Verizon v. NH PUC CV—04— 65—PB 08/17/05
UNITED STATES DISTRICT COURT FOR THE DISTRICT OF NEW HAMPSHIRE
Verizon New England, Inc.
v. Case No. 04-CV-65-PB Opinion No. 2005 DNH 119
New Hampshire Public Utilities Commission
MEMORANDUM AND ORDER
Verizon New England, Inc.1 ("Verizon") owns and operates a vast telecommunications network in the state of New Hampshire. This network consists of various elements such as loops (wires that connect telephones, fax machines, and modems to switches), switches (devices that direct communications to destinations), and transport trunks (wires and cables that connect switches to other switches). See AT&T Corp. v. Iowa Util. B d., 525 U.S. 366, 371 (1999) (describing elements of a local telecommunications network).
1 Verizon New England is a subsidiary of Verizon Communications, Inc. In New Hampshire, Verizon New England does business as Verizon New Hampshire.
Verizon is required by the Telecommunications Act of 1996, Pub. L. 104-104, 110 Stat. 56 ("Telecommunications Act" or "Act"), to provide competing telecommunications carriers with access to the elements of its network on an unbundled basis. 47 U.S.C. § 251(c)(3). The Act, in turn, authorizes Verizon to charge a "just and reasonable" rate for access to such elements. See 47 U.S.C. § 252(d)(1). One of the components of a just and reasonable rate is an allocation for "cost of capital." See 47 C.F.R. § 51.505(b)(2). The Act's implementing regulations specify that cost of capital must be "forward-looking," i d ., but otherwise leave the concept undefined.
On January 16, 2004, the New Hampshire Public Utilities Commission ("PUC") issued an order setting Verizon's cost of capital for all purposes at 8.2%. See Order Establishing Cost of Capital ("Cost of Capital Order") at 71. Verizon challenges the order to the extent that it applies to the rates that Verizon will be permitted to charge for access to its unbundled network elements ("UNEs") because it contends that the PUC failed to use the forward-looking methodology that the Act and its implementing regulations require. Because I find this argument persuasive, I
3
vacate the PUC order.
I. The Cost of Capital Order The Cost of Capital Order states that a utilities' weighted average cost of capital "is determined by multiplying the cost of equity by the percentage of equity in the company's capital structure, and adding that number to the cost of debt, similarly multiplied by the percentage of debt in the capital structure." Cost of Capital Order at 4. Following this approach, the PUC proceeded to identify the capital structure, the cost of debt, and the cost of equity that it would use in determining Verizon's cost of capital.
The PUC determined that Verizon's capital structure should be 55% debt (comprised of 53% long-term debt and 2% short-term debt) and 45% equity. See i d . at 57. It based this determination on the average of Verizon New England's reported capital structure at year-end 2000 and 2001, and as of June 30 and September 30, 2002. See i d . at 50-51, 16. The Commission used book values for Verizon New England because the company did not maintain separate books for its New Hampshire operations.
4
See i d . at 48-51.
The PUC determined that Verizon's cost of debt was 2% for short-term debt2 and 7.051% for long-term debt. See i d . at 57. It explained that the short-term debt rate was undisputed and it drew the 7.051% long-term debt rate directly from the "embedded cost of debt for Verizon New England as of the balance sheet for June 30, 2 0 02." I d . at 57.
The Commission set Verizon's cost of equity at 9.82%. See i d . at 70. It used a three-stage version of the "Discounted Cash Flow" ("DCF") method to arrive at this figure. It described the DCF method by stating that it can be explained as
K = Do(1 + g) + g "where K is the cost of equity. Do PO
is the current annual dividend on one share of common stock, Po is the current stock price and g is the anticipated growth rate."3 I d . at 4. The Commission drew its inputs for stock
2 The PUC apparently arrived at the 2000 short-term debt figure by taking reports of Verizon New England's average daily short-term debt balances for the 13-month period ending December 31, 2002 (4.35%) and making a downward adjustment to account for short-term volatility. See i d . at 56.
3 For a more detailed description of the DCF method, see Roger A. Morin, Regulatory Finance: Utilities Cost of Capital (1994) 99-129.
5
price, annual dividend, and growth rate from a composite of two telecommunications companies that it determined were comparable to Verizon New England in "risk profiles, [and] positive dividend earnings growth on average over the last five years. . . Id. at 31, 61.
The Commission rejected Verizon's proposal to add a 5.48% risk premium to its cost of capital. See i d . at 47. Thus, applying Verizon's cost of debt (2% for short-term debt, 7.051% for long-term debt) and its cost of equity (9.82%) and using the approved capital structure (55% debt and 45% equity), the Commission determined that Verizon's weighted average cost of capital was 8.2%. See i d . at 70.
II. ANALYSIS
Verizon argues that the Cost of Capital Order cannot stand because the PUC improperly based the order primarily on historical data rather than the forward-looking cost of capital that a hypothetical business would incur if it were to offer access to UNEs in a competitive market. The PUC defends the order primarily by arguing that it was entitled to use historical
6
data because it supportably found that Verizon's historical cost of capital is a reliable proxy for its forward-looking cost of capital. To resolve this dispute, I begin by taking a closer look at what the Federal Communications Commission ("FCC") likely meant when it required state commissions to set cost of capital by using a forward-looking methodology. I then examine the Cost of Capital Order to determine whether the PUC used the correct methodology.
