Except that the big-brand CEOs don't always pan out, and the compensation paid is sometimes not worth the results. Also, because the performance compensation packages are usually based on short-term results, there is an incentive there for a CEO with a big performance package to goose the short-term results at the expense of the long-term results in order to maximize his/her compensation at the expense of the long-term shareholders (and other stakeholders) in the corporation.
One of the culprits here is, IMHO, the tax code, which overtaxes corporate profits distributed to shareholders as dividends and, comparatively, undertaxes the gains a shareholder realizes by selling his/her shares. As a result, fewer corporations pay dividends than would otherwise do so, and more corporations undertake activities designed to boost the share price, even though that can only be done for a limited subset of the universe of shareholders because boosting the short-term price usually comes at the expense of the long-term price. And one of the ways this is done is by bringing on board a big-name CEO who is touted as being able to, essentially, drive up the value of the stock. That also over-incentivizes the use of stock option and restricted stock compensation because it ties pay to the rise in stock value as opposed to, say, the amount of net profit that could be dividended to stockholders. Measuring a CEO's worth at the time of hiring by his/her likely effect on future stock values - a promise that can always be pushed off to "next year" if this year proved disappointing - is a much more speculative way of compensating the CEO than basing comp on actual annual returns to the shareholders as such (and not as market participants in the securities markets). A CEO either did, or did not, drive the corporation to earn enough to pay a dividend of a given amount - there isn't much guesswork there at the end of the year - but a CEO who didn't quite meet the stock price goals for this year can still claim that, over the long term, he (or she) will eventually pull that stock price up and what do the shareholders care anyways because the stock price really isn't that meaningful until a shareholder goes to sell, and that is both wholly up to the shareholder and something that each shareholder does on his/her own, without reference to what every other shareholder is doing. Dividends, by contrast, are not under the control of the shareholder, accrue in determinate periods - annually - and are received by all shareholders at the same time, i.e., collectively rather than in isolation.
This is, essentially, the MO of many so-called "growth" stocks - the promise is that the value of the stock will grow, not that the net profits to shareholders will grow. That might seem a little odd, but it isn't when one takes into account that the value of a stock in the market is as much a matter of psychology as of underlying fundamentals; because of that, a stock's value will usually include a psychological discount or premium not attributable to the actual net earnings of a corporation, which if retained would increase the value of the stock because those retained earnings would either be dividended out or would be reinvested in such a way that the liquidation value of the corporation increased.
For example, assume that the value of a share of X-corp stock is, solely on economic fundamentals, worth $100 today. If X-corp has retained earnings of $10 per share at the end of this year, the value of a share of stock should rise to $110 solely on economic fundamentals, because the retained earnings represent cash funds that can either be dividended out - i.e., dividends that did not exist at the beginning of the year - or else reinvested in the business in such a way that, based on rational economic forecasting, will produce a return on investment commensurate with a present value of $10 per share. However, once the psychology of the market comes into play, so-called behavioural finance - things can come unmoored from rational economic fundamentals. Apropos to the subject, if a corporation spends that $10 per share hiring a Jack Welch whom everyone believes will end up raising the price of the stock in a few years, that stock will continue to trade - to have a value - at something greater than $100 per share, even though that $10 per share hasn't been reinvested in the business in the conventional sense, but simply paid out as compensation for services (to be performed) by a CEO who may or may not perform and who, if he fails, will leave the corporation without anything to show for its investment - the typical reinvestment into the business could at least be expected to leave the corporation with some capital assets that could be sold off if the new investment fails, thereby recovering at least some of that $10 per share that was invested in the new line of business. By contrast, if that hypothetical Jack Welch fails to perform as promised, none of that $10 per share that was "invested" in buying his services can be recouped.
Because the tax system distorts the shareholder's choice of how to realize a return on her investment in a corporation by favoring stock price growth over retained earnings reinvested or dividend growth, the tax system tends to incentivize corporations to spend too much (from a pre-tax economic perspective) on hiring "talent" that will boost the stock price rather than just the underlying earnings fundamentals of the corporation.
I believe that the qualified dividend regime, under which some dividends are taxed at capital gains rates, has shown some evidence of this because there are some more corporations paying out earnings as dividends than before; however, because a dividend is still effectively taxed to the corporation - because the corp doesn't get a deduction for dividends paid to its shareholders - while compensation paid to a big-name CEO is not taxed to the corporation because it can be deducted for tax purposes, the tax system still incentivizes shareholders to look for investment returns from increases in the market value of their stock and disincentivizes them to look for investment returns in the form of dividends or reinvestments of earnings in capital assets and businesses, and that in turn means that the tax system still over-incentivizes corporations to pay outlandish compensation in the vicious game of trying to hire the biggest and baddest brand-name CEOs.
Remove the distortions caused by the tax system and you might see CEO compensation start to realign with the real performance of the underlying corporate business and less with the market performance of the stock price.