A Response to Dr - Business Valuation Resources



A Response to Dr. Shannon Pratt’s Critique of my Work on Marketability Discounts

In his Editor’s Column in the February issue of Shannon Pratt’s Business Valuation Update,( Dr. Shannon Pratt critiques my coauthored paper, “Firm Value and Marketability Discounts,” forthcoming in the Journal of Corporation Law.[1], [2] I would like to take this opportunity to rectify Dr. Pratt’s various misunderstandings and misinterpretations about my research and to highlight the key issues that practitioners are concerned about.

Dr. Pratt begins by implying that my research position is a biased one in favor of the IRS. For instance, he states that “Drs John Kania [an IRS economist] and Mukesh Bajaj are at it again, espousing discounts for lack of marketability for stocks of closely held companies that are a fraction of those encountered in the real world.” A little later in his column, he states that “they are doing real damage because they have influence on the Internal Revenue Service. Dr. Kania is employed by the IRS as an economist and Dr. Bajaj has been retained by the IRS on many occasions as a valuation expert.”

This characterization is misleading. As an independent expert, the majority of my work has actually been on behalf of taxpayers’ counsel. Yet the IRS has also retained me on numerous occasions. I believe that this client diversity in fact reaffirms my independence and credibility as an expert. As a minor aside: Dr. Kania is an IRS employee whereas I am not. Now, let us move on to the more substantive and important issues.

REEVALUATING THE COMMON APPROACHES TO DETERMINE MARKETABILITY DISCOUNTS

Appraisers usually rely on two approaches to determine marketability discounts. The first of these two approaches, the one seemingly preferred by Dr. Pratt, is the IPO approach. The second is the restricted stock approach. Let us reevaluate both.

The IPO Approach for Determining the Marketability Discount

The IPO approach attempts to quantify the “marketability” discount by comparing share prices in IPO transactions with (lower) transaction prices in the same shares prior to the IPO. John Emory and Dr. Pratt’s firm, Willamette Management Associates (WMA), have each conducted such IPO studies. Emory finds an average marketability discount of approximately 45%. WMA also found marketability discounts that are similar to those found in the Emory studies.

There are four reasons that indicate that such discounts do not reflect the lack of marketability as Dr. Pratt advocates. First, the discounts are so large that they cannot possibly reflect what discounts investors can expect on an ex-ante basis. If a discount of 45% were available to an investor willing to endure lack of liquidity similar to that inherent in pre-IPO transactions, such an investor could purchase a non-publicly-traded share for $5.50, then sell it for $10, earning an 82% rate of return. ($4.50 profit on an investment of $5.50 represents a return of $4.50/$5.50 = 82%.) Assuming that the private transaction takes place six months prior to the IPO (approximately the period of time in the Emory studies), this implies an annualized rate of return of 231%). To say that people can expect to earn an incremental annual rate of return of 231% simply by being willing to forego liquidity makes no sense. There are trillions of dollars invested in pension funds, university endowments, and other long-term investment vehicles. Competition for such high rates of return will drive the illiquidity premium to a more sustainable level. If saying that a 231% return premium for lack of liquidity cannot be correct makes me results oriented as Dr. Pratt alleges, then I am guilty as charged.

Second, pre-IPO buyers are usually insiders, such as venture capitalists, who provide expertise and services to the firm. Thus, the IPO discount in part may reflect compensation for these services rather than compensation for the lack of marketability. WMA alleges that it called and verified that the pre-IPO transactions were at arm’s-length prices and received an affirmative answer. What had it expected to hear? Do we know whether any of the buyers in fact provided any services to the firms in which they purchased shares? In fact, we do know from the Emory studies, which reach similar discount conclusions, that most of these transactions actually represented issuance of stock options. Emory compared the exercise price of the option to the IPO price to conclude a “marketability discount”!

