Update on Study of Discount for Lack of Marketability
The concept of investment marketability relates to the liquidity of an asset, property or business interest—that is, how quickly and certainly it can be converted into cash at the owner's discretion. Investors value liquidity. Rational investors will pay a price premium for liquidity. Conversely, rational investors will demand a price discount for the lack of liquidity. For example, the common stock of a closely held company is not as liquid as the otherwise comparable common stock of a publicly traded company. The closely held company stock does not have the same degree of marketability as the otherwise comparable publicly traded company stock. All other factors held constant, the closely held stock is less valuable than the comparable publicly traded stock, reflecting a price discount for lack of marketability.
Empirical research quantifies that this discount for lack of marketability (DLOM) for closely held stock generally falls into two fields of analysis: (1) studies of sale transactions in the restricted stock of publicly traded companies (also known as restricted stock studies) and (2) studies of private sale transactions in the stock of closely held companies that subsequently implement successful initial public offerings (also known as pre-IPO studies).
Willamette Management Associates (WMA) analysts previously published empirical research pre-IPO studies analyzing the DLOM for the years 1975-97. WMA analysts recently completed an additional pre-initial public offering (pre-IPO) DLOM study covering the five years of 1998-2002. This discussion summarizes the results of the WMA pre-IPO DLOM study for the years 1998-2002.
Restricted-stock DLOM Studies
Studies of the sale of restricted shares of the stock of publicly traded companies quantify one indication of the DLOM related to closely held securities. Typically, restricted stock is identical in all respects to the freely traded stock of the same public corporation, except that the restricted stock cannot be traded on a stock exchange for a specific time period.
Numerous restricted-stock DLOM studies have been performed (and the study results published) since the late 1960s. These restricted-stock studies have concluded an average DLOM for restricted shares (compared to otherwise identical freely traded shares) in the range of 13-45 percent. The DLOM concluded by the more recent restricted stock studies are smaller than the DLOM concluded by the older restricted stock studies. One explanation for this phenomenon is the increase in volume of privately placed stock under Securities and Exchange Commission (SEC) Rule 144(a) in the past several years. Also, a change in the minimum investment holding period required by the SEC under Rule 144—from two years to only one year—took place as of April 29, 1997.
The DLOM conclusions indicated by the published restricted stock studies are typically lower than the DLOM conclusions indicated by the published pre-IPO studies. Restricted shares of public corporation stock are not able to be traded directly on a stock exchange temporarily. However, the investor has certainty that, in a relatively short time period, the trading restrictions will lapse. However, the shares of stock of a closely held corporation may never be traded directly on a stock exchange. The prospect of any level of efficient marketability is much lower for closely held corporation shares compared to restricted public corporation shares. Therefore, the appropriate DLOM related to closely held corporation shares (or to similar closely held investment securities) is generally considered to be greater than the DLOM indicated by published restricted stock studies.
Pre-initial Public Offering (Pre-IPO) DLOM Studies
A second type of empirical analysis that quantifies the appropriate DLOM is the pre-IPO study. A pre-IPO study examines arm's-length sale transactions in the stock of a closely held company that has subsequently achieved a successful initial public offering of its stock. In a pre-IPO study, the DLOM is quantified by analyzing (with various adjustments) the difference between (1) the public market price at which a stock was issued at the time of the IPO and (2) the private market price at which a stock was sold (in an arm's-length transaction) prior to the IPO.
WMA Pre-IPO Studies
WMA analysts previously published the results of 18 pre-IPO DLOM studies covering the period of 1975-97. We have recently completed an additional pre-IPO DLOM study encompassing the five years from 1998-2002. As in the previous pre-IPO studies, we include only private-market stock sale transactions that were conducted on an arm's-length basis in the 1998-2002 study. The transactional data analyzed in the 1998-2002 pre-IPO study included (1) sales of closely held corporation stock in private placements and (2) repurchases of treasury stock by the closely held corporation. All transactions involving the granting of employee, executive or other compensation-related stock options were eliminated from consideration in the 1998-2002 study. All transactions involving stock sales to corporate insiders or other related parties were eliminated from consideration in the 1998-2002 study unless WMA could verify (by a telephone interview with at least one principal party) that the stock sale transaction was, in fact, a bona fide, arm's-length transaction.1
In the current pre-IPO study, WMA obtained the list of initial public offerings from the Corporate New Issues database, provided by Thomson Financial Securities Data. Eliminated from the analyses were all IPOs involving financial institutions, real estate investment trusts (REITs) and natural-resource companies. Also eliminated were all offerings priced at $1 or less per share and all offerings on units or warrants. Consistent with the procedures performed in previous WMA pre-IPO studies, these data were eliminated because such offerings may have unique investment characteristics. The source documents used to identify the potential pre-IPO private-market sale transactions were complete SEC registration statements, primarily Forms S-1. The private-market sale transactions analyzed occurred during the 36-month period prior to the IPO. If a corporation had more than one private market transaction that met the study's criteria, all such private market transactions in that company were included in the analysis. Also, multiple private-market transactions occurring within 12 months at the same share price were treated as a single transaction in the analysis.
