i
Before the
FEDERAL COMMUNICATIONS COMMISSION
Washington, DC 20554
In the Matter of
Connect America Fund
A National Broadband Plan for Our Future
Establishing Just and Reasonable Rates for Local
Exchange Carriers
High-Cost Universal Service Support
Developing an Unified Intercarrier
Compensation Regime
Federal-State Joint Board on Universal Service
Lifeline and Link-Up
Universal Service Reform – Mobility Fund
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
WC Docket No. 10-90
GN Docket No. 09-51
WC Docket No. 07-135
WC Docket No. 05-337
CC Docket No. 01-92
CC Docket No. 96-45
WC Docket No. 03-109
WT Docket No. 10-208
COMMENTS
of
PEÑASCO VALLEY TELEPHONE COOPERATIVE, INC.
ii
TABLE OF CONTENTS
INTRODUCTION AND SUMMARY ..................................................................................... iii
I. ANALYSIS PERFORMED BY PEÑASCO VALLEY TELEPHONE
COOPERATIVE, INC. ..................................................................................................1
II. THE FCC’S REGRESSION ANALYSIS UTILIZES INCORRECT DATA ...........2
III. THE MODEL DOES NOT YIELD CONSISTENT RESULTS FOR
SIMILARLY SITUATED COMPANIES.....................................................................3
IV. THE REGRESSION MODEL IS OVERLY COMPLEX AND
UNPREDICTABLE THUS DISCOURAGING FUTURE INVESTMENT ..............6
V. THE FCC’S REGRESSION ANALYSIS DOES NOT CONSIDER THE
IMPACTS OF DEPRECIATION RESERVE .............................................................7
VI. THE LIMITATIONS ARE APPLIED INCORRECTLY TO THE HIGH
COST LOOP SUPPORT ALGORITHM .....................................................................7
VII. THE LIMITATIONS ARE MISSING CRITICAL COMPONENTS .......................9
VIII. THE FCC’S REGRESSION ANALYSIS DOES NOT APPROPRIATELY
CALCULATE LIMITATIONS ON DEPRECIATION EXPENSE ........................11
IX. CONCLUSIONS ...........................................................................................................12
iii
Introduction and Summary
The Federal Communications Commission’s (Commission or FCC) Report and Order
and FNPRM1 in the above captioned proceeding requests comment on proposed changes to the
existing Universal Service Fund (USF) and Intercarrier Compensation (ICC) mechanisms for
rural rate-of-return carriers, among other issues. Specifically, the FCC requests comments on
Sections XVII.A-K of the FNPRM, which address a wide variety of USF related issues.
Peñasco Valley Telephone Cooperative, Inc. 2 (PVT) submits these comments for the
FCC’s consideration. PVT is a rural telecommunications provider serving 2,823 voice access
lines and 1,223 broadband customers in the State of New Mexico. The following characteristics
are true of PVT:
PVT is the Carrier of Last Resort designated by the New Mexico Public Regulation
Commission, which legally obligates the company to provide telecommunications service
to all requesting customers within its service territory.
1 In the Matter of Connect America Fund, WC Docket No. 10-90, A National Broadband Plan for Our Future, GN
Docket No. 09-51, Establishing Just and Reasonable Rates for Local Exchange Carriers, WC Docket No. 07-135,
High-Cost Universal Service Support, WC Docket No. 05-337, Developing an Unified Intercarrier Compensation
Regime, CC Docket No. 01-92, Federal-State Joint Board on Universal Service, CC Docket No. 96-45, Lifeline and
Link-Up, WC Docket No. 03-109, Universal Service Reform – Mobility Fund, WT Docket No. 10-208, Report and
Order and Further Notice of Proposed Rulemaking and Further Notice of Proposed Rulemaking, FCC 11-161 (rel.
November 18, 2011) (Report and Order and FNPRM).
2 PVT is located in Southeastern New Mexico with a service territory of 4,651 square miles. The terrain is extremely
rocky and mountainous and very little water. This rocky and mountainous terrain makes it extremely expensive to
plow/rocksaw cable and fiber in PVT’s territory. With 2,350 customers in the 4,651 square miles of service territory,
the density per square mile is only 0.5 customers per square mile. The economic drivers are farming, ranching and
petroleum.
iv
PVT is the Eligible Telecommunications Carrier (ETC) determined by the New Mexico
Public Regulation Commission to provide universal service within the company’s
designated service territory.
