FOIA Confidential Treatment Request
Pursuant to 17 C.F.R. 200.83
[LOGO]
March 24, 2006
Mr. Jim B. Rosenberg
Senior Assistant Chief Accountant
Division of Corporation Finance
United States Securities and Exchange Commission
Washington, D.C. 20549
| Re: | Biovail Corporation Form 20-F for Fiscal Year Ended December 31, 2004 Form 6-K filed August 12, 2005 File No. 000-22358 |
Dear Mr. Rosenberg:
Biovail Corporation (the "Company", "Biovail", "we", "us" or "our") acknowledges the receipt of the letter from the staff of the United States Securities and Exchange Commission dated February 6, 2006. Please find below our detailed responses to the staff's comments on our Form 20-F for the fiscal year ended December 31, 2004. For ease of reference, we have provided the text of each comment prior to our response.
Comment:
Response:
We will provide the disclosures requested in the staff's prior comments four, six, ten and twelve in our amended Form 20-F for the year ended December 31, 2004. However, we reiterate to the staff that we do not believe we can provide the additional disclosures requested in prior comments two and thirteen, based on the information that was available to us as of March 30, 2005 (the date of the discussion and analysis in our Management's Discussion and Analysis ["MD&A"]). Please see our response below to the staff's comment 7 regarding the date of our MD&A.
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-1
Comment:
Item 5. Operating and Financial Review and Prospects, page 56
MD&A in Accordance with U.S. GAAP, page 57
Critical Accounting Policies and Estimates, page 58
Revenue recognition, page 59
Response:
We believe that our methodology provides us with a reasonable basis to estimate returns. We also believe that in making our estimates we have applied reasonable judgment and assumptions with respect to the factors that impact our methodology.
Pre-Acquisition Returns, Rebates and Chargebacks
It appears that our response to the staff's prior comment one in our letter dated November 25, 2005 may have inadvertently created a misimpression that we could not estimate returns, rebates and chargebacks related to Teveten, Vasotec®, Vaseretic® and Cardizem® CD. We want to assure you that this is not the case. Specifically, we were not responsible for any pre-acquisition returns, rebates or chargebacks in accordance with the respective purchase agreements related to our acquisitions of Teveten and Cardizem® CD. We were responsible for returns of Vasotec® and Vaseretic® that were processed after May 2004 (a date two years post-acquisition). At the time we acquired Vasotec® and Vaseretic® in May 2002, we considered the likelihood that we would be required to accept a material amount of returns related to pre-acquisition sales to be remote, as these products have 24-month dating. Consequently, we did not consider it necessary, in accordance with paragraph 4 of FASB Statement No. 141,Business Combinations (FAS 141), to record a liability for pre-acquisition returns at the date of acquisition. The total amount of returns that we assumed related to sales made prior to our acquisition was immaterial (approximately $14,000) and, as a result, we did not consider it necessary to subsequently adjust our original purchase price allocation.
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-2
Based on the due diligence we conducted in advance of these acquisitions, which included a review of sales data including gross-to-net adjustments for at least the three-year period prior to the dates of our acquisitions, we believe that Solvay Pharmaceuticals Marketing & Licensing AG ["Solvey"], Merck & Co., Inc. ["Merck"], and Aventis Pharmaceuticals Inc. ["Aventis"] from whom we had acquired Teveten, Vasotec® and Vaseretic®, and Cardizem® CD, respectively, had a reasonable basis to estimate their provisions for returns, rebates and chargebacks. Our business approach to commercializing these products, in terms of sales, marketing and distribution activities, was not dissimilar to those employed by the predecessor companies and, as such, we believe that utilizing their experience rates was a reasonable basis for initially estimating our provisions for returns, rebates and chargebacks related to these products.
Returns Background
Our returns exposure is primarily related to our promoted and non-promoted (genericized) products that we sell directly to wholesale and warehousing chain customers in the United States. These direct product sales account for a proportion of our total products sales revenue, as we out-license a significant portion of our primary-care products to strategic partners such as GlaxoSmithKline, plc ["GSK"] (Wellbutrin XL®), Forest Laboratories Inc. (Tiazac®) and Teva Pharmaceutical Industries Ltd. (generics) for marketing and/or distribution in the United States, for which returns are only permitted in very limited circumstances.
The following table displays the key promoted and non-promoted products that we marketed directly in the United States during the period 2002 to 2004, their respective launch dates and, in the case of non-promoted products, the first generic launch date, as well as the date that we acquired these products.
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-3
| Brand | Launch Date | Generic Launch Date | Acquired By Biovail | |||
|---|---|---|---|---|---|---|
| Promoted Products | ||||||
| Cardizem® LA | April 2003(1) | NA | NA(2),(3) | |||
| Teveten | October 1999 | NA | March 2002(4) | |||
| Teveten HCT | March 2003(1) | NA | March 2002(4) | |||
| Zovirax Ointment | September 1982 | NA | January 2002 | |||
| Zovirax Cream | July 2003(1) | NA | January 2002 | |||
Non-Promoted Products | ||||||
| Ativan® | February 1975 | August 1985 | May 2003 | |||
| Cardizem® CD | January 1992 | June 1999 | January 2001 | |||
| Isordil® | January 1959 | October 1975 | May 2003 | |||
| Vaseretic® | February 1987 | September 2001 | May 2002 | |||
| Vasotec® | January 1986 | August 2000 | May 2002 | |||
Methodology for Estimating Returns
Our direct product sales meet all the criteria of paragraph 6 of FASB Statement No. 48,Revenue Recognition When Right of Return Exists (FAS 48), at the time the product is delivered to our customers. Pursuant to paragraph 7 of FAS 48, we recognize revenue from product sales net of provisions for estimated returns. In developing our methodology for estimating returns, we utilize the guidance provided in FAS 48 and Staff Accounting Bulletin No. 104,Revenue Recognition (SAB 104), which superseded Staff Accounting Bulletin No. 101,Revenue Recognition in Financial Statements.
