VIA EDGAR
November 29, 2007
Ms. Keira Ino
Staff Accountant
Division of Corporation Finance
United States Securities and Exchange Commission
Mail Stop 6010
Washington, D.C. 20549
| | |
| RE: | | Avanir Pharmaceuticals
Item 4.02 Form 8-K
Filed on November 21, 2007
File No. 001-15803
|
Dear Ms. Ino:
Thank you for your comments on November 26, 2007 in response to our Form 8-K filed on November 21, 2007. We submit to you the following information in reply. We note that we intend to file our amended quarterly reports on Form 10-Q for the periods ended March 31, 2007 and June 30, 2007 no later than December 5, 2007. Accordingly, to the extent that you have additional questions or concerns regarding these matters, we respectfully request that you contact us as soon as possible so that we may discuss and resolve these issues.
| | 1. | | Question: In order to assist us in better understanding your proposed restatement, please provide the following information: |
| | a. | | Definition of weighted average forfeiture rate; |
Answer:There is no specific definition of “weighted average forfeiture rate” in either FAS 123R or the glossary included in our purchased software system for computation of FAS 123R. We believe that “weighted average forfeiture rate”, as used by our software system, is based on applying the forfeiture rate to each vesting tranche of an option to determine the expected number of options that will vest over the life of the option. The expected number of options expected to vest for all tranches is then divided by the total options granted. The “cumulative expected forfeiture amount” is 100% minus the portion expected to vest; this amount is referred to as the “weighted average forfeiture rate” in the software system. A simplified example is as follows:
1
Example A :
Option grant : 750 shares
Vest schedule : Annually over a 3 year period
Forfeiture rate: 30%
| | | | | | | | | | | | |
| |
| | | | | Shares | | | | | | | |
| | | | | Vested per | | | Expected Shares | | | | |
| | | | | Vest | | | Forfeited Based on | | | Expected Shares to Vest | |
| | | | | Schedule | | | 30% Forfeiture Rate | | | Net of Forfeitures | |
| | Tranche 1 | | | 250 shares | | | 250 X 30% = 75 shares | | | 250 x .70 = 175 shares | |
| | Tranche 2 | | | 250 shares | | | 175 X 30% = 52 shares | | | 250 X .70 X .70 = 122 shares | |
| | Tranche 3 | | | 250 shares | | | 123 X 30% = 36 shares | | | 250 X .70 X .70 X .70 = 85 shares | |
| | | | | | | | | | | | |
| | TOTAL | | | 750 shares | | | | | | 382 shares | |
| |
Percent Expected to Vest = 51% due to the iterations from tranche to tranche (i.e. 382 shares/750 shares)
Weighted Average Forfeiture Rate = 49% [i.e. 1 — Percent Expected to Vest (51%)]
| | b. | | Question: The difference of how weighted average forfeiture rates are calculated between the new system and the old system; |
Answer:The old system applied a weighted average forfeiture rate to determine the expected shares to vest over the life of the whole option grant. In Example A, the weighted average forfeiture rate calculated by the system was 49% over the term of the option grant.
