SEC Company Response from ENERPLUS RESOURCES FUND (CIK 0001126874) — Nov 22, 2005
ENERPLUS RESOURCES FUND (CIK 0001126874)
Date: Nov. 22, 2005 · CIK: 0001126874 · Accession: 0001279569-05-001049
AI Filing Summary & Sentiment
Letter
Ms. April Sifford United States Securities and Exchange Commission F Street N.E. Washington D.C. 20549
Dear Ms. Sifford:
Re: Accounting for stock-based compensation
This letter is in response to your request in late October 2005 regarding “fair value” option pricing calculations for our trust unit rights plan.
Black-Scholes Option Pricing Model
In consultation with an independent options pricing expert, we have used a modified Black-Scholes option pricing model to measure the “fair value” of rights granted under our plan. However, we must emphasize that this model has some apparent shortcomings with respect to the function for which we are using it. Firstly, the Black-Scholes model was not designed for high dividend paying entities (and Enerplus distributes 70-90% of its cash flow every month). We also believe that this model is not designed for entities with a naturally declining asset base where a portion of the distributions may be a return of capital (such as an oil and gas trust). We understand that it is also not intended to measure non-transferable options (and all of our employee options are non-transferable). Finally, it is designed for European style options that vest on one day (and our employee options vest over an extended time period). Despite these shortcomings we have yet to uncover a better alternative. Our frustrations were echoed by Mr. Christopher Cox, Chairman of the SEC, when he recently issued a statement encouraging the private sector to design new market instruments that can more accurately measure the cost of employee stock options.
The preliminary results of our modeling and corresponding comparison to actual expenses recorded under the intrinsic method are attached as Appendix A.
A discussion of the variables utilized in the model is provided below.
Black-Scholes Model Variables
1.
Number of rights
We calculated the fair value of all rights issued on or after January 1, 2003 through to September 30, 2005 on a quarterly basis. The quarterly calculations were based on the actual number of rights issued in that quarter.
2.
Average expected life of rights
Our rights vest over three years commencing on the first anniversary date of the grant and have a seven year contractual life. Based on actual historical exercise patterns, we determined that the expected average life of our rights is approximately 2.5 years.
3.
Stock price and exercise price
In each quarter, the exercise price was the same as the market price of the trust units on the grant date.
4.
Volatility
We employed a historical volatility measure using the preceding 2.5 year annualized volatility, which ranged between 22% and 23%. This three year term is generally commensurate with the expected life of our rights.
5.
Risk free interest rate
The Government of Canada 3 year bond rate was utilized.
6.
Vesting period
Rights vest over three years commencing on the first anniversary date of the grant. Therefore, the fair value was amortized to unit based compensation expense evenly over a three year period, beginning with the grant date.
7.
Dividends/Distributions
The model does not allow us to reduce the exercise price for the strike price reduction as this is a quarterly determination. Consequently, the strike price reduction was applied to the distribution amount.
We understand that this was the appropriate treatment as it is consistent with the concept of “dividend protection”. Under the concept of “dividend protection”, option exercise prices are reduced for dividends paid on the underlying stock, to leave the option holder in an equivalent economic position as an actual stockholder. For modeling purposes, dividend protection reduces the dividend included in the model.
We are uncomfortable forecasting future distributions and forecasting future strike price reductions and caution you of the measurement uncertainty in this regard.
Nevertheless, in an effort to utilize the prescribed model we used the actual distribution rate at the end of each quarter as the proxy for future distributions. Furthermore, we used the actual strike price reduction at the end of each quarter as the proxy for future strike price reductions. As a consequence, we assumed that the distribution and quarterly strike price reduction would be held constant over the expected life of the right.
8. Declining Strike Price vs. Fixed Strike Price
Our model assumes that all rights holders will elect to exercise using the reduced strike price. Although recent history indicates that most employees have chosen the fixed strike price option (to take advantage of the S. 110(1)(d) personal tax deduction) we had difficulty assuming a fixed strike price in the model. We believe the historical pattern was more a reflection on the recent step change in oil and gas prices. Without the benefit of hindsight, our assumptions in and 2004 would have been that employees will elect to exercise using the reduced strike price.
Going forward, without the benefit of ever increasing commodity prices, it is more reasonable that employees would make use of the declining strike price feature, especially for an organization with a naturally declining asset base that is paying out 70- 90% of its cash flow. The declining strike price feature was designed and implemented with the expectation that employees would make use of this feature. As a result, we believed it was prudent to model the rights plan with this feature intact.