Third, the discount on pre-IPO shares contains another important component that is completely unrelated to marketability. This is the discount for possible failure of the business enterprise. When an investor buys shares in a private, non-traded enterprise, he has no assurance that this enterprise will be one of the few to successfully emerge into the publicly traded market. Indeed, extant empirical evidence confirms that the odds are clearly stacked against achieving enough success to undertake a successful IPO. Therefore, when setting the price for a pre-IPO purchase, investors will rationally bid down the price of the share to reflect the probability that the firm may fail. If the firm actually succeeds and undertakes an IPO, the price of its shares no longer has to reflect the possibility of failure. Therefore, the issue price of the firm will usually be higher than the prices at which pre-IPO transactions occurred. However, this difference partly reflects the “discount” demanded by investors prior to the IPO for the possibility of failure, which is distinct from the lack of marketability.[3]

An example will help clarify the intuition underlying the sample selection problem. Suppose that investors know that the value of a share in the near future will be either $20 or $0 with equal probability. If the value turns out to be $20, the firm will issue shares via an IPO. Otherwise, it will not. Assuming no marketability discount, rational investors will price the shares today at their expected value, $10 (expected value = 50% ( $20 + 50% ( $0 = $10). An analyst using the IPO approach to estimate the marketability discount would observe an IPO only if the value of the firm turned out to be $20. Hence, the analyst would conclude that the marketability discount was 50% even though the true discount was zero. Extensive research on business survival rates show that the risk of business failure is indeed high. Dunne, Roberts and Samuelson (1988) examined the Census of Manufacturers for 1967, 1972, 1977, and 1982 and found that, on average, 79.6% of firms exit within just 10 years.[4] “Simply staying in business long enough to get old is a challenge.” [Popkin and Kirchhoff (1991) page 63.]

Fourth, the Emory IPO studies cited are simply wrong and biased. In at least one study, I found significant data errors that result in overestimation of the “marketability discount.” After reviewing 732 prospectuses, the Emory study identified 91 IPOs between November 1995 and April 1997 that matched specified criteria. Only 22 of these involved a pre-IPO transaction in company stock. For the remaining 69 observations, Emory used an option’s exercise price to estimate the shares’ fair market value. I examined the prospectus accompanying the firm’s IPO for all 91 cases. In 7 firms out of 22 involving stock transactions, I found at least one transaction that had occurred within five months of the IPO, in which the price was higher than the transaction price chosen by Emory to calculate the discount. The Hambrecht and Quist Group (H&Q) observation illustrates the magnitude of the bias. H&Q’s IPO in August 1996 was at a price of $15. The Emory study used a private transaction in March 1996 at $6.52 to compute its IPO discount of 56.53%. Curiously, Emory ignored the fact that H&Q had also sold shares privately in March 1996 for $13.06 and in June 1996 for $13.13, which would imply a discount of less than 13% compared to the IPO price.

In fact, the bias is even greater for the 69 pre-IPO transactions that involved a sale of options. In 31 of these cases, I found at least one option transaction within five months prior to the IPO where the stock price was greater than the price used by the Emory study. Consider the case of Impath, which conducted its IPO in February 1996 at an issue price of $13.00. The Emory study cites an option transaction in October 1995, in which the exercise price was set at $9.50, and uses this exercise price to compute the discount associated with Impath’s shares. However, the study ignored another transaction in December 1995 in which the fair market value of the share was stated to be $13.00, which would have implied a zero discount in Emory’s methodology.

Even if one is disposed to consider the entire IPO discount as a marketability discount – a position I reject – the degree of bias I encountered in this study makes it clearly unreliable. WMA has never disclosed its data. As such, an independent assessment of its claims is impossible. A reasonable debate can hardly ensue under these circumstances. While the courts have accepted the results of these studies in the past, I doubt they will do so once they learn of the biases.

The Restricted Stock Approach for Determining the Marketability Discount

The restricted stock approach is another approach to quantify the marketability discount. Privately placed securities may be either registered shares or unregistered shares that are considered restricted securities under SEC Rule 144.[5] Consequently, buyers of unregistered shares must wait for a specified period of time before the shares become marketable.[6] Unregistered shares placed privately are typically sold at a discount.

Average discounts on unregistered shares are sizable, ranging from 20% to 35%.[7] But these average figures do not reveal the tremendous range (from -15% to 80%) in these discounts across different firms. Silber (1991) found that those firms in the high (above median) discount group had (i) a higher number of shares issued as a fraction of total shares outstanding, (ii) smaller dollar size of the restricted stock issue, (iii) lower earnings, (iv) lower revenues, and (v) smaller market value of equity. This suggests that it is incorrect to simply apply a constant discount for all restricted stock issues. Rather, it appears that the firm’s characteristics and the restricted stock issue’s size affect the price difference observed between restricted and publicly traded shares.