To expand on Example A and demonstrate how the weighted average forfeiture rate affects the calculation of share-based compensation expense, the table below shows how expense is calculated using the weighted average forfeiture rate under the old method used by the software system:
EXAMPLE A — WITH Wtd Avg Forfeiture Rate and using Old System
(assumed annual share-based compensation expense of $1,000 annually)
| | | | | | | | | | | | | | | | | |
| |
| | | | | | | | Shares | | | Total | | | | |
| | | | | | | | Vested | | | Expense | | | | |
| | | | | Vest | | | per Vest | | | (excluding | | | Expected Shares to Vest Net of | |
| | | | | Date | | | Schedule | | | forfeitures) | | | Forfeitures | |
| | Tranche 1 | | | 1/1/00 | | | 250 shares | | | | | $1,000 | | | 250 x .70 = 175 shares | |
| | Tranche 2 | | | 1/1/01 | | | 250 shares | | | | | $1,000 | | | 250 X .70 X .70 = 122 shares | |
| | Tranche 3 | | | 1/1/02 | | | 250 shares | | | | | $1,000 | | | 250 X .70 X .70 X .70 = 85 shares | |
| | | | | | | | | | | | | | | | | |
| | TOTAL | | | | | | 750 shares | | | | | $3,000 | | | 382 shares | |
| |
If calculating expense on 1/1/00, total expense to recognize = $1,000 X 51% = $510.00
2
EXAMPLE A — WITHOUT Wtd Avg Forfeiture Rate and using New System
(assumed annual share-based compensation expense of $1,000 annually)
| | | | | | | | | | | | | | | | | |
| |
| | | | | | | | Shares | | | Total | | | | |
| | | | | | | | Vested | | | Expense | | | | |
| | | | | Vest | | | per Vest | | | (excluding | | | Expected Shares to Vest Net of | |
| | | | | Date | | | Schedule | | | forfeitures) | | | Forfeitures | |
| | Tranche 1 | | | 1/1/00 | | | 250 shares | | | | | $1,000 | | | 250 x .70 = 175 shares | |
| | Tranche 2 | | | 1/1/01 | | | 250 shares | | | | | $1,000 | | | 250 X .70 X .70 = 122 shares | |
| | Tranche 3 | | | 1/1/02 | | | 250 shares | | | | | $1,000 | | | 250 X .70 X .70 X .70 = 85 shares | |
| | | | | | | | | | | | | | | | | |
| | TOTAL | | | | | | 750 shares | | | | | $3,000 | | | 382 shares | |
| |
If calculating expense on 1/1/00, total expense to recognize = $1,000 X 100% = $1,000.
There are two noticeable changes to the calculation of expense in the new system: 1) the expense for tranches that have fully vested are trued-up1 and 2) the expense for the remaining tranches are based upon the Percent Expected To Vest of the remaining unvested tranches, not the Percent Expected To Vest for all tranches. The old system did not true up and the expense for the remaining tranches was based upon the Percent Expected to vest of all tranches.
Under the new software system, which does not apply a weighted average forfeiture rate, the system takes 100% of the expense when each tranche has vested. Since Tranche 1 vested on 1/1/00, 100% of the expense is recognized. Under the old software system, even if a tranche had achieved its vesting date, a true-up adjustment is only applied at the end of the options full three year vesting period. This error in the software system was fixed in the newest version which was released in 2007. In Example A, $510 of expense is recognized under the old system as opposed to $1,000 expense recognized under the new system.
The second difference between the new system and the old system is that, under the new system, as tranches vest, the unadjusted expense from the remaining tranches is multiplied by the Percent Expense To Vest for each of the remaining tranches. Thus, as only longer vesting tranches remain, the remaining Percent Expected To Vest declines over the remaining vesting period of the option. In the old system, a single weighted average Percent Expected To Vest is applied to all vested and unvested tranche expense.
The largest difference between the new system and old system is that the new system applies a true-up adjustment to fully recognize expenses on vested tranches. The second difference has a less significant impact on the timing of expense recognition.
The differences in the expense calculated under both systems can be dramatic in the period when changes in forfeiture rates are significant.
| | c. | | Question: How weighted average forfeiture rate is applied in the way that share-based compensation expenses are recognized less during the earlier vesting periods than in the later periods; |
Answer:The weighted-average methodology lowers the rate of expense recognition in the first part of the vesting period, because of how it adjusts for forfeitures and how the software performs “true-ups” for vested options. The weighted-average methodology applies a weighted-average forfeiture based on all Tranches of the award, whether the requisite service had been rendered or not. The application of a weighting factor to all Tranches is not problematic on its own, but when an option vests, the software algorithms do not true-up cumulative expense to at least equal the value of options vested to date. The overall effect of the weighted-average methodology and the software’s true-up algorithm results in a lower rate of expense recognition.