Reasonableness Test - Binomial Model
We were advised by our independent expert that a simple binomial or lattice model is not appropriate for valuing our rights due to our high dividend yield. Specialized binomial modeling software is required in such a case. Our expert was not able to share their proprietary binomial model with us; however they did calculate values for Enerplus rights using their advanced modeling software that were not materially different from those calculated using our modified Black-Scholes model.
Accounting Treatment
As shown in Appendix A, the difference in our unit based compensation expense between the intrinsic and the fair value method was $943,000 (or 0.27% of net income) for the year ended December 31, 2003 and $3.1 million (or 1.26% of net income) for the year ended December 31, 2004. We believe that these amounts are immaterial and we propose to not make changes to our financial reporting for 2003 or 2004. Refer to Appendix B for a discussion on materiality with reference to SAB Topic 1:M.
For fiscal 2005, we believe that a change from the intrinsic to the fair value method can be accomplished in two ways:
1.
Early Adoption of FAS 123R
We could adopt, through modified prospective application, FAS 123R as at January 1, 2005. Our December 31, 2005 financial statements will reflect the fair value expense associated with unvested rights granted prior to January 1, 2005 and all rights granted on and after January 1, 2005. As shown in Appendix A, this amount was $1.97 million as at September 30, 2005. This accounting treatment would result in no retroactive restatement.
2.
Change in Accounting Estimate
Alternatively, we can apply the change to the fair value method as a change in accounting estimate resulting from additional information being available to apply the fair value methodology as set out in FAS 123 paragraph 22. This information includes additional experience with the rights plan and trends with respect to distributions and strike price reductions along with additional guidance from the SEC and FASB. In this case, prospective application would apply.
Recommended Disclosure
Although we believe a case may be made for a change in accounting estimate, our preference is to proceed with the adoption of the modified prospective application of FAS 123R. For purposes of our December 31, 2005 financial statements (included in Form 40-F) we propose the following:
•
In accordance with the modified prospective application of FAS 123R, there is no adjustment to prior year financial statements.
•
Our December 31, 2005 financial statements will indicate that we adopted the fair value method as at January 1, 2005 using modified prospective application.
Examples of the MD&A and Canadian note disclosure are attached in Appendix C. We are proposing a similar accounting treatment for Canadian GAAP and consequently we do not expect any disclosure or adjustment to reconcile Canadian to U.S. GAAP.
We will also be enhancing our critical accounting estimates note disclosure with respect to measurement uncertainty. Furthermore, the readers will be warned that the values calculated by option pricing models and expensed on the Income Statement will likely not reflect the actual value realized by employees.
You will note that in our MD&A we have referenced our correspondence with both the Alberta and U.S. Securities Commissions. We want to ensure the readers understand the change in accounting method, which increases earnings, is a legitimate accounting policy change and not an attempt by the Fund to inflate earnings.
We continue to believe that we have applied GAAP correctly in good conscience and faith. At the same time we acknowledge your desire to improve comparability of financial statements among reporting issuers. In the spirit of cooperation, we believe our proposal achieves your intended outcome by transferring to the fair value methodology in a manner that is consistent with the materiality of this issue to our unitholders.
This submission has been discussed with both the Audit Committee of Enerplus Resources and our auditors, Deloitte & Touche LLP.
We trust this response addresses your concerns. Should you have any additional concerns or comments please contact me directly at 403-298-1295.
Sincerely,
Robert
J.
Waters, C.A.
Show Raw Text
CORRESP
1
filename1.htm
Corresp
Enerplus
Resources Fund
The
Dome
Tower
3000,
333-7th Avenue SW
Calgary,
Alberta
T2P 2Z1
Tel
403.298.1295
Fax
403.298.2299
www.enerplus.com
November
18, 2005
BY
FAX:
202-772-9220
Ms.
April
Sifford
United
States Securities and Exchange Commission
100
F
Street N.E.
Washington
D.C. 20549
Dear
Ms.
Sifford:
Re:
Accounting for stock-based compensation
This
letter is in response to your request in late October 2005 regarding “fair
value” option pricing calculations for our trust unit rights plan.