We have refined our methodology for estimating returns over time, based on new and better information becoming available to us. Some of the principal factors we consider in estimating returns are:
October 2000 - 2002
Biovail directly entered the United States market in October 2000 upon the acquisition of DJ Pharma, Inc. ["DJ Pharma"]. From that date through 2002, we applied the methodology for estimating returns that had been developed and utilized by DJ Pharma. Under this methodology, we determined, based on actual historical returns data, a weighted average experience rate as a percentage of gross sales by individual product. In the case of acquired products, we initially utilized the experience rates of the companies from whom we had acquired the products.
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-4
We also conducted an assessment of known events or factors that could impact the historical experience rate on a go forward basis, such as the impact of generic competition and the level of inventory in the distribution channel. We utilized prescription and trade inventory data obtained from IMS Health ["IMS"], an independent pharmaceutical market research firm. IMS data was the recognized industry standard and considered to be the best information available for tracking prescription and trade inventory levels, and we believed that this data was accurate and reliable. We continually monitored processed returns relative to our provision and revised our estimates based on actual experience.
2003
At December 31, 2003, we obtained inventory data directly from two of our major wholesalers, Cardinal Health, Inc. ["Cardinal"] and McKesson Corporation ["McKesson"]. We used this information to verify the accuracy of the wholesaler inventory data supplied by IMS. Our assessment of trade inventory levels up until this time was based on the average months of supply on hand at the wholesalersand warehousing chains, but our experience revealed that most of the returns were originating from the wholesalers. Therefore, we refined our estimation process to consider only months of supply on hand in the wholesale distribution channel.
In 2003, we recorded an adjustment of $28.1 million to our provision for returns. This adjustment was the result of the following developments related to business decisions made in 2003, or based on new information available in 2003:
With regard to generic competition, we believed until 2002 that the analogs we were using that indicated a brand product could be expected to retain approximately * * of its pre-genericization market share were sufficient. However, during 2003 there was an increasing trend by private and public benefit programs, such as Medicaid, to require the substitution of lower-priced generic products in place of brand name prescriptions. As a result, we believed analogs indicating that brand products would generally retain only * * of its pre-genericization volumes were warranted and, accordingly, we increased our returns estimate for Cardizem® CD.
With regard to the introduction of Cardizem® LA, we initially recognized that its eventual launch and promotion would likely have an effect on Cardizem® CD returns because fewer patients would be initially prescribed Cardizem® CD and because existing patients would be converted to Cardizem® LA. We also had to consider a number of other factors such as the timing of Food and Drug Administration ["FDA"] approval, our ability to successfully scale-up and manufacture the product, and the success of our physician detailing, sampling and promotion activities. During 2003, the conversion of patients from Cardizem® CD to Cardizem® LA was higher than we had originally anticipated in 2002 and 2001, which we factored into our 2003 adjustment.
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-5
Lastly, in May 2002, we received regulatory approval to extend the dating on Cardizem® CD from 24 months to 36 months. Initially, we did not believe this would impact our returns experience. However, we became aware of a higher level of 24-month dated product returns that had been sold during 2001 and 2002, as wholesalers ordered and shipped new product with the extended shelf life, and unexpectedly returned the product with the shorter shelf life. It therefore became apparent to us that the wholesalers were not managing their inventories of Cardizem® CD at all times on a first-in, first-out basis.
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-6
2004 — Present
In 2004, IMS stopped providing inventory data. As a result, during 2004 we obtained this data directly from our three major wholesalers (Cardinal, McKesson and AmerisourceBergen Corporation ["ABC"]).
We continue to enhance our methodology for estimating returns by factoring into our calculations the potential returns liability that remains to be processed by the year of sale. To do this, we take a sample of actual returns by product and examine lot numbers to determine the year in which the product was originally sold. We believe this calculation provides a reasonable indication of the percentage of product that we expect will be returned in each year following the date of original sale. Combining this calculation with our weighted average experience rate, we can estimate the potential returns liability that remains to be processed at each year-end. For example, we may determine that the historical weighted average experience rate is 5% of gross product sales. Based on our sample of returns by lot, we may determine (for a product with a 24-month shelf life) that 10%, 60% and 30% of returns occur in years two, three and four, respectively. If gross sales for a particular product were $100 million in year one, we would expect a $5 million liability for product returns at the end of that year. At the end of years two, three and four, we would expect remaining returns liability would be $4.5 million (90%) and $1.5 million (30%) and nil, respectively. We adjust the remaining returns liability related to prior years for any subsequent changes in the weighted average experience rate.
We conduct a qualitative analysis by comparing the returns liability with the inventory levels of our products in the distribution channel. Using internal and external information, we assess whether these inventory levels are reflective of current and projected pull-through sales and known trends in demand. If we determine that those inventory levels are higher than anticipated demand, we ensure that our returns liability is adequate to cover the expected amount of product that may be returned due to expiration.