| | d. | | Question: Clarify whether the manual adjustment of $2.6 million was to reflect 30% forfeiture rate that was not yet reflected by the system. If so, explain why an adjustment to 30% forfeiture rate resulted in the income of $1,669,000 during the three-month period ended March 31, 2007; |
| | |
| 1 | | True-Up adjustment refers to the recognition of additional expense for a tranche of shares that has fully vested that was originally expected to have a less than 100 percent likelihood of vesting. For example, if Tranche 1 had only a 70% chance of vesting, after 364 days, the expense recognized for that tranche would be $1,000 x 70% x (364/365). After one year, Tranche 1 fully vests all 250 shares, meaning that the expense than must ultimately be recognized for Tranche 1 is $1,000. In the new system, an adjusting expense of $1,000 x 30% x (365/365) is added to the normally recognized expense of $1,000 x 70% x (365/365) to insure that all of the expense is recognized. The old system only provided for true-up adjustments at the end of the option vest term, not the tranche vest term. |
3
Answer:The adjustment of $2.6 million was intended to reflect a change in the forfeiture rates that we had been applying historically since adoption of FAS 123R. In fiscal year 2006, we used a forfeiture rate of 8% for directors and officers and 13% for employees. In fiscal 2007, prior to our Q2 2007 re-examination of forfeiture rates, we used a 4% rate across the board.
Due to significant changes in our employee base resulting from the closure of certain research and development activities and the sale of a major product line, in the first half of fiscal 2007. We hired an outside consulting firm to perform a detailed analysis of our actual historical forfeiture rates since 2006. Their analysis resulted in an average historical forfeiture rate of approximately 30% for fiscal 2006. We decided to re-assess our share-based compensation expense from adoption to date with the new forfeiture rate of 30%. This analysis, using a much higher forfeiture rate of 30%, resulted in a credit adjustment of $2,600,000 to share-based compensation expense and net $1,669,000 in share-based compensation income for the quarter ended March 31, 2007.
| | e. | | Question: Describe the error you made in applying FAS123R. |
Answer:As described above, in the second quarter of fiscal 2007 we performed a re-assessment of our forfeiture rates and increased the rate to 30% from adoption of FAS 123R to date. In fiscal 2006, we applied a rate of 8% for directors and officers and 13% for employees. In fiscal 2007, prior to our re-assessment, we applied a rate of 4% in our expense calculation. In order to re-calculate expense using the 30% forfeiture rate, we contacted our software provider for guidance. We implemented the following steps:
Step 1: When re-calculating expense in the software system, the system requests inputs for “Prior Forfeiture Rate” and “Forfeiture Rate”. We verified with our software’s technical support department that the correct inputs for fiscal 2006 should be “Prior Forfeiture Rate = 8% or 13%” (depending on whether the award was to a director/officer or an employee) and “Forfeiture Rate = 30%”. Using these inputs in the system, we re-ran the calculation of expense.
Step 2: To re-calculate fiscal 2007, we verified with our software provider that the correct inputs for fiscal 2007 should be “Prior Forfeiture Rate = 4%” and “Forfeiture Rate = 30%”. Using these inputs in the system, we re-ran the calculation of expense.
Step 3: We recorded an adjustment of $2,600,000 to reflect the expense calculated in steps 1 and 2 above for fiscal 2006 and 2007.
In October 2007, when we performed our detailed review of FAS 123R (in connection with the installation of the new version of our purchased software system) from adoption to date using the new method discussed in 1.b. above, we discovered that a portion of the difference was related to the recording of our forfeiture rate adjustment in Q2 2007. In Step #2, we should have entered a “Prior Forfeiture Rate = 30%” and “Forfeiture Rate = 30%”. The reason why the Prior Forfeiture Rate is 30% when running the fiscal 2007 calculation is because after performing Step #1, the forfeiture rate adjustment to 30% had already been accounted for pursuant to the algorithms in the software system. Thus, when performing Step #2, since the adjustment to 30% had already been completed for fiscal 2006, the inputs should have instead been 30% for both Prior Forfeiture Rate and Forfeiture Rate.