PVT receives High Cost Support from the Federal Universal Service Fund. This support
totaled $5,728,317 in 20103 and comprised over 50.86% of PVT revenues in 2010.
Support came from the following sources:
o High Cost Loop Support (HCLS) $3,085,887
o Interstate Common Line Support (ICLS) $2,104,134
o Local Switching Support (LSS) $538,296
PVT generates revenues from providing intrastate switched access and reciprocal
compensation services. In 2010 intrastate switched access and net reciprocal
compensation revenues totaled $219,983.
PVT provides voice and broadband services to schools, libraries, rural health care
facilities, governmental agencies, and/or other anchor institutions within its service
territory.
PVT is one of the three largest employers in the company’s rural service territory,
providing 85 jobs and financial stability in rural areas of Southeastern New Mexico.
PVT has deployed substantial financial and human resources to provide voice and
broadband services under the existing rate of return rules prescribed by the FCC and by
3 2010 revenues are used throughout these comments because final 2011 numbers are not yet known. We believe
that 2010 revenues are reasonably representative of 2011.
v
the New Mexico Public Regulation Commission. In 2010, PVT had gross regulated
investment of $68,070,083
PVT would not have had the financial resources to deploy and maintain either voice or
broadband services without rate of return regulation and the support of the Universal
Service Fund under the existing rules.
PVT is very concerned with the potential financial implications of the Report and Order
and FNPRM and the impact they will have on PVT’s ability to continue to provide high
quality voice and broadband services at the public interest standards established by the
Commission.
In these comments, PVT outlines the impacts that adoption of the limitations on capital
and operating expenses, as proposed in the Report and Order and FNPRM, would have on its
financial results.
1
Before the
FEDERAL COMMUNICATIONS COMMISSION
Washington, DC 20554
In the Matter of
Connect America Fund
A National Broadband Plan for Our
Future
Establishing Just and Reasonable Rates
for Local Exchange Carriers
High-Cost Universal Service Support
Developing an Unified Intercarrier
Compensation Regime
Federal-State Joint Board on Universal
Service
Lifeline and Link-Up
Universal Service Reform – Mobility
Fund
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
WC Docket No. 10-90
GN Docket No. 09-51
WC Docket No. 07-135
WC Docket No. 05-337
CC Docket No. 01-92
CC Docket No. 96-45
WC Docket No. 03-109
WT Docket No. 10-208
COMMENTS
of
PEÑASCO VALLEY TELEPHONE COOPERATIVE, INC.
I. Analysis Performed by Peñasco Valley Telephone Cooperative, Inc.
In order to provide relevant financial context to the FCC in these comments, PVT
engaged Moss Adams LLP4 to perform a detailed financial analysis of the potential impacts of
the limitations on capital and operating expenses proposed in the Report and Order and FNPRM.
This analysis primarily focused on the impacts of the proposed regression analysis identified in
4 Moss Adams LLP (Moss Adams) is the 11th largest accounting and consulting firm in the United States, with more
than 225 partners and 1,800 staff. Moss Adams’ Telecom Group has served the telecommunications industry since
1957. Today, they provide audit, tax, and consulting services to more than 80 small and mid-sized
telecommunications carriers throughout the United States and its territories.
2
Appendix H to the Report and Order and FNPRM. This analysis was performed using PVT data
used by, and provided by, the FCC in the development of its regression analysis. In doing so,
Moss Adams recreated the regression analysis performed by the FCC and reproduced the same
results. In addition, Moss Adams also utilized other information generally available from PVT
in the analysis. The following comments include our overall assessment of the FCC’s regression
analysis and provide a summary overview of the financial impacts on PVT, including the
impacts of changes in the analysis proposed by PVT.
II. The FCC’s Regression Analysis Utilizes Incorrect Data
The census data that the Commission uses as inputs to its model in the Report and Order
and FNPRM are subject to a substantial degree of error. In any model, where there are errors or
inaccuracies in the inputs, those data flaws will also create errors or inaccuracies in the outputs
of the model. The Report and Order and FNPRM relies on significantly flawed data, and
therefore produces similarly flawed results.