A. Forward-Looking Cost of Capital Neither the Telecommunications Act nor its implementing regulations explain what it is that qualifies a method for determining cost of capital as "forward-looking." We know, however, that the Act provides that state commissions must base access rates for UNEs on "cost" and that cost must be determined "without reference to a rate-of-return or other rate-based proceeding."4 47 U.S.C. § 252(d)(1)(A)(1). Because cost of
4 For a detailed discussion of rate-of-return regulation see Verizon Communications, Inc. v. F CC, 535 U.S. 467, 480-88 (2002). For a comparison of rate-of-return regulation with alternative pricing methodologies, see Jonathan E. Nuerchterlein & Philip J. Weiser, Digital Crossroads. American Telecommunications Policy in the Internet Age (2005), Appendix A.
7
capital is a component of an incumbent local exchange carrier's ("ILEC") recoverable cost. See 47 C.F.R. § 51.505(b)(2), it is at least evident that a forward-looking method for determining capital cost must be something other than rate-of-return regulation under a different name. Thus, because the forbidden rate-of-return method of rate setting looks to an ILEC's historical costs as a starting point, see Verizon. 535 U.S. at 500, it is reasonable to assume that, as the term "forward- looking" implies, the FCC intended state commissions to identify an ILEC's anticipated future cost of capital rather than merely to adopt its historical costs.
It also seems reasonably clear that the FCC intended state commissions to adopt certain assumptions that are described in the Act's implementing regulations when setting a cost of capital. The regulations identify a forward-looking cost of capital as a component of the "total element long run incremental cost" ("TERLIC") method of rate setting that the FCC adopted in place of traditional rate-of-return regulation. See i d . at 496 (identifying forward-looking cost of capital as a component of TELRIC). TELRIC, in turn, requires state commissions to base access rates for UNEs on the cost of operating a hypothetical network that is constructed "using the most efficient telecommunications technology currently available and the lowest cost network configuration given the existing location of an incumbent EEC's wire centers." 47 C.F.R. § 51.505(b)(1). It follows, therefore, that when calculating cost of capital under TELRIC, state commissions must attempt to determine the capital cost of operating the hypothetical network that TELRIC envisions rather than the network as it currently exists.
It is also important to bear in mind that the Telecommunications Act was designed to promote competition in the local telecommunications marketplace. Verizon, 535 U.S. at 488- 89. TELRIC encourages competition over an ILEC's existing network by requiring state commissions to set access rates based on the cost of a hypothetical state-of-the-art network rather than the presumably higher costs that the ILEC actually incurred in building its network. The FCC has recognized, however, that competitors will have no incentive to engage in the type of facilities-based competition that is the Act's ultimate aim if the allowed cost of capital is too low. As it has explained:
9
To calculate rates based on an assumption of a forward-
looking network that uses the most efficient technology (i.e. the network that would be employed in a competitive market), without also compensating for the risks associated with investment in such a network, would reduce artificially the value of the incumbent LEG network and send improper pricing signals to competitors. Establishing UNE prices based on an unreasonably low cost of capital would discourage competitive LECs from investing in their own facilities and thus slow the development of facilities-based competition.
In the Matter of Review of the Section 251 Unbundling Obligations of Incumbent Local Exchange Carriers ("Triennial Review Order"), 2003 WL 22175730 *17396-97 5 682 (2003). To address this concern, the FCC required state commissions to adopt certain assumptions when setting a forward-looking cost of capital. Most significantly, such commissions must assume that an ILEC is offering to lease network elements in an environment in which there is facilities-based competition. I d . at 5 680. This assumption is vital, the FCC reasoned, because facilities-based competition increases risk and increased risk in turn results in an increased cost of capital. I d . at 5 681. Accordingly, a forward-looking process for determining cost of capital must attempt to identify the hypothetical cost of capital that a
10
competing local exchange carrier ("LEG") would face in building and operating a network in an environment in which there is facilities-based competition. Further, the process must account for the FCC's determination that a market with facilities-based competition will produce greater risk and hence a higher cost of capital than a market without such competition.
To summarize, the forward-looking method for calculating cost of capital envisioned by the Telecommunications Act and its implementing regulations requires an assessment of anticipated future costs rather than historical costs, it requires an assessment of the cost of capital that a competing LEG would incur in building and operating the hypothetical network that TELRIC assumes, and it requires that this assessment be made in an environment in which there is facilities-based competition.
