The Importance of Insider Ownership
NEW YORK, NY (September 24, 1999) -- As the number of companies issuing stock options soars, there has been an increasingly lively debate relating to how their cost is accounted for -- or, in most cases, not accounted for (for more on this topic, see the excellent five-part series by the Fool's Louis Corrigan, as well as Warren Buffett's critique in his most recent annual letter; click here to read excerpts).
Defenders of stock options argue that they align the incentives of companies' managers and employees with those of shareholders. Does this result in better stock performance? Intuitively, I think it would, so when I analyze a company I always like to see a high degree of insider ownership. However, since I have seen little data to confirm this assumption, I did my own analysis to fill the gap.
Using Value Line, I sorted the companies in the S&P 500 by insider ownership and then looked at 10-year compounded annual stock returns. The initial results were quite striking:
10 Year Returns
Click here to see the data behind this graph and the one below
Without exception, increasing insider ownership correlated with higher average stock returns. Great! Perfect confirmation of my assumption, right? Not so fastï¿½
Using Value Line's data for three- and five-year returns, I split the ten-year period into three periods by calculating for each stock the average returns for the past three years, the two years before that, and the five years before that (click here if you want to see an example of the math). Here is the same chart, with data added for these three periods:
Returns Over Different Time Periods
Click here to see the data for all 388 stocks.
Uh oh. Now the correlation doesn't look as strong, especially over the past three years and the two years before that (though the lockstep correlation remains intact in the period 6-10 years ago). What's going on?
Given the small sample size -- fewer than 50 data points in all but two of the cohorts --and how many factors affect a stock's price over time, I'm actually surprised to see any correlation (I'm sure a statistician would have apoplexy because I haven't run a multi-variable regression). And to the extent that there might be a correlation, I would have expected it to appear only over a long period of time, which is in fact what happened. Over a two-, three-, or even five-year period, a few stocks can show extremely high growth rates -- which can skew the average, especially with such small sample sizes -- but high growth rates are much less likely over a ten-year period.
It's also important to keep in mind that correlation does not prove causation. For example, it's been well-documented that years in which the AFC wins the Super Bowl, the stock market does much worse on average than years in which the NFC wins. Correlation? Yes. Causation? No. It's entirely possible that the correlation we see is due to the fact that the best-performing sectors of the economy over the past ten years (such as high technology) are also the ones where there are high degrees of insider ownership.
Yet there is one set of numbers that leaps out at me: while the differences in returns aren't that great among most of the cohorts (for example, the 10-year return for stocks with 20-29% insider ownership is only 3% more than stocks with 2-4% insider ownership), in every time period there is a huge jump in returns when insider ownership is 30% or more. This could imply three things:
- Insider ownership may only really matter when it's very high, or
- The key variable is not the degree of insider ownership, but rather whether a company's founders still hold large stakes (since insider ownership of 30% or more generally only happens when the founders hold on to their equity), or
- With only 15 data points, which include a few phenomenal performers such as Dell, Microsoft, and Charles Schwab, the sample size could be too small for the results to mean anything.
My conclusions: while these analyses are far from definitive, they give some reinforcement to my belief that insider ownership is a plus -- and more is better -- but modest differences are not going to influence my decision whether to buy one stock versus another. The data provides more support for my belief that large insider holdings, especially by founders, are a big plus. In fact, the weighted average insider ownership of stocks in my fund is 29%, which does not mean that insider ownership is the dominant factor in choosing the stocks I buy -- rather, it tends to correlate highly with many other criteria I consider.
Whitney Tilson appreciates your feedback at Tilsonfund@aol.com.
The Fool is hiring. Answer the call.