Part of the input error is created by the Commission’s use of the Tele Atlas tool to define
study areas. This tool is notably flawed for PVT. While the Commission’s model assumes
PVT’s study area encompassing 2,331 square miles, PVT’s actual study area encompasses
approximately 4,651 square miles – an extraordinary error by any measure and one that is certain
to produce a flawed result.
This is extremely significant to PVT. Prior to regression caps, 2010 PVT data at a
national average cost per loop of $509.06, yielded estimated High Cost Loop Support (HCLS) of
$3,207,529. The proposed regression caps, which included the incorrect square miles above,
reduced HCLS to $3,000,821, a reduction of $206,708 or 6.44%. Increasing the square miles in
the regression analysis to the correct amount of 4,651 results in HCLS of $3,182,726, a much
3
more palatable reduction of $24,803. Should the Commission continue down this regression
path, the data for PVT and all carriers needs to be accurate. PVT serves as an example that the
census data which the Commission uses as inputs to its model are subject to a substantial degree
of error, and therefore is not appropriately used to cap PVT’s costs.
III. The Model Does Not Yield Consistent Results for Similarly Situated
Companies
In defining similarly situated companies, PVT notes that the FCC’s model used to
perform the regression analysis does not take into account the length of loops – a major factor
leading to high loop costs. In PVT’s case, it has 1.08 access lines per mile of loop plant. The
model also does not take into account unusual terrain conditions such as rock that PVT must cut
through with heavy equipment (rocksaw) or directional bore to bury cable plant. PVT noted that
these conditions often cause significant delays and cost increases to place cable plant. PVT
points out that when comparing PVT to similarly situated companies like those listed in the table
below, companies with higher concentrations of loops had significantly higher cable and wire
facility (CWF) caps per mile than PVT. Noted elsewhere in these comments, the limitation on
PVT’s CWF plant yields significant impacts to the company and limits their ability to upgrade
the plant facilities in order to meet the FCC's stated broadband requirements.
Using the FCC land area data, PVT’s service area has 1.26 loops per square mile (if the
land area in the regression model were corrected to the 4,651 sq. miles, then PVT serves 0.65
loops per square mile). PVT has a ceiling or cap on its cable and wire loop facilities of $13,498
per mile. The table provided compares loops, land area served and the resulting caps under the
proposed regression model. The table supports PVT’s position that areas with few subscribers,
or loops, per mile necessitate higher values under the regression caps and that caps should be
comparable for similarly situated companies. However, the table shows contrary results.
4
The table demonstrates that of the six companies listed, all of them have more loops per
mile, (i.e., have a higher population density than PVT), however four out of the five have
significantly higher caps per mile. Specifically, Example E with 2.4 loops per mile, which is very
similar to the total loops of PVT (2,865 for Example E versus 2,948 for PVT), yields a
significantly higher cap. The regression model yields a cap of $20,537 per mile for Example E,
$7,039 more per mile than PVT even though Company E has a more densely populated area.
PVT points out that utilizing the same $20,537 per mile cap for its area would yield a loop cable
and wire facility investment cap of $47,875,323. This is an increase of $16,408,800, or 52.15%,
higher than its current cap and much more consistent with PVT’s actual investment in cable and
wire facilities. PVT also points out that Example A which has a density very consistent with
PVT’s density (1.38 loops per mile for Example A), has a cap that is also lower than the other
companies depicted in the table provided with more loops per mile. This is yet another example
of the regression analysis not producing consistent results for similarly situated companies.
5
Loops
Land
Area
AL1 ‐ FCC Cap
(Est.) Cap/Mile Loops/Mile
Example A
3,227
2,344
29,487,336
12,581
1.38
PVT
2,948
2,331
31,466,523
13,498
1.26
Example B
2,947
170
16,553,652
97,289
17.32
Example C
2,675
1,189
19,188,671
16,139
2.25
Example D
2,973
139
15,490,782
111,125
21.33
Example E
2,865
1,194
24,516,302
20,537
2.40
Example E Cap/Mile for PVT =
47,875,323
There are several ways to improve the FCC’s regression model. PVT and other carriers
can provide the FCC with average loop lengths and other relevant data similar to what they do
for the United States Department of Agriculture Rural Development Utilities Program Form 479.