For each private-market transaction for which meaningful earnings data were available in the registration statement as of both the private transaction and the IPO dates, we compared the price/earnings multiple of each private transaction with the subsequent IPO price/earnings multiple. Corporations that had no meaningful earnings data as of either the private transaction date or the IPO date were eliminated from the analysis.
Because the private-market transactions occurred over a time period up to three years prior to the IPO, we made adjustments to reflect differences in the stock market conditions of the respective industries between the time of each private transaction and the time of each subsequent IPO. We adjusted the price/earnings multiples for differences in the industry average price/earnings multiple between the time of each private transaction and the time of each subsequent IPO. Both the calculation of company and industry price/earnings multiples and the adjustment for temporal changes in stock market conditions within each industry are consistent with the procedures performed in previous WMA pre-IPO studies.
The results of the recent WMA pre-IPO study, and the results of the previous WMA pre-IPO studies, are present in Table 1.
In addition to the above-mentioned WMA pre-IPO DLOM studies, other recent pre-IPO DLOM studies include analyses performed by John D. Emory and Valuation Advisors LLC. The results of these recent studies are summarized below.
John D. Emory Pre-IPO Studies
John D. Emory, president of Emory Business Valuation LLC, conducted a series of pre-IPO DLOM studies from 1980 to 2000. His most recent pre-IPO study in the series covers the time period of May 1997 to December 2000.2 For this time period, Emory concluded a mean DLOM of 48 percent and a median DLOM of 44 percent. These recent DLOM conclusions are consistent with previously published Emory pre-IPO studies. The previously published Emory pre-IPO studies concluded a median DLOM ranging from 40 percent in the 1989-90 and 1990-92 studies, to 66 percent in the 1980-81 study.
Emory also performed (1) a dot.com pre-IPO study and (2) an expanded pre-IPO study during the same time period. The Emory dot.com pre-IPO study3 included initial public offering companies that had ".com" in their names. These types of companies typically involve Internet-related businesses. With the 92 dot.com IPOs in the May 1997 to December 2000 time period, Emory found 53 private market stock sale transactions to analyze in this pre-IPO study. In this particular study, Emory concluded that the mean DLOM and the median DLOM conclusion were both 54 percent.
In the expanded pre-IPO study,4 Emory included companies that were not included in the general pre-IPO study due to their uncertain financial condition. In the Emory expanded pre-IPO study, companies that were not determined to be "reasonably sound" were included in the analysis. The DLOM conclusions in the Emory expanded pre-IPO study were greater than the DLOM conclusions in the general pre-IPO study. In the expanded pre-IPO, Emory concluded a mean DLOM of 50 percent and a median DLOM of 52 percent.
Valuation Advisors Pre-IPO Study
Valuation Advisors LLC (VA) performed a pre-IPO DLOM study for each of the years 1999, 2000 and 2001. The conclusion of the VA 1999 pre-IPO study indicated a mean one-year DLOM of 51.91 percent.5 In the 2000 pre-IPO study, the mean one-year DLOM was concluded to be 47.07 percent.6 The mean one-year DLOM concluded in the VA 2001 pre-IPO study was 22.41 percent.7 However, the mean DLOM became 40.84 percent when a narrowed DLOM range of 10-90 percent was used. This narrowed DLOM range was considered in order to reduce the influence of (1) cheap stock or stock options and (2) price premiums due to changing stock market conditions.
WMA 1998 DLOM Conclusions
The mean and median WMA pre-IPO study DLOM conclusions for 1998 are generally consistent with the DLOM conclusions of our previously published studies. In 1998, the indicated mean DLOM was 35.0 percent and the indicated median was 49.4 percent. These DLOM conclusions are entirely consistent with the long-term (e.g., the decade prior to 1998) trends in prior WMA studies and the DLOM conclusions of other analysts.
WMA 1999-2000 DLOM Conclusions
The DLOM conclusions for 1999 and 2000 are significantly lower than the DLOM conclusions of our previously published pre-IPO studies. In addition, the DLOM conclusions for 1999 and 2000 are significantly lower than the DLOM conclusions reported by the VA study for the same years and the DLOM conclusions reported by John Emory for the period from May 1997 through December 2000.