We also consider any new facts and circumstances that may indicate that historical returns experience may not be indicative of current and/or future results, and assess whether an adjustment to the returns liability is warranted. These facts and circumstances require us to make subjective judgments based primarily on our evaluation of current market conditions related to our products, and may result in an increase or decrease in the experience rate that is applied to current and future sales, or an adjustment related to past sales, or both. For example, a business decision to apply for a change to the NDCs of a product line could result in higher returns of current and future sales of product still labeled with the old NDC, as well as higher than originally anticipated returns related to past sales of product in the distribution channel. Consequently, we may need to increase our experience rate related to current and future sales, as well as record an adjustment based on the amount of product in the distribution channel.
In 2004, we recorded an adjustment of * * million to our provision for returns. This adjustment was the result of the following developments related to business decisions made in 2004, or based on new information available in 2004:
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-7
Final Considerations
In determining our estimates for returns, we are required to make certain assumptions regarding the timing of the introduction of new products and the potential of these products to capture market share. In addition, we make certain assumptions with respect to the extent and pattern of decline associated with generic competition. To make these assessments we utilize market data for similar products as analogs for our estimations. We use our best judgment to formulate these assumptions based on past experience and information available to us at the time. We continually reassess and make the appropriate changes to our estimates and assumptions as new information becomes available to us.
We believe that the adjustments we have made in 2003 and 2004 arose from business decisions that were not contemplated at the time we originally recorded the product sales, or were based on the receipt of additional information that was not available at that time and, therefore, represent a change in estimate. Now that we have worked down our trade inventory levels, we do not anticipate these levels of adjustments in the future due to the terms of the distribution service agreements we have entered into with our three major wholesalers. These agreements limit the amount of inventory these wholesalers can own and eliminates their incentive to purchase inventory in advance of a price increase, as well as provide more timely and complete information with respect to inventory levels held and better data regarding sales and marketplace activity. We also anticipate the overall returns experience rates to decline significantly.
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-8
Our methodology for estimating returns provides a reasonable indication of the percentage of returns that will be returned in each year following the date of original sale by tracking the lot numbers of actual product returns. In addition, beginning in 2005 we are receiving "Morgue Reports" from our three major wholesalers that detail inventory within 12 months of expiration, which we can factor into our estimate for returns.
Our Ability to Make Reasonable Estimates
In connection with our methodology for estimating returns, we thoroughly analyze whether there are any facts and circumstances that impair our ability to reasonably estimate returns including technological obsolescence, changes in demand, length of return periods, reliability of historical data and nature of our products.
Paragraph 8 of FAS 48 lists the following factors that may impair the ability to make a reasonable estimate:
Our products are proven medications that have been on the market for many years and are characterized by physician and patient loyalty resulting from long-term usage and established reputations for safety and efficacy. To a degree, that reputation insulates these products from substitution and changes in demand. That is, once a patient is placed on a medication and appears to tolerate it well, physicians are reluctant to substitute a new medication in its place, as this may entail unforeseen complications to the patient's health.
Our products are susceptible to changes in demand due to the introduction of generic competition. In 2003, we noted that the decline in demand for our non-promoted products was more rapid than we had originally anticipated at the time we acquired these products, due mainly to lower utilization of our products through private and public benefit plans. However, following this change in marketplace dynamics, we believe that we can reasonably predict the decline in demand due to generic competition, as a large amount of internal and external data exists to support the trend in prescription demand for genericized brand drugs. This data reveals that once generic versions enter the market, the brand usually loses the majority of its pre-genericization market share within the first six to 12 months following the loss of exclusivity or patent expiration. However, over time, and after this initial period of rapid share decline, the drop in utilization of the brand tends to incrementally slow and level off. As a result of brand name awareness and physician and patient loyalty, these products continue to generate revenue, albeit declining, for a very long time even though promotional efforts have stopped.
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-9
In estimating our provisions for returns related to our promoted products, we assess the likelihood of generic competition to these products during the return period (as described below under ii). Both Cardizem® LA and Teveten are patent protected and, therefore, a generic competitor would first have to prove that its formulation did not infringe on our patents. Regardless, the FDA would generally be precluded from granting final approval to the generic formulation for 30 months under the Hatch Waxman Act. As a result, we did not consider that generic competition to Cardizem® LA and Teveten was likely, as there were no known generic filers and the 30-month stay of approval would effectively cover the return periods for these products based on their respective shelf lives of 18 months and 24 months. With respect to Zovirax, although this product has lost patent protection (and therefore not subject to the 30-month stay), we do not presently anticipate generic competitors in the United States due to the difficulty in proving bioequivalence for a topical medication.
Our non-promoted products are at different stages of their respective life cycles and have lost market exclusivity over the past four to 30 years. We acquired these products after their initial post-genericization period of rapid decline and volatility. These products are following a typical pattern of incremental slowing decline in utilization, characterized as the "long tail" or the end of the bell curve. We do not believe that the entry of one or more additional generic competitors to these products would have a significant impact, as most of the pre-genericization market share for brand products is lost following the entrance of the first two or three generic competitors.
Consistent with industry practice, we allow customers to return product within an 18-month period commencing six months prior to and continuing up to 12 months after its expiration date. Our products have long shelf lives that range from 18 months to 60 months and, therefore, there is an extended period during which each product can be sold through to the end-customer. The introduction of additional generic competition does not give our customers the right to return product outside of this established policy. Under our methodology for estimating returns, we consider the percentage of returns that we can expect to receive each year between the period of sale and the latest period in which the product can be returned.
