If that is your answer, however, you would be wrong. A stock price is simply the consensus of investor expectations about the future performance of the company, and linking compensation to it is an incentive for executives to focus more on raising investor expectations than improving actual performance.

While we might imagine that real performance drives expectations of future performance, the link is exceedingly tenuous.

Just ask the executives of Microsoft. Last month, it reported a blow-out fourth quarter with sales up 22 per cent and profits up 45 per cent. The stock? It jumped a mere 2.9 per cent on the announcement. And that wasn’t because there had been a recent big run-up on the stock in expectation of a strong quarter. With a few brief exceptions, Microsoft has traded (adjusted for splits) in a narrow range between $20 and $30 per share for the past 10 years. During that decade, revenue and profit nearly tripled but the stock has remained flat. Had a dutiful executive been given a generous grant of 100,000 options on January 2 2001 at $21.69 and held on to them to today, the executive would be able to exercise those for a profit of a mere $440,000 (at the current price of about $26 per share) – after 10 years of hard slog to triple the company’s real performance.

Of course, this is one example and there are examples all over the map – including those in which expectations track real performance exactly. But that is precisely the point: the relationship between real performance and expectations is all over the map.

The only way an executive can be sure to realise a return from the incentive compensation provided is to work first, foremost and directly on raising expectations from the current level – the only thing that makes a stock price rise – often at the expense of improving the actual underlying value and performance of the company. There are much easier ways to accomplish that objective than working for a decade to triple the revenue and bottom line. It is much easier to go to the City and hype your stock. Or change your accounting treatment to appear to produce a jump in performance. Or make stupid acquisitions to appear like a fast-growing company.

Stock-based compensation was originally conceived as a way to align the interests of senior executives with those of the shareholders. Interestingly, it has created a wonderful alignment between segments of each: bloody-minded executives and hedge fund investors. Both profit most from expectations volatility. A bloody-minded executive bent on doing whatever is necessary to maximise stock-based compensation earnings will happily drive down expectations in order to get more low-priced stock compensation (whether options, stock or phantom stock) and then drive expectations back up to realise huge gains, then repeat the process until fired. Hedge funds, meanwhile, make all their money from volatility – the rise and fall of expectations – so they are totally aligned with and actually help out the bloody-minded executives in producing and profiting from volatility.

One might ask, what is the harm in all of this: a little hyping, some accounting hanky-panky, a few acquisitions that might not have been needed, executive compensation going through the roof, hedge funds making extraordinary profits. The problem is that this is a short-term game. Expectations cannot be made to rise forever – Jack Welch was the last chief executive to master that ancient art form – so executives need to raise expectations as precipitously as possible, and then simultaneously get out and cash out.

More than anything else, stock-based incentive compensation is responsible for short-termism in the modern corporation and the shrinking average tenure of today’s chief executives. It is an incentive for manipulating expectations rather improving real performance.

The solution is to replace stock-based compensation with incentives that affect underlying value – whether that is increasing revenues, profitability, market share, customer service or, optimally, a combination of all of these. And for longer-term incentives based on the actual market not the expectations market, use royalties on real results, as are given to designers, inventors and musicians. The bottom line is that if you want to skew reality, use stock-based compensation. But if you want to build the real company, use incentive compensation anchored in reality-based measures.

Roger Martin is dean of the Rotman School of Management and author of a forthcoming book on the dangers of stock-based compensation, to be published by Harvard Business School Press