Introduction

This eleventh post in the Deja Vu series involving restricted stock studies addresses an issue that is rarely mentioned in the context of the studies – of the impact of dividends on restricted stock discounts (RSDs). Shareholder returns come in two forms, current income or dividends (or distributions) and capital appreciation, as can be seen in this simple graphic.

It is important to understand the impact of dividends on the value of illiquid minority interests of companies because many closely held and family businesses pay dividends. I recently found a 2008 version of the FMV Opinions Restricted Stock Database that I purchased back then on our system. It is similar to the 2004 version I used to write Deja Vu #8: Review of the FMV/Stout Restricted Stock Database. This 2008 version had information on 477 restricted stock transactions, up from 430 transactions in the 2004 version. All the additions were for 2004 and earlier years, so it is essentially the same database. There are 244 pre-April 1997 transactions in the 2008 version, down from 248 transactions in the 2004 version. Perhaps a few transactions were scrubbed.

The FMV/Stout database is the only restricted stock study that enables the analysis of dividends since this is one of the fields in the database. There is simply insufficient detail in all the other studies discussed in this Deja Vu series. Of these 244 transactions, only 24 involved companies that paid dividends, or less than 10% of the transactions.

There were 231 transactions after April 1997, when the SEC’s period of restriction was reduced from two years to one year (up from 182 transactions in the 2004 version). Only two of these companies paid dividends, rendering this portion of the database useless in analyzing the impact of dividends on restricted stock discounts.

Conclusions prior to the analysis: dividend paying companies:

• Had lower RSDs than non-dividend paying companies, with a median of 13.2% versus 21.9% for non-dividend payers. Given two otherwise identical investments where one pay a dividend and the other does not, the dividend payer should be more valuable (and have a lower RSD).
• Were also bigger and had higher market capitalizations and better operating performance, on average. So relatively more attractive companies that pay dividends tend to have lower RSDs than non-dividend payers. Excluding specific information on dividends, that was the conclusion of the Silber Study, which was published in 1997, or more than thirty years ago, and was the subject of Deja Vu #6: The Silber Restricted Stock Study (1981-1988 Transactions).
• With only 24 transactions involving dividends that occurred between 1981 and 1995, the Stout/FMV database is not useful to help estimate the impact of dividends on RSDs.

Dividend-Payers Have Lower Restricted Stock Discounts

We pulled the 24 dividend payers out of the pre-1997 portion of the database. This left 220 non-dividend-paying transactions in the two-year category, which we used as a basis to compare the dividend-paying transactions. What did we learn? We will look at the analysis in a series of tables to focus on specific aspects of the transactions.

From the first figure, we make a few observations. First and most obvious, look at the bottom right of the figure. The median RSD for dividend-paying companies was 13.2%, which is significantly lower than the median RSD for the non-dividend-paying companies.

• The transactions occurred between 1981 and 1995 (or many, many years ago).
• At least ten of the transactions involved banks, insurance companies, or other financial entities, which are very unlike the remainder of the companies in the non-dividend paying portion of the database.
• At least three transactions involved real estate companies. So financials and real estate account for more than half of the 24 transactions.
• Seven of the companies paying dividends traded on the New York Stock Exchange, while only three of the 220 non-dividend paying companies traded there.
• The median and average restricted stock discounts for the dividend paying stocks were about 13%, while the comparable discounts for non-dividend paying stocks were 22%-23% (bottom right of the figure).

Market Pricing and Restricted Stock Discounts

The next figure examines the offer amounts, size of the transactions, and market pricing of the dividend-paying restricted stock issuers.

We make observations from the second figure.