In the case of our acquired product lines, we initially based our provisions for returns related to these products on the historical experience rates of the predecessor companies. As our sales, marketing and distribution practices were similar to those employed by those predecessor companies, we believed that utilizing their experience rates was a reasonable basis for initially estimating returns related to these products. In the case of a product that is an extension of one of our existing products, we initially base our provision for returns on the historical experience rate of the predecessor drug. Over time, we adjust these experience rates as appropriate to reflect our actual experience with those acquired or extension products.
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-10
In late 2004 and early 2005, we entered into the aforementioned distribution service agreements with our three major wholesalers. By establishing limits on inventory levels owned by these wholesalers, these agreements eliminate forward buying that can result in sales fluctuations unrelated to end-customer demand. Lower inventories at the wholesale level, combined with improved inventory management and reporting practices by the wholesalers, has resulted in reductions in the amount of our actual returns.
We generate product sales from a large pool of homogeneous transactions that are based on normal and customary sales terms whereby title to the product and substantially all of the risks and rewards transfer to the customer upon delivery. We have published price lists for our products, as well as standard trade policies that are consistent with pharmaceutical industry practices. These policies include terms related to delivery, payment, cash discounts, chargebacks and returned goods.
Question 1 of SAB 104 A.4.b. lists the following additional factors that may impair the ability to make a reasonable estimate:
Our methodology for estimating returns includes a comparison of inventory levels in the distribution channel to the rate of pull-through based on prescription demand. In situations where we become aware of a build-up of inventory in the distribution channel, we analyze the situation and if the analysis indicates that returns will be higher than expected, we adjust our original estimate for product returns accordingly.
Prior to 2004, drug wholesalers operated under a business model that was largely predicated on inventory appreciation from manufacturer price increases. This model resulted in fluctuations in the levels of inventory owned by these wholesalers depending on the degree to which these wholesalers speculated on the timing of price increases.
The distribution service agreements with our three major wholesalers allow these wholesalers to own between two weeks and 11/2 months of inventory of our products at any given time. In addition, one of the services we receive through these agreements is a single point distribution whereby the major wholesalers are responsible for rerouting product through its network of regional distribution centres to meet demand. These factors has allowed us to reduce the inventory levels of our products in the distribution channel to approximately two months on hand at December 31, 2004, and approximately one month on hand at December 31, 2005.
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-11
Prior to 2004, we relied on data obtained from IMS to evaluate inventory levels in the distribution channel. IMS data was the recognized industry standard and considered to the best available information. Since 2003, IMS no longer provides this service. As a result, during 2004 we acquired this data directly from our three major wholesalers. Currently, the distribution service agreements require our three major wholesalers to provide us with more extensive data with respect to the sales and inventory levels of our products, such as inventory levels on-hand and on order, aggregate sales out and product returns. We also obtain prescription data for our products from IMS. We use this data to identify sales trends based on end-customer demand and to estimate inventory requirements.
The impact of the introduction of new products can be difficult to assess, as the acceptance of a new product is dependent on the ability of the developer to successfully scale up, manufacture, promote, market and distribute the product. As discussed above under i, we do not anticipate that our brand products will be displaced by new drugs in the near future. However, as our intent is to maintain inventory levels of two weeks to 11/2 months in the distribution channel, we do not believe that the introductions of new products would have a significant effect on our ability to estimate returns. That is, even new products with markedly improved therapeutic benefits will not immediately displace our products, due to the reluctance on the part of physicians to alter a patient's medication that has proven to be effective and well tolerated. In addition, using the information that is available to us from IMS and our major wholesalers, we can continually assess the market demand for our products, and adjust our sales into the distribution channel to compensate for reduced pull-through due to declining prescription demand.
Our direct product sales in the United States are made to drug wholesalers and warehousing chains. Three national players, Cardinal, McKesson and ABC, dominate the drug wholesale market in the United States. These three wholesalers accounted for approximately two-thirds of our direct product sales in 2004. While each of these wholesalers represents a major customer of ours, the actual function of drug delivery performed by each is very similar. As a result, we are not dependent on any one of these wholesalers.
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-12
With the exception of Cardizem® LA, Teveten HCT and Zovirax Cream, which were launched by us, the products that we directly market in the United States are all well-established pharmaceuticals with long histories. For those products that we acquired, we initially based our estimate for returns related to these products on the historical experience rates of the predecessor companies. For the products that we launched, we based our estimate for returns primarily on the product returns experience of our existing products within the same category of drug, as follows:
Our directly marketed products in the United States fall into the following classes of drugs: Benzodiazepine (Ativan®), CCB (Cardizem® CD and Cardizem® LA), ARB (Teveten and Teveten HCT), angiotensin converting enzyme inhibitors (Vasotec® and Vaseretic®) and topical anti-viral (Zovirax Ointment and Zovirax Cream). These therapies have well-established reputations for safety and efficacy. We do not anticipate that new or alternate technologies or classes of drugs will supplant these therapies in the foreseeable future.
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-13
Our non-promoted products face significant competition from more cost-effective generic substitutes. Post-genericization, current analogs indicate that brand name drugs such as ours generally retain 5% – 10% of their pre-genericization volumes. The degree of generic substitution generally follows a typical pattern, which results in a fairly predictable revenue stream. This reasonably predictable pattern permits us to anticipate the impact that generic competition will have on prescription trends for our mature products, which we can then factor into our estimate for returns. However, the rate of decline in market share is variable due mainly to the number of generic competitors that enter the market for a particular product, which may be as few as one or two. Conversely, for some products, a very large number of generic competitors may enter the market. Therefore, we continually monitor the impact of generic competition on the demand for our products as higher than anticipated competition, in terms of the number of generic competitors, could result in higher than expected inventory levels in the distribution channel. However, all of our non-promoted products are well past the date of initial genericization and, therefore, the impact of additional generic competition to these products is not expected to have a material impact on their remaining market share. In addition, we know from six months to more than 30 months in advance of a generic competitor's entrance into the market, which allows us to adjust our forecasts and estimates accordingly.
