April 25, 2012
Mr. Jim B. Rosenberg
Senior Assistant Chief Accountant
United States Securities and Exchange Commission
Division of Corporate Finance
Mail Stop 6010
100 F Street, N.E.
Washington, D.C. 20549
| Re: | Affirmative Insurance Holdings, Inc. | |
| Form 10-K for Fiscal Year Ended December 31, 2010 | ||
| Filed March 31, 2011 | ||
| File No. 000-50795 |
Dear Mr. Rosenberg:
This letter is submitted in response to the follow-up questions you communicated via phone call on March 28, 2012 in response to our written response dated March 16, 2012 to the comments contained in your letter dated February 16, 2012 concerning the above-referenced filing.
Item 7. Management’s Discussion and Analysis of Financial Condition and Results of Operations
Critical Accounting Policies
Valuation of goodwill and other intangible assets, page 34
The Staff continues to believe that quoted market prices are the best evidence of fair value. In addition, the Staff believes that the stock is actively traded and, as such, market capitalization was the best representation of fair value as of December 31, 2010.
Response:
While it could be interpreted that our stock trades in an active market due to its listing on the NASDAQ national market, we do not believe that fact alone leads to the conclusion that market capitalization is the best representation of fair value. For a number of reasons, we continue to believe market capitalization is not the best representation of the fair value of our company. There are several unique factors that cause the market capitalization of our stock to not be the best representation of fair value.
First, our company is controlled by a single shareholder, and approximately two-thirds of our outstanding common shares are held by three shareholders. So even though we are a public company, we are essentially closely held. Minority shareholders are therefore at a disadvantage compared to more widely-held public companies in terms of their ability to influence decisions that may affect the value of the Company and their ownership interests. As a result, the market price of our company’s stock is not driven by pure market forces as it would be with a more widely-held public company. Investors are naturally going to place a higher minority discount on the market price of the Company’s stock.
Second, our stock lacks liquidity as evidenced by extremely low trading volume. For the time period between January 1, 2010 through February 29, 2012, the average daily volume of trades in our stock
Mr. Jim B. Rosenberg
United States Securities and Exchange Commission
April 25, 2012
Page 2
was just 0.05% of shares outstanding. The highest day of trading during this twenty-six month period was when 192,300 shares, or 1.25% of shares outstanding, traded in 2011. During this twenty-six month period, there were a total of 75 days that our shares did not trade at all.
Given that there are usually 21 trading days in a month, that means for 3.57 months out of the twenty-six month period from January 2010 through February 2012 our shares did not trade at all. This represented 14% of trading days during the twenty-six month period. While we do not have long gaps of days that our shares did not trade, as documented above and in previous responses, on the days when our shares do trade there is very limited trading volume and, as such, a lack of liquidity.
Third, our stock lacks liquidity as evidenced by a high bid-ask spread. Highly liquid assets have a spread of a few basis points at most. Using the bid-ask prices at the market close over the twenty-six month period, our average bid-ask spread for the years ended December 31, 2010 and 2011 and the two-month period ended February 29, 2012 were 2.50%, 5.50% and 10.02%, respectively. In comparison, Infinity Property and Casualty Corp. (IPCC), a peer company of ours, had average bid-ask spreads at the market close for the years ended December 31, 2010 and 2011 and the two-month period ended February 29, 2012 of 0.17%, 0.16% and 0.18%, respectively. We believe that our common stock’s bid-ask spreads are very large for a publicly-traded company, which are indicative of a market that is less liquid and harder to trade.
Fourth, we have been told anecdotal stories by investors that it is extremely difficult to trade our stock in any significant size due to the lack of liquidity for the stock. The ability to have an active market is particularly welcome for large institutional investors as they deal in greater volumes and as a result present a greater risk of their transactions distorting the price. In a market that is not very liquid, significant sales of assets can cause the market price to fall, which may limit or even undo the benefits of selling at that particular moment. Similarly, trying to buy a significant amount of stock in an illiquid market can drive the price up beyond that which the buyer was willing to pay.
Fifth, there is limited information on the Company available to investors. We generally provide little additional information to the public beyond what is required by U.S. generally accepted accounting principles, U.S. Securities and Exchange Commission regulations, NASDAQ listing rules and other laws and regulations governing public disclosure. Since management does not hold investor calls and the Company is not followed by any security analysts, investors do not have the same level of information typically available on more actively traded companies. As a result, an investor’s ability to accurately assess the market value of the Company’s stock is more limited.
Finally, significant information regarding the Company typically has limited effect on our stock price or the level of trading in the stock. For example, we announced on September 22, 2010 that Gary Kusumi would be replacing Kevin Callahan as CEO around noon that day. The stock did not trade for the remainder of the day or the next day as well. The following represents the trading activity for five days prior to, the day of and five days after the announcement.