Because loop lengths are a major component of loop cost, it is a critical component that must be
included in order for the model to work accurately. Terrain data is also not included in the
model, other than percentage of water in the study area, which is not a significant driver of loop
costs.
These peer companies also have a much higher cap/mile related to AL 15 - Network
Support Expenses than PVT. Under the current regression model, PVT is allowed $142 per sq.
mile for Network Support Expense while Example E is capped at $205 per sq. mile. PVT is
much less dense in terms of population, has several exchanges that are very spread out and
therefore has a greater need for a higher cap per sq. mile of land area. This is another example
that the current FCC regression model does not take into account that companies with a far
greater land area to serve will have a need for much higher network cost than companies serving
a more dense area.
6
IV. The Regression Model is Overly Complex and Unpredictable Thus
Discouraging Future Investment.
The fact the model relies on flawed data, combined with the reality it does not generate
consistent results for similarly situated companies is troubling and needs to be resolved. Other
comments will follow on the application of regression caps, however, PVT feels one of the most
critical issues as it relates to us is the fact the model and approach taken is so complex and
unpredictable that we cannot adequately plan for the future. PVT recently secured a $28 million
loan from the Rural Utilities Service (RUS) and $9.5 million loan/grant from ARRA-BIP to
modernize the network and comply with the new FCC broadband requirements. However, as it
stands today, PVT has no straight forward means of understanding the results relied upon by the
Commission and predicting capped values in the future. Use of the tools that the FCC utilized in
developing its regression analysis, such as the Tele Atlas Telecommunications Suite5 and Stata6
software is costly and requires a high level of sophistication to develop and modify inputs, run
the models and analyze the results.
The Commission’s approach is not one that can be easily predicted or replicated, and as a
consequence, we cannot adequately plan for the future. Because cost recovery in this scenario is
not predictable, we are seriously considering delaying and potentially not drawing down the loan
funds from RUS and AARA-BIP funds to invest in future capital expenditures for fear we will
not be able to repay the debt. This outcome is contrary to the Commission’s intent; to deploy 4
Mbps downstream/1 Mbps upstream broadband services to all areas of the country. However, in
order to avoid this outcome, and assuming that regression caps will be utilized, underlying data
5 Tele Atlas Telecommunications Suite 2010.6 is the FCC’s source of study area boundaries used in the regression
analysis. See Appendix H, paragraph 5 and footnote 10.
6 Stata is the software used by the FCC to run the regression analysis. This is not referenced in the Report and Order
and FNPRM, however, we were informed of the need to purchase the Stata software to replicate the analysis
performed by the FCC, using the data set provided by the FCC.
7
and calculations must be readily available to carriers like PVT to allow for adequate financial
and strategic planning. We believe that a minimum of five years of data should be made
available to PVT, so that we can make appropriate financial decisions based on known cost
recovery mechanisms.
V. The FCC’s Regression Analysis Does Not Consider the Impacts of
Depreciation Reserve
The FCC’s model used to perform the regression analysis does not take the depreciation
reserve of the plant being limited into account; it is purely analyzed on a gross plant value.
Companies like PVT deployed the network years ago and, like many, face the need to upgrade
facilities as the plant is reaching the end of its useful life. In addition, PVT has and will soon
need to make more necessary network changes to meet the Commission’s 4 Mbps downstream/1
Mbps upstream broadband requirements, and this will require significant investment. The
regression model, as proposed does not allow for this, and its failure to recognize the impacts of
depreciation reserve is another significant flaw in the model.
VI. The Limitations Are Applied Incorrectly to the High Cost Loop Support
Algorithm
PVT believes there are three accounting issues that must be addressed in the calculation
and application of the proposed regression-based limitations. First, the High Cost Loop Support
(“HCLS”) data inputs (“data lines” or “DL”) should be limited, not the outputs (“algorithm
lines” or “AL”). Second, the limitations must take into account the impact of accumulated
depreciation and other Part 32 accounts on the calculation of support. Third, the methodology
used to calculate the limitations on depreciation expense must be modified.