A factor we consider in preparing our estimates for returns is the impact of changes to the expiration dating related to our products. The initial shelf life for prescription drugs usually falls between 18 and 24 months, which can be extended based on satisfactory long-term stability data accumulated years after the original approval. The longer the shelf life of a product, generally the lower the likelihood that a product will be returned due to expiration. However, a change to increase the dating on a product can also have a significant temporary impact on the level of returns, since the change in dating will only apply to product manufactured and packaged after that change is approved. Therefore, product in the distribution channel at the point in time that the change in dating is approved will have a shorter shelf life than new product being manufactured. As a result, our customers may order new product with the extended shelf life, and return the product with the shorter shelf life when it expires. This can result in a period of significantly higher returns related to the product with the shorter shelf life. We experienced such an outcome after we increased the dating on Cardizem® CD from 24 months to 36 months in May 2002, which contributed to returns of this product during 2003.
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-14
Another factor we consider is the impact of changes to the NDCs related to our products. As the NDC identifies the company that distributes the drug, we are required to change the NDC for our acquired products from the predecessor companies to Biovail. Changes to the NDC can result in a period of significantly higher returns related to the product with the old NDC, due to the limitations of our customers' inventory systems. We anticipated such an eventuality when we changed the NDCs for certain of our acquired products from the predecessor companies to Biovail in 2003 and 2004.
Based on the above factors outlined in paragraph 8 of FAS 48 and Question 1 of SAB 104 A.4.b., we believe we have a reasonable basis to estimate returns.
Rebates and Chargebacks
We are subject to rebates on sales made under governmental pricing programs and managed care contracts. The largest of these rebates are associated with sales covered by Medicaid. We participate in state government-managed Medicaid programs, as well as certain other qualifying federal and state government programs whereby discounts and rebates are provided to participating government entities.
Chargebacks relate to our contractual agreements to sell products to group purchasing organizations and other indirect customers at contractual prices that are lower than the list prices we charge wholesalers. When these group purchasing organizations or other indirect customers purchase our products through wholesalers at these reduced prices, the wholesaler charges us for the difference between the prices they pay us and the prices they sold the products to the indirect customers.
Our methodology for estimating provisions for rebates and chargebacks considers relevant statutes with respect to governmental pricing programs, and existing contractual sales terms with group purchasing organizations and managed care providers. We estimate the amount of our product sales subject to these programs based on historical utilization levels.
We account for rebates and chargebacks in accordance with EITF Issue No. 01-9,Accounting for Consideration Given by a Vendor to a Customer (Including a Reseller of the Vendor's Products) (EITF 01-9). In accordance with paragraphs 22 and 30 of EITF 01-9, we recognize the cost of rebates and chargebacks concurrently with the recognition of product sales revenue. We occasionally offer performance-based incentive programs that are based on certain purchase volumes. In those cases, we assess the likelihood of the customer attaining the required purchase volumes and record a liability based on the expected incentive amount; however, if we cannot reasonably assess the volume of the customer's future purchases, we record the maximum incentive amount. Paragraphs 23 and 30 of EITF 01-9 list the following factors that may impair the ability to make a reasonable and reliable estimate:
Rebates and chargebacks are determined based on contractual sales terms or relevant governmental statutes at the date of purchase by the group purchasing organization or other indirect customer, or after the dispensing of our products by a pharmacy to a private or public benefit plan participant.
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-15
Due to the time to process claims, Medicaid rebates are typically billed around 180 days after the product is shipped, but can be up to 270 days after the quarter in which the product is dispensed to the Medicare participant. As a result, our methodology for estimating our Medicaid rebate provision considers the amount of outstanding claims for end-customer sales that occurred but for which the related claim has not been billed, and the amount of future claims that will be made when our inventory in the distribution channel is sold through to Medicaid participants. Periodically, we adjust the Medicaid rebate provision based on actual claims paid. Due to the delay in billing, adjustments for actual results may incorporate revisions of this rebate provision for several periods. Other rebates and chargebacks are generally settled within a few months of original sale.
Changes in the level of utilization of our products through private or public benefit plans and group purchasing organizations will affect the amount of rebates and chargebacks that we owe. We continually update these factors based on new contractual or statutory requirements, and significant changes in sales trends that may impact the percentage of our products subject to rebates or chargebacks. With respect to rebates to private and public benefit plans, these programs tend to require the substitution of lower-priced generic products in place of brand name prescriptions. For example, managed care providers often provide a disincentive to consumers using more expensive branded products by requiring the consumer to "co-pay" for a portion of the price of the brand product. Also, many state administered Medicaid programs require that the generic version of a drug be dispensed to participants when one is available. Therefore, as generic competition to our brand products increases, the utilization of our products by private and public benefit plans will generally decline.
Rebates and chargebacks are product-specific and, therefore, can also be impacted by the mix of products sold. For example, in 2004 our overall experience rate for Medicaid rebates declined following our introduction of Zovirax Cream in July 2003. The Medicaid discount on Zovirax Cream is approximately * * percent, compared with approximately * * percent on Zovirax Ointment. In 2004, Zovirax Cream accounted for approximately * * percent(1) of the topical anti-herpes market, while Zovirax Ointment declined approximately six-percentage points(1) between 2003 and 2004. As a result, our overall experience rate with respect to Medicaid was lower in 2004, compared with 2003.