Even after such issue and issuer characteristics are considered, it is premature to label the resulting discount a “marketability discount.” For example, private placements often involve key investors who provide the firm with advice, oversee its management, and commit to providing capital in the future if the firm meets a set of predetermined performance targets. Consequently, at least a portion of the price discount observed in private equity placements might reflect compensation to these investors for future services rendered, rather than compensation for the lack of marketability.

Even though private placements could also involve registered shares, early restricted stock studies only examined private placements of restricted stock issues, not other private placements of equity.[8] They were, therefore, incapable of distinguishing whether a privately placed unregistered stock was sold at a greater discount because it was not registered compared to other registered privately placed stocks. Academic studies by Wruck (1989)[9] and Hertzel and Smith (1993)[10] do compare restricted stock to similar issues of registered shares. Wruck found that the difference in median discounts between the restricted shares in her study and registered shares was 10.4%. Since both registered and restricted shares may have been given (in part) as compensation, but only restricted shares have marketability impairments, the difference in the discounts between registered and unregistered shares is a more appropriate measure of the marketability discount.

Hertzel and Smith provided a more comprehensive examination of the private placement market in 1993. Their study indicates that private placements are often undertaken by firms with limited tangible assets, those engaged in speculative development of new products, and those in financial distress, and that discounts tend to be higher for these types of firms.[11] Hertzel and Smith suggest that discounts are required on the stock of these firms to serve as compensation for the higher information and monitoring costs associated with the investments.[12] Therefore, discounts are not solely compensation for future services rendered or due to the lack of marketability, but are also caused by these factors. After controlling for nonmarketability determinants of private placement discounts in a multivariate regression framework, Hertzel and Smith found that the discount for restricted shares was only 13.5% greater than that of registered shares.[13] This is further evidence that the marketability discounts suggested by early restricted stock studies were too high.

INCREMENTAL CONTRIBUTION OF BAJAJ, DENIS, FERIS AND SARIN STUDY

In my work with my coauthors, we had three objectives. First, we wanted to bring the insights from recent scholarly research to the attention of practitioners. Second, we wanted to update the restricted stock study performed by Hertzel and Smith on more recent data (from 1990 to 1995). Third, we wanted to provide some guidance on the implications of this academic literature for practitioners.

There is a detailed literature review in our study. Dr. Pratt has picked up on some of the discussion and, in my view, criticized it incorrectly. However, I will not address those criticisms that do not further the professional discourse. Therefore, I will confine the rest of my comments on how I believe the results of the recent literature are applicable for valuation of closely held firms.

IMPLICATIONS FOR DETERMINING MARKETABILITY DISCOUNTS ON CLOSELY HELD FIRMS

Let us first consider valuation of a closely held operating entity. Do the results discussed above imply that, on average, the marketability discount should reflect the difference between observed discounts on registered and unregistered private placements, after adjusting for other factors? (This number, from the Hertzel and Smith study, would be about 13.5% and about 7% for the 1990-95 period). The answer, in my opinion, is not necessarily. When valuing a closely held operating company, a prospective buyer will take all elements of the marketability discount into account. Therefore, the fact that we can say that some of these discounts reflect information and monitoring cost is helpful in our understanding of economics of privately held companies, but it does not necessarily imply that only the narrowly construed notion of liquidity impairment is relevant. If a hypothetical buyer will incur information and monitoring-related costs, he will expect to be fairly compensated for such services in the form of a discount. It is because of this reasoning that I had considered the total private placement discount (adjusted for various characteristics of the firm) when I determined the marketability discount in the Gross case (T.C. Memo 1999-254) on behalf of the IRS and the Heck case (T.C. Memo 2002-34) on behalf of the taxpayer in the U.S. Tax Court.

However, consider an example in which an interest in a closely held business is contributed to a family limited partnership. Assume that the operations and management of the business are not changed as a result. Also assume that the valuation of the business interest contributed to the partnership already reflects the appropriate marketability discount for valuation of similar businesses. To value an interest in this partnership, should we apply the full marketability discount based on restricted stock studies? In my opinion, the answer is no. Placing an interest in a closely held business in a partnership makes such interest somewhat less marketable. However, there is no question of additional information and monitoring costs in this situation if holding this business through a partnership did not affect its operation or management. In such a situation, only the narrowly construed marketability discount is applicable as a second layer of discount (above the one that has already been applied in valuing the closely held shares placed in the partnership).