4
| | 2. | | Question: In light of your proposed restatements, please tell us whether you have re-evaluated the effectiveness of your disclosure controls and procedures for the quarterly periods ended March 31, 2007 and June 30, 2007. Provide us a disclosure you propose to include in your restatedForm 10-Q for the periods then ended. |
Answer:We have re-evaluated the effectiveness of our disclosure controls and procedures and have concluded that our disclosure controls and procedures were not effective as of March 31, 2007 and June 30, 2007 due to a material weakness in internal controls over financial reporting. We have prepared the following disclosure to be included in our restated Form 10-Qs. We have included the disclosure for the second quarter 2007 Form 10-Q/A below. The disclosure for the third quarter 2007 Form 10-Q/A is consistent with the second quarter 2007 Form 10-Q/A.
Item 4. CONTROLS AND PROCEDURES
Evaluation of Disclosure Controls and Procedures
In connection with the preparation of our Quarterly Report on Form 10-Q, for the period ended March 31, 2007, originally filed on May 10, 2007, an evaluation was performed under the supervision of our management, including our Chief Executive Officer (“CEO”) and Interim Chief Financial Officer (“CFO”), of the effectiveness of the design and operation of our disclosure controls and procedures (as defined in Rule 13a-15(e) under the Securities Exchange Act of 1934, as amended (“Exchange Act”)). Based on that evaluation, our CEO and CFO concluded at that time that our disclosure controls and procedures were effective as of the end of the period covered by that Quarterly Report. Subsequently, the Company determined that it was necessary to restate the Company’s financial statements for the fiscal quarter ended March 31, 2007 and that the financial statements for those periods should no longer be relied upon.
These restatements had no impact on the Company’s previously reported revenues, cash flow from operations or total cash and cash equivalents shown in the Company’s financial statements for the fiscal quarter ended March 31, 2007.
In connection with the restatement, as of the end of the period covered by this report, we carried out an evaluation, under the supervision and with the participation of the Company’s management, including the Company’s CEO and Interim CFO of the effectiveness of the design and operation of our disclosure controls and procedures (as defined in Rules 13a-15 (e) under the Exchange Act). Based upon that evaluation, our CEO and Interim CFO concluded that our disclosure controls and procedures were not effective as of March 31, 2007 because of a material weakness in internal controls over financial reporting.
The Company did not have effective internal controls over financial reporting for the calculation of share-based compensation expense. Specifically, the Company’s calculation of share-based compensation expense was being calculated improperly due to the application of a weighted average forfeiture rate as opposed to the correct method which is to apply an actual forfeiture rate. The incorrect application of weighted average forfeiture rate in the calculation of share-based compensation expense resulted in the understatement of share-based compensation expense. The
5
Company’s corporate monitoring controls failed to operate at a sufficient level of precision to detect the understatement of share-based compensation expense and the material misstatement of operating expenses and net loss.
Changes in Internal Controls over Financial Reporting
There has been no change in the Company’s internal control over financial reporting (as defined in Rule 13a-15(f) under the Exchange Act) during the Company’s fiscal quarter ended March 31, 2007, that has materially affected, or is reasonably likely to materially affect the Company’s internal control over financial reporting. However, in October 2007 the Company determined that its purchased software system had released a new version which resolved the issue regarding the application of weighted average forfeiture rate as described above under “Evaluation of Disclosure Controls and Procedures”.
The Company is in the process of evaluating remediation efforts to address this issue affecting the calculation of share-based compensation expense. During the preparation of our Form 10-K the Company performed detailed review procedures to ensure that fourth quarter 2007 and fiscal 2007 share-based compensation expense was calculated correctly. The Company will continue remediation efforts in the first quarter of 2008 and anticipates full remediation by 9/30/08. Additional measures may be forthcoming as the Company evaluates the effectiveness of these efforts. We cannot assure you that these efforts will be successful or that our internal control over financial reporting will be effective in accomplishing all control objectives all of the time.
Sincerely,
/s/ Christine G. Ocampo
Christine G. Ocampo
Corporate Controller
cc: Stephen G. Austin, Chairman of Avanir Pharmaceuticals Audit Committee
6
