New england roe update

This a summary of briefs filed in the return on equity cases at FERC involving the New England Transmission Owners.  This now reflects all briefs. 
  
Background:
Currently there are four pending complaints against the New England Transmission Owner’s (NETOs) ROE that go back to 2011.  Each complaint has been fully litigated before an Administrative Law Judge (ALJ) with only the first complaint resulting in a FERC Commission decision (Opinion 531). The D.C. Circuit of Appeals vacated the Commission’s determinations in its order on the First Complaint (Opinion No. 531).  In the meantime, the NETOs are continuing to collect their 10.57% base ROE from Opinion 531, although the Commission has indicated that it will exercise its “broad remedial authority” to correct its legal error to make whatever ROE it sets on remand effective as of the date of that Order. 
 
On October 16, 2018, FERC issued an order in these complaint cases.  In its Order, FERC set forth its methodology for addressing ROE complaints while considering the remand from the DC Court.  In its proposal, FERC gives equal weight to the results of the four financial models in the record in the NETO cases, instead of primarily relying on the DCF model. In relying on a broader range of record evidence to estimate the NETOs’ cost of equity, FERC states that this will ensure that the selected ROE is based on substantial evidence and bring its methodology into closer alignment with how investors make investment decisions. 
 
To determine whether an existing ROE remains just and reasonable (i.e., the first prong of the FPA section 206 analysis), FERC proposes (1) relying on the three financial models that produce zones of reasonableness—the DCF, CAPM, and Expected Earnings models—to establish a composite zone of reasonableness[1]; and (2) relying on that composite zone of reasonableness as an evidentiary tool to identify a range of presumptively just and reasonable ROEs for utilities with a similar risk profile to the targeted utility.  Under this approach, FERC intends to dismiss an ROE complaint if the targeted utility’s existing ROE falls within the range of presumptively just and reasonable ROEs for a utility of its risk profile—unless that presumption is sufficiently rebutted. 
 
Where the existing ROE has been shown to be unjust and unreasonable and therefore, requiring that FERC move to the second prong of the FPA section 206 analysis, FERC proposes to rely on all four financial models in the record—i.e., the three listed above, plus the Expected Earnings model—to produce four separate cost of equity estimates. FERC proposes to give them equal weight by averaging the four estimates to produce the just and reasonable ROE.  For each of the DCF, CAPM, and Expected Earnings models, FERC proposes to use the central tendency of the respective zones of reasonableness as the cost of equity estimate for average risk utilities.  FERC would then average those three midpoint/median figures with the sole numerical figure produced by the Risk Premium model to determine the ROE of average risk utilities.  FERC would use the midpoint/medians of the resulting lower and upper halves of the zone of reasonableness to determine ROEs for below or above average risk utilities, respectively. Because its current policy is to cap a utility’s total ROE (the base ROE plus incentive ROE adders) at the top of the zone of reasonableness, FERC proposes to use the composite zone of reasonableness produced by the DCF, CAPM, and Expected Earnings to establish the cap on a utility’s total ROE.
The October 16 Order evaluated the justness and reasonableness of existing ROE of the NETOs by dividing a “composite zone of reasonableness” bounded by the average of the highest values obtained in the DCF, CAPM and Expected Earnings analysis, and the average of the three lowest values obtained in those analyses, into quartiles. The Order compares the pre-complaint NETO ROE of 11.14% to a “middle quartile” extending from 37.5% (three-eighths) to 62.5% (five-eighths) of a range from 7.51% to 13.08%, or from 9.6% to 10.99%.  Since the NETOs’ Opinion No. 489 ROE of 11.14% exceeds 10.99%, FERC would find it unjust and unreasonable, and reduce it to 10.41%.  FERC would also set the total ROE cap at 13.08%. 
 