Black-Scholes
Option Pricing Model
In
consultation with an independent options pricing expert, we have used a modified
Black-Scholes option pricing model to measure the “fair value” of rights granted
under our plan. However, we must emphasize that this model has some apparent
shortcomings with respect to the function for which we are using it. Firstly,
the Black-Scholes model was not designed for high dividend paying entities
(and
Enerplus distributes 70-90% of its cash flow every month). We also believe
that
this model is not designed for entities with a naturally declining asset base
where a portion of the distributions may be a return of capital (such as an
oil
and gas trust). We understand that it is also not intended to measure
non-transferable options (and all of our employee options are non-transferable).
Finally, it is designed for European style options that vest on one day (and
our
employee options vest over an extended time period). Despite these shortcomings
we have yet to uncover a better alternative. Our frustrations were echoed by
Mr.
Christopher Cox, Chairman of the SEC, when he recently issued a statement
encouraging the private sector to design new market instruments that can more
accurately measure the cost of employee stock options.
The
preliminary results of our modeling and corresponding comparison to actual
expenses recorded under the intrinsic method are attached as Appendix A.
A
discussion of the variables utilized in the model is provided below.
Black-Scholes
Model Variables
1.
Number
of rights
We
calculated the fair value of all rights issued on or after January 1, 2003
through to September 30, 2005 on a quarterly basis. The quarterly calculations
were based on the actual number of rights issued in that quarter.
2.
Average
expected life of rights
Our
rights vest over three years commencing on the first anniversary date of the
grant and have a seven year contractual life. Based on actual historical
exercise patterns, we determined that the expected average life of our rights
is
approximately 2.5 years.
3.
Stock
price and exercise price
In
each
quarter, the exercise price was the same as the market price of the trust units
on the grant date.
4.
Volatility
We
employed a historical volatility measure using the preceding 2.5 year annualized
volatility, which ranged between 22% and 23%. This three year term is generally
commensurate with the expected life of our rights.
5.
Risk
free interest rate
The
Government of Canada 3 year bond rate was utilized.
6.
Vesting
period
Rights
vest over three years commencing on the first anniversary date of the grant.
Therefore, the fair value was amortized to unit based compensation expense
evenly over a three year period,
beginning with the grant date.
7.
Dividends/Distributions
The
model
does not allow us to reduce the exercise price for the strike price reduction
as
this is a quarterly determination. Consequently, the strike price reduction
was
applied to the distribution amount.
We
understand that this was the appropriate treatment as it is consistent with
the
concept of “dividend protection”. Under the concept of “dividend protection”,
option exercise prices are reduced for dividends paid on the underlying stock,
to leave the option holder in an equivalent economic position as an actual
stockholder. For modeling purposes, dividend protection reduces the dividend
included in the model.
We
are
uncomfortable forecasting future distributions and forecasting future strike
price reductions and caution you of the measurement uncertainty in this regard.
Nevertheless,
in an effort to utilize the prescribed model we used the actual distribution
rate at the end of each quarter as the proxy for future distributions.
Furthermore, we used the actual strike price reduction at the end of each
quarter as the proxy for future strike price reductions. As a consequence,
we
assumed that the distribution and quarterly strike price reduction would be
held
constant over the expected life of the right.
8.
Declining
Strike Price vs. Fixed Strike Price
Our
model
assumes that all rights holders will elect to exercise using the reduced strike
price. Although recent history indicates that most employees have chosen the
fixed strike price option (to take advantage of the S. 110(1)(d) personal tax
deduction) we had difficulty assuming a fixed strike price in the model. We
believe the historical pattern was more a reflection on the recent step change
in oil and gas prices. Without the benefit of hindsight, our assumptions in
2003
and 2004 would have been that employees will elect to exercise using the reduced
strike price.
Going
forward, without the benefit of ever increasing commodity prices, it is more
reasonable that employees would make use of the declining strike price feature,
especially for an organization with a naturally declining asset base that is
paying out 70- 90% of its cash flow. The declining strike price feature was
designed and implemented with the expectation that employees would make use
of
this feature. As a result, we believed it was prudent to model the rights plan
with this feature intact.
Reasonableness
Test - Binomial Model
We
were
advised by our independent expert that a simple binomial or lattice model is
not
appropriate for valuing our rights due to our high dividend yield. Specialized
binomial modeling software is required in such a case. Our expert was not able
to share their proprietary binomial model with us; however they did calculate
values for Enerplus rights using their advanced modeling software that were
not
materially different from those calculated using our modified Black-Scholes
model.
Accounting
Treatment
As
shown
in Appendix A, the difference in our unit based compensation expense between
the
intrinsic and the fair value method was $943,000 (or 0.27% of net income) for
the year ended December 31, 2003 and $3.1 million (or 1.26% of net income)
for
the year ended December 31, 2004. We believe that these amounts are immaterial
and we propose to not make changes to our financial reporting for 2003 or 2004.