Mr. Jim B. Rosenberg
United States Securities and Exchange Commission
April 25, 2012
Page 3
| Prior to Announcement | Announcement | After Announcement | ||||||||||||||||||||||||||||||||||||||||||
Date | 9/15 | 9/16 | 9/17 | 9/20 | 9/21 | 9/22 | 9/23 | 9/24 | 9/27 | 9/28 | 9/29 | |||||||||||||||||||||||||||||||||
Share Volume | — | — | 5.900 | 5.200 | — | 7.000 | — | 17.400 | 3.000 | 1.300 | 2.200 | |||||||||||||||||||||||||||||||||
Closing Price | $ | 3.63 | $ | 3.63 | $ | 3.87 | $ | 3.77 | $ | 3.77 | $ | 3.75 | $ | 3.75 | $ | 3.82 | $ | 3.70 | $ | 3.70 | $ | 3.70 | ||||||||||||||||||||||
Change | $ | — | $ | — | $ | 0.24 | $ | (0.10 | ) | $ | — | $ | (0.02 | ) | $ | — | $ | 0.07 | $ | (0.12 | ) | $ | — | $ | — | |||||||||||||||||||
There have been numerous academic research papers published on the impact of CEO turnover on a company’s stock. Furtado and Karan (1990) cite 10 studies that estimated the announcement effect of CEO changes. They find that abnormal returns around the announcement are typically 25 to 50 basis points for all changes. Bonnier and Bruner (1989) found a stronger positive return of about 2.5% when they attempt to isolate the real effect by focusing on turnovers at distressed firms (where an announcement is unlikely to signal significant information about the state of the firm). Clayton, Hartzell and Rosenberg (2003) looked at 872 CEO turnovers during the 1979-95 period. They found that stock volatility increases following a CEO turnover, even when the CEO leaves voluntarily and is replaced by someone inside the company. As a whole, this evidence suggests that there can be positive real effects, but that the information effect can dominate. For this reason, we believe that the lack of trading in our stock for a full day and a half following our CEO turnover announcement, coupled with the lack of significant trading activity in any of the other five days following the announcement is a compelling indication that our stock price is not be the best indication of fair value.
Other occasions on which there was very limited trading following the release of significant information occurred when we filed our 2010 10-K and 10-Q’s that accompanied our release of earnings. During the 5 trading days after our release of earnings and filing of our December 31, 2010 10-K and September 30, 2010 10-Q, the average daily volume was 4,180 shares, or 0.027% of the outstanding shares, and 2,000 shares, or 0.013% of the outstanding shares, respectively. The day after we released our December 31, 2010 earnings disclosing an $89 million loss for the year and significant reserve strengthening and filed our 10-K (events that all occurred after the prior day’s market close) the Company’s shares did not trade at all.
A recent occasion that occurred was when we filed our 2011 10-K on March 30, 2012, which reported a $164 million loss. During the 5 days after the release of this information, our average daily volume was 4,420 shares, or 0.027% of the outstanding shares and our closing share price ranged from $0.52 to $0.54 during this period.
Provide an evaluation of an appropriate control premium to reconcile the market capitalization of the company with the estimate of the fair value at December 31, 2010.
Response:
Given the factors noted above and considering the nature of our stock (i.e. limited trading, large bid/ask spread, etc), there is a lack of available information that would be considered relevant to determine a control premium. The basis for applying a control premium is market capitalization, which as stated above is not considered to be the most relevant indication of fair value. In order to assume a control premium, it must be assumed that the foundation is appropriate, and as we have stated above, our market capitalization is not considered the best indicator of fair value.
Mr. Jim B. Rosenberg
United States Securities and Exchange Commission
April 25, 2012
Page 4
Fundamental to the position that the most appropriate method in estimating the fair value of the company is to adjust the quoted market price of our stock by an appropriate control premium are two things: first, that the starting point of market capitalization is relevant; and second, that an adequate volume of supporting data upon which to estimate a control premium is available. We do not believe that using market capitalization is the best measure to estimate the fair value of our company. In addition, our research did not yield a sufficient number of peer companies with similar control structures. The insurance industry transactions we reviewed were not comparable to a hypothetical transaction involving the Company, and factors that may influence the prices paid and related control premiums were difficult to determine from publicly available information. Accordingly, we do not believe there was sufficient or relevant information available from which to reasonably estimate a control premium.
Please explain how U.S. generally accepted accounting principles allows the use of multiple methods to estimate the fair value as of December 31, 2010.