NETOs’ Brief:
NETOs support the overall approach proposed by FERC in the October 2018 Order and recommend limited changes to ensure that it is statutorily and procedurally sound and consistent with the D.C. Circuit’s opinion in Emera Maine.  The NETOs request that the Commission adopt the following results for these proceedings:

Table A
Case        Prior ROE       Presumptive Range         ZOR           New ROR
I                  11.14%          9.60%-10.99%        7.51%-13.08%     10.41%

II                 10.41%          9.85%-11.21%        Dismiss - N/A           N/A

III                10.41%          9.62%-11.10%        Dismiss - N/A           N/A

IV                10.41%          9.42%-10.88%        Dismiss - N/A           N/A
 
 
With some limited modifications, FERC’s approach satisfies the two steps of FPA Section 206 and the two holdings of the Court’s Emera Maine decision.  First, it accords utilities the “statutory protection” that the Court mandated by ensuring that the Commission will not exercise its FPA Section 206 authority unless it satisfies the “condition precedent” of showing that the existing rate is outside “a broad range of potentially lawful ROEs.”  Second, it provides the rational connection between the “record evidence” that undermined the reliability of the DCF analysis and the Commission’s “placement of the base ROE.”  In developing a new framework that addresses the Court’s directives, FERC also achieves the “careful balance that attracts sufficient transmission investment but doesn’t impose undue burdens on consumers.”
 
The NETOs request the following clarifications:

1. The range of presumptively just and reasonable ROEs is distinct from the statutory zone of reasonableness—i.e., the smaller (single quartile) range used to establish a presumption of whether an existing ROE is just and reasonable is solely a screening tool to allow the Commission to evaluate an existing base ROE in step one of its FPA Section 206 inquiry.
2. The range of presumptively just and reasonable ROEs does not affect the Commission’s use of the wider statutory zone of reasonableness under step two of FPA Section 206 to set a cap on overall ROE, including incentives.
3. Complaints should not be automatically granted where an ROE falls between the high end of the narrower quartile range established for step one purposes and the high end of the wider statutory zone of reasonableness. Rather, there should be a rebuttable presumption that such ROEs may not be just and reasonable. This rebuttable presumption would not be sufficient by itself to support an automatic ruling that grants a complaint; instead it would have the effect of shifting the burden of proof so that TOs, in responding to a complaint, would be required to identify specific circumstances or other evidence demonstrating that a base ROE above the quartile range remains just and reasonable in order to rebut this presumption, or that the quartile range is incorrectly identified. 
4. A complaining party may only overcome the presumption that a base ROE within the range of the designated quartile is just and reasonable through clear and convincing evidence submitted in the Complaint – that FERC does not intend to set every complaint for hearing on the issue of whether a complainant can overcome the rebuttable presumption that existing rates within the designated quartile are just and reasonable.
5. FERC should not reduce ROE incentives in these ROE complaint cases.   

 
 
Trial Staff Brief:
Recommends the following:

1. Eliminate the Expected Earnings approach as it does not measure the market cost of equity.
2. Recommends three adjustments to the CAPM proposed in the October Order:  1) utilize the two-step growth rate in calculating the market return versus the one-step; 2) remove the size premium adjustment; and 3) implement screening criteria for growth rates (remove companies with short-term IBES growth rates or composite growth rates that are zero and less and short-term growth rates that are 20 percent and higher).
3. Eliminate from the Risk Premium method: 1) ROEs that were not based on a market analysis at the time they were authorized; 2) the six-month average Moody’s “Baa” Public Utility Bond Yields for the months October to December 2008 and from January to August of 2009 because those yields either suddenly rose 143 basis points or those yields were in steady decline from that sudden rise; and 3) projected bond yields believing FERC inadvertently used  projected bond yields in the Risk Premium analysis as FERC recently rejected the use of projected bond yields as being speculative and less reliable than historical yields.
4. Trial Staff recommends the following base ROEs and ROE Caps for each of the four complaints[2]:

Table B
Complaint      Base ROE     With Expected Earnings     ROE Cap      With Expected Earnings
I                       9.29%                     9.57%                      10.82%                   11.72%
II                      9.26%                     9.68%                      10.13%                   11.62%
III                     9.14%                     9.58%                      11.06%                   12.02%
IV                     8.45%                     9.28%                      10.46%                   11.63%

 
EMCOS Brief:
Eastern Massachusetts Consumer-Owned entities (EMCOs) state that the DCF remains a sound and reliable method to determine the market cost of equity.  FERC has not used other ROE methodologies to same degree as DCF, so other methodologies lack maturity in application by FERC.  FERC must exercise extreme caution when employing CAPM and Risk Premium.  FERC should not use Expected Earnings as it is based upon accounting data and not reflective of the market cost of equity.  FERC’s proposed Expected Earnings analysis is impervious to the market cost of equity capital and heavily anchors ROE results in past regulatory decisions.  The proposed averaging of extreme ends of ranges of implied costs of equity exaggerates the skew that results from reliance on the midpoint, as opposed to the median.  The results of the NETOs’ non-DCF analyses are substantially overstated in ways that the limited adjustments undertaken in developing the October 16 Order do not address.  Failure to deploy a predictable and statistically effective screen for high-end outliers means that the resulting ranges continue to be skewed toward higher than reasonable results. 
 