8
PVT believes that the limitations should be applied to the HCLS data lines instead of the
algorithm lines, which would allow the 26 step algorithm to work as designed. The current
limitation of the algorithm lines does not account for the interrelationship between many of the
data lines used in the calculation of support. It should be noted that all of the algorithm lines are
calculations based on various data lines, so any proposed limitations can also be accomplished
by adjusting the data lines. As currently proposed, the FCC’s regression model limits outputs,
rather than limiting inputs and allowing the inputs to be run through the model. An excellent
example of this is AL 3, also referred to as the “A” Factor, which is calculated as Cable and Wire
Facilities (CWF) divided by Total CWF. The “A” Factor is used in the allocation of expenses
associated with CWF. AL 3 is one of several algorithm steps that uses both AL and DL inputs to
produce the result; in this case AL1, DL 255 (Account 2400 - Total CWF) and DL 815 (Account
2680 – Amortizable Tangible Assets – CWF). The FCC’s proposed treatment only limits the
AL1 amount, however, neither DL 255 (which includes AL1) nor DL 815 are adjusted. As a
result, the algorithm is not allowed to calculate support as it was intended and produces an
incorrect result.
This is extremely significant to PVT. Prior to regression caps, 2010 PVT data at a
national average cost per loop of $509.06, yielded estimated HCLS of $3,207,529. The proposed
regression, taking only AL1 – CWF assigned to Cat 1.0 into consideration, caps PVT by
$2,817,688, which reduced HCLS to $2,887,726, a reduction of $319,803 or 11.35%. A
significant driver in this reduction was the fact that AL3 was erroneously adjusted. AL3, prior to
the regression caps, was 79.86%, which dropped to 73.30% after the caps. AL3, referred to as
the “A” factor, represents the CWF loop percentage of the carrier. The higher this amount, the
more plant and expenses are allocated to a carrier’s study area cost per loop, increasing their
support. The FCC’s approach incorrectly impacts this factor and significantly impacts PVT.
9
Again, AL3 is just one example of this inappropriate result and PVT firmly believes that
applying caps to the DL lines will resolve this error.
While further comments will show that other HCLS inputs, such as accumulated
depreciation, need to be taken into consideration while applying the proposed FCC regression
caps, the following example shows the specific impact of removing the capped investment for all
of the appropriate data lines and allowing the HCLS algorithm to work as intended. Instead of
removing $2,817,688 from AL1- CWF assigned to Cat 1.0, AL1 should not be adjusted and the
following Data Lines should be reduced instead by this same $2,817,688:
DL 160 – Account 2001 – Total Plant in Service
DL 255 – Account 2410 – Total CWF
DL 700 – Cost Study Average CWF – Total Account 2410
DL 710 – Cost Study Average CWF Cat 1 – Total Subscriber Line Plant
Doing so results in adjusted HCLS of $2,926,582, $38,000 higher than the FCC’s AL 1
capped amount. AL3 is a significant driver of this reduction. This inappropriate limitation of
AL1 reduces AL3 to 73.30%. If the data lines were capped, AL3 would be a much more
appropriate 78.42%.
VII. The Limitations Are Missing Critical Components –
As mentioned above, accumulated depreciation and other Part 32 accounts must be taken
into consideration if the FCC is going to limit the 11 proposed algorithm lines, or follow the
approach to limiting the data lines described above. The FCC’s proposed regression analysis
does not limit the accumulated depreciation related to PVT CWF assets removed via AL1, nor
does it remove amounts from other associated accounts. If the FCC is going to limit AL1, the
following data lines should also be limited:
10
DL 160 – Account 2001 – Total Plant in Service
DL 190 – Account 3100 – Accumulated Depreciation
DL 255 – Account 2410 – Total CWF
DL 280 – Account 3124 – CWF Accumulated Depreciation
DL 700 – Cost Study Average CWF – Total Account 2410
DL 710 – Cost Study Average CWF Cat 1 – Total Subscriber Line Plant
By not limiting these data lines, the FCC’s regression analysis yields flawed and punitive
results on PVT. In addition, as discussed above, limiting the algorithm lines and not the data
lines does not allow the HCLS algorithm to work as designed. There could be some question as
to how to appropriately limit the accumulated depreciation reported on DL 190 and DL 280, but
this could be handled one of two ways. First, a ratio of the limited investment in the associated
plant account to the total plant account could be developed and applied to the accumulated
depreciation. Alternatively, the limited plant could be handled as a retirement, in which case
Part 32 for retirement accounting would treat the investment as fully depreciated. Whichever
method is selected would be more appropriate than the current approach of ignoring depreciation
reserve and other associated accounts in the algorithm. The limitation of algorithm lines rather
than data lines yields inappropriate results and ignores the net book value of the assets being
removed.