Because the WMA DLOM conclusions indicated for 1999 and 2000 are lower than indicated in our previous studies, one could ask: "Is there a downward trend for the DLOM?" Although this conclusion is possible, we believe that there are several reasons why the DLOM results for 1999 and 2000 are atypical of both (1) long-term trends and (2) the conclusions of other analysts.
First, there were relatively few IPO companies and relatively few private sale transactions that qualified for inclusion in the WMA pre-IPO study in 1999 and 2000. The result of a DLOM study based on a larger data set of private sale transactions is likely to be more statistically valid. Second, 1999 and 2000 were not typical years with regard to stock market price trends. This conclusion is illustrated by Chart 1, the monthly NASDAQ Composite Index for the period of January 1993 through December 2002. The years 1999 and 2000 clearly represent a "spike" in stock market prices when compared to both prior years and subsequent years.
Second, the height of the dot.com "bubble" occurred during this time frame. The 1999-2000 period encompassed both a record number of dot.com IPOs and a record with regard to dot.com company pricing multiples. These pricing multiples are often unexplainable based on the dot.com company financial fundamentals. Of course, a large number of dot.com companies consummated IPOs around this time, and many of these dot.com companies subsequently failed.
Third, the average first-day returns for IPO stocks, measured from the offer price to the first-listed close price, were extraordinarily high in 1999 and 2000. These average first-day return data are presented on Table 2. One possible explanation for this return phenomenon is that many of the 1999-2000 IPOs may have been underpriced. There have also been allegations of IPO "spinning" on the part of certain investment banking (I/B) firms. In IPO "spinning," IPO "hot" offerings are allocated by the underwriter to corporate executives—so that the IPO shares could be sold or "spun" for a quick profit. The allegation is that such allocations were made by the underwriter in exchange for future I/B business. Furthermore, underwriters are alleged to have charged excessive brokerage commissions to investors to whom lucrative IPO share allocations were made and required customers to buy shares of the new stock after the IPO date at much higher prices. Due to all of these influences, the IPO share prices during 1999 and 2000 may not be the best indication of the intrinsic value of the subject stocks. To the extent that IPO prices were artificially low—allowing some underwriters and corporate executives to make excessive profits—the DLOM conclusions calculated in our 1999-2000 pre-IPO study may be understated.
WMA 2001-02 DLOM Conclusions
The WMA DLOM conclusions for 2001 and 2002 are also inconsistent with the conclusions of our previous pre-IPO studies. The mean DLOM conclusions for 2001 is -195.8 percent (indicating a DLOM price premium). This conclusion is completely counterintuitive both in terms of magnitude and direction (i.e., a price premium instead of a price discount). The mean DLOM conclusion for 2002 is 55.8 percent. This conclusion is abnormally high compared to the mean DLOM conclusions of, say, the decade period prior to 1999.
Accordingly, we recommend that analysts do not rely on the indicated DLOM conclusions for these two years 2001 and 2002 because (1) there are too few private market stock sale transactions in the study for the results to be statistically meaningful, and (2) the DLOM indications of the individual private market transactions are too extreme (both positive and negative) for the results to be practically meaningful.
For 2001, there were only two private market-sale transactions that met the selection criteria to be included in the WMA pre-IPO study. In the year 2002, there were only seven private market-sale transactions that met the selection criteria to be included in the WMA pre-IPO study. Due to the low number of private-market sale transactions included in the study in these two years, any extreme/aberrational prices from individual private sale transactions have the affect of disproportionately influencing the overall DLOM conclusions.
With regard to the selection of an appropriate DLOM, we recommend that analysts either:
- ignore the 2001 and 2002 results of the WMA pre-IPO study as not being representative of long-term trends with regard to the DLOM, or
- consider the 2001 and 2002 results of the WMA pre-IPO study only within the context of an analysis of longer-term five-year or 10-year period DLOM data (e.g., the consideration of the total private market sale transaction data with regard to the five-year period 1998 through 2002, as presented in Table 1).
Both industry-specific and general capital market events in the years 2001 and 2002 could explain the unusual results of our DLOM study for those years. The year 2001 included much of the impact of the "burst" of the "dot.com" bubble. This "burst" ended the several years of apparent "irrational exuberance" both in terms of (1) the volume of IPO transactions and (2) the inflated prices of dot.com and many other stocks. The year 2001 also experienced a significant decline in both (1) the volume of merger and acquisition (M&A) transaction activity and (2) M&A transaction pricing multiples. These unusual capital market events likely affected the DLOM conclusions with regard to 2001 and 2002.