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-16
Historically, a significant proportion of our direct product sales have been subject to one form or another of rebate or chargeback. This provides us a substantial volume of historical experience upon which we can base our current estimates with respect to the utilization rates related to our products.
Based on the above factors outlined in paragraphs 23 and 30 of EITF 01-9, we believe that we have a reasonable ability to estimate rebates and chargebacks.
Comment:
In addition, please clarify how having no basis to estimate relates to you apparently having a basis to assert "that the returns levels for these products were higher overall than the historical experience of the companies from whom [you] acquired these products would have indicated".
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-17
Response:
As clarified above in our response to the staff's comment 2, we were not responsible for any pre-acquisition returns in accordance with the respective purchase agreements related to Teveten and Cardizem® CD. The amount of returns that we were responsible for related to Vasotec® and Vaseretic® was not expected to be, nor was in fact, material (approximately $14,000). Accordingly, we did not record a liability at the date of acquisition related to this amount in accordance with paragraph 4 of FAS 141, nor amend our purchase price allocation. Our response to the staff's prior comment one stated that we had processed returns on behalf of Solvay and Aventis related to pre-acquisition sales of Teveten and Cardizem® CD, respectively. However, the cost of those returns were borne solely by Solvay and Aventis and, accordingly, those companies reimbursed us for any amounts paid by us for returns related to sales made prior to the dates of our acquisitions. Therefore, we had no liability for pre-acquisition returns.
Based on the due diligence we conducted in advance of these acquisitions, we believe that Solvay, Merck, and Aventis had a reasonable basis to estimate their provisions for returns related to Teveten, Vasotec® and Vaseretic®, and Cardizem® CD, respectively. As a result, we believed that no material returns liabilities existed at the dates of acquisition and, therefore, we believed that it was appropriate to rely on the historical experience rates of the predecessor companies to initially estimate our future returns.
Based on the above clarification of the facts, we recognize that we need to modify the proposed disclosures in our amended Form 20-F for the year ended December 31, 2004 to remove the reference to the impact of returns related to pre-acquisition sales.
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-18
Comment:
Response:
As clarified above in our response to the staff's comment 2, our methodology for estimating returns utilizes the guidance provided in SAB 104 A.4.b. In determining our estimates for returns, we are required to make certain assumptions regarding the timing of the introduction of new products and the potential of these products to capture market share. In addition, we make certain assumptions with respect to the extent and pattern of decline associated with generic competition. We use our best judgment to formulate these assumptions based on past experience and information available to us at the time. We continually reassess and make the appropriate changes to our estimates and assumptions as new information become available to us. For example, in 2003 we adjusted our estimate for returns to reflect changes by a number of state-managed Medicaid programs to require mandatory generic substitution.
Comment:
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-19
Response:
As clarified above in our response to the staff's comment 2, the cost of rebates and chargebacks related to sales made prior to our acquisition of Teveten, Vasotec®, Vaseretic®, and Cardizem® CD were borne solely by the companies from whom we had acquired these products. Accordingly, we were not required to record a liability at the date of acquisition related these amounts in accordance with paragraph 4 of FAS 141.
As previously communicated to the staff, the adjustment made to reduce the rebates provision in 2003 resulted from the availability of additional information related to the Medicaid utilization of the products we had acquired, including Ativan®, Cardizem® CD, Teveten, Vasotec® and Zovirax. No portion of this adjustment was related to sales made by the predecessor companies. As previously discussed, this decline in utilization was prompted by the increasing requirements of many state programs that the generic version of a drug be dispensed to Medicaid participants when one is available. Under these mandatory generic substitution policies, pharmacists are required to dispense generics first unless a doctor specifies the use of brand name drugs instead. As state governments continue to look for cost savings, some have implemented even tougher policies, which require a doctor to explain why, in writing, and get permission from the Medicaid program in order to force dispensing of a brand drug instead of its equivalent generic. As a result of increasing generic substitution for our products, we noted that actual utilization of our products under Medicaid programs was lower than historical experience. We believe that the adjustment that we recorded to reduce our rebate liability was the result of new information with respect to declining utilization rates and, therefore, represents a change in estimate, not the correction of an error.
Comment:
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-20
Response:
We are able to disclose the total amount of inventory in sales dollars that could be potentially returned as of the balance sheet date in table format by product.
The information we receive from our three major wholesalers (Cardinal, McKesson and ABC) pursuant to the distribution service agreements includes inventory data in dollars. These wholesalers accounted for approximately two-thirds of our direct product sales in 2004. Based on this information, we can disclose the total dollar value of our products owned by these wholesalers, that could potentially be returned, as set forth in the following table at December 31, 2004.
* *
For the staff's consideration, we are prepared to include the above table in our proposed disclosures in our Form 20-F for the year ended December 31, 2004.
Comment:
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-21
Response:
The 10% change we referred to in our response to the staff's prior comment one was a hypothetical amount. The most subjective judgments we are required to make in estimating our provisions for returns, rebates and chargebacks are related to the expected impact of increases or decreases in the level of inventory of our products in the distribution channel. Our intent is to maintain trade inventory levels in the range of 2 weeks to 11/2 months of supply. Accordingly, we can reasonably expect one-month fluctuations from time to time in these inventory levels.