While Dr. Pratt has cast my work with my coauthors as being biased in favor of the IRS, I believe we have simply tried to bring the benefits of academic research to this important issue. I have no axe to grind in this debate. In fact, I have worked for taxpayers as well as the IRS on several tax controversies, and I have never changed my opinion on the basis of who is paying my bills. My job as an expert, above all, is to assist the courts in reaching an informed decision. Clients engage me because what they need from their expert is an independent and unbiased opinion—not partisan attacks.

If readers would like to get a copy of my article or make comments, they can reach me at mukesh_bajaj@.

Sincerely;

Mukesh Bajaj

Managing Director/Finance and Damages

LECG, LLC

-----------------------

[1] Mukesh Bajaj, David Denis, Stephen Feris and Atulya Sarin, “Firm Value and Marketability Discount,” Journal of Corporation Law, Vol. 27, No. 1.

[2] Dr. Pratt’s editorial misspells the title of the paper as, “Fair Value and Marketability Discounts.”

[3] Dr. Pratt, in talking about the hapless employees of Enron, mixes the discount for marketability with risk of failure. However, risk of failure should be accounted for in valuation before considering the marketability discount.

[4] Dunne, T., M. J. Roberts, and L. Samuelson, “Patterns of Firm Entry and Exit in U.S. Manufacturing Industries,” Rand Journal of Economics 19, 1988, pp. 495-515.

[5]. 17 C.F.R. § 230.144(a)(3) (2001).

[6]. Prior to February 1997, investors had to wait for two years for their shares to become marketable. As of February 1997, this waiting period has been reduced to one year. See Revision of Holding Period Requirements in Rules 144 and 145, Securities Act Release No. 33,7390, 63 SEC Docket 2077 (Feb. 20, 1997).

[7]. Institutional Investor Study: Report of the Securities and Exchange Commission (Washington, DC: Government Printing Office, 1971): 2444-2456, Document No. 92-64, Part 5; M. Gelman, “An Economist-Financial Analysts Approach to Valuing Stock of a Closely Held Company,” 36 Journal of Taxation 353, June 1972, pp. 353-54; R. R. Trout, “Estimation of the Discount Associated with the Transfer of Restricted Securities,” 55 Taxes 381, June 1977, pp. 381-85; J. M. Maher, “Discounts for Lack of Marketability for Closely-Held Business Interests,” 54 Taxes 562, Sept. 1976, pp. 562-71; W. L. Silber, “Discounts on Restricted Stock: The Impact of Illiquidity on Stock Prices,” 47 Financial Analysts Journal 60, July-Aug. 1991, pp. 60-64; B. A. Johnson, “Quantitative Support for Discounts for Lack of Marketability,” 16 Business Valuation Review. 152, Sept. 1999, pp. 152-55.

[8]. See Thomas Frieblob, “What are the Effects of Differing Types of Restrictions on Closely-Held Stocks?” 58 Journal of Taxation 240, 1983; Richard D. Johnson & George A. Racette, “Discounts on Letter Stocks Do Not Appear to be a Good Base on Which to Estimate Discounts for Lack of Marketability on Closely Held Stocks,” 59 Taxes 574, 1981.

[9]. Karen H. Wruck, “Equity Ownership Concentration and Firm Value: Evidence from Private Equity Financings,” 23 Journal of Financial Economics 3, 1989, p. 17.

[10]. Michael Hertzel & Richard L. Smith, “Market Discounts and Shareholder Gains for Placing Equity Privately,” 48 Journal of Finance 459, 1993, pp. 459-469.

[11]. Id. at 480.

[12]. Id. at 484.

[13]. In fact, Hertzel and Smith argue that even the 13.5% difference in discount between registered and restricted stock probably overstates the costs of illiquidity. In addition, they argue that discounts of this magnitude would provide strong incentives for firms to register shares prior to issue or to commit to quickly register shares following the private sale. Moreover, because buyers of the restricted shares tend to be institutional investors that do not value liquidity highly (e.g., life insurance companies and pension funds), it seems unlikely that such investors would require substantial marketability discounts for the commitment not to resell quickly. Id. at 480.

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download