In a prior article we summarised some of the possible consequences of requiring CEOs to own shares. We also suggested alternative methods that may be considered in an August 2005 article HERE. In that article, we analysed situations when a shareholding policy may be considered. This time, we report on original research on the extent of CEO share ownership and Australian company performance over 1, 5, and 10-year periods.
The results are conclusive. There is, indeed, a strong relationship between CEO share ownership and shareholder returns. But before you rush out to invest in companies where the CEO has a big stake, or dumping companies where there is not the case, or consider introducing your own company’s CEO shareholding requirement policy, we urge you to read our research. There are differences between industries, size of companies and extent of shareholdings.
In recent times, there was concern with the level of non-compliance in the timely reporting of changes in directors’ interests to the ASX. There are very sound reasons for this. Excellent US research found that there was a high probability that if a CEO sold shares then share price would lose significant value shortly thereafter. Conversely, if a CEO purchased shares then there was a high probability that share price would soon increase significantly in value. Guerdon Associates would love to replicate this research for Australia. Sadly, we still lack confidence in the extent of compliance and hence data reliability to conduct a valid study.
In other recent research Guerdon Associates have established that both fixed and at-risk remuneration are more strongly related to market capitalisation than to any measure of performance. That article can be located HERE.
In this study we wanted to examine whether the size of a CEO’s shareholding produced any discernable difference in company performance over time.
We extracted remuneration and shareholding information – as reported in the most recent annual report – for CEOs in the GuerdonData® database. Companies with a market capitalisation of less then $100 million were excluded, as were executives who had only served part of the reporting year.
We valued share holdings in the following three ways:
- the dollar value of the holding (we multiplied the holding by the share price as at May 15, 2006);
- the value as a percentage of Total Fixed Remuneration (TFR); and
- the value as a percentage of market capitalisation (as at May 15, 2006).
In all, we identified 230 CEOs of whom 185 (80%) held shares in the company they manage.
We analysed the group as a whole and also broke the sample down by industry and company size.
Our analysis relied heavily on correlation statistics to establish whether a relationship exists and its strength. Of course, correlation does not necessarily imply causation, but it is difficult to imagine that a large share holding would not influence CEO‘s behaviour. The questions we wanted to answer were:
- How much do CEOs own?
- Is the size of the holding a function of company size or remuneration?
- Is there a relationship between share holdings and performance?
- How strong are these relationships?
- Is the relationship stronger for short or long term performance?
- What other variables influence the relationship?
How much do CEOs own?
CEOs own a surprising amount of the companies they manage. The pie chart below breaks down the amount of value CEOs have tied up in the companies they manage. Only 20% of CEOs do not own any of the company they are employed to manage. But a significant 66% have $500,000 or more of their wealth tied up in company shares. 13% percent own more than $50 million in shares of the company they manage.
The table below breaks down the statistics of share ownership in more detail. It shows not only the statistics for the actual dollar value of ownership, but also ownership as a % of their fixed pay and ownership as a percent of the total value of the company.
Value of Shareholding and Company Size
Overall, there is no correlation between the value of the CEO’s shareholding and the market capitalisation of the associated listed company. That is, the CEO of a small company would be just as likely to have a million dollars tied up in company shares as the CEO of a larger company.
The same held true when we looked at how much of a CEO’s wealth was tied up in the company as a percentage of his/her pay. That is, the CEOs who held at least one year of their fixed pay in company shares were just as likely in the smaller company as the larger company.
However, there was a negative relationship between the shareholding as a percentage of market capitalisation and market capitalisation. That is, larger company CEOs owned less of the company than smaller company CEOs. This is not surprising since CEOs are more likely to own a greater proportion of the company if it is small than if it is large.
Overall there is a weak negative relationship between remuneration and the value of the share holding. The strongest relationship is between the value of the holding as a percentage of market capitalisation and CEO TFR. The correlation was -0.24. This suggests that the greater the proportion of the market capitalisation of the company that is held by the CEO, the lower his/her TFR is likely to be.