Response:
In instances where market capitalization is not the best representation of fair value, we understand that U.S. generally accepted accounting principles provide the following measures to value a company:
ASC 820-10 recognizes three techniques for proper measurement of fair value:
| 1. | Market Approach. The market approach uses prices and other relevant information generated by market transactions involving identical or comparable assets or liabilities (including a business). |
| 2. | Income Approach. The income approach uses valuation techniques to convert estimated future amounts (for example, estimated cash flows or earnings) to a single present amount (discounted). The measurement is based on the value indicated by current market expectations about those future amounts. |
| 3. | Cost Approach. The cost approach is based on the amount that currently would be required to replace the service capacity of an asset (often referred to as current replacement cost). |
ASC 350-20-35-22 states that fair value refers to the price that would be received to sell the entity in an orderly transaction between market participants at the measurement date. Quoted market prices in active markets are the best evidence of fair value and shall be used as the basis for the measurement, if available. The market price of an individual equity security (and thus the market capitalization of a reporting unit with publicly traded equity securities) may not be representative of the fair value of the reporting unit as a whole.
In estimating the fair value of a reporting unit, a valuation technique based on multiples of earnings or revenue or a similar performance measure may be used if that technique is consistent with the objective of measuring fair value. Use of multiples of earnings or revenue in determining the fair value of a
Mr. Jim B. Rosenberg
United States Securities and Exchange Commission
April 25, 2012
Page 5
reporting unit may be appropriate, for example, when the fair value of an entity that has comparable operations and economic characteristics is observable and the relevant multiples of the comparable entity are known. Conversely, use of multiples would not be appropriate in situations in which the operations or activities of an entity for which the multiples are known are not of a comparable nature, scope, or size as the reporting unit for which fair value is being estimated.
Management considered multiple approaches including both the market approach and the income approach, as summarized below:
| • | Transactions:an approach utilizing observable data from industry transactions. |
| • | Peer company data:an approach utilizing observable market data, including market capitalization, revenues, earnings, book values, and debt to equity ratios in relation to the respective companies’ per share equity values based on their stock market price. |
| • | Market capitalization: approach using observable market data. |
| • | Discounted cash flows:an approach using unobservable data, based on the Company’s forecast. |
For the annual impairment analysis, management estimated the fair value of the Company by eliminating the high and low values resulting from the four methods and averaging the remaining values. These two remaining values consist of one based on observable market inputs from published peer data and one based on discounted cash flows of the Company. The variance between the median and mean values is simply based on the transactions and multiples of peer data reviewed. The result of our goodwill impairment test would have been the same, i.e. the fair value of the company exceeded its book equity, regardless of whether the median or mean values were used.
We reviewed the public disclosures of several companies and noted that most of them disclosed the use of multiple methods of valuation, including peer analysis and a discounted cash flow based on management’s internal forecast. The disclosures of peer companies included the following (two of which also have only one reporting unit):
| • | Infinity Property & Casualty Corporation disclosed that they used a variety of methods, including discounted cash flow and peer market value comparisons. |
| • | First Acceptance Corporation disclosed that their valuation methods included the use of a discounted cash flow model using revenue and profit forecasts and recent industry transaction and trading multiples of their peers. |
| • | Hallmark Financial Services disclosed that the determination of fair value was based on multiple valuation approaches including an income approach utilizing discounted cash flows and a market approach utilizing observable key ratios of peer companies. |
| • | Mercury General Corporation disclosed the use of widely accepted valuation techniques and certain assumptions and estimates related to future earnings, long-term strategies, and its annual planning and forecasting process. |
We believe that we maximized the use of observable inputs by using peer data based on the fundamentals of those entities that a market participant would analyze if a transaction were being
Mr. Jim B. Rosenberg
United States Securities and Exchange Commission
April 25, 2012
Page 6
considered. We believe that the combination of observable market data of peer companies combined with the discounted cash flows of our company based on management’s five-year forecast was a balanced and realistic estimate of the fair value of our company as of December 31, 2010. At year-end 2010, based solely on the discounted cash flow model, the Company passed step 1 of the goodwill impairment test.
*******
The Company specifically acknowledges that:
| • | the Company is responsible for the adequacy and accuracy of the disclosure in the filing; |
| • | staff comment or changes to disclosure in response to staff comment do not foreclose the Commission from taking any actions with respect to the filing; and |
| • | the Company may not assert staff comment as a defense in any proceeding initiated by the Commission or any person under the federal securities laws of the United States. |
If you have any questions or comments, please feel free to contact the undersigned directly at 630.560.7205.
| Sincerely, |
| /s/ Michael J. McClure |
| Michael J. McClure |
| Executive Vice President and |
| Chief Financial Officer |