• The dividend-paying stocks are not penny stocks. The average stock price was $17.60 per share, while the average for non-dividend-paying stocks was $7.43 per share. Eight of the non-dividend paying stocks traded at less than $1.00 per share. Another 28 traded in the range of $1-$2 per share. 59 stocks traded from $2-$5 per share, and another 40 stocks traded between $5-$7 per share. This is not a hard analysis, but readers can get the picture that the non-dividend paying stocks were, for the most part, speculative businesses.
• The median offer amount (amount of the transaction) was $14.7 million for the dividend-paying stocks, while the median for non-dividend paying stocks was $4.0 million. The corresponding averages were $17.5 million and $10.1 million, respectively. We made a separate calculation excluding five large transactions involving Republic Industries (not shown). Excluding those transactions, the median and average discounts for non-dividend-paying stocks were $4.0 million and $6.9 million, respectively. The median excluding the five transactions remained the same; however, the average discount fell from $10.1 million to $6.9 million.
• The average and median percentage of shares owned by the restricted stock investors after the transaction was about 10.5% for dividend paying stocks, while these statistics for non-dividend paying stocks were about 11% to 13%, or not too dissimilar. However, we just saw that the dividend paying stocks received more dollars from their offerings than did the non-dividend paying stocks.
• The median market value of equity (MVE) for dividend-paying stocks was $133 million, and the average MVE was $248 million. For non-dividend paying stocks, the median MVE was $46 million, while the average was $122 million. Excluding the five transactions involving Republic, the average MVE fell to $74 million. The non-dividend paying stocks were considerably smaller than the dividend paying stocks in terms of MVE.
• The differing natures of the two groups of transactions can be seen when looking at the price/book value multiples. The median P/B multiple for dividend paying stocks was 1.7x, while the average was 2.2x. For non-dividend paying stocks, the median P/B multiple was 5.9x, and the average P/B multiple was 27x, indicating a range of speculative pricing for the non-dividend payers.

Earnings, Risk and Dividends, and Restricted Stock Discounts

We examine earnings, risk factors, and dividends in the third figure.

In this final figure, we look at other aspects of performance and risk for the dividend-paying companies relative to the non-dividend paying companies. We make a number of observations.

• Dividend paying companies are larger in terms of revenue than non-dividend paying companies. Median revenues for the dividend-payers was $118 million, compared with only $10.5 million for non-dividend payers.
• The dividend-paying companies are much more profitable, on average than non-dividend paying companies. The median pre-tax income for dividend-payers was $9.3 million, with a median operating margin of 22.5%. The median pre-tax income for the non-dividend payers was a loss of $1.2 million.  
• Stock price volatility for the dividend-paying companies was 0.37 (median) versus 0.73 (median) for the non-dividend payers. The smaller, less profitable companies had higher volatilities than did the larger, more profitable dividend-payers.
• Finally, The dividend paying stocks had a median dividend of 2.3%, while the non-dividend paying stocks had a yield of 0%.

Conclusion

We started with a figure indicating that the total return for an illiquid (or any) investment is the sum of the expected dividend yield and expected capital appreciation. For illiquid investments, these expectations are over the reasonably expected holding periods for investments. It is the expectation of future returns that give present value to investments. We have just seen very little information about dividends in the Stout/FMV Restricted Stock Database. The available information suggests that the presence of dividends, larger size and better performance yielded lower RSDs for dividend-paying companies relative to those who did not pay dividends.

It would appear that the expectation of future dividends has an inverse relationship to marketability discounts. Some level of dividends in an investment would mitigate the marketability discount for otherwise similar investments with no dividends. The expectation of larger future dividends relative to smaller future dividends would mitigate the marketability discount for the higher dividend-paying investment. This is just common sense.

Unfortunately, in large part, the business appraisal profession has ignored this common sense when estimating marketability discounts since Shannon Pratt reported on four restricted stock studies in his first (1981) edition of Valuing a Business (link is to 6th edition).

Any study or analysis that does not directly consider the impact of expected future dividends on the value of illiquid minority interests of companies is flawed. Readers who have made it this far are directed to Deja Vu #10: Valuation Theory is the Same for Businesses and Business Interests: V =f(CF, G, and R). A more in-depth resource would be Business Valuation: An Integrated Theory Third Edition (Mercer and Harms).

I’m going to close with something of an analogy. Estimating the value of illiquid minority interest using restricted stock studies (i.e., estimating the marketability discount) is like estimating the value of a business without considering the expected future income stream. Refer to the initial figure. There is no information in any restricted stock study to help business appraisers estimate the value of expected future dividends. And what about the terminal value that gives rise to capital appreciation? That is also an expected future cash flow. It is time for many appraisers to rethink their methods for developing marketability discounts.