Refer to Appendix B for a discussion on materiality with reference to SAB Topic
1:M.
For
fiscal 2005, we believe that a change from the intrinsic to the fair value
method can be accomplished in two ways:
1.
Early
Adoption of FAS 123R
We
could
adopt, through modified prospective application, FAS 123R as at January 1,
2005.
Our December 31, 2005 financial statements will reflect the fair value expense
associated with unvested rights granted prior to January 1, 2005 and all rights
granted on and after January 1, 2005. As shown in Appendix A, this amount was
$1.97 million as at September 30, 2005. This accounting treatment would result
in no retroactive restatement.
2.
Change
in Accounting Estimate
Alternatively,
we can apply the change to the fair value method as a change in accounting
estimate resulting from additional information being available to apply the
fair
value methodology as set out in FAS 123 paragraph 22. This information includes
additional experience with the rights plan and trends with respect to
distributions and strike price reductions along with additional guidance from
the SEC and FASB. In this case, prospective application would
apply.
Recommended
Disclosure
Although
we believe a case may be made for a change in accounting estimate, our
preference is to proceed with the adoption of the modified prospective
application of FAS 123R. For purposes of our December 31, 2005 financial
statements (included in Form 40-F) we propose the following:
•
In
accordance with the modified prospective application of FAS 123R,
there is
no adjustment to prior year financial
statements.
•
Our
December 31, 2005 financial statements will indicate that we adopted
the
fair value method as at January 1, 2005 using modified prospective
application.
Examples
of the MD&A and Canadian note disclosure are attached in Appendix C. We are
proposing a similar accounting treatment for Canadian GAAP and consequently
we
do not expect any disclosure or adjustment to reconcile Canadian to U.S.
GAAP.
We
will
also be enhancing our critical accounting estimates note disclosure with respect
to measurement uncertainty. Furthermore, the readers will be warned that the
values calculated by option pricing models and expensed on the Income Statement
will likely not reflect the actual value realized by employees.
You
will
note that in our MD&A we have referenced our correspondence with both the
Alberta and U.S. Securities Commissions. We want to ensure the readers
understand the change in accounting method, which increases earnings, is a
legitimate accounting policy change and not an attempt by the Fund to inflate
earnings.
We
continue to believe that we have applied GAAP correctly in good conscience
and
faith. At the same time we acknowledge your desire to improve comparability
of
financial statements among reporting issuers. In the spirit of cooperation,
we
believe our proposal achieves your intended outcome by transferring to the
fair
value methodology in a manner that is consistent with the materiality of this
issue to our unitholders.
This
submission has been discussed with both the Audit Committee of Enerplus
Resources and our auditors, Deloitte & Touche LLP.
We
trust
this response addresses your concerns. Should you have any additional concerns
or comments please contact me directly at 403-298-1295.
Sincerely,
Robert
J.
Waters, C.A.
cc: Gordon
J.
Kerr, C.A.
President and CE
Robert
Normand, C.A. Chairman Enerplus Audit Committee
Ward
Zimmer, C.A. Deloitte
& Touche LLP
Appendix
A - Analysis of Fair Value vs. Intrinsic method
(CDN$
thousands)
Reporting
period
Intrinsic
Expense
Fair
Value expense
Difference*
Reported
net income
Difference
as a %
of net income
Reported
total
assets
Difference
as a %
of total assets
Q1
2003
4
20
-16
94,836
-0.02
%
2,644,938
0.00
%
Q2
2003
101
50
51
53,423
0.10
%
2,577,968
0.00
%
Q3
2003
362
78
284
59,205
0.48
%
2,556,280
0.01
%
Q4
2003
897
273
624
50,852
1.23
%
2,661,765
0.02
%
Total
2003
1,364
421
943
258,316
0.37
%
2,661,765
0.04
%
Q1
2004
359
291
68
45,166
0.15
%
2,716,732
0.00
%
Q2
2004
523
313
210
48,034
0.44
%
3,181,214
0.01
%
Q3
2004
1,458
353
1,105
50,604
2.18
%
3,148,116
0.04
%
Q4
2004
2,328
575
1,753
104,242
1.68
%
3,180,748
0.06
%
Total
2004
4,668
1,532
3,136
248,046
1.26
%
3,180,748
0.10
%
Q1