face similar risks to the 
pipeline whose ROE is being determined, and any differences in risk 
should be recognized in determining where to place the pipeline in the 
proxy group range of reasonable returns.
---------------------------------------------------------------------------

    \31\ Petal Gas Storage, LLC v. FERC, 496 F.3d 695 (DC Cir. 2007) 
(Petal v. FERC).
    \32\ Petal v. FERC, 496 F.3d at 697, quoting Canadian 
Association of Petroleum Producers v. FERC, 254 F.3d 289 (DC Cir. 
2001).
---------------------------------------------------------------------------

    25. The court recognized that changes in the gas pipeline industry 
compel a change in the Commission's traditional approach to determining 
the proxy group, and the court stated that ``controversy about how it 
should change has been bubbling up in a number of recent cases,'' 
citing both Williston II and Opinion No. 486. But the court found that 
the cases on appeal ``seem[] to represent an arrival point of sorts for 
the Commission,'' pointing out that Opinion No. 486 had reversed an

[[Page 23226]]

administrative law judge for deviating from the HIOS proxy group.\33\
---------------------------------------------------------------------------

    \33\ Opinion No. 486 reversed the ALJ's inclusion of the two 
financially troubled pipelines in the proxy group.
---------------------------------------------------------------------------

    26. The court held that the Commission had not shown that the proxy 
group arrangements it approved in Petal and HIOS were risk-appropriate. 
The court pointed out that the Commission had rejected the inclusion of 
MLPs in the proxy group on the ground that MLP distributions, unlike 
dividends, might provide returns of equity as well as returns on 
equity. While stating that this proposition is not ``self-evident,'' 
the court accepted it for the sake of argument. Nonetheless, the court 
stated that nothing in the Commission's decision explained why the 
companies selected by the Commission for inclusion in the proxy group 
are risk-comparable to HIOS. The court stated that when the goal is a 
proxy group of comparable companies, it is not clear that natural gas 
companies with highly different risk profiles should be regarded as 
comparable.
    27. The court further stated that in placing Petal and HIOS in the 
middle of the proxy group in terms of return on equity, the Commission 
expressly relied on the assumption that pipelines generally fall into a 
broad range of average risk as compared to other pipelines. However, 
the court stated, this assumption is decisive only given a proxy group 
composed of other pipelines. Thus, the court reasoned that if gas 
distribution companies generally face lower risk than gas 
pipelines,\34\ a risk-appropriate placement would be at the high end of 
the group. The court stated that the Commission erred by failing to 
explain how its proxy group arrangements were based on the principle of 
relative risk.
---------------------------------------------------------------------------

    \34\ The court noted that this seems likely.
---------------------------------------------------------------------------

    28. Therefore, the court vacated the Commission's orders with 
respect to the proxy group issue. The court stated that on remand, it 
did not require any particular proxy group arrangement, but stated that 
the overall arrangement must make sense in terms of the relative risk 
and in terms of the statutory command to set just and reasonable rates 
that are commensurate with returns on investments in other enterprises 
having corresponding risks.

II. The Proposed Policy Statement

    29. A month before the court's decision in Petal v. FERC, the 
Commission reached a similar conclusion that its proxy group 
arrangements for gas and oil pipelines must be reexamined. Accordingly, 
on July 19, 2007, the Commission issued a Proposed Policy Statement, in 
which it proposed to modify its policy to allow MLPs to be included in 
the proxy group. The Proposed Policy Statement found that:

    Cost of service ratemaking requires that firms in the proxy 
group be of comparable risk to the firm whose equity cost of capital 
is being determined in a particular rate proceeding. If the proxy 
group is less than clearly representative, this may require the 
Commission to adjust for the difference in risk by adjusting the 
equity cost-of-capital, a difficult undertaking requiring detailed 
support from the contending parties and detailed case-by-case 
analysis by the Commission. Expanding the proxy group to include 
MLPs whose business is more narrowly focused on pipeline activities 
would help provide a more representative proxy group.\35\
---------------------------------------------------------------------------

    \35\ Proposed Policy Statement, 120 FERC ] 61,068 at P 17.

    30. However, the Commission proposed to cap the cash distribution 
used to determine an MLP's return under the DCF method at the MLP's 
reported earnings. The Commission found that this was necessary to 
exclude that portion of an MLP's distributions constituting return of 
equity. The Commission provides for the return of equity through a 
depreciation allowance. Therefore, the Commission stated that the cash 
flows used in the DCF analysis should be limited to those which reflect 
a return on equity. The concern was the pipeline could double recover 
its depreciation expense. The Commission also proposed to require a 
showing that the MLP has had stable earnings over a multi-year period, 
so as to justify a finding that it will be able to maintain the current 
level of cash distributions in future years. The Proposed Policy 
Statement found that these requirements should render the MLP's cash 
distribution comparable to a corporation's dividend for purposes of the 
DCF analysis.
    31. Under the Proposed Policy Statement, the Commission would leave 
to individual cases the determination of which specific MLPs and 
corporations should be included in the proxy group. The Commission 
proposed to apply its final policy statement to all gas and oil cases 
that have not completed the hearing phase as of the date the Commission 
issues its final policy statement. The Commission stated that it would 
consider on a case-by-case basis whether to apply the final policy 
statement in cases that have completed the hearing phase.

III. The Record in the Policy Statement Proceeding

A. Pre-Technical Conference Comments

    32. Twenty-two initial comments and thirteen reply comments were 
filed in response to the Proposed Policy Statement \36\ and fall into 
two categories: (1) Those of gas and oil pipelines and the related 
trade associations (Pipeline Interests),\37\ and (2) those of gas and 
oil producers and shippers, public and municipal utilities, state 
public service commissions, and related trade associations (Customer 
Interests).\38\ Two comments were also submitted by individuals in 
their business or personal capacity.\39\
---------------------------------------------------------------------------

    \36\ Comments related to the technical conference are discussed 
infra and are characterized as conference comments or conference 
reply comments.
    \37\ The Pipeline Interests include: the Association of Oil Pipe 
Lines (AOPL); El Paso Corporation (El Paso); Enbridge Energy 
Partners, L.P. (Enbridge); the Interstate Natural Gas Association of 
America (INGAA); MidAmerican Energy Pipeline Group (MidAmerican); 
the National Association of Publicly Traded Partnerships (NAPTP); 
Panhandle Energy Pipelines (Panhandle); Spectra Energy Transmission, 
LLC (Spectra); TransCanada Corporation (TransCanada); and Williston 
Basin Interstate Pipeline Company (Williston).
    \38\ The Customer Interests include: The American Gas 
Association (AGA); the America Public Gas Association (APGA); the 
Air Transport Association of America; the Canadian Association of 
Petroleum Producers (CAPP); Indicated Shippers (consisting of Area 
Energy, LLC, Anadarko E&P Company LP, Anadarko Petroleum 
Corporation, Chevron USA Inc., Coral Energy Resources LP, Occidental 
Energy Marketing Inc., and Shell Rocky Mountain Production, LLC); 
the Natural Gas Supply Association (NGSA); the Process Gas Consumers 
Group; the Public Service Commission of New York (PSCNY); Tesoro 
Refining and Marketing Company (Tesoro); the Northern Municipal 
Distributors Group (NMDG) and the Midwest Region Gas Task Force 
Association filing jointly; and the Society for the Preservation of 
Oil Shippers (Society).
    \39\ The individual comments include Crowley Energy Consulting, 
supporting the Customer Interests, and Barry Gleicher, supporting 
the Pipeline Interests.
---------------------------------------------------------------------------

    33. The comments focus on three issues: (1) Whether MLPs should be 
included in the gas pipeline proxy group at all; (2) whether the 
proposed cap on the MLP cash distributions used in the DCF analysis is 
necessary or adequate; and (3) whether the short- and long-term growth 
component of the DCF model should be modified given the financial 
practices of MLPs. Secondary points include the potential distorting 
effects of: MLP tax treatment, the large payouts by MLPs, the general 
partner's incentive distribution rights (IDRs), and the relative 
returns to the limited and general partners.
    34. All parties recognize that MLPs are the only available entities 
for inclusion in the oil pipeline proxy group. The Pipeline Interests 
also all assert that the Commission correctly

[[Page 23227]]

proposed to include MLPs in the gas pipeline proxy group. In contrast, 
most of the Customer Interests assert that there are enough 
corporations available for inclusion in the gas pipeline proxy group 
and that there is no need to include MLPs.
    35. Both the Pipeline and Customer Interests question the proposed 
earnings cap on MLP distributions, with the Pipeline Interests 
asserting the cap is unnecessary and the Customer Interests asserting 
the cap should be lower. The Pipeline Interests assert that an MLP's 
share price reflects investors' projection of all cash flows it will 
receive from the MLP, including distributions in excess of earnings. 
Therefore, any cap on the distributions while still using a dividend 
yield reflecting the full share price would lead to distorted 
results.\40\ The Customer Interests agree that the adjustment to MLP 
distributions is necessary to remove a double count attributed to 
depreciation, but they also uniformly assert that the proposed 
adjustment is inadequate to compensate for a wide range of financial 
factors that distinguish MLPs from Schedule C corporations.
---------------------------------------------------------------------------

    \40\ AOPL initial comments at 8, 10; INGAA initial comments at 
13-14; Spectra initial comments at 4; NAPTP initial comments at 4; 
NAPTP initial comments at 4.
---------------------------------------------------------------------------

    36. On the growth rate issue, the Pipeline Interests in their 
initial comments generally agree that, if MLPs have greater 
distributions than a corporation, then the MLP may have less growth 
potential than a corporation. However, they argue that this fact does 
not require any additional adjustment, since any lower growth potential 
would be reflected in a reduced IBES growth forecast. The Pipeline 
Interests also state that distributions in excess of earnings do not 
prevent reinvestment or organic growth. They assert that pipeline MLPs 
have ready access to capital markets given their stable cash flows and 
the projected expansion of the pipeline system, which can be the basis 
for organic growth.\41\
---------------------------------------------------------------------------

    \41\ AOPL comments at 21-24 and attachments; Enbridge Energy 
reply comments at 5; INGAA comments at 22-24; TransCanada reply 
comments at 8-10.
---------------------------------------------------------------------------

    37. In contrast, the Customer Interests assert that MLPs have 
significantly lower growth potential than corporations due to their 
distributions in excess of earnings, particularly over the long 
term.\42\ They cite studies by established investment firms suggesting 
that the long term growth potential of MLPs is less than the long term 
growth factor now included in the DCF model. Moreover, they argue that 
given the high level of MLP distributions and declining opportunities 
for acquisitions with high returns, MLP growth must now come from 
investment of external funds in projects that will enhance organic 
growth of existing business lines.\43\
---------------------------------------------------------------------------

    \42\ APGA reply comments at 11-15; CAPP initial comments at 1; 
CAPP reply comments at 6-7, and attachment at 3-4; NYPSC initial 
comments at 19-21, 23, including attachments of financial materials 
from major investment houses; NYPSC reply comments at 4-7; Tesoro 
reply comments at 25-27.
    \43\ Id.
---------------------------------------------------------------------------

    38. Some of the Customer Interests further argue that there are 
inadequate investment opportunities to support capital investment, and 
in the relatively near future the present level of MLP distributions 
will be maintained only by borrowing or issuing additional limited 
partners' units.\44\ Therefore, they argue, sustainability of MLP 
growth is a major issue that must be examined in rate proceedings as 
this implies a lower equity cost-of-capital component in the pipeline's 
rate structure.\45\ The Customer Interests also assert that the 
Commission's traditional DCF model has never permitted the inclusion of 
externally generated funds in the growth component of the model. Thus, 
to the extent the IBES projections include such external funds, they 
assert that this compromises the forecasts.
---------------------------------------------------------------------------

    \44\ Crowley Energy Consultant initial comments; Society at 5-6.
    \45\ Id.
---------------------------------------------------------------------------

    39. Finally, NGSA urge the Commission to initiate a new proceeding 
to consider alternatives to the DCF methodology for determining gas 
pipeline ROEs. AGA requests a technical conference to discuss the 
issues further, which as noted, the Commission granted with regard to 
the growth factors.\46\ Two commenters assert that any change in policy 
should apply prospectively and should not apply to proceedings for 
which the hearing record is completed, e.g., the Kern River 
proceeding.\47\
---------------------------------------------------------------------------

    \46\ AGA initial comments at 8.
    \47\ Id. at 8, 25; NGSA initial comments at 3, 11.
---------------------------------------------------------------------------