 
Recommend that FERC: 

1. Eliminate Expected Earnings analysis.
2. Average the CAPM and DCF results for each proxy group member, and then using the high and low ends to delineate the “composite” range of implied costs of equity rather than averaging the result of each ROE method.  This modification would ensure that the “composite” range provides more comprehensive information on the dispersion of implied costs of equity among proxy group members.
3. Use appropriate measures to avoid manipulating the CAPM and Risk Premium. 
4. Replace the 150% high outlier screen with the Grubbs Test, a familiar, widely accepted and demonstrably reliable statistical screen for elimination of high-end outliers.
5. Clarify that the October 16 Order’s “zone of reasonableness quartiles” proposal does not create a “zone of immunity” for any ROE value within the zone or in any way prejudice or preclude a demonstration that circumstances render a rate within the relevant quartile unjust and unreasonable.
6. Recognize that the NETOs are not of average risk. On the contrary, various aspects of the NETOs’ arrangements under the ISO New England Tariff— use of a formula rate with cost-plus equity return and income tax allowances, the use of equity-heavy capital structures, and limited exposure to customer challenges to cost recovery under that formula rate—place the NETOs substantially below average in terms of business and financial risk.
7. Complaint I - Adopt a base ROE of 8.07% (the product of two-thirds weighted median value of a composite DCF-CAPM range and one-third weighted risk premium) and that any incentive adders be capped at 11.75%, the top of the DCF zone of reasonableness.
8. Complaint II - Adopt a base ROE of 8.32% (the product of two-thirds weighted median value of a composite DCF-CAPM range and one-third weighted risk premium) and that any incentive adders be capped at 9.72%, the top of the DCF zone of reasonableness.
9. Complaint III - Adopt a base ROE of 8.58% (the product of two-thirds weighted median value of a composite DCF-CAPM range and one-third weighted risk premium) and that any incentive adders be capped at 10.32%, the top of the DCF zone of reasonableness.
10. Complaint IV – Adopt a base ROE of 7.67% (the product of a two-thirds weighted median value of a composite DCF-CAPM range and a one-third weighted risk premium) and that any incentive adders be capped at 10.00%.

 
 
 
CAPS Brief:
Consumer Aligned Parties (CAPS) state that FERC’s proposal to determine a zone of presumptively just and reasonable ROEs that would be used to determine if the existing ROE is just and reasonable is: 1) inconsistent with the Federal Power Act’s consumer protection purpose; 2) creates an asymmetry between Section 205 and 206 cases in that a TO can request a higher ROE if its calculations demonstrate one higher than the existing ROE – on the contrary, in a Section 206 case, the complainant would have to show that the calculated ROE is below the presumptive range; 3) the DC Court did not contemplate, let alone require, a presumption that an above-cost ROE remains just and reasonable unless it exceeds the cost-based level by more than one-eighth of the composite range; 4) the specifics of the proposed presumptive range are arbitrary - the composite range is arbitrary, and the non-DCF methods used in identifying the composite range are unreasonable, if applied as the Order does; and 5) the presumption range departs without justification from precedent specific to New England transmission ROEs in Opinion 489 where FERC required refunds for any NETO which had an ROE different than the new, determined ROE. 
 
CAPS disagree with FERC that the outcome of the prior complaint being deemed the “existing” rate for purposes of the next complaint.  CAPs state that the existing ROE for purposes of the second, third, and fourth complaints must be the ROE that is charged or collected by the NETOs—that is, the allowed ROE that was either (a) actually charged when a Section 206 complaint is filed, or (b) determined and fixed by a FERC order making it effective at such time.  FERC’s approach here ignores the extended timeline associated with the complaints in these proceedings, where the period between filing of the complaint and Commission action has well exceeded the fifteen-month refund protection afforded by Section 206 and the existing rate will not always be identical to the outcome of the prior Complaint.    
 