B. The Cost of Capital Order A careful review of the Cost of Capital Order leaves no doubt that the PUC used a historical method rather than a forward-looking method to determine all three of the essential components of Verizon's cost of capital. The Commission relied directly on Verizon New England's historical capital structure in
11
selecting a capital structure for Verizon's New Hampshire operations. It also explained that it had relied on data from Verizon New England rather than Verizon New Hampshire only because Verizon New England did not maintain separate books for its New Hampshire business. The Commission also based its long term debt rate directly on Verizon New England's embedded debt costs. Finally, although it did not use inputs from Verizon New England for its cost of equity calculation, the Commission selected inputs from two other telecommunications companies with risk profiles similar to Verizon's. Given this approach, it is difficult to see how the PUC can credibly maintain that it calculated Verizon's cost of capital "without reference to a rate-of-return or other rate-based proceeding," 47 U.S.C. § 252(d)(1)(A)(i), as the Telecommunications Act requires.
In response, the PUC offers only straw man arguments to support its claim that it used a forward-looking methodology. For example, it argued in the Cost of Capital Order that its methodology was appropriate because TELRIC does not bar a state commission from using the same cost of capital for both an ILEC's retail and UNE rates. Cost of Capital Order at 43-44. While
12
this may well be true in theory, it fails to address Verizon's contention that the PUC's methodology was not forward-looking. New Hampshire law requires the PUC to use a rate-of-return methodology to set Verizon's retail rates. See Appeal of Chester Bridge Corp., 126 N.H. 425, 431 (1985) (describing rate setting method required under N.H. Rev. Stat. Ann. § 378:7). The Telecommunications Act, in contrast, requires that access rates for UNEs be set "without reference to a rate-of-return or other rate-based proceeding." 47 U.S.C. § 252(d)(1)(A)(i). Thus, while it is conceivable that these two methods could produce the same cost of capital in certain cases, this theoretical possibility does not relieve the PUC of its obligation to set UNE rates using the methodology required by federal law.
The Commission also makes much of the fact that it used the DCF method to determine Verizon's cost of equity. Def.'s Mem. at 13-14. The DCF method, however, is neither inherently forward- looking nor inherently backward-looking. It is the selection of inputs that makes the difference. If, as was the case here, a commission selects inputs that seek to replicate the ILEC's historical stock price, dividend and growth rate, without
13
accounting for the risk that a competing LEG would face in offering access to UNEs in the kind of market that TELRIC assumes, its use of the DCF method is historical rather than forward-looking.
The PUC alternatively argues that it reasonably relied on a historical method for computing Verizon's cost of capital because Verizon failed to prove "that it would face greater risks in a fully competitive wholesale market than it does in the provision of retail services." Def's Mem. at 9. In light of this failure of proof, the Commission reasons that Verizon's historical cost of capital is an acceptable substitute for its forward-looking cost of capital. I reject this argument because it misstates Verison's burden of proof and fails to properly account for the assumptions about risk and its effect on cost of capital that are an essential part of a forward-looking methodology.
As I have already noted, the FCC has concluded that the provision of UNEs in a market in which there is facilities-based competition necessarily involves greater risk than ILEC's currently face in the highly regulated retail markets in which they operate. It has also determined that increased risk
14
necessarily results in an increased cost of capital. In the face of these determinations, the PUC cannot justify its reliance on Verizon New England's historical cost of capital as a proxy for its forward-looking capital cost merely by claiming that Verizon has failed to prove what the Telecommunication Act's implementing regulations require the PUC to assume. While the Telecommunications Act does not flatly prohibit embedded cost methods such as the one that the PUC used in this case, as the Supreme Court has observed, "it seems safe to say that the statutory language places a heavy presumption against any method resembling the traditional embedded cost-of-service model of rate-setting." Verizon, 535 U.S. at 512. The PUC has failed to overcome this presumption merely by claiming that Verizon has not proved that it will face greater competitive risk in the kind of market that TELRIC assumes.
IV. CONCLUSION
Because each of the variables relied upon by the PUC to calculate Verizon's overall cost of capital were calculated using an improper methodology, the Cost of Capital Order must be set
15
aside. Verizon's motion for summary judgment (Doc. No. 36) is therefore granted, the PUC's motion for summary judgment (Doc. No. 38) is denied, and the clerk is instructed to enter judgment accordingly.5 SO ORDERED.
/s/Paul Barbadoro________ Paul Barbadoro
United States District Judge
August 17, 2005
cc: Lynn R. Charytan, Esq.
Thomas J. Donovan, Esq.
Suzanne M. Gorman, Esq.
Daniel J. Mullen, Esq.
5 The PUC has made the additional argument that this case must be dismissed because it was not filed in a timely manner. Absent the existence of an explicit limitations period, civil claims that arise under federal statutes enacted after December 1, 1990 are subject to 28 U.S.C. § 1658(a) which imposes a fouryear limitations period on such actions. See Peiepscot Indus. Park. Inc. v. Maine Cent. R.R. Co.. 215 F.3d 195, 203 n.5 (1st Cir. 2000). This case was brought under the Telecommunications Act of 1996, a statute enacted after December 1, 1990 without any explicit limitations period. Section 1658(a) therefore applies. This case would thus have had to have been filed on January 16, 2008, four years after the PUC rendered its order, for it to be barred. Instead, Verizon's suit was filed on February 19, 2004, well within the statutory period. The PUC's claim that the statute had run prior to the date on which Verizon filed this action therefore has no merit.