In many companies, decisions about the level and timing of share repurchases are left mainly to the management. But given their importance, corporate directors should probably pay closer attention.

Capital allocation is a significant function for company directors. How much of the company’s profits gets reinvested in the business rather than distributed to shareholders through cash dividends or share repurchases is a critical decision companies must make. Boards of directors typically approve a dividend policy and precise amounts for each quarter: Everyone knows that cutting the dividend will result in a sharp decline in the share price.

Yet in many companies, decisions about the level and timing of share repurchases are left to management. That stems partly from differences in legal requirements: The board must formally approve the amount of the company’s quarterly dividend but not its repurchases. Moreover, the implementation of the repurchase program is heavily influenced by the company’s actual cash flows.

Nevertheless, share repurchases are something to which directors should pay more attention. Specifically, directors should carefully consider the capital allocated to repurchases relative to the company’s realistic opportunities for value creation through internal development or external acquisitions. They should be highly skeptical of large repurchase programs that are financed by selling debt rather than paid for out of company profits.

From 2014 through 2016, distributions to shareholders — dividends and repurchases together — consistently exceeded 100% of the net income of the companies in the S&P 500. During the same period, share repurchases for the Russell 1000 companies (excluding financial and real estate companies) ranged from 62% to 71% of the free cash flow (net income minus capital expenditures). These trends seem to reflect a slowly growing global economy, together with the availability of very cheap debt. According to ASR Research, roughly half of all share buybacks were financed by debt rather than profits.

Share repurchases are sometimes justified as a way to maintain and increase a company’s share price. However, this view is not supported by the data. The 100 companies with the highest buybacks in the S&P 1500 underperformed that index from 2005-2016.

Why have these companies underperformed? First, sophisticated shareholders know that share buybacks increase earnings per share (EPS) by spreading the same amount of revenue over a reduced number of shares. To these shareholders, buybacks are seen as a form of financial engineering for companies with weak growth prospects.

Second, executives are notoriously bad at timing their share repurchases — they do a lot of buying when the company’s stock price is high and relatively little when the price is low. That’s why, between 2004 and 2016, companies reduced their share count by roughly 25% but increased their EPS by only 12%.

Third, share repurchases reduce the relative market cap of companies in market-weighted indexes such as the S&P 500. As a result, the giant index funds based on the S&P 500 are effectively forced to rebalance by selling the stock of companies with large repurchase programs.

So how should directors evaluate various uses for a company’s cash flow? To begin with, some buybacks are quite sensible. For example, directors should endorse share buybacks sufficient to fund the company’s plans for stock options and restricted shares for employees. Buybacks at this level would minimize the dilution effects of such plans on the company’s public shareholders. Similarly, directors should support capital expenditures necessary to maintain the company’s asset base. In addition, companies should have enough cash on hand to cope with the vagaries of the business, especially if share repurchases are financed from new debt rather than current profits.

Once these priorities are taken care of, directors must address the issue of capital allocation. Does the company have internal products or research projects that are likely to deliver a return that’s higher than its cost of capital? Alternatively, if the company makes a significant acquisition, will the additional revenues and earnings over time justify the deployment of cash debt capacity? These are the questions that we hear more and more from the large institutional investors such as BlackRock and Vanguard, which hold a majority of the shares in most large, public U.S. companies. Given the difficulties these large investors have trading in and out of big blocks of stocks, they tend to be more interested in long-term value creation than brief run-ups in a stock price.

Nevertheless, some boards capitulate to activist investors on share repurchases without polling their long-term shareholders. For example, having bought less than 1% of the stock of General Motors in 2015, investor Harry J. Wilson persuaded GM’s board to invest $5 billion in cash to repurchase shares. Although the share buybacks didn’t seem consistent with long-term value creation (particularly for a company that had recently emerged from bankruptcy), shareholders never got an opportunity to vote on this proposal.

Share repurchases are too important to be left to the discretion of company management.

Boards should set the level of annual repurchases after carefully considering the internal and external opportunities available for the company’s capital as well as the objectives of its long-term investor base. To the extent feasible, directors should support corporate strategies to increase the company’s revenues and profits over several years, and they should look askance at share repurchases that are financed by debt to maintain EPS in the next quarter.