Individuals who own large proportions of the company stock are often company founders. They have a significant amount of personal wealth invested in the company and will sometimes forego market related levels of fixed pay in order to achieve the company’s goals. We have examined several annual reports that state this explicitly.
Ownership and Performance – Managing for the Long Term
There is a significant relationship between the size of the shareholding and long-term performance. The relationship between shareholding and long term company performance is stronger than the relationship between shareholding value and short-term performance.
This suggests that sound long-term decisions are more likely if the CEO has a substantial holding in the company.
The correlation between the dollar value of the stock held and ten-year TSR performance is 0.46. This contrasts with a weak correlation of 0.10 with one-year TSR performance. The following two graphs illustrate the different relationship between the average ten-year TSR and one-year TSR as the dollar value of the shares held increases.
The following graph illustrates this trend, regardless of how we measured the size of the holding. Note that the Y-axis represents the correlation (or strength of the relationship) between the holding and TSR performance, not the actual TSR performance level as in the prior graphs.
The correlation between the dollar value of the holding and ten-year TSR is 0.46. This produces an R-squared statistic of 0.21. In other words, 21% of the variability in the ten-year TSR for the 185 companies can be explained by the size of the CEO’s shareholding.
Figure 5 indicates the correlation between the dollar value of the shares held and the one, three, five and ten-year TSR performance. (Some of the industry groups have been excluded because the sample sizes became too small for the correlations to be meaningful.)
Findings by Industry
The following table illustrates the variability of share holdings by industry.
When the data was broken down by industry the relationship between share holdings and company performance changed.
The Health Care, Industrial and Material sectors have a clear relationship between share ownership and long-term (ten year) performance.
The Financial sector relates most strongly with medium-term (five year) performance. This is diverse group of companies including banks, insurance companies and property related companies.
The Energy and Consumer Discretionary industries appear to relate most strongly to short-term (one year) performance.
With the exception of the energy sector, this is almost exactly what we would expect given the event horizons of these industries. That is, mining, industrial and healthcare have longer investment cycles than financials, which have longer investment cycles than the consumer sector.
Findings by Market Capitalisation
Although we found little or no correlation between the size of the shareholding and the market capitalisation of the company, we did find that the strength of the relationship between shareholdings and TSR varied as market capitalisation changed. Additionally, the average TSR tended to rise as the market capitalisation rose. That is, the bigger the stake and bigger the company, the higher the shareholder returns.
The following graph illustrates how much stronger the relationship is for very large companies than for smaller companies. This is true regardless of how we measured the size of the holding.
The correlation between the dollar value of the share holding and ten year TSR for companies with market capitalisation over five billion is 0.7293. This translates to an R squared statistic of 0.5319. In other words, 53% of the variability in ten year TSR for very large companies can be explained in terms of the size of the CEO’s share holding.
Overall having a CEO with a substantial holding in company shares will likely result in higher shareholder returns over the longer term. This is particularly true for larger companies.
However, it becomes problematic if we infer that the share ownership is causing the TSR performance to be higher.
Most CEOs in the sample would receive some form of share-based remuneration. These grants are increasingly being tied to company performance, particularly relative TSR. As a consequence of this, we would expect that CEOs in the highest performing companies will receive more share value than other CEOs. We would also expect that well performing company shares, or shares in a company about to realise long term potential, are less likely to be sold than under performing shares, or a company with limited potential. It may be that the performance is causing the share ownership, not the other way around.
It is probably safe to say that, to some extent, both of the above are true. That is, share ownership contributes to company performance and company performance contributes to share ownership.
The implications for remuneration policy are both significant and complex. Should a board require a CEO to hold company shares? If such a policy is standard policy, it may assist in the “self selection” process for the CEO role when it is vacant. That is, potential recruits who have more confidence in the match between their abilities and company potential would embrace the policy, while those that harboured doubts may withdraw their interest. The outcome may be better for both the CEO and shareholders.