B. Technical Conference and Post-Technical Conference Comments

    40. After review of the initial comments summarized above, the 
Commission issued a supplemental notice on November 15, 2007, 
requesting additional comments solely on the issue of MLP growth rates, 
and establishing a technical conference to discuss that issue. The 
technical conference was held on January 23, 2008. The Commission 
concluded that supplementing the record before the Commission could 
resolve the issue of how to project MLP growth rates assuming that the 
Commission ultimately decides to permit the use of MLPs in the proxy 
group. The Commission focused the technical conference on the 
appropriate method for determining MLP growth and, in particular, that 
which should be used if the Commission did not cap the distributions 
used to determine the dividend yield. Thus, whether to include MLPs in 
the proxy group or to limit the distributions to earnings were not 
issues before the technical conference. The technical conference was 
transcribed for use in the record herein.
    41. Thirteen parties submitted comments in response to the November 
15 notice, on three main topics: (1) The short-term growth component; 
(2) the long-term growth component; and (3) the weighting of these two 
components.\48\ Of these, eight parties requested to participate on the 
panels and the Commission accepted all of the individuals proffered by 
these parties.\49\ To summarize, two of the panelists represented 
parties that continued to assert that MLPs should not be included in 
the ROE proxy group.\50\ More consistent with the premise of the 
conference, three panelists stated that there needed to be an 
adjustment to the long term GDP component the Commission currently uses 
in its DCF model.\51\ Two stated that MLPs would grow at a slower rate 
than corporations in the long-term phase of growth. However, six other 
panelists asserted that an MLP as a whole could grow as fast as a 
corporation in the terminal phase, but most conceded that the use of 
incentive distribution rights (IDRs) \52\ would cause the limited 
partnership interests to grow at slower rate than the

[[Page 23228]]

MLP as a whole.\53\ In addition, three panelists questioned the 
reliability of the IBES forecasts for use in developing the short-term 
projection\54\ and one stated that the longer term growth component of 
the formula should be weighted at no greater than 10 percent.\55\
---------------------------------------------------------------------------

    \48\ APGA, AOPL, CAPP, Enbridge, INGAA, MidAmerica, NAPTP, NGSA, 
PSNYC, State of Alaska, Tesoro, TransCanada, and Williston.
    \49\ Professor J. Peter Williamson on behalf of the Association 
of Oil Pipelines, Mr. J. Bertram Solomon on behalf of the American 
Public Gas Association, Mr. Michael J. Vilbert on behalf of the 
Interstate Natural Gas Association of America, Mr. Park Shaper and 
Mr. Yves Siegel on behalf of the National Association of Publicly 
Traded Partnerships, Mr. Patrick Barry on behalf of the Public 
Service Commission of New York, Mr. Thomas Horst on behalf of the 
State of Alaska, and Mr. Paul Moul on behalf of TransCanada 
Corporation.
    \50\ PSCNY and APGA. CAPP, NGSA, and Tesoro supported this 
position but did not participate on the panel.
    \51\ PSCNY, APGA, and State of Alaska as well as the NGSA.
    \52\ As discussed further below, an incentive distribution 
provision in an MLP partnership agreement provides for an increasing 
large percentage of distributions to the general partner as the cash 
distributions per limited partnership share increase over time. The 
maximum incentive distribution to the general partner varies with 
the partnership agreement, but may be as high as 47 percent.
    \53\ Two spoke for NAPTP and one each for AOPL, INGAA, the State 
of Alaska, and TransCanada. Williston, Enbridge, and MidAmerican 
also asserted that there is no reason to conclude the growth would 
not at least equal GDP. They did not speak to the issue of the 
limited partner growth rate that might be lower as a result of the 
incentive distributions to the general partner.
    \54\ APGA, PSCNY, and State of Alaska.
    \55\ TransCanada, Additional Comments dated December 21 at 12.
---------------------------------------------------------------------------

IV. Discussion

    42. Based on its review of all the comments and the record of the 
technical conference, the Commission is adopting the following policy 
concerning the composition of the natural gas pipeline and oil pipeline 
proxy groups: (1) Consistent with the Proposed Policy Statement, the 
Commission will permit MLPs to be included in the proxy group for both 
gas and oil pipelines; (2) the proposed earnings cap on the MLPs' 
distributions will not be adopted; and (3) the Commission will use the 
same DCF analysis for MLPs as for corporations, except that the long-
term growth projection for MLPs shall be 50 percent of projected growth 
in GDP.

A. Whether To Include MLPs in the Gas and Oil Pipeline Proxy Groups

1. Comments
    43. The first issue is whether to include MLPs in the proxy group 
used to determine a pipeline's return on equity. No commenter contests 
the Commission's statement that, in oil pipeline proceedings, MLPs are 
the only firms available for inclusion in the proxy group.\56\ In 
addition, the Pipeline Interests all assert that the Commission 
correctly proposed to include MLPs in the gas pipeline proxy group. 
They agree with the Commission that this will result in a more 
representative proxy group that reflects long-term trends within the 
gas pipeline industry and assert that the resulting returns will 
encourage further investment in both the gas and oil pipeline 
industries. Including MLPs in the proxy group would reduce the need for 
difficult adjustments to projected equity returns to accommodate 
differences in risk among the different types of firms that might 
reasonably be included in the proxy group.
---------------------------------------------------------------------------

    \56\ AOPL initial comments at 5. Tesoro initial comments at 2. 
See also Society initial comments addressing the possible inclusion 
of oil pipeline MLPs in the proxy group.
---------------------------------------------------------------------------

    44. In contrast, most of the commenters representing the Customer 
Interests assert that there are enough corporations available for 
inclusion in the gas pipeline proxy group that there is no need to 
include MLPs. They further argue that the differences between the MLP 
and corporate business model render any use of MLPs inconsistent with 
the DCF model. APGA expressly states that the Commission should abandon 
the Proposed Policy Statement.\57\
---------------------------------------------------------------------------

    \57\ APGA initial comments at 14.
---------------------------------------------------------------------------

    45. The NMDG asserts that the Commission has not established that 
there is any reason to issue the Policy Statement or to relieve a 
pipeline applicant of the burden of establishing why any MLPs should be 
included in the proxy group. In this vein, Indicated Shippers assert 
that the Commission should consider alternative procedures for defining 
the proxy group, and that the improvement in El Paso Natural 
Corporation's and the William Company's financial situation and the 
creation of the Spectra Group suggest that the corporate gas proxy 
group is becoming more representative.
    46. Finally, NGSA urges the Commission to initiate a new proceeding 
to consider alternatives to the DCF methodology for determining gas 
pipeline ROEs. NGSA generally supports including MLPs in the proxy 
group, subject to adjustments, as a means of continuing to use the DCF 
method on a temporary basis. But it argues that a better long-term 
solution to determining gas pipeline ROEs would be to stop using the 
DCF method, and instead adopt a risk premium approach to determining 
ROE. It asserts that the risk premium approach is used in Canada and 
does not require adjustments to account for variations in corporate 
structure.\58\ INGAA states in its reply comments that the DCF 
methodology is not necessarily the only financial model that may be 
used, and asks the Commission to clarify that parties may propose other 
approaches in individual rate cases.\59\
---------------------------------------------------------------------------

    \58\ NGSA initial comments at 13-15.
    \59\ INGAA reply comments at 18.
---------------------------------------------------------------------------

2. Discussion
    47. As the Commission pointed out in the proposed policy statement, 
the Supreme Court has held that ``the return to the equity owner should 
be commensurate with the return on investment in other enterprises 
having corresponding risks. That return, moreover, should be sufficient 
to assure confidence in the financial integrity of the enterprise, so 
as to maintain its credit and to attract capital.'' \60\ In order to 
attract capital, ``a utility must offer a risk-adjusted expected rate 
of return sufficient to attract investors.'' \61\ In other words, the 
utility must compete in the equity markets to obtain capital.
---------------------------------------------------------------------------

    \60\ FPC v. Hope Natural Gas Co., 320 U.S. 591, 603 (1044).
    \61\ CAPP, 254 F.3d at 293.
---------------------------------------------------------------------------

    48. The Commission performs a DCF analysis of publicly-traded proxy 
firms to determine the return on equity that markets require a pipeline 
to give its investors in order for them to invest their capital in the 
pipeline. As the court explained in Petal Gas Storage, LLC v. FERC, the 
purpose of the proxy group is to ``provide market-determined stock and 
dividend figures from public companies comparable to a target company 
for which those figures are unavailable. Market-determined stock 
figures reflect a company's risk level and when combined with dividend 
values, permit calculation of the `risk-adjusted expected rate of 
return sufficient to attract investors.' '' \62\ It is thus crucial 
that the firms in the proxy group be comparable to the regulated firm 
whose rate is being determined. In other words, as the court emphasized 
in Petal, the proxy group must be ``risk-appropriate.'' \63\
---------------------------------------------------------------------------

    \62\ Petal, 496 F.3d at 697, quoting Canadian Association of 
Petroleum Producers v. FERC, 254 F.3d 289 (DC Cir. 2001).
    \63\ Id. 6.
---------------------------------------------------------------------------

    49. The Commission continues to believe that including MLPs in the 
gas and oil proxy groups will, as required by Petal, make those proxy 
groups more representative of the business risks of the regulated firm 
whose rates are at issue. While there has been some modest expansion of 
the number of publicly-traded diversified natural gas companies that 
could be included in the proxy group, this does not change one basic 
fact. This is that more and more gas pipeline assets are being 
transferred to publicly-traded MLPs, whose business is narrowly focused 
on pipeline activities. As a result, these MLPs are likely to be more 
representative of predominantly pipeline firms than the diversified gas 
corporations still available for inclusion in a proxy group. As such, 
including MLPs in the gas pipeline proxy group should render the proxy 
group more ``risk-appropriate,'' consistent with Petal. Moreover, MLPs 
are the only publicly traded ownership form for oil pipelines and are 
the most representative group for determining the equity cost of 
capital for oil pipelines.

[[Page 23229]]

    50. As the court also emphasized in Petal, when a proxy group is 
less than clearly representative, there may be a need for the 
Commission to adjust for the difference in risk by adjusting the equity 
cost-of-capital, a difficult undertaking requiring detailed support 
from the contending parties and detailed case-by-case analysis by the 
Commission. Expanding a proxy group to include MLPs whose business is 
more narrowly focused on pipeline activities should help minimize the 
need to make adjustments, because the proxy group should be more 
representative of the regulated firms whose rates are at issue.
    51. While this Policy Statement modifies Commission policy to 
permit MLPs to be included in the proxy group, the Commission is making 
no findings at this time as to which particular corporations and/or 
MLPs should be included in the gas or oil proxy groups. The Commission 
leaves that determination to each individual rate case. In order to 
assist the Commission in determining the most representative possible 
proxy group in those cases, the parties and other participants should 
provide as much information as possible regarding the business 
activities of each firm they propose to include in the proxy group, 
including their recent annual SEC filings and investor service analyses 
of the firms. This information should help the Commission determine 
whether the interstate natural gas or oil pipeline business is a 
primary focus of the firm and whether investors view an investment in 
the firm as essentially an investment in that business. While the 
Commission is not precluding use of diversified corporations or MLPs in 
the proxy group, the probable difference in the risk of the natural gas 
pipeline business and the risk profile of a diversified gas corporation 
with substantial local distribution activities has been highlighted by 
the parties and specifically recognized by the court in Petal.\64\
---------------------------------------------------------------------------

    \64\ Id. at 6-7.
---------------------------------------------------------------------------

    52. As discussed further below, the Commission recognizes that 
there are significant differences in the cash flows to investors and 
growth rates of corporations and MLPs. However, as discussed below, the 
Commission believes that those issues may be accounted for in a 
correctly performed DCF analysis, and therefore these differences do 
not preclude inclusion of MLPs in the proxy group.
    53. Finally, the Commission has concluded that it will not explore 
other methods of determining the equity cost of capital at this time. 
The DCF model is a well established method of determining the equity 
cost of capital,\65\ and other methods such as the risk premium model 
have not been used by the Commission for almost two decades. In the 
Commission's judgment, the uncertainty that would be created by 
reopening its procedures to include other approaches outweighs any 
limitations in its current pragmatic approach to the financial 
characteristics of MLPs. Therefore the alternatives suggested by 
certain of the parties will not be pursued further here. Nothing 
submitted at the January 23rd technical conference warrants different 
conclusions.
---------------------------------------------------------------------------

    \65\ See Illinois Bell Telephone Co. v. FCC, 988 F.2d 1254, 1259 
n. 6 (DC Cir. 1993), stating, ``The DCF method `has become the most 
popular technique of estimating the cost of equity, and it is 
generally accepted by most commissions. Virtually all cost of 
capital witnesses use this method, and most of them consider it 
their primary technique.' '' quoting J. Bonbright et al., Principles 
of Public Utility Regulation 318 (2d ed. 1988).
---------------------------------------------------------------------------

B. The Proposed Adjustment to MLP Cash Distributions

1. Comments
    54. Both the Pipeline and Customer Interests attack the proposed 
earnings cap on MLP distributions, with the Pipeline Interests 
asserting the cap is unnecessary and the Customer Interests asserting 
the cap should be lower. The Pipeline Interests assert that there is no 
need to adjust the distributions included in the DCF model. They argue 
that investors include all cash flows that are generated by an MLP in 
applying a DCF model and do not distinguish between a return of 
investment and a return on investment \66\ since depreciation is an 
accounting concept that is used to calculate an MLP's earnings that is 
not relevant to determining the cash flows included in a DCF 
analysis.\67\ The Pipeline Interests further assert that an unadjusted 
DCF calculation does not result in the double recovery of the 
depreciation component of an MLP's cost-of-service.\68\
---------------------------------------------------------------------------