This situation is critical to PVT and we continue our example from above to show the
impacts on the company. As before, instead of removing PVT’s $2,817,688 from AL1- CWF
assigned to Cat 1.0, AL1 should not be adjusted and the following Data Lines reduced instead by
this same $2,817,688:
160 – Account 2001 – Total Plant in Service
255 – Account 2410 – Total CWF
11
700 – Cost Study Average CWF – Total Accounts 2410
710 – Cost Study Average CWF Cat 1 – Total Subscriber Line Plant
The next step is to reduce CWF accumulated depreciation reported on DL 280 and total
company accumulated depreciation reported on DL 190 to reflect the net book value of the assets
removed. For purposes of this example, this was done via a ratio. PVT reported $25,270,378 of
CWF accumulated depreciation on DL 280 and $42,927,115 of total CWF on DL 255,
consequently, PVT’s CWF plant is 58.87% depreciated. Multiplying this 58.87% ratio by the
$2,817,688 of CWF limited results in an accumulated depreciation figure of $1,658,719, which is
the figure used to reduce DL 190 and DL 280. Doing so results in adjusted HCLS of $3,035,148
for PVT, $147,421 higher than the FCC’s original HCLS limit on PVT.
VIII. The FCC’s Regression Analysis Does Not Appropriately Calculate
Limitations on Depreciation Expense
Depreciation expenses have not been properly accounted for in the FCC’s regression
model. Specifically, depreciation expenses should not be analyzed independently via regression,
as they are a byproduct of the associated plant investment. Instead, depreciation expenses should
be reflected as a function of the asset values removed. The FCC’s current, regression-based
approach results in limitations on depreciation expenses on AL 17 that are excessive and
inconsistent with Part 32 accounting principles. PVT’s depreciation rates are approved by the
New Mexico Public Regulation Commission and are therefore not subject to unilateral
adjustment by the company. Finally, we are audited annually by an independent CPA firm that
verifies the proper use of the approved depreciation rates, thus there is minimal risk of improper
application. Therefore, we recommend that regression not be used to limit depreciation expense.
Instead, we believe that depreciation expense limitations should be computed as the percentage
of limitation of the associated plant investment multiplied by depreciation expense. While this is
12
not currently impacting PVT, internal projections indicate that with the future investments that
are needed in order to reach the 4 Mbps downstream/1 Mbps upstream broadband requirements
mandated by the FCC, that the current regression methodology could cap AL 17 substantially
while having no correlation to the amount of associated plant investment that is capped.
IX. Conclusions
PVT believes the FCC’s order is fundamentally flawed in many ways. It takes PVT’s
ability to plan for the future away. PVT’s data in the regression analysis model is materially
flawed. The order does not take the customer density, terrain (rock and mountains) and loop
length into account. The model’s methodology is significantly flawed in general. Based on the
FCC’s order as written, PVT would be capped and therefore would be required to stop investing
in the network. By not investing in the network, the existing plant will deteriorate and PVT
would not be able to meet the mandate of providing 4/1 Mbps for broadband data to all of the
customers located in the 4,651 square miles of service territory.
January 18, 2012
Respectfully Submitted,
/s/ Glenn Lovelace
PEÑASCO VALLEY TELEPHONE COOPERATIVE, INC.
By:
Glenn Lovelace, CEO
4011 West Main
Artesia, NM 88210
575.748.1241
Glovelace@pvt.com