CAPs recommend the following changes:

1. For the CAPM analysis, use a lower growth component of the market risk premium - use a second-stage GDP-based constraint in determining the market risk premium to which the adjusted betas are multiplied - and eliminate the size adjustment.
2. Eliminate the use of Expected Earnings as it does not measure the market cost of equity – if it is used, the expected earnings per book equity should be adjusted to account for how the proxies’ equity ratios differ from those of NETOs.
3. The Risk Premium analysis can accentuate erratic market conditions and tends to overemphasize recent market changes.  It should not be used – if used, the NETO calculations on which the Order relies is riddled with errors and therefore it must be modified to use actual bond yields and not projected bond or treasury yields, not use outdated ROES (use those granted by state commissions for T&D utilities), recognize the relatively low risk of the NETOS and should not assume that the risk premium increases linearly, as bond yields decline.
4. The three proxy-based methods (DCF, CAPM, and Expected Earnings) should not be treated as if they studied three different proxy groups. If those three studies are all to be referenced and given equal weight, then their individual-proxy outputs should be averaged to produce a composite return estimate for each proxy, before forming the composite range (same recommendation as EMCOs).
5. The NETOs are lower than average risks because they have formula rates and higher TOs equity ratios.
6. The DCF approach should continue to be relied upon by FERC;
7. The new base ROE should be set at the median and not the midpoint for the NETOs.
8. Any ROE method that relies on ranges, and especially ones that set the base ROE on midpoints, should be stringent and transparent in excluding both low and high outliers - the specific outlier tests suggested in the Order are neither stringent nor transparent.  FERC should replace its 150% high end outlier test with a two standard deviation test.
9. Without a rulemaking, any approaches to ROE other than DCF should not be used to determine the ROE Cap.
10. CAPs recommend the following ROEs:

 
Start              End        Recommendation                    Alternative
10/1/2011    12/26/2012               8.91%                                    9.64%
12/27/2012   12/31/2012             8.79%                                    9.64%
1/1/2013         3/27/2014              8.79%                                    9.79%
3/28/2014       7/30/2014               11.14%                                  11.14%
7/31/2014     10/15/2014             8.64%                                    9.64%
10/16/2014   10/30/2015             8.64%                                    9.64%
10/31/2015     4/28/2016              8.91%                                   9.64%
4/29/2016       9/29/2016              8.33%                                    9.19%
9/30/2016       7/28/2017              8.64%                                    9.64%
7/29/2017       9/29/2018              8.91%                                    9.64%
9/30/2018      continuing               8.33%                                    9.19%
 
 
 
 
 
AEP and EEI Comments:
Both entities request that FERC not decide how to determine the central tendency (midpoint versus median) for a single transmission owner in the NETO proceeding which involves several New England transmission owners. 
 
Southern California Edison Comments:
Suggests that the issue of central tendency for a single-filing utility rate be left for a single utility ROE case where it can be briefed more thoroughly (same point as AEP and EEI) and that FERC consider modifying its proxy group selection criteria to ensure that a small proxy group does not negatively impact ROEs by allowing individual transmission owners who have small proxy groups to propose alternative methods for determining an appropriate proxy group.  The practice of setting the low-end threshold 100 basis points above the utility bond yield does not contemplate that the spread between utility bond yields and the cost of utility equity can change over time, and thus the 100-basis point
spread may be too low.
 
Louisiana Public Service Commission Comments:
Requests late-intervention as they are concerned that they will not have the ability to influence FERC’s direction on ROE in other proceedings involving transmission owners doing business in Louisiana and therefore request that ability here.


[1] Each of these three methodologies relies on a proxy group to determine a zone of reasonableness, and thus the top and bottom of the zone of reasonableness produced by each methodology can be averaged to determine a single composite zone of reasonableness. After determining the composite zone of reasonableness, FERC will then calculate the lower, middle and upper ranges of that composite zone. The presumptively just and reasonable ROEs for below-average, average, and above-average risk utilities will then be the quartile of the respective zone. 
[2] FERC Trial Staff concludes in Complaint I that the existing 11.14% ROE is unjust and unreasonable and therefore FERC needs to determine the just and reasonable ROE, and they recommend 9.29%.  They do not reach such conclusions in the other complaints.  They simply provide the information needed for FERC to determine if the existing ROE is within the presumptive just and reasonable range and provide a new ROE if FERC determines that the existing ROE is unjust and unreasonable.  Table B provided herein contains Staff’s recommendations if FERC were to reset the ROE and ROE Cap in each complaint case.