    \66\ AOPL initial comments at 16, 18; Spectra Energy initial 
comments at 14; NAPTP initial comments at 3.
    \67\ INGAA initial comments at 5-6, 15-18; NAPTP initial 
comments at 4-5; MidAmerican initial comments at 5; Panhandle 
initial comments at 3 and attachment; Williston initial comments at 
11.
    \68\ INGAA initial comments at 15-17 and 20-21.
---------------------------------------------------------------------------

    55. Moreover, the Pipeline Interests assert that, because all parts 
of the DCF model are linked, if the distribution component is reduced, 
this will necessarily affect the growth component of the model. They 
assert that any adjustment limiting the distributions used to earnings 
will result in below market returns to investors and thus any such 
adjustment is arbitrary.\69\ As an alternative, they suggest that if an 
MLP's distributions are unrepresentative, it is wiser to exclude that 
MLP from the sample as an outlier.\70\ They further assert there have 
been corporations in the proxy group that have distributed dividends in 
excess of earnings for years and the Commission has never required an 
adjustment.\71\ They claim that in any event there are practical 
problems with an earnings cap because earnings are reported quarterly 
(unlike distributions which are reported monthly) and such reports are 
unedited and may require seasonal adjustments.\72\
---------------------------------------------------------------------------

    \69\ AOPL initial comments at 8, 10; INGAA initial comments at 
13-14; Spectra initial comments at 4; PAPTP initial comments at 4.
    \70\ INGAA initial comments at 13; Spectra Energy initial 
comments at 5, 19-20.
    \71\ INGAA initial comments at 18; MidAmerica initial comments 
at 6.
    \72\ AOPL initial comments at 24-25; Spectra Energy initial 
comments at 17-18.
---------------------------------------------------------------------------

    56. The Customer Interests support the Commission's initial 
conclusion that an adjustment to MLP distributions is necessary to 
remove a double count attributed to depreciation, but they also 
uniformly assert that the proposed adjustment is inadequate to 
compensate for a wide range of financial factors that distinguish MLPs 
from Schedule C corporations. Thus, they assert that further 
adjustments to the distributions should be made to reflect the tax 
advantages that flow to MLPs,\73\ the alleged distortions that result 
from incentive distributions to the general partner,\74\ and the fact 
that distributions may also include cash derived from the sale of 
assets, bond issues, and the issuance of further limited partnership 
units.\75\ Several also assert that for an MLP's distribution to be 
comparable to that of a corporation, the percentage of the MLP's 
distribution included in the DCF model should be no higher than the 
percentage of earnings corporations typically include in their dividend 
payments, or about 60 percent.\76\ Finally, to the extent that INGAA 
and others assert that depreciation is not a direct source of cash flow 
for distribution, the Customer Interests cite to investor literature 
and MLP filings

[[Page 23230]]

with the SEC disclosure that state exactly the opposite.\77\
---------------------------------------------------------------------------

    \73\ Crowley Energy at 2; Indicated Shippers initial comments at 
24; PSCNY initial comments at 12-13; Society initial comments, 
passim.
    \74\ APGA at 7-8; Crowley Energy at 2; Indicated Shippers 
comments at 24; NGSA at 6; Society initial comments passim.
    \75\ Crowley Energy initial comments; Society, passim; Tesoro 
reply comments at 26.
    \76\ CAPP initial comments at 3, 6; Indicated Shippers initial 
comments at 23; PSCNY initial comments at 6; Tesoro initial comments 
at 15.
    \77\ APGA initial comments at 11; CAPP reply comments at 3-4; 
NGSA reply comments at 9-10; Tesoro reply comments at 19-21.
---------------------------------------------------------------------------

2. Discussion
    57. The Commission concludes that a proposed earnings cap on the 
MLP distributions that would be included in the DCF model should not be 
adopted. On further review, the Commission concludes that its concern 
with the distinction between return on capital and return of capital 
improperly conflates cost-of-service rate-making techniques with the 
market-driven DCF method used for determining the pipeline's cost of 
obtaining capital in the equity markets. This is inconsistent with the 
DCF model's internal structure.
    58. The fundamental premise of the DCF model is that a firm's stock 
price should equal the present value of its future cash flows, 
discounted at a market rate commensurate with the stock's risk. No 
commenter seriously contends that an investor would distinguish between 
cash flows attributable to return on capital, and those attributable to 
return of capital, in performing a DCF analysis. In short, under the 
DCF model, all cash flows, whatever their source, contribute to the 
value of stock. The Commission agrees that, since the DCF model uses 
the total unadjusted cash flows to determine a stock's value, it is 
theoretically inconsistent to use lower adjusted cash flows when using 
the DCF model to determine the return required by investors purchasing 
the stock.
    59. More specifically, the investor first determines what risk 
should be attributed to a prospective investment and the related return 
that would be required in order to make the investment. For example, 
the investor may conclude that the minimum return from the investment 
must be 10 percent on equity. The investor then looks at the total cash 
flows from all sources over time, including the current distribution 
(or dividend) and its projected growth. The DCF model yields a price 
for the share that reflects the present value of those cash flows at 
the discount rate.
    60. In contrast, the Commission solves the DCF formula for the 
return required by the investor, not the price of the stock. This 
results in the Commission calculating the proxy firm's ROE as the sum 
of (1) the proxy firm's dividend yield and (2) the projected growth 
rate. The Commission determines dividend yield by dividing the proxy 
firm's cash distribution (or dividend) by its current stock price. As 
the court in Petal pointed out, both the stock price and distribution 
(or dividend) figures of the proxy firms are market-determined. 
Moreover, an investor's projection of the MLP's growth prospects would 
be affected by the actual level of its distributions, with 
distributions in excess of earnings generally perceived as reducing the 
growth projection because less cash flow is available for reinvestment 
in the firm.\78\ The pipeline industry generally acknowledged this fact 
in earlier rate proceedings as well as in this proceeding, or at least 
until its later phases.\79\ As illustrated in Appendix B to this Policy 
Statement, a DCF analysis using market-determined inputs for each of 
the variables in the DCF formula appropriately determines, consistent 
with Petal, the percentage return on equity a pipeline must offer in 
the equity market in order to attract investors, whether the proxy 
firms are corporations or MLPs.
---------------------------------------------------------------------------

    \78\ Because a corporation typically retains a portion of its 
earnings, general financial theory suggests that it is able to use 
internally generated funds to obtain a higher growth rate. An MLP's 
higher level of distributions theoretically produces a lower 
projected growth rate. In fact, the most recent IBES projections for 
the four corporations included in the gas pipeline proxy group in 
Appendix A average 10.5 percent, while the IBES growth projections 
for the six MLPs average only 6.67 percent.
    \79\ See AOPL Initial Comments, Williamson Aff. at 6-7; AOPL 
Reply Comments at 6-7; Panhandle Initial Comments, Attachment dated 
August 30, 2007, Analysis of the Use of MLPs in the Group of Proxy 
Companies Used For Determining Gas and Oil Pipeline Return on Equity 
at 10-11; Transwestern Pipeline Company, LLC, Docket No. RP06-614-
000, Ex. TW-56 filed September 29, 2006, at 23-24; High Island 
Offshore System, LLC, Docket No. RP96-540-000, Ex. HIO-73 filed 
August 26, 2006 at 28-29; Texaco Refining and Marketing Inc, et al. 
v. SFPP, L.P., Docket No. OR96-2-012, Ex. SEP SFPP-56 dated February 
14, 2005 at 9-10; Mojave Pipeline Company, Docket No. RP07-310-000, 
Ex. MPC-70 dated February 2, 2007 at 28-32 (including tables and 
charts on the relative growth rates of corporations and MLPs); Kern 
River Gas Transmission Company, Docket No. RP04-274-000, Ex. KR-107 
at 17.
---------------------------------------------------------------------------

    61. If the Commission were to cap the distribution used to 
determine an MLP's dividend yield at below the market-determined level, 
but use the actual market price of the MLP's publicly traded units and 
a growth projection reflecting the actual level of distributions, the 
DCF analysis would fail to achieve its intended purpose of determining 
the return the equity market requires in order to justify an investment 
in the pipeline. That is because there would be a mismatch among the 
inputs the Commission used for the variables in the DCF formula. The 
DCF analysis presumes that the market value of an MLP's units is a 
function of the entire present and future cash flow provided by an 
investment in those units. Given this interlocking nature of the 
variables in the DCF formula, INGAA and the other pipeline commenters 
are correct that limiting the distribution input to earnings, while 
using market values for the other inputs to the DCF formula, would 
result in the calculation of a return below that implied in the share 
price.\80\
---------------------------------------------------------------------------

    \80\ The earnings cap on the distribution would artificially 
reduce an MLP's dividend yield below that assumed by the investor in 
valuing the stock. Adding the artificially reduced dividend yield to 
a growth projection that reflects the MLP's reduced growth prospects 
due to its high actual distributions would inevitably result in an 
ROE lower than that actually required by the market.
---------------------------------------------------------------------------

    62. In addition, use of a proxy MLP's full distribution in 
determining ROE will not cause a double recovery of the depreciation 
component included in the pipeline's cost-of-service rates. In a rate 
case, the Commission determines the dollar amount of the ROE component 
of the cost-of-service of the pipeline filing the rate case by 
multiplying (1) the percentage return on equity required by the market 
by (2) the actual rate base of the pipeline in question. Having found 
that use of a proxy MLP's full distribution is necessary for the DCF 
analysis to accurately determine the percentage return on equity 
required by the equity markets, it necessarily follows that the same 
percentage should be used in determining the dollar amount of the ROE 
component of the pipeline's cost of service. Awarding the pipeline an 
ROE allowance based on that percentage of its own rate base will give 
the pipeline an opportunity to provide its investors with the return on 
their investment required by the market. Such an ROE allowance does not 
implicate the separate depreciation allowance the Commission also 
includes in a pipeline's cost of service to provide for return of 
investment.
    63. The Commission therefore concludes that it is not analytically 
sound to cap the distributions to be included in the DCF model by the 
MLP's earnings. As discussed below, the record is more convincing that 
if any adjustment is required, this issue centers on the projected 
growth of the MLPs. Given this, it is not necessary to discuss the 
appropriate level for any earnings cap.
    64. Having concluded that an earnings cap adjustment would be 
inappropriate, the Commission also concludes that it is not necessary 
to address the long term sustainability of MLPs as a whole, or those of 
the particular MLP whose rates are under review. As has been discussed, 
the DCF model has two components. One is the cash distribution in the 
current period and

[[Page 23231]]

the second is the discounted value of the anticipated growth in that 
distribution. The increase in distribution is driven by the anticipated 
growth in earnings that generates the cash to be used for the 
distribution. If projected earnings suggest that the distribution 
cannot be sustained, this will be reflected in the projected cash flow 
for the firm and ultimately the MLP unit price.\81\ In this regard, 
some MLPs will inevitably do better and others not as well, and from 
the Commission's point of view, this will be reflected in the required 
rate of return developed by the DCF model.
---------------------------------------------------------------------------

    \81\ The investor requires a minimum return that reflects the 
perceived risk of the investment. Thus, if the cash flows decline, 
so will the price of the stock assuming the percentage return 
required remains the same.
---------------------------------------------------------------------------

    65. For this reason, as the Pipeline Interests suggest, if an MLP's 
financial condition or growth rate is outside the norm for the 
industry, or is unrepresentative, the best way to deal with this issue 
is to exclude that particular MLP from the proxy group sample, just as 
the Commission has done with unrepresentative diversified gas 
corporations. Finally, the Commission has previously held that the 
issue of whether MLPs are an appropriate investment vehicle for the 
pipeline industry as a whole is a matter that is best left for 
Congress, the body that authorized MLPs in the first instance. Thus the 
Commission will not address that issue, or the appropriateness of the 
tax deferral aspects of MLPs further in this proceeding.\82\ Nothing 
presented at the technical conference warrants different conclusions.
---------------------------------------------------------------------------

    \82\ See SFPP, L.P., 121 FERC ] 61,240, at P 20-61 (2007) for an 
extensive discussion of these income tax allowance and tax deferral 
policy issues relating to MLPs. Moreover, any tax advantages are 
normally reflected in the MLP unit price. See also INGAA Reply 
Comments at 12-13; MidAmerica, Reply Comments at 4-5; AOPL Reply 
Comments at 11-12; Tr. 121-22; AOPL Post-Technical Conference 
Comments at 14.
---------------------------------------------------------------------------

    66. The Commission now turns to the issue of how to project the 
growth rates of MLPs. For the reasons discussed below, the Commission 
finds that the differences between MLPs and corporations, and 
particularly the MLPs' lower growth prospects due to their 
distributions in excess of earnings, are appropriately accounted for in 
the growth projection component of the DCF model.

