From Capital Returns (the excerpt below was from a Marathon letter in February 2012):
While the increasing influence of the proxy advisory services is welcome, the prescriptive, rule-based approach of these organizations does not suit every case, particularly when it comes to executive remuneration. What is the optimal incentive scheme, then? The answer is it depends on the circumstances. Remuneration structures based on earnings per share (EPS) growth and total shareholder return (TSR) performance measures are increasingly commonplace. Yet they suffer from the problem identified long ago by the management guru, Peter Drucker, who observed that the search for the right performance measure is “not only likely to be as unproductive as the quest for the philosopher’s stone; it is certain to do harm and to misdirect.” This is particularly the case when pay is linked to EPS – a particular bête noire for Marathon over many years.
The earnings per share measure is prone to manipulation by unscrupulous executives; it takes no account of risk and encourages value destroying acquisitions and buybacks, especially when interest rates are low. It also encourages the quarterly EPS guessing game beloved by the sell-side. At times, it seems that meeting the EPS target has become the main strategic purpose of the company. This is regrettable. Corporate strategy should be about how best to allocate resources. If a turnaround requires a three-year investment phase, management may not pursue the optimal business plan if their compensation is linked to interim EPS results. While these inter-temporal issues can be partly resolved by phasing in performance rewards over a period of years, investor myopia and management’s own interest tend to lead to an exclusive focus on the calendar year EPS, which bears no relation to long-term value creation.
Linking compensation to total shareholder return (TSR), the most common share price-based measure, is better than EPS, as it forces management to think about what drives shares prices over the medium term. Such schemes suffer from point-to-point measurement, which can be distorted if the stock price at either the start or end date is inflated by takeover speculation or by general overvaluation in the stock market. Then, there are questions over what time frame to measure the returns; also, whether the benchmark should be absolute or relative – both have their merits, neither is perfect. In the case of relative schemes, should the benchmark be provided by a peer group or by the broader market index? Sir Martin Sorrell, the head of advertising giant WPP, has become a very wealthy man thanks to his ability to outperform a small group of marketing service companies. Unfortunately, this wealth creation has not been shared with the company’s owners due to the under-performance of this sector over many years.
For this reason, we normally prefer corporate incentives schemes to be benchmarked against the stock market index, in line with our own performance fees. Company managers might feel aggrieved that they have no control on performance relative to a broad index, which may be driven by moves in some heavily-weighted sector, such as mining or pharmaceuticals in the FTSE 100. Some companies have come to us seeking to switch from a relative TSR scheme to an absolute one – often after a period of relative outperformance which presumably management believes will end imminently.
As regards the time frame over which performance should be measured, here one runs into the problem of investor myopia. Since the average holding period for European shares is down to 12 months (see Chart 3.2), the “average” investor has little interest in the performance of a company over a five-year period. We prefer longer measurement periods, with multiyear phasing in of benefits to encourage long-range strategic thinking. The views of high frequency traders and investors obsessed with quarterly EPS should be given a very low weight by management. Time frames may also need to vary by sector. In the capital goods and extractive industries, project terms may be well in excess of five years (for aero engines, product life cycles can be decades).
Given that each measure has pros and cons, it is not surprising that remuneration consultants seek a compromise, bundling together a mixture of measures in the incentive scheme. But so-called “balanced” approaches, such as those which mix an EPS target with a return on capital overlay and a TSR override, are likely to confuse both management and investors and, even worse, can encourage sophisticated gaming strategies.
Insider ownership has always seemed to us as the most direct way to deal with the principal-agent problem, which arises with the separation of corporate management from ownership. Our portfolios have tended to be skewed towards companies where successful entrepreneurs run their companies and retain sizeable shareholdings. Pleasingly, a number of companies have followed the example set by Reckitt Benckiser, where executives are required to build up significant shareholdings. Along similar lines, HSBC has recently revamped its incentives so that generous deferred share awards, which vest after five years, have to be held until retirement. Long-term share ownership is probably the best way of concentrating the minds of management on the true drivers of value. The manager’s instinct for wealth protection should guard against excessive risk-taking, the unfortunate counterexample of Lehman’s Dick Fuld notwithstanding.