Because of both (1) the unusual events in the stock markets and in the M&A capital markets in this period and (2) the aberrational quantitative conclusions of the 2001/2002 analyses compared to previously published studies, we believe that the quantitative conclusions for 2001 and 2002 are not reliable indicators of the current DLOM.
WMA 1999-2002 Conclusions Compared to Other Studies
For the period 1999-2002, the WMA pre-IPO DLOM conclusions differ from the DLOM conclusions of the Valuation Advisors and the Emory studies for two primary reasons. First, the selection criteria for inclusion of private-market transactions in the respective pre-IPO studies vary. Second, and more importantly, changes in subject company earnings and in the prices/earning multiples within the subject company industry (between the time of the private transaction and the IPO) are included in the WMA pre-IPO DLOM calculation. The DLOM conclusions in the VA and in the Emory pre-IPO studies are calculated by simply comparing the private transaction (pre-IPO) price to the IPO price with adjustment for changes in industry-specific stock market conditions.
Summary and Conclusion
This discussion focused on the concept of marketability (or the lack thereof) in business/security valuations performed for bankruptcy purposes. This discussion summarized the general types of empirical research that have been conducted with regard to estimating a market-derived DLOM to be used in the valuation of closely held investments. This discussion also presented the results of the most recent WMA pre-IPO DLOM study within the context of other pre-IPO studies.
With regard to our 1998-2002 DLOM study, the DLOM conclusions for 1998 are a mean discount of 35.0 percent and a median discount of 49.4 percent. These conclusions are consistent with our previous studies and with the results of other published DLOM studies. The DLOM conclusions for 1999 and 2000 are lower than we would expect based on our previous studies and on the results of other published DLOM studies. As explained above, there are several capital market reasons why the 1999 and 2000 results of our DLOM study are suspect. The DLOM conclusions for 2001 and 2002 are simply not reasonable. As explained above, there are several capital market reasons the 2001 and 2002 results of our DLOM study are unreliable. In addition to these reasons, the extremely small number of observations analyzed in 2001 and 2002 would suggest concern in the reliability of the DLOM conclusions for these two years.
Therefore, we recommend that analysts do not rely on the discrete results of our DLOM study for the years 2001 and 2002. Rather, we recommend that analysts rely on the collective results of the five-year period 1998-2002 as a reasonable indication of the DLOM for the contemporary period. As presented in Table 1, the mean DLOM for the five-year period 1998-2002 is 23.9 percent, and the median DLOM for the five-year period 1998-2002 is 36.1 percent.
In estimating the appropriate DLOM, an analyst should consider all of the facts and circumstances relevant to the subject business/security valuation. Based on the unique facts of a specific analysis, there are times when the restricted stock DLOM studies may be more relevant, and there are times when the pre-IPO DLOM studies may be more relevant. This is because marketability and lack of marketability are relative (and not absolute) terms.
The concept of marketability of a closely held security is often intertwined with the concept of ownership control of a closely held security. However, these two investment valuation concepts (while generally related) are distinct in both theory and practice. Consideration of how marketability and ownership control (or the lack of either of these two investment attributes) can affect value is an integral procedure in the estimation of business/security valuation discounts or premiums. The investment attributes of marketability and ownership control are fundamental considerations in valuations performed for bankruptcy purposes. However, these investment attributes should also be considered in valuations performed for other purposes.
1 The specific analytical procedures performed in the various WMA pre-IPO DLOM studies are detailed on pages 408-411 of Valuing a Business, fourth edition, by Shannon P. Pratt, Robert F. Reilly and Robert P. Schweihs, published in 2000 by McGraw-Hill. Return to article
2 Emory, John D. Sr., ASA, Dengel, F.R. III, and Emory, John D. Jr., "The Value of Marketability as Illustrated in Initial Public Offerings of Common Stock May 1997 through December 2000," Business Valuation Review, September 2001, p. 15-19. Return to article
3 Emory, John D. Sr., ASA, Dengel, F.R. III, and Emory, John D. Jr. "The Value of Marketability as Illustrated in Initial Public Offerings of Dot-Com Companies May 1997 through March 2000," Business Valuation Review, September 2000, p. 111-121. Return to article
4 Emory, John D. Sr., ASA, Dengel, F.R. III, and Emory, John D. Jr., "Expanded Study of the Value of Marketability as Illustrated in Initial Public Offerings of Common Stock May 1997 through December 2000," Business Valuation Review, December 2001, p. 4-20. Return to article