C. The Short-Term Growth Component

    67. This section of the Policy Statement discusses whether changes 
should be made to the short-term growth component of the DCF model. For 
the short-term growth estimate the Commission currently uses security 
analysts' five-year forecasts for each company in the proxy group, as 
published by IBES. IBES is a service that monitors the earnings 
estimates on over 18,000 companies of interest to institutional 
investors. More than 850 firms contribute data to IBES to be used in 
its projections and the information is provided on a subscription 
basis.
1. Comments
    68. The Pipeline Interests support the continued use of five-year 
IBES forecasts for short-term growth projections in the DCF model with 
regard to MLPs. In general, they argue that, while no growth forecast 
is perfect, IBES provides the best available information regarding what 
investors expect in companies. They state that IBES estimates are 
unbiased and publicly available. They add that since IBES estimates are 
company-specific, they already adjust for any differences among the 
entities analyzed, including whether the company is organized as an MLP 
or corporation.
    69. For example, NAPTP supports the IBES estimates because the 
various items that may affect the growth rate expected by the market, 
such as the effect of IDRs to the general partner, are already factored 
into IBES projections.\83\ Williston Basin argues that since IBES data 
is drawn from many financial analysts, and since the information is 
widely accepted in the financial industry, use of IBES helps reduce 
subjectivity when estimating appropriate short-term growth 
forecasts.\84\ TransCanada acknowledges that IBES may underestimate 
short-term growth for MLPs, but argues that modifying IBES would only 
further understate short-term growth rates and compound any problems 
brought on by trying to estimate growth for MLPs.\85\ The AOPL 
similarly argues that studies have shown that IBES estimates understate 
short-term growth rates for MLPs and therefore the growth projections 
are conservative.\86\
---------------------------------------------------------------------------

    \83\ NAPTP, Initial Technical Conference Comments at 3.
    \84\ Williston, Additional Comments dated December 21 at 2.
    \85\ TransCanada, Additional Comments dated December 21 at 12-
13.
    \86\ AOPL, Initial Technical Conference Comments at 5, 
Williamson Post-Technical Conference Aff. at 3, 8.
---------------------------------------------------------------------------

    70. However, certain parties recommend that the Commission 
discontinue using IBES estimates for MLPs to project short-term growth 
rates in its DCF model. These parties argue there is considerable 
uncertainty of whether the individual forecasts IBES is reporting 
reflect earnings growth or distribution growth. The State of Alaska 
asserts that IBES growth estimates of distributions per share are 
incomplete and unreliable for use in the DCF calculation. It argues 
that there are not a sufficient number of stock analysts providing IBES 
with distribution per share growth estimates to get a reliable estimate 
for the purposes of calculating the cost of equity for pipeline 
companies. Speaking for the State of Alaska, Dr. Thomas Horst notes 
that of the 37 gas and oil companies he examined data for, there was 
not a single case where IBES received two or more estimates of 
distributions per share growth rates.\87\
---------------------------------------------------------------------------

    \87\ State of Alaska, Reply Comments dated February 20 at 5.
---------------------------------------------------------------------------

    71. APGA states that through communications with personnel at 
Thompson Financial, the owner of IBES and the publisher of its 
forecasts, it verified that the five-year analysts' growth rate 
projections reported by IBES for MLPs are projections of earnings per 
unit, and not distributions per unit.\88\ PSCNY also considers IBES 
projections unreliable, since they do not account for such parameters 
as IDRs. It questions whether analysts can truly estimate MLP growth 
beyond two years. It also questions whether lower earnings retention 
necessarily would translate into lower short-term IBES growth rates 
relative to corporations.\89\ CAPP expresses concerns that the analysts 
that produce IBES growth estimates continue to be concentrated within 
the same financial institutions that also underwrite the securities of 
the subject companies, invest in those securities, and furnish other 
financial services to the subject enterprises \90\ and also notes the 
uncertainty of whether the forecasts are for earnings or 
distributions.\91\
---------------------------------------------------------------------------

    \88\ APGA, Reply Technical Conference Comments at 5-6.
    \89\ NYPSC Initial Technical Conference Comments at 5-6.
    \90\ CAPP Supplemental Comments dated December 21 at 3-4.
    \91\ CAPP Initial Technical Conference Comments at 7.
---------------------------------------------------------------------------

    72. However AOPL maintains that historical records confirm that 
what analysts actually report to IBES is distribution growth. It adds 
that Yves Siegel, Wachovia's representative, confirmed that Wachovia 
provides projected MLP distribution growth to IBES, and not earnings 
growth.\92\ NAPTP asserts that, for projecting the short-term growth 
rates of MLPs, the Commission should use analysts'

[[Page 23232]]

forecasts of growth in the MLP's distributable cash flow for all of its 
equity holders and that, while not perfect, this is the best 
information that is available.\93\
---------------------------------------------------------------------------

    \92\ AOPL Initial Technical Conference Comments at 4-5.
    \93\ NAPTP Post-Technical Conference Comments at 1-3.
---------------------------------------------------------------------------

2. Discussion
    73. The Commission's longstanding policy is to use security 
analysts' five-year growth forecasts as reported by IBES to determine 
the short-term growth rates for each proxy company. In Opinion No 414-
A,\94\ the Commission explained that the growth rate to be used in the 
DCF model is the growth rate expected by the market. Thus, the 
Commission seeks to base its growth projections on ``the best evidence 
of the growth rates actually expected by the investment community.'' 
\95\ Moreover, the Commission stated, the growth rate expected by the 
investment community is not, quoting a Transco witness, ``necessarily a 
correct growth forecast; the market may be wrong. But the cost of 
common equity to a regulated enterprise depends upon what the market 
expects not upon precisely what is going to happen.'' \96\
---------------------------------------------------------------------------

    \94\ 85 FERC ] 61,323 at 62,268-9.
    \95\ Id. at 62,269.
    \96\ Id.
---------------------------------------------------------------------------

    74. The Commission held that the IBES five-year growth forecasts 
for each company in the proxy group are the best available evidence of 
the short-term growth rates expected by the investment community. It 
cited evidence that (1) those forecasts are provided to IBES by 
professional security analysts, (2) IBES reports the forecast for each 
firm as a service to investors, and (3) the IBES reports are well known 
in the investment community and used by investors. The Commission has 
also rejected the suggestion that the IBES analysts are biased and 
stated that ``in fact the analysts have a significant incentive to make 
their analyses as accurate as possible to meet the needs of their 
clients since those investors will not utilize brokerage firms whose 
analysts repeatedly overstate the growth potential of companies.'' \97\
---------------------------------------------------------------------------

    \97\ Transcontinental Gas Pipe Line Corp., 90 FERC ] 61,279, at 
61,932 (2000).
---------------------------------------------------------------------------

    75. Based on the comments, the Commission concludes that the IBES 
five-year growth forecasts should also be used for any MLP included in 
the proxy group. While the Commission recognizes that there may be some 
statistical limitations to the IBES projections, the record here 
demonstrates that it remains the best and most reliable source of 
growth information available. IBES publishes security analysts' five-
year growth forecasts for MLPs in the same manner as for corporations. 
No party questions the Commission's findings in past cases that 
investors rely on the IBES projections in making investment decisions, 
because they are widely available and generally reflect the input of a 
number of financial analysts. Also, since IBES projections are company-
specific, they should already adjust for any differences among the 
entities analyzed, including any reduced growth prospects investors 
expect due to the fact an MLP makes distributions in excess of 
earnings. In fact, the most recent IBES projections for the seven MLPs 
included in the gas pipeline proxy group in Appendix A, Table 1, 
average 6.86 percent, while the IBES growth projections for the four 
corporations average 10.75 percent. Thus, those MLP growth projections 
are about 400 basis points below those for the corporations.
    76. As discussed above, several parties assert that the security 
analysts' five-year growth forecasts appear generally to be forecasts 
of growth in earnings, rather than distributions. They point out that 
the relevant cash flows for the DCF model are the MLP's distributions 
to the limited partners, and therefore the growth projections used in 
the DCF analysis should be growth in distributions, not earnings. 
Despite these concerns, the Commission again concludes that the IBES 
short-term growth projections provide the best estimate of short-term 
growth rates for MLP distributions. Professor J. Peter Williamson, on 
behalf of AOPL, reviewed historical IBES five-year growth forecasts for 
five oil pipeline MLPs since the mid-1990s. IBES had published five to 
nine growth forecasts for each of the MLPs, with a total of 39 
forecasts. Williamson compared each of these 39 forecasts to the MLP's 
actual growth in earnings and distributions during the subsequent five-
year period. He found that 29 of the 39 IBES five-year forecasts, or 74 
percent, were closer to the actual average distribution growths over 
that time span than the actual earnings growths. In his study, 
Williamson also found that historical records fail to support any 
claims that the IBES forecasts are biased or tend to overstate future 
growth.\98\ In fact, 22 of the 39 forecasts were lower than the actual 
distribution growth, and 17 were higher. Thus, far from showing a 
pattern of overestimating actual growth in distributions, the IBES 
growth projections underestimated growth in distributions 56 percent of 
the time, a conservative result. Accordingly, regardless of whether 
financial analysts stated they are reporting projected earnings growth 
or projected distribution growth for MLPs, the Commission finds the 
five-year growth rates that IBES reports are acceptable since they 
closely approximate distribution growth for MLPs, which is the short-
term input for the DCF model.
---------------------------------------------------------------------------

    \98\ AOPL, Post-Technical Conference Comments, Williamson Aff. 
at 2-6.
---------------------------------------------------------------------------

    77. As noted, the State of Alaska expresses concerns that there are 
an insufficient number of stock analysts providing IBES with estimates 
which are expressly identified at forecasts of MLP distribution per 
share growth to obtain reliable short-term growth projections for MLPs. 
At the technical conference, Mr. Horst presented a chart showing the 
number of IBES report counts for 37 oil and gas pipeline companies--
both corporations and MLPs. The chart breaks the analyst report counts 
down into earnings reports and distribution reports. It shows that 
analysts made an average of 3.1 earnings reports for each MLP and an 
average of 0.8 distribution reports for each MLP.\99\ However, as 
discussed above, Williamson's analysis of a historical period suggests 
that actual MLP growth in the short term tracks IBES earnings 
projections better than distribution projections. Moreover, Mr. Horst's 
averages include many smaller, less frequently traded MLPs and thus 
understate the number of analysts that are likely to follow the larger, 
more established pipeline MLPs likely to be included in a proxy group. 
The Commission therefore concludes that the number of reports made by 
analysts for oil and gas companies MLPs is acceptable for use in the 
DCF model.
---------------------------------------------------------------------------

    \99\ State of Alaska, Comments dated December 21, Second Horst 
Aff. at 4-5; Reply Comments dated February 20 at 5, Third Horst Aff. 
at 16-17, 21.
---------------------------------------------------------------------------

    78. Some of the Customer Interests are agreeable to the continued 
use of IBES forecasts, but only under certain conditions. Specifically, 
PSCNY contends that, should the Commission continue to use IBES 
forecasts in its DCF model, any MLP the Commission allows in a proxy 
group must be market-tested and representative of a natural gas 
pipeline company. PSCNY contends that IBES would be acceptable if the 
MLP is tracked by Value Line, has been in operation for at least five 
years as an MLP, and derives 50-percent of its operating income from, 
or has 50 percent of its assets devoted to, interstate natural gas 
transportation operations. PSCNY also contends that the Commission 
should exclude MLPs from proxy groups when their growth

[[Page 23233]]

projections are illogical or anomalous.\100\
---------------------------------------------------------------------------

    \100\ PSCNY Supplemental Comments dated Dec. 21 at 3-5.
---------------------------------------------------------------------------

    79. The Commission agrees in principle with PSCNY's position that 
IBES forecasts should only be used for an MLP that is tracked by Value 
Line, has been in operation for at least five years as an MLP, and 
derives at least 50 percent of its operating income from, or 50 percent 
of its assets devoted to, interstate operations. Thus, when developing 
its proxy group, a pipeline should select MLPs that are well 
established and have assets that are predominantly gas and oil 
pipelines. Such pipelines are those most likely to have risk comparable 
to the pipeline seeking to justify its rates. However, there may be 
particular MLPs that do not satisfy these criteria, but are still 
appropriate for inclusion in the proxy group. The pipeline must justify 
including such an MLP in its proxy group. Thus, while the Commission 
encourages pipelines to follow the guidelines suggested by PSCNY, it 
will not make them a condition of including a particular MLP in the 
proxy group. As suggested by the parties, the Commission will continue 
to exclude an MLP from the proxy groups if its growth projection is 
illogical or anomalous.
    80. Two parties state that, should the Commission continue to use 
IBES projections to estimate short-term growth rates in its DCF model 
for MLPs, it must modify the estimated rates. Tesoro states that, if 
the Commission makes no adjustments to dividend distributions of MLPs, 
it should significantly reduce its IBES short-term growth estimates to 
recognize the fact that an MLP cannot indefinitely sustain its 
operations when distributions consistently exceed earnings. It argues 
that, if the Commission caps MLP distributions at earnings, it would 
still have to reduce IBES rates in order to recognize the fact that 
proxy group members would not be reinvesting retained earnings in 
ongoing operations, thereby achieving lower growth rates. Tesoro only 
recommends no adjustments to short-term growth estimates if the 
Commission caps distributions at a level below earnings, offering 65-
percent of earnings as an example.\101\
---------------------------------------------------------------------------