From a quantitative prospective, it is very difficult to assess the impact that a one-month change in inventory levels may have on our returns provision, as we have to consider a number of factors that could cause the increase. For example, an increase in inventory levels due to increasing generic competition may be an indication that returns may be higher than originally anticipated due to increased product expiration. However, an increase in inventory levels prior to a price increase may indicate that the wholesalers are forward buying, which may not impact the amount of future returns. Therefore, we do not believe that quantitative information with respect to the sensitivity of returns is reasonably available, except in hypothetical terms. In any event, we do not believe that a one-month increase in inventory would be reasonably expected to have a material impact on our results of operations, financial condition or cash flows. Therefore, we propose to include the following alternate disclosure to address the qualitative factors that we consider in determining our provision for returns:
Our estimate for returns may be impacted by a number of factors, but the principal factor relates to the level of inventory in the distribution channel. When we are aware of an increase in the level of inventory of our products in the distribution channel, we consider the reasons for the increase to determine if the increase may be temporary or other-than-temporary. Increases in inventory levels assessed as temporary will not result in an adjustment to our provision for returns. Other-than-temporary increases in inventory levels, however, may be indication that future product returns could be higher than originally anticipated and, accordingly, we may need to adjust our estimated product returns provision.
Some of the factors that may be an indication that an increase in inventory levels will be temporary include:
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-22
Conversely, factors that may be an indication that an increase in inventory levels will be other-than-temporary include:
For rebates and chargebacks, we believe we can provide a quantitative assessment of the impact of a one-month change in inventory levels based on our historical weighted average utilization and discount rates. Accordingly, we propose the following alternate disclosure:
Our estimate for rebates and chargebacks may be may be impacted by a number of factors, but the principal factor relates to the level of inventory in the distribution channel. If the level of inventory of our products in the distribution channel increased or decreased by one-month supply, our provision for rebates and chargebacks would increase or decrease by approximately * * million to * * million.
Comment:
Intangible Assets, page 60
In so doing, please discuss the sale of the Teveten® product line, the agreements related to Cardizem® LA, and the concurrent restructuring of your U.S. commercial operations, as your response indicated that your strategic review partially culminated, in May 2005, with these transactions. When discussing these transactions, please disclose what you believed, as of June 30, 2005, to be the reasonably likely effect of these transactions on your future operations, cash flow and financial position. Similarly, please describe the intangible assets affected and disclose their carrying amounts as of the latest balance sheet presented and the associated revenue for each of the periods presented.
Response:
As a Canadian corporation we are required under National Instrument 51-102 to file our annual audited financial statements, and related MD&A, within 90 days of the end of the financial year to which those statements relate. Since we have a December 31 year-end, the MD&A must be filed no later than March 30th each year. The date of the MD&A must not be earlier than the date of the auditors' report. Accordingly, the date of the MD&A could have been any date from March 8, 2005 (the date of the auditors' report) to March 30, 2005 (the date of filing). We selected March 30, 2005 as the date of the MD&A. In order to avoid disparate disclosure in the United States and Canada, we chose to submit the identical MD&A in our Form 20-F for the year December 31, 2004 as had already been filed under Canadian securities laws. This explains why the discussion and analysis in our MD&A speaks as of March 30, 2005, instead of as of June 30, 2005.
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-23
As discussed above, there is no requirement under the Form 20-F rules that the discussion and analysis in our MD&A speak as of a date later than December 31, 2004, other than as required in Item 5(D) of Form 20-F, which states in relevant part: "The company also should discuss, for at least the current financial year, any known trends, uncertainties, demands, commitments or events that are reasonably likely to have a material effect on the company's net sales or revenues, income from continuing operations, profitability, liquidity or capital resources, or that would cause reported financial information not necessarily to be indicative of future operating results or financial condition".
We did include disclosure in our Form 20-F of the sale of the Teveten product rights, the agreements related to Cardizem® LA and the concurrent restructuring of our U.S. commercial operations (see Item 4.B., "Business Overview" on pages 15-17, "Three-Year History — Material Developments" on page 50 and Item 8.B., "Significant Changes" on page 134). As a result, we do not believe that disclosure additional to that which we have already made in the Form 20-F is necessary or appropriate.
Comment:
Financial Statements Completed in Accordance with U.S. GAAP, page F-2
Notes to Consolidated Financial Statements, page F-9
3. Disposition and Acquisitions of Intangible Assets, page F-17
Zovirax, page F-21
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-24
In addition, please clarify for us whether you believe that paragraph 8 of SFAS 5 applies to the amount and why, as we had referenced it in our comment but your response did not appear to address it. If it is applicable, please justify why it was appropriate to defer the value of the reduction in the supply price when, according to your response, you could not assess the likelihood of receiving FDA approval for Wellbutrin XL as probable; otherwise, please clarify why you believed that not receiving the approval was probable.
To the extent that neither SFAS 5 or EITF 02-16 is applicable, please tell us what alternative to this accounting and what other authoritative guidance you considered and how, in light of those alternatives, you concluded your accounting was appropriate.