    \101\ Tesoro, Comments on Growth dated December 21 at 3-4, 5-7.
---------------------------------------------------------------------------

    81. The State of Alaska recommends that if a pipeline company's 
distributions per share exceed its earnings per share (as is frequently 
the case with pipeline MLPs), then the expected growth rate of the 
pipeline's distributions per share should be adjusted to equal (1) the 
expected growth of its earnings per share, multiplied by (2) the ratio 
of the pipeline's earnings per share to its distributions per share. 
According to Alaska, if a pipeline company distributes more cash than 
its current earnings, then the projected growth in earnings per share 
should also be adjusted by the ratio of the pipeline's earnings per 
share to its distributions per share.\102\
---------------------------------------------------------------------------

    \102\ State of Alaska, Comments dated Dec. 21 at 3-4; Second 
Horst Aff. at 2-3, 5-11.
---------------------------------------------------------------------------

    82. The Commission rejects these proposals by Tesoro and the State 
of Alaska. As already discussed, to the extent investors expect an 
MLP's distributions in excess of earnings to reduce its growth 
prospects, that fact should be reflected in the IBES five-year growth 
projections themselves, without the need for any further adjustment. 
MLPs must publicly report their earnings and distribution levels. 
Therefore, the security analysts are aware of the degree to which each 
MLP is making distributions in excess of earnings. The security 
analysts presumably take that information, together with all other 
available information concerning the MLP, into account when making 
their projections. Moreover, these proposals would have a similar 
effect as capping the distributions used to calculate dividend yield at 
or below the level of the MLP's earnings. For the reasons previously 
discussed, the Commission finds that any cap on an MLP's distributions 
used in the DCF model at a level below the actual distribution is 
inconsistent with the basic operation of the DCF model. Thus, using a 
straight IBES five-year projection without modification presents the 
best method of estimating an MLP's short-term growth rate.
    83. APGA further suggests revising IBES growth rates by averaging 
them with the comparable growth forecasts reported by Zacks Investment. 
It states that this averaging could help remove anomalous or outlying 
growth rates. It offers as an example, on December 10, 2007, IBES 
projected a five-year growth rate of 7.60 percent for Kinder Morgan 
Energy Partners (KMEP), whereas Zacks Investment projected a 33.70 
percent growth rate for that company. APGA argues that the Commission 
should also use Value Line reports to test the reasonableness of 
projected growth rates for MLPs.\103\
---------------------------------------------------------------------------

    \103\ APGA, Additional Comments dated Dec. 21 at 3, 9-10.
---------------------------------------------------------------------------

    84. The Commission will not require that IBES growth rates be 
averaged with the corresponding company's growth rates as reported for 
Zacks Investment at this time, or that Value Line reports be used to 
test the reasonableness of projected growth rates for MLPs. Finally, 
PSCNY requests that the Commission clarify that Thomson Financial Data 
posted on Yahoo.com may be used in the DCF formula, since Thomson 
Financial owns IBES.\104\ The Commission clarifies that the growth 
projections to be used in the DCF model are those reported by IBES. If 
they are the same growth projections posted by Thomson Financial Data 
on Yahoo.com, then they are acceptable for the DCF model.
---------------------------------------------------------------------------

    \104\ PSCNY, Supplemental Comments dated Dec. 21 at 5.
---------------------------------------------------------------------------

D. The Long Term Growth Component

1. Comments
    85. At this point the critical issue is whether the long term 
growth component of the Commission's DCF methodology should be modified 
in determining the equity cost of capital for an MLP. As has been 
discussed, for more than a decade the Commission has required that 
projected long-term growth in GDP be used as the corporate long term 
(terminal) growth component of the DCF calculation. The discussion at 
the technical conference disclosed four general positions. The 
AOPL,\105\ NAPTP,\106\ INGAA,\107\ and TransCanada \108\ asserted that 
the use of long term GDP is equally applicable to MLPs as to 
corporations.\109\ However, the APGA,\110\ PSCNY,\111\ and the State of 
Alaska \112\ all made suggestions for a reduction to the GDP growth 
projection to reflect the different retention and investment practices 
of MLPs.\113\ In a different vein, INGAA suggested the use of the 
average of the projected long term inflation rate and projected long 
term

[[Page 23234]]

GDP as a proxy for the lower growth rate of the limited partnership 
interests, but only if the Commission concluded that some reduction in 
the MLP long term growth rate was warranted.\114\ NAPTP further argued 
that there must be an upward adjustment of the limited partnership 
growth rate to reflect the equity cost of capital of the limited and 
general partners, and thus that of the entire firm.\115\
---------------------------------------------------------------------------

    \105\ AOPL, Post-Technical Conference Comments at 7-9, 13.
    \106\ NAPTP Additional Comments dated Dec. 21 at 1, 10-11; Post-
Technical Conference Comments at 4-8.
    \107\ INGAA, Additional Initial Comments dated Dec. 21 at 2-3; 
Post-Technical Conference Reply Comments at 3-6.
    \108\ TransCanada Post-Technical Comments at 2-5.
    \109\ MidAmerican and Williston supported this position.
    \110\ APGA Additional Comments dated Dec. 21 at 4, 7-8; Initial 
Post-Technical Comments at 2, J. Bertram Solomon Aff. at 4-8.
    \111\ PSCNY, Supplemental Comments dated Dec. 21 at 5, 8-9 and 
appended Prepared Statement of Patrick J. Barry for the January 23, 
2008 Technical Conference; Initial Post-Technical Conference 
Comments at 14-16.
    \112\ State of Alaska, Comments dated Dec. 21 at 3-4 and Second 
Horst Aff. at 3, 5-7. Reply Comments dated February 20, 2008 at 6.
    \113\ NGPA and Tesoro also supported a lower long term growth 
rate for MLPs.
    \114\ INGAA Additional Initial Comments dated Dec. 21 at 3-4 and 
Vilbert Report attached thereto, passim.
    \115\ NAPTP Reply Comments dated Sept. 19 at 2-4; Additional 
Comments dated Dec. 21 at 9-12.
---------------------------------------------------------------------------

    86. The Pipeline Interests also generally assert that an MLP's 
terminal growth can be at least equal to that of a corporation, and 
perhaps exceed it. They assert that MLPs are able to raise external 
capital in a tax efficient manner. Because an MLP does not retain cash 
it does not immediately need and can distribute without the tax 
penalty, it is under less pressure to invest idle capital. Rather, an 
MLP can wait until sounder investment opportunities are available and 
pursue them more discreetly, which results in a more consistent return 
from the projects selected.\116\ Moreover, while the computation is 
very complicated, the tax-deferral aspects of MLP limited partnership 
interest normally result in a higher per unit price when issued and 
thus a lower cost of equity capital to the issuing MLP. For these 
reasons the Pipeline Interests conclude that MLPs should readily find 
profitable investment opportunities despite their lower retention 
ratios.\117\
---------------------------------------------------------------------------

    \116\ NAPTP Post-Technical Conference Comments at 9; TransCanada 
Post Technical Conference Comments at 8-9.
    \117\ NAPTP, id. 2, 5-6. TransCanada, id.
---------------------------------------------------------------------------

    87. The Pipeline Interests further assert that the record 
demonstrates that MLPs have a long term history of growing 
distributions and an overall growth rate that has at times been higher 
than that of corporations.\118\ They cite to the example of KMEP in 
particular and that KMEP has been able to grow its distributions in 
good or poor financial environments.\119\ They therefore conclude that 
there is no reason to conclude that MLPs cannot continue to grow at 
least as fast as corporations or that the relatively high distribution 
growth rate for the industry as a whole will not be sustained.\120\ 
However, INGAA concedes that even if an MLP as a whole can grow as fast 
as a corporation, the limited partnership interests would grow less 
rapidly than the MLP as a whole because of the IDRs \121\ most MLPs 
have granted their general partners.\122\ The Pipeline Interests also 
argue that investors will not invest in enterprises that have a 
projected growth rate that is less than GDP and that such firms are 
likely to fail.\123\
---------------------------------------------------------------------------

    \118\ NAPTP Additional Comments dated Dec. 21 at 4-8.
    \119\ NAPTP Additional Comments dated December 21 at 8.
    \120\ NAPTP and Post-Technical Conference Comments at 11-12 AOPL 
Post-Technical Conference at 9-10 and Williamson Post Technical 
Conf. Aff. Ex. at 1 and 2.
    \121\ IDRs operate as follows. Most MLP agreements provide that 
the limited partners own 98 percent of the equity when the firm is 
first created and the general partner 2 percent. Thus, given a 
distributable cash of $1,000, the limited partners would obtain $980 
(98 percent) and the general partner $20.00 (2 percent). The 
partnership agreement also provides that as the total cash available 
for distribution increases, a greater share goes to the general 
partner, including that which would be available in liquidation. For 
example, the partnership agreement may provide that once 
distributable cash is $3,000, the general partner will receive 2 
percent based on its partnership interest and 48 percent based on 
the IDRs.
    At that point the limited partners' share of the distribution is 
$1,500 (50 percent) and the general partner's share is also $1,500 
(50 percent). Thus, while the limited partners' distribution has 
grown in the relevant time frame (by 50 percent), it has not grown 
as fast as it would have absent the general partner's IDR. Absent 
the IDR the general partner's share would only be $60. Since a 
proportionately smaller share of future value flows to the limited 
partners in the initial years, the projected long term growth rate 
for a limited partnership interest will be lower. Therefore the 
limited partnership interests have lower return than that of the 
general partner.
    \122\ INGAA Additional Initial Comments dated December 21 at 5; 
TransCanada.
    \123\ AOPL, Post-Technical Comments at 7-8. TransCanada, 
Additional Comments dated Dec. 21 at 2, 4-5.
---------------------------------------------------------------------------

2. Discussion
    a. Should the MLP long-term growth projection be lower than 
projected growth in GDP?
    88. As discussed in the previous section, in determining the 
appropriate growth projections to use in its DCF analysis, the 
Commission seeks to approximate the growth projections investors would 
rely upon in making their investment decisions. This principle applies 
equally to the long-term growth projection, as to the short-term growth 
projection. When the Commission first established its policy of basing 
the long-term growth projections on projected growth in GDP in Opinion 
No. 396-B and Williston I, the Commission stated in both cases, ``The 
purpose of using the DCF analysis in this proceeding is to approximate 
the rate of return an investor would reasonably expect from a pipeline 
company.'' \124\ The Commission found, ``the record shows that Merrill 
Lynch and Prudential Bache do not attempt to make long-term growth 
projections for specific industries or companies in doing DCF analyses. 
Instead they use the long-term growth of the United States economy as a 
whole as the long-term growth forecast for all firms, including 
regulated businesses.'' \125\ The Commission thus relied heavily on 
evidence concerning investment house long-term growth projections in 
deciding to base its long-term growth projections for corporations that 
were properly included in the proxy group on the long-term growth of 
GDP. In affirming this aspect of Williston I, the DC Circuit similarly 
relied on the fact that the record ``demonstrated that major investment 
houses used an economy-wide approach to projecting long-term growth * * 
* and that existing industry-specific approaches reflected investor 
expectations and many unfounded economic assumptions.'' \126\
---------------------------------------------------------------------------

    \124\ Opinion No. 396-B, 79 FERC ] 61,309 at 62,383. Williston 
I, 79 FERC at 62,389.
    \125\ Opinion No. 396-B, 79 FERC ] 61,309 at 62,382. Williston 
I, 79 FERC ] 61,311 at 62,389. As the Commission pointed out in a 
subsequent case, the exhibits in both the Opinion No. 396-B 
proceeding and Williston I, describing Prudential Bache's 
methodology stated that it used a lower long-term growth projection 
for electric utilities, because of their high payout ratios. System 
Energy Resources, Inc., 92 FERC ] 61,119, at 61,445 n.23 (2000).
    \126\ Williston Basin Interstate Pipeline Co. v. FERC, 165 F.3d 
54 (DC Cir. 1999).
---------------------------------------------------------------------------

    89. Consistent with this precedent, the key question in deciding 
what long-term growth projection the Commission should use in its DCF 
analysis of MLPs is whether investors expect MLP long-term growth rates 
to be less than projections of growth in GDP. The record established 
here shows that at least two major investment houses project terminal 
growth rates for MLPs that are notably lower than the current 4.43 
percent projected growth in GDP. Citicorp Smith Barney (Citicorp) \127\ 
projects a 1 percent terminal growth rate for pipeline MLPs. Wachovia 
projects terminal growth rates for individual MLPs that vary from zero 
to 3.5 percent.\128\ The Wachovia projection for each MLP which the 
Commission is likely to include in a proxy group \129\ is for a 2.5 
percent terminal growth rate.\130\

[[Page 23235]]

The Pipeline Interests did not submit any evidence of a major 
investment house projecting long-term growth rates for MLPs equal to or 
above the growth in GDP. Thus, applying the same approach as that in 
Opinion No. 396-B and Williston I, the record supports a finding that 
investors project MLP growth rates significantly below the growth in 
GDP.
---------------------------------------------------------------------------

    \127\ Society, Reply Comments at 11, citing: Citicorp Master 
Limited Partnership Monitor and Reference Book, Citigroup Investment 
Research (March 2007) at 28, Figure 24.
    \128\ Comments of Enbridge Energy Partners, L.P., Attachment A, 
Wachovia Equity Research Paper dated August 20, 2007 at 9-12; 
Wachovia Equity Research dated January 30, 2008, MLP Outlook 2008: 
Cautious Optimism at 39-44.
    \129\ These are the MLPs listed in Tables 1 and 2.
    \130\ NAPTA, in its Post-Technical Conference Comments, provided 
a publication by Morgan Stanley Research which, among other things, 
reported on our January 23, 2008 technical conference. That 
publication, at page 3, states, ``At Morgan Stanley, we assume an 
MLP will increase its cash flow--1.5%-3.0% per year beyond 2012. 
Importantly we make the same assumption in forecasting long-term 
growth for our C-Corp companies.'' Pipeline MLPs: What's in the 
Pipeline, Morgan Stanley Research at 3. These projections are also 
less than the current projection of 4.43 percent long-term growth in 
the economy as a whole. However, we give greater weight to the 
Citigroup and Wachovia publications, because those publications 
include specific long-term growth projections for individual MLPs, 
whereas the Morgan Stanley publication simply sets forth a general 
range it uses without specifying how that range is distributed among 
individual firms. Also, the Citigroup and Wachovia analyses were not 
issued in response to the technical conference.
---------------------------------------------------------------------------

    90. To counter this conclusion, the Pipeline Interests argue that 
these lower figures reflect the investment houses' desire to use 
``conservative'' estimates in order to prevent unrealistic investor 
expectations. However, as discussed above, the Commission has found in 
earlier cases that investment houses try to give the most accurate 
information to their investors. In any event, it is appropriate for the 
Commission to use growth estimates that reflect the investment houses' 
view of what investors should realistically expect from an investment 
in an MLP. Moreover, the fact that some MLPs have grown rapidly in the 
past does not mean necessarily that they will maintain the same growth 
rate in the future. In fact, KMEP's projected growth rate is expected 
to drop in future years.\131\ This record also demonstrates that a rate 
of long term growth is dependent on the base years selected. Thus, the 
Customer Interests focus on more recent years to show that the growth 
rate has slowed for many MLPs.\132\
---------------------------------------------------------------------------

    \131\ APGA, Post-Technical Conference Reply Comments, Solomon 
Aff. at 4.
    \132\ APGA, Post-Technical Conference Reply Comments at 4-5 and 
attached Solomon Aff. at 4-9.
---------------------------------------------------------------------------

    91. The Pipeline Interests also argue that investors will not 
invest in entities with a projected long term growth rate that is less 
than the long-term growth in GDP.\133\ However, the fact is that, 
despite major investment houses advising their clients that MLPs will 
have long-term growth rates below GDP, investors have continued to 
invest in MLPs, and in increasing amounts through 2007. Historically 
this was true even though the Commission's analyses continue to 
indicate that the IBES five-year growth projections for MLPs are lower 
than those for corporations.\134\
---------------------------------------------------------------------------

    \133\ TransCanada, Additional Comments at 5; AOPL Post-Technical 
Conference Comments at 8.
    \134\ See Appendix A, which displays in part the comparative 
corporate and MLP short term growth projections. Cf. PSCNY Post 
Technical Conference Comments at 7-8.
---------------------------------------------------------------------------

    92. At bottom, the key financial assumption advanced by the 
Pipeline Interests is that MLPs and corporations have equal access to 
capital. However, the Customer Interests advance credible reasons why 
MLPs may not have as ready access to capital markets in the future 
given the MLPs' unique financial structure. This would reduce the total 
capital pool available to the MLPs, thus reducing their growth 
prospects. These include a greater exposure to interest rate risk,\135\ 
the increased cost of capital that a high level of IDRs imposes on an 
MLP,\136\ and lower future returns from either acquisitions or organic 
investments as the MLP industry matures.\137\ This latter point is of 
greater importance to MLPs because they are limited by law to a 
narrower range of investment opportunities than a schedule C 
corporation. These arguments suggest why the long term forecasts by 
investment houses investors rely on could conclude that the long term 
growth rate for MLPs would be less than the long term GDP the 
Commission uses for corporations. Each addresses the consistency of 
investment opportunities and as such consistency of access to capital 
markets that MLPs are dependent on to maintain long term growth.
---------------------------------------------------------------------------

    \135\ Indicated Shippers Initial Comments at 21, citing Citicorp 
Smith Barney; AGPA Reply Comments at 5; Wachovia August 20, 2007 
Report, supra, at 1-2;
    \136\ PSCNY Supplemental Comments at 3, n. 8 and Initial Post-
Technical Conference Comments at 12.
    \137\ PSCNY Initial Post-Technical Conference Comments at 9-10 
and cited Value Line attachments; Reply Comments at 5-6 citing 
Merrill Lynch, n. 16.
---------------------------------------------------------------------------

    93. In particular, the Commission concludes that corporations (1) 
have greater opportunities for diversification because their investment 
opportunities are not limited to those that meet the tax qualifying 
standards for an MLP and (2) are able to assume greater risk at the 
margin because of less pressure to maintain a high payout ratio. It is 
a corporation's higher retention ratio that allows this greater 
flexibility. This is consistent with the fact that Prudential Bache 
projected the long-term growth rates of electric utilities to be less 
than that of the economy as a whole because of their greater dividend 
payouts and lower retention ratios.\138\ Therefore, investors would 
quite reasonably conclude that MLP long term growth rates would be 
lower than that of tax paying corporations, because MLPs have fewer 
opportunities to participate in the broad economy that underpins the 
Commission's current use of long-term growth in GDP.
---------------------------------------------------------------------------

    \138\ System Energy Resources, Inc., 92 FERC ] 61,119, at 61,445 
n. 23 (2000).
---------------------------------------------------------------------------

    94. Thus, while it is true that the Commission uses GDP as a proxy 
for long term growth, the point here is not whether some firms, 
including MLPs may have a growth rate that is more or less than the 
proxy over time. The issue is whether MLPs have the same relative 
potential as the corporate based economy that has been the basis for 
the Commission's assumption that a mature firm will grow at the same 
rate as the economy as a whole. For the reasons stated, the Commission 
concludes that the collective long term growth rate for MLPs will be 
less than that of schedule C corporations regardless of the past 
performance of MLPs the Pipeline Interests have inserted in the record.
    b. What specific projection should be used for MLPs?
    95. We now turn to the issue of exactly what long-term growth 
projection below GDP should be used in MLP pipeline rate cases. As the 
Commission recognized when it established its policy of giving the 
long-term growth projection only one-third weight, while giving the 
short-term growth projection two-thirds weight, ``long-term growth 
projections are inherently more difficult to make, and thus less 
reliable, than short-term projections.'' \139\ Thus, as the Commission 
has stated with respect to the other aspects of its long-term growth 
projection policy, the Commission is ``required to choose from among 
imperfect alternatives'' \140\ in deciding what specific long-term 
growth projection should be used for MLPs.
---------------------------------------------------------------------------

    \139\ Transcontinental Gas Pipe Line Corp., 84 FERC ] 61,084, at 
61,423 (1998).
    \140\ Northwest Pipeline Corp., 88 FERC ] 61,298, at 61,911 
(1999).
---------------------------------------------------------------------------

    96. The technical conference panelists advanced four methods of 
determining long-term growth projections for MLPs which are less than 
the growth in GDP. After reviewing all four, the Commission adopts the 
APGA proposal to use a long-term growth projection for MLPs equal to 50 
percent of long term GDP.\141\ At present, that proposal results

[[Page 23236]]

in a long-term growth projection of 2.22 percent. This is within the 
range of long-term growth projections used by investment houses for 
MLPs discussed in the preceding section. For example, Wachovia projects 
terminal growth rates for individual MLPs that vary from zero to 3.5 
percent,\142\ and its projection for each MLP which the Commission is 
likely to include in a proxy group is for a 2.5 percent terminal growth 
rate.\143\ Therefore, in light of the inherent difficulty of projecting 
long-term growth, the 50 percent of GDP proposal would appear to result 
in a long-term growth projection that falls within any reasonable 
margin of error for such projections, while giving recognition to the 
fact that investors expect MLPs' long-term growth to be less than that 
of GDP.\144\
---------------------------------------------------------------------------

    \141\ APGA Additional Comments dated Dec. 21 at 2-3, 8; Outline 
for the Presentation of Bertrand Solomon on the Behalf of APGA dated 
January 23, 2008 at 3; Initial Post-Technical Conference Comments. 
J. Bertrand Solomon Aff. at 3-4, 6-7 and supporting exhibits.
    \142\ Comments of Enbridge Energy Partners, L.P., Attachment A, 
Wachovia Equity Research Paper dated August 20, 2007 at 9-12; 
Wachovia Equity Research dated January 30, 2008, MLP Outlook 2008: 
Cautious Optimism at 39-44.
    \143\ The Commission will not use the specific long-term MLP 
growth projections of the investment houses to determine the cost of 
equity for specific firms for the same reasons we have not done so 
with respect to the projections of long-term growth in GDP the 
Commission uses for corporations. As the Commission explained in 
Michigan Gas Storage Co., 87 FERC ] 61,038, at 61,162-5 (1999) and 
Williston Basin Interstate Pipeline Co., 87 FERC ] 61,264, at 
62,005-6 (1999), there is no evidence as to how the investment house 
figures were derived which limits their utility in determining the 
cost of equity for an individual firm. However, as here, the 
Commission has relied on the perceptions of the investment community 
in developing a generic long term growth rate. See also Opinion No. 
396-B, 79 FERC ] 61,309 at 62,384.
    \144\ As the DC Circuit stated with respect to our choice of the 
relative weighting of the short- and long-term growth projections, 
the choice of the long-term growth component is also an exercise 
``hard to limit by strict rules.'' CAPP v. FERC, 254 F.3d at 290.
---------------------------------------------------------------------------

    97. The Commission also concludes that the other three proposed 
methods of projecting MLP long-term growth rates all have flaws 
justifying their rejection. The State of Alaska and the NYPSC propose 
methods which would result in varying long-term growth projections for 
each MLP, based upon financial information for each of the MLPs to be 
included in a proxy group. These proposals are contrary to the 
Commission's policy of using a single long-term growth projection for 
all corporations, based on the fact that it is not possible to make 
reliable company-by-company long-term growth projections.\145\ The 
State of Alaska and NYPSC have provided no basis to conclude that they 
have provided a more reliable way to make long-term growth projections 
for individual MLPs. Their difficulty in doing so reinforces the 
Commission's traditional practice in this regard.
---------------------------------------------------------------------------

    \145\ Opinion No. 396-B, 79 FERC ] 61,309 at 62,382.
---------------------------------------------------------------------------

    98. The State of Alaska suggests adjusting the GDP long term growth 
projection used for each MLP based on its current positive or negative 
retention ratio.\146\ Thus, if an MLP's retention ratio was positive, 
then 100 percent of long term growth in GDP would be used. If the 
retention ratio was less than one, then the long term growth in GDP 
would be reduced accordingly. This theory essentially caps the long 
term growth rate at the earnings of the entities involved. As such, it 
suffers from the same weakness as the original proposal to cap the 
distribution component included in the model at earnings. Consistent 
with the premise of the DCF model that a stock is worth the present 
value of all future cash flows to be received from the investment, 
investors base their DCF analyses on the MLP's entire cash 
distributions, including projected cash flows generated by external 
investments, which to date is the bulk of the investment for the MLP 
model. In addition, because MLPs rely substantially on external capital 
to finance growth, the fact one MLP currently pays out more of its 
earnings than another MLP does not necessarily mean that the first 
MLP's long-term growth prospects are less than the second MLP's. 
Moreover, Alaska's proposed method assumes each MLP's current retention 
ratio will continue indefinitely into the future, without any support 
for the accuracy of such an assumption.
---------------------------------------------------------------------------

    \146\ State of Alaska, Comments dated December 21 at 3-4 and 
Second Horst Aff. at 3, 5-7. Reply Comments dated February 20, 2008 
at 6.
---------------------------------------------------------------------------

    99. The NYPSC recommends use of a modified form of the sustainable 
growth model the Commission uses to determine electric return on 
equity.\147\ Under that method, the Commission determines growth based 
on a formula under which growth = br + sv, where b is the expected 
retention ratio, r is the expected earned rate of return on common 
equity, s is the percent of common equity expected to be issued 
annually as new common stock, and v is the equity accretion rate. The 
br component of this formula projects a utility's growth from the 
investment of retained earnings, and the sv component estimates growth 
from external capital raised by the sale of additional units. The NYPSC 
would assume zero growth from investment of retained earnings (the br 
component) and then base the long-term growth projection for each MLP 
on projected growth from external capital resulting from the sv 
component of the br + sv formula.
---------------------------------------------------------------------------

    \147\ PSCNY, Supplemental Comments dated Dec. 21 at 5, 8-9 and 
appended Prepared Statement of Patrick J. Barry for the January 23, 
2008 Technical Conference; Initial Post-Technical Conference 
Comments at 14-16.
---------------------------------------------------------------------------