Response:
As we discussed in our response to your prior comment eight, we acquired the exclusive distribution rights to Zovirax from GSK effective January 1, 2002. GSK will continue to manufacture and supply Zovirax to us over the term of the distribution agreement. The original supply price to be paid by us to GSK was equal to * * of our published list price to our wholesale customers for Zovirax (the "Original Supply Price"). Effective October 1, 2002, we amended the terms of the Zovirax distribution agreement with GSK, including a reduction in the supply price for this product purchased by us on or after that date. The reduced supply price was equal to * * of our published list price to our wholesale customers for Zovirax (the "Reduced Supply Price"). To be clear, the difference between the Original Supply Price and the Reduced Supply Price (the "Supply Price Reduction") was credited by GSK to us, not payable by us to GSK. However, the Supply Price Reduction was subject to reimbursement by us to GSK if Wellbutrin XL was not approved by the Food and Drug Administration ["FDA"] and GSK terminated the development agreement for this product.
We deferred the benefit of the Supply Price Reduction by charging cost of goods sold based on the Original Supply Price, and recording the cumulative difference between that price and the Reduced Supply Price paid to GSK in accrued liabilities. We believed that paragraph 7 of EITF 02-16 was applicable to the Supply Price Reduction because this Reduction represented cash consideration (as defined in EITF 02-16B to include credits that a customer can apply against trade amounts owed to the vendor) received by us (the reseller) from GSK (the vendor) in the form of a price reduction. We considered that the receipt of FDA approval for Wellbutrin XL was analogous to the milestone events specifically addressed by paragraph 7. Accordingly, we did not believe we had earned the Supply Price Reduction until we received an Approvable Letter for Wellbutrin XL from the FDA on June 24, 2003. Prior to that date, we could not assess the likelihood of receiving FDA approval to be probable, as there is a high degree of risk and uncertainty associated with every New Drug Application. At the date of the Approvable Letter, we believed that approval was probable and the amount of the Supply Price Reduction was known, as it represented the cumulative difference between the Original Supply Price and Reduced Supply Price multiplied by the actual quantity of Zovirax purchased by us from GSK between October 1, 2002 and June 24, 2003. Accordingly, at the date of the Approvable Letter, we recognized the Supply Price Reduction as a reduction to cost of goods sold in accordance with paragraph 1 of EITF 02-16.
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-25
We did not believe that paragraph 8 of SFAS 5 was applicable to the Supply Price Reduction as referenced in your prior comment eight. As the proceeding discussion indicates, we did not recognize the Supply Price Reduction in earnings until it was probable that the amount had been earned by us. Consequently, there was no cause to recognize a loss contingency. If, hypothetically, we had recognized the Supply Price Reduction in earnings as the cash consideration was received rather than deferring it as we did, then the provisions of paragraph 8 of SFAS 5 may have been applicable. That is, the reimbursement of the Supply Price Reduction may have constituted a contingent loss that would have been resolved when FDA approval for Wellbutrin XL was either received or denied.
An alternative method of accounting for the Supply Price Reduction that we considered was to recognize the benefit of this Reduction as the cash consideration was received from GSK. However, we did not believe that this method met the aforementioned recognition criteria in paragraph 7 of EITF 02-16 as the cash consideration had to be both probable and reasonably estimable. In addition, we did not believe that this method met the criteria in paragraph 17 of SFAS 5 with respect to the accounting for gain contingencies. That paragraph states that contingencies that might result in gains usually are not reflected in the accounts since to do so might recognize revenue prior to its realization. We believed that the contingency in this case was FDA approval of Wellbutrin XL. We also believed that this contingency was resolved with the receipt of the Approvable Letter on June 24, 2003. In addition, after the receipt of the Approvable Letter, it would not be possible for GSK to terminate the Wellbutrin XL development agreement. As the cash consideration from the Supply Price Reduction had already been received from GSK prior to the date of this Approvable Letter, we believed the contingent gain was realized as of that date.
Comment:
12. Other Assets, page F-28
Interest Rate Swaps, page F-29
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-26
Response:
As we discussed in our response to your prior comment eleven, we treated the fair value adjustment to our Senior Subordinated Notes (the "Notes") of $15.1 million at April 1, 2003 (the date the hedging relationship failed to meet the criteria for a highly effective hedge), as a premium to be accreted to reduce interest expense over the remaining term of the Notes on an effective-yield basis. On October 1, 2003, we determined that the hedging relationship again qualified a highly effective hedge. FAS 133 provided no specific guidance to indicate whether this accretion should be discontinued if, and when, the hedging relationship again qualified as a highly effective hedge. Accordingly, we believed that there were two alternative accounting treatments available with respect to the unamortized fair value adjustment at that date, as follows: (a) continue to accrete the fair value adjustment to reduce interest expense over the remaining term of the Notes; or (b) recognize the fair value adjustment as a one-time immediate benefit to earnings. We concluded that alternative (b) was not consistent with the guidance in paragraph 24 of FAS 133, which states that an adjustment of the carrying amount of a hedged interest-bearing financial instrument shall be amortized to earnings. In addition, we believed that alternative (a) was the more conservative of the two treatments, as the accretion (benefit) to earnings occurs over the remaining life of the Notes rather than immediately on October 1, 2003.
Shortly following the completion of our discussions with the staff concerning the foregoing, we would anticipate filing an amended Form 20-F for the fiscal year ended December 31, 2004, which would incorporate these additional disclosures, as well as those disclosures requested by the staff in your letter dated September 30, 2005.
Please direct any additional questions or comments to me at (908) 927-1799.
Sincerely,
Charles A. Rowland, Jr.
Senior Vice President and
Chief Financial Officer
Cc: Michael Van Every, Chairman, Audit Committee
Cynthia Orr, Partner, Ernst & Young LLP
Rob Scullion, Partner, Ernst & Young LLP
Confidential Treatment Requested.
Information for which Confidential Treatment has been requested is omitted and is noted within "* *".
BV-27