    100. A fundamental problem with this approach is that the 
Commission has consistently held that the br + sv formula only produces 
a projection of short-term growth, similar to the IBES 
projections.\148\ This follows from the fact that the inputs used in 
the formula are all drawn from Value Line data and projections reaching 
no more than five years into the future. In addition, there would be 
great uncertainties in projecting any of the inputs to the formula, 
such as the retention ratio, the amount and timing of equity sales, and 
the projected price of the sale for any longer period. Moreover, 
setting the br component at zero assumes that an MLP can only grow 
through the use of external capital. This does not reflect accurately 
the retention and investment flexibility vested in an MLP's general 
partners or the fact that some MLPs may reinvest a fairly high 
proportion of the free cash available. Therefore this methodology does 
not appropriately adjust the long term GDP component that the 
Commission now uses for corporations.
---------------------------------------------------------------------------

    \148\ See Southern California Edison Co., 92 FERC ] 61,070, at 
61,262-3 (2000).
---------------------------------------------------------------------------

    101. Finally, INGAA provided a complex model designed to calculate 
the equity cost of capital for an MLP as a whole.\149\ This model was 
developed by Mr. Vilbert and attempts to calculate the equity cost of 
capital for both the limited and the general partners. At their 
inception, MLPs establish agreements between the general and limited 
partners, which define how the partnership's cash flow is to be divided 
between the general and limited partners. Such agreements give the 
general partners IDRs, which provide for them to receive increasingly 
higher percentages of the overall distribution, if the general partners 
are able to increase that distribution above defined levels. The INGAA 
model recognizes that, as a result of these incentive distribution 
rights, a DCF analysis of the MLP as a whole should (1) include higher 
projected growth rates for the general partner interest than for the 
limited partner interest and (2) a correspondingly higher value for 
general partner interests than the MLP units which would, in turn, 
reduce the

[[Page 23237]]

general partner's current ``dividend'' yield. However, since there are 
relatively few publicly traded general partner interests, in most cases 
the estimated equity cost of capital for the general partner can only 
be derived through various assumptions that markup the limited 
partner's cost of capital.
---------------------------------------------------------------------------

    \149\ INGAA, Additional Initial Comments dated Dec. 21 at 4-5 
and Report on the Terminal Growth Rate for MLPs for Use in the DCF 
Model by Michael J. Vilbert dated December 21, 2007 (Vilbert 
Report), particularly at 10.
---------------------------------------------------------------------------

    102. INGAA drew two significant conclusions from Mr. Vilbert's 
analysis. First, application of the Commission's existing DCF 
methodology solely to the limited partner interest in the MLP would 
generate returns relatively close to those that would be required to 
reflect the growth rate, and cost of equity capital, for the MLP as a 
whole. Second, if the Commission remains concerned that a DCF analysis 
using data solely for the limited partner interest,\150\ together with 
a long-term growth rate equal to the growth in GDP, may overstate the 
appropriate return based on the limited partners' projected growth, the 
long-term growth projection could be adjusted by averaging projected 
long term GDP and the projected long term inflation rate.\151\ The 
latter would have to be updated regularly to test its accuracy.
---------------------------------------------------------------------------

    \150\ In such a DCF analysis the dividend yield would be 
calculated by dividing the distribution to the limited partner by 
the limited partner share price.
    \151\ INGAA Additional Initial Comments dated Dec. 21 at 4-6; 
Vilbert Report at 18-19.
---------------------------------------------------------------------------

    103. Mr. Horst, the witness for the State of Alaska, responded that 
the INGAA model was mathematically correct, but that the model's 
assumptions about the rate of growth and incentive distributions were 
open to question and the results would overstate the equity for the MLP 
as a whole.\152\ INGAA filed a reply to Mr. Horst's arguments by Mr. 
Vilbert that first calculates the actual DCF values for eight publicly 
traded general partner interests.\153\ Mr. Vilbert then compares the 
resulting value of the general partner interests for the same eight 
firms generated by the model. The results calibrate more closely to the 
eight market samples than the analysis produced by Mr. Horst but, like 
Mr. Horst's analysis, tend to overstate the value of the general 
partner interest.
---------------------------------------------------------------------------

    \152\ State of Alaska, Reply Comments dated February 20, 2008 at 
6 and Third Horst Aff. at 6-15.
    \153\ INGAA, Post-Technical Supplemental Comments dated March 
12, 2008 at 2-4 and Vilbert Aff. attached thereto, passim. The 
Commission will accept INGAA's March 12 filing because INGAA had no 
earlier opportunity to reply to the material contained in the State 
of Alaska's February 20, 2008 filing.
---------------------------------------------------------------------------

    104. The Commission will not use the INGAA model for several 
reasons. First, the internal operations of the model are relatively 
opaque, and the model appears to have a relatively wide range of error. 
Second, as the court stated in Petal Gas Storage, LLC v. FERC,\154\ the 
purpose of the proxy group is to ``provide market-determined stock and 
dividend figures from public companies comparable to a target company 
for which those figures are unavailable.'' While INGAA used eight 
publicly traded general partner interests to test the validity of the 
model, most of those interests are not related to MLPs that have been 
proffered in rate proceedings before the Commission. In the absence of 
such market-determined figures for the general partner interest of the 
MLPs to be included in the proxy group, use of the INGAA model would 
necessarily entail deriving an estimated equity cost of capital for the 
general partner through various assumptions that markup the limited 
partner's cost of capital. In these circumstances, use of the INGAA 
model would be inconsistent with the purpose of the proxy group of 
providing a fully market-based estimated cost of capital.
---------------------------------------------------------------------------

    \154\ 496 F. 3d 695 at 699.
---------------------------------------------------------------------------

    105. INGAA alternatively suggested that the returns from the 
current methodology be reduced somewhat to reflect the admittedly lower 
growth rate of a MLP's limited partnership interests. However, its 
proposal to do that by averaging GDP growth projections with the 
Federal Reserve's target inflation rate appears to have no analytical 
basis. Therefore, INGAA's recommendations will not be accepted 
here.\155\
---------------------------------------------------------------------------

    \155\ See AOPL Post-Technical Comments at 3-4, which suggest 
that the complexity of Mr. Vilbert's model and the use of its 
assumption indicate that it is more appropriate to rely on the 
limited partners' distributions in a DCF analysis.
---------------------------------------------------------------------------

    106. Based upon the above discussion, the Commission concludes that 
the long term growth component for an MLPs equity cost of capital 
should be 50 percent of long term GDP, rather than the full long term 
GDP currently used for corporations.
    c. Proposed upward adjustments to the long term component
    107. NAPTP asserted that the Commission should increase rather than 
decrease the long term growth component used to determine an MLP's 
equity cost of capital to reflect the general partner component of an 
MLP's equity.\156\ It asserts that equity cost of capital must be 
determined for the MLP as a whole, not just for the limited partners. 
NAPTP asserts that the return, and hence the projected growth rate, 
must generate sufficient cash flows to support the IDRs provided the 
general partner under most MLP agreements. To this end, it marked up 
the growth rate of the limited partners to reflect the portion of the 
equity effectively controlled by the general partner through its IDRs. 
Thus, if the growth rate for the limited partners was 10 percent and 
the general partner received a total of 50 percent of the 
distributions, the growth rate for the general partner could be as high 
as 20 percent. The Shipper Interest partners argued that this only 
rewarded the general partner for its excessive distributions and would 
inordinately increase the MLPs equity cost of capital.
---------------------------------------------------------------------------

    \156\ NAPTP Additional Comments dated Dec. 21 at 3-4.
---------------------------------------------------------------------------

    108. Both INGAA's witness Vilbert and the State of Alaska's witness 
Horst rejected the NAPTP approach on mathematical grounds. Both argue 
that the gross-up fails to properly value the general partner's 
interest at multiples that reflect the general partner interest's 
relative risk to that of the limited partners.\157\ Furthermore, 
Vilbert argues that the general partner's risk, while always greater 
than that of the limited partner, declines as the MLP matures and the 
general partner's share of distributions increases.\158\ As this 
occurs, the growth rate of the general partner's interest slows and 
approaches that of the limited partner. Failure to adjust for both 
facts means that the general partner's interest is undervalued using 
the NAPTP method, thus overstating the yield, and thus the return, that 
would be incorporated in the DCF model. As such, the NAPTP approach is 
inappropriate.
---------------------------------------------------------------------------

    \157\ State of Alaska, Reply Comments dated February 20, 2008 at 
6 and Third Horst Aff. at 2, 4-5.
    \158\ INGAA, Post-Technical Supplemental Comments dated March 
12, 2008 at 2-4 and Vilbert Aff. at 6-12.
---------------------------------------------------------------------------

    109. The Commission agrees that the NAPTP method is mathematically 
and conceptually flawed. Moreover, it has the same basic limitation as 
the INGAA model in that there is simply not enough publicly generated, 
transparent information at this time to support developing an equity 
cost of capital for the MLP as a whole. INGAA likewise attempted to 
develop an approach that would reflect the growth rate, and the return, 
of the MLP as a whole. The Commission has previously concluded that 
this approach has too many practical limits. Therefore the Commission 
will not pursue this issue further here.

E. The Weighting of the Growth Components

    110. The third issue is whether to change the weighting of the 
short-term and long-term components now used in

[[Page 23238]]

the Commission's DCF model. As has been discussed, the Commission's 
existing policy is to provide two-thirds of the weight to the short-
term component and one-third to the long-term component. TransCanada 
suggested changing the weighting, so that the 90 percent of the weight 
should be to the short-term component.\159\ MidAmerica recommended the 
use of a single stage model and abandoning the long-term component 
completely.\160\ However, these suggestions received no support from 
the other parties and would serve to increase the overall returns by 
sharply diminishing or eliminating the long-term component of the DCF.
---------------------------------------------------------------------------

    \159\ TransCanada, Reply Comments at 13-14; Additional Comments 
dated December 21 at 9-12.
    \160\ MidAmerican Response to Request for Additional Comments 
dated December 21 at 9-11.
---------------------------------------------------------------------------

    111. As discussed in the previous section, the Commission's 
longstanding policy is that the growth component of the DCF analysis of 
gas and oil proxy companies must include a projection of long-term 
growth, and the court affirmed that policy in Williston I. As the 
Commission has explained in numerous orders, the DCF methodology 
requires that a long-term evaluation be taken into account. In the 
preceding section, the Commission has fully discussed why the long-term 
growth projection for MLPs should be 50 percent of projected long-term 
growth of GDP.
    112. The Commission established its policy of giving the long-term 
growth projection one-third weight, while the short-term growth 
projection is given two-thirds weight, in Opinion Nos. 414-A. The 
Commission explained its weighting policy as follows:

    While determining the cost of equity nevertheless requires that 
a long-term evaluation be taken into account, long-term projections 
are inherently more difficult to make, and thus less reliable, than 
short-term projections. Over a longer period, there is a greater 
likelihood for unanticipated developments to occur affecting the 
projection. Given the greater reliability of the short-term 
projection, we believe it appropriate to give it greater weight. 
However, continuing to give some effect to the long-term growth 
projection will aid in normalizing any distortions that might be 
reflected in short-term data limited to a narrow segment of the 
economy.\161\

    \161\ Opinion No. 414-A, 84 FERC at 61,423.

The court affirmed this policy in CAPP v. FERC,\162\ stating that ``in 
an exercise so hard to limit by strict rules, it would likely be 
difficult to show that the Commission abused its discretion in the 
weighting choice.''
---------------------------------------------------------------------------

    \162\ 254 F.3d at 289.

    113. The need to normalize any distortions that may be reflected in 
short-term data limited to a narrow segment of the economy applies 
equally to the IBES five-year growth projections for MLPs as for 
corporations. At the same time, the two-thirds weighting for the short-
term growth projections recognizes their greater reliability. Moreover, 
TransCanada does not establish why the MLP short-term growth 
projections should be accorded a greater weight than that of 
corporations. In fact, as was discussed in the previous section, the 
record reasonably shows that investment houses include a long-term 
growth component in their DCF analyses of MLPs, and use a long-term 
growth projection that is lower than the projected long-term growth in 
GDP. Therefore the Commission will not modify the two-thirds to one-
third ratio it now uses in its DCF model and will apply that ratio to 
all pending cases.

V. Pending Proceedings

    114. The procedural issue here is whether this Policy Statement 
should be applied to all proceedings that are now before the Commission 
for which the ROE issue has not been resolved with finality. NGSA 
asserts that any new policy should apply only prospectively and not to 
cases now pending before the Commission. Indicated Shippers take the 
same position, asserting that application of the Policy Statement to 
pending proceedings would be administratively inefficient and would 
materially delay instituting new rates in the Kern River proceeding, 
which is now before the Commission on rehearing. Indicated Shippers 
further argue that in Kern River the Commission addressed and rejected 
the use of MLPs without some adjustment to reflect the fact that MLP 
distributions involve both a return of and return on equity. They also 
argue that there would be no inequity because Kern River could always 
file a new section 4 rate case if the existing proceeding proved 
unsatisfactory. Finally, Indicated Shippers assert that a policy change 
should not be applied retroactively because it does not have the force 
of law\163\ and because policy statements are considered ``statements 
issued by the agency to advise the public prospectively of the manner 
in which the agency proposes to exercise a discretionary power.'' \164\
---------------------------------------------------------------------------

    \163\ Citing Consolidated Edison of New York, et al., v. FERC, 
315 F.3d 316, 323-24 (DC Cir. 2003) (Consolidated Edison).
    \164\ Citing American Bus Assn. v. ICC, 627 F.2d 525, 529 (DC 
Cir. 1980).
---------------------------------------------------------------------------

    115. MidAmerica answered that the Policy Statement must be applied 
to all pending cases and Kern River in particular for two reasons. It 
states that in Petal the court both seriously questioned the 
Commission's analysis regarding MLPs and held that it was improper to 
include an entity of higher risk (a pipeline) and one of lower risk, 
such as a diversified natural gas company, in the same sample without 
adjusting the returns. MidAmerica argues that application of the 
Williston doctrine\165\ requires that it be given an opportunity to 
address the return on equity issue further. This is particularly the 
case since the court suggested applying the upper end of the range of 
reasonableness as a way of compensating for the difference in risk. 
MidAmerica asserts that application of either this suggestion or use of 
the unadjusted MLP sample Kern River advanced at hearing would result 
in the same return on equity.
---------------------------------------------------------------------------

    \165\ See Williston Basis Interstate Pipeline Co. v. FERC, 165 
F.3d 54 (DC Cir. 1999) (Williston). MidAmerica cites to the related 
administrative proceeding, Williston Basin Interstate Pipeline Co., 
104 FERC ] 61,036 (2003), but the principles are the same. The cited 
Commission case was in response to the remand in cited court 
decision.
---------------------------------------------------------------------------

    116. The Commission concludes that the instant Policy Statement 
must be applied to all proceedings now pending at hearing before an ALJ 
or before the Commission for which the ROE issue has not been resolved 
with finality. In Petal v. FERC, the court vacated and remanded the 
Commission's orders on the ROE issue in both Petal and HIOS. In both 
those cases, the Commission applied its current policy of using a proxy 
group based on the corporations listed in the Value Line Investment 
Survey's list of diversified natural gas firms that own Commission-
regulated natural gas pipelines, without regard to what portion of the 
company's business comprises pipeline operations. The court found that 
the Commission had not shown that the proxy group arrangements used in 
those cases were risk-appropriate. In this Policy Statement we have 
reexamined our proxy group policy in light of the Petal v. FERC remand 
as well as current trends in the gas and oil pipeline industries, and 
determined we must modify our policy as discussed above. Therefore, 
because the Commission's current proxy group policies as applied in 
prior cases have not withstood court review, the Commission cannot and 
will not apply them in currently pending

[[Page 23239]]

cases in which there has been no final determination of ROE issues.
    The Commission orders:
    (A) The Commission adopts the Policy Statement and supporting 
analysis contained in the body of this order.
    (B) This Policy Statement is effective the date issued and shall 
apply to all oil and gas pipelines then pending before the Commission 
in which there has been no final determination of ROE issues.

    By the Commission.
Nathaniel J. Davis, Sr.,
Deputy Secretary.

Appendix A

      Table 1.--DCF Analysis for Selected Corporations and MLPs Owning Jurisdictional Natural Gas Pipelines
                                 [Six-month period ended 03/31/2008, in percent]
----------------------------------------------------------------------------------------------------------------
                                                     (2)       (3) Growth      (4)
                                     (1) 6-mos. -------------     rate    -------------     (5)          (6)
                                        avg                     (``g'')                   Adjusted    Estimated
              Company                 dividend    IBES  (03/ -------------                dividend     cost of
                                       yield         08)       GDP (1/22/   Composite      yield        equity
                                                                  08)
----------------------------------------------------------------------------------------------------------------
Spectra Energy Corp...............         3.65            6         4.43         5.48         3.75         9.23
El Paso Corp......................         0.96           11         4.43         8.81         1.00         9.81
Oneok Partners, LP................         6.66            5         2.22         4.07         6.80        10.87
Boardwalk Pipeline Partners, LP...         6.29            6         2.22         4.74         6.44        11.18
Oneok, Inc........................         3.10           10         4.43         8.14         3.23        11.37
TC Pipelines, LP..................         7.46            5         2.22         4.07         7.61        11.68
TEPPCO Partners, LP...............         7.31            6         2.22         4.74         7.48        12.22
Spectra Energy Partners...........         5.00           10         2.22         7.41         5.18        12.59
Enterprise Products Partners, LP..         6.45            8         2.22         6.07         6.64        12.71
Kinder Morgan Energy Partners, LP.         6.69            8         2.22         6.07         6.89        12.96
Williams Companies................         1.17           16         4.43        12.14         1.24        13.38
----------------------------------------------------------------------------------------------------------------
Column (1) is taken from individual company analysis.
Column (2) is taken from I/B/E/S Monthly Summary Data, U.S. Edition.
Column (3) is calculated from three sources: BA, Global Insight, and SSA.
Column (4) = Column(2)*\2/3\ + Column(3)*\1/3\.
Column (5) = Column(1)*(1 + 0.5*Column(4)).
Column (6) = Column(4) + Column(5).


    Note: This Appendix is for illustrative purposes only and does 
not prejudge what would be an appropriate proxy group for use in 
individual proceedings.


                  Table 2.--DCF Analysis for Selected MLPs Owning Jurisdictional Oil Pipelines
                                 [Six-month period ended 03/31/2008, in percent]
----------------------------------------------------------------------------------------------------------------
                                                     (2)       (3) Growth      (4)
                                     (1) 6-mos. -------------     rate    -------------     (5)          (6)
                                        avg                     (``g'')                   Adjusted    Estimated
              Company                 dividend    IBES  (03/ -------------                dividend     cost of
                                       yield         08)      50% GDP (1/   Composite      yield        equity
                                                                 22/08)
----------------------------------------------------------------------------------------------------------------
Buckeye Partners, LP..............         6.72            5         2.22         4.07         6.86        10.93
Magellan Midstream Partners, LP...         6.16            6         2.22         4.74         6.30        11.04
NuStar Energy, LP.................         7.07            6         2.22         4.74         7.24        11.98
TEPPCO Partners, LP...............         7.31            6         2.22         4.74         7.48        12.22
Plains All American Pipelines, LP.         6.74            7         2.22         5.41         6.93        12.33
Enbridge Energy Partners, LP......         7.58            6         2.22         4.74         7.76        12.50
Enterprise Products Partners, LP..         6.45            8         2.22         6.07         6.64        12.71
Kinder Morgan Energy Partners, LP.         6.69            8         2.22         6.07         6.89        12.96
----------------------------------------------------------------------------------------------------------------
Column (1) is taken from individual company analysis.
Column (2) is taken from I/B/E/S Monthly Summary Data, U.S. Edition.
Column (3) is calculated from three sources: BA, Global Insight, and SSA.
Column (4) = Column(2)*\2/3\ + Column(3)*\1/3\.
Column (5) = Column(1)*(1 + 0.5*Column(4)).
Column (6) = Column(4) + Column(5).


    Note: This Appendix is for illustrative purposes only and does 
not prejudge what would be an appropriate proxy group for use in 
individual proceedings.

Appendix B

    In this Appendix, we illustrate with a simplified numerical 
example why a DCF analysis using a proxy MLP's full distribution, 
including any return of equity, does not lead to the award of an 
excess ROE in a pipeline rate case or the double recovery of 
depreciation.
    In this example, we compare the results of a DCF analysis for 
two firms included in a proxy group, one a corporation and the other 
an MLP. We initially assume that the theoretical basis of the DCF 
methodology is sound. In other words, the DCF formula will lead to 
valid results for investors in pricing shares and returns. We 
further assume that each proxy firm engages only in jurisdictional 
interstate natural gas pipeline business. Therefore, each proxy firm 
charges cost-of-service rates determined by the Commission in the 
proxy firm's last rate case. We also assume that the Commission 
awarded the same 10 percent ROE to each proxy firm in its last rate 
case.
    Based on these assumptions and the additional facts set forth 
below illustrating the typical differences between corporations and 
MLPs, we first set forth the DCF analysis

[[Page 23240]]

an investor would perform to determine the value of the 
corporation's stock and the MLP's limited partner units. We then 
assume, consistent with the underlying premise of the DCF model, 
that the results of the investor's DCF analysis represent the actual 
share prices of the two proxy firms. Using those share prices, we 
then apply the DCF formula used in rate cases to determine the ROEs 
of the two proxy firms. As illustrated below, that DCF analysis 
arrives at the same 10 percent ROE for the proxy MLP, as for the 
proxy corporation, despite the fact the MLP's distribution includes 
a return of equity. Thus, the inclusion of return of equity in the 
MLP's distribution does not improperly distort the rate case DCF 
analysis.

Assumed Facts

    The proxy corporation's rate base is $100. In its last rate 
case, the Commission awarded the proxy corporation an ROE of 10 
percent, and found that its depreciable life is 25 years. So the 
proxy corporation's cost of service includes $10 for ROE, and $4 for 
depreciation. We assume that in its most recent year of operations, 
the corporation actually collected those amounts from its customers, 
and paid a dividend of $6.50, i.e., a dividend equal to 65 percent 
of its annual earnings. The corporation thus retains $7.50 in cash 
flow, which it reinvests the following year. This reflects the fact 
that corporations typically pay out less than earnings in their 
dividends. We also assume that the corporation's composite growth 
rate is 8 percent.
    The facts with respect to the MLP are the same, with two 
exceptions. First, the MLP paid its unit holders a distribution of 
$13, i.e., a distribution equal to 130 percent of earnings. The 
remaining $1 is distributed to the general partner of the MLP. 
Second, the MLP's composite growth rate is only 5 percent.

DCF Analysis of Proxy Corporation

    As discussed at P 2 of the notice, an investor uses the 
following DCF formula to determine share price (with simplifying 
assumptions):

D/(ROE-g) = P
where P is the price of the stock at the relevant time, D is the 
current dividend, ROE is the discount rate or rate of return, and g 
is the expected constant growth in dividend income to be reflected 
in capital appreciation. Using that formula, investors would 
determine the rational stock price for the proxy corporation as 
follows:

$6.50 dividend/(ROE of .10-growth of .08) = Stock Price of $325

That is, investors would sell shares at a price above $325, and buy 
shares until the price reached $325. In a rate case for another 
pipeline, the Commission will determine the ROE of the proxy firm by 
solving the above formula for ROE, instead of share price. This 
rearranges the formula so that:

D/P + g = ROE

    Using that formula and assuming the proxy corporation's actual 
stock price is $325, the Commission would determine the proxy 
corporation's ROE as follows:

$6.50 dividend/$325 stock price + growth of .08 = ROE of .10

    Therefore, if the corporation was included in the proxy group 
for purposes of determining another firm's ROE in a new rate case, 
we would find, under the assumed facts, that the proxy corporation 
has the same 10 percent ROE as we awarded in its last rate case.

DCF Analysis of Proxy MLP

    We now go through the same exercise for the proxy MLP to 
determine whether its distribution in excess of earnings distorts 
its DCF analysis so as to improperly inflate its ROE. Using the D/ 
(ROE - g) = P formula described above, investors would determine the 
proxy MLP's share price as follows:

$13 distribution/ (ROE of .10 - growth of .05) = Share price of $260

    Assuming that the actual price of units in the proxy MLP is 
$260, we now determine the ROE of the proxy MLP, using the DCF 
formula used in rate cases (D/P + g = ROE). Under that formula, we 
would calculate the proxy MLP's ROE as follows:

$13 distribution/$260 unit price + growth of .05 = ROE of .10

    Therefore, if the MLP was included in the proxy group for 
purposes of determining another firm's ROE in a new rate case, we 
would, under the assumed facts, reach the same result as we reached 
for above proxy corporation: That the proxy MLP has the same 10 
percent ROE as we awarded in its last rate case.
    By contrast, if the Commission capped the proxy MLP's 
distribution at its $10 in earnings but continued to use the $260 
share price, the ROE calculated for the proxy MLP would be only 
about 8.8 percent, and thus less than the 10 percent ROE the 
Commission awarded the proxy MLP in its last rate case and less than 
the results for the proxy corporation:

$10 distribution/$260 unit price + growth of .05 = ROE of .088

Conclusion

    As shown by the above illustrative calculations, an MLP may be 
included in the proxy group and its full distribution used in the 
DCF analysis without distorting the results. This is because the 
level of an MLP's distributions affects both its share price and its 
projected growth rate. The MLP's inclusion of a return of equity in 
its distribution causes its share price to be higher than it 
otherwise would be and its growth rate to be lower. These facts 
offset the effect of the higher distribution on the DCF calculation 
of the MLP's ROE. Indeed, capping the MLP's distribution at earnings 
would lead to a distorted result. This is because there would be 
mismatch between the market-determined share price, which reflects 
the actual, higher uncapped distribution, and the lower earnings-
capped distribution.

[FR Doc. E8-9186 Filed 4-28-08; 8:45 am]

BILLING CODE 6717-01-P
