An op ed piece in today's Financial Times claims that share repurchases destroy value because they are often done at market peaks. In other words firms are paying too much when they repurchase stock.
Unfortunately, the article is a little sloppy in that it only looks at aggregate data and then only considers a very short time period. More importantly though, I think that the article draws the wrong conclusion.
When we look at the the academic evidence on share repurchases we find support for firms timing their repurchases for periods when the firm's stock is low relative to the stock's fundamental value. For example, a paper by Ikenberry, Lakonishok and Vermaelen in 1995 finds that value stocks in particular seem to do very well after repurchases - consistent with managers timing the repurchase when stock prices are low. The opposite result is usually found for share issuances - these usually occur when the firm's stock value is high and subsequent returns are low. Both of these are manifestations of the "Market Timing" theory of capital structure - something that I've done a bit of research on. In general there is pretty strong evidence that managers time stock issuances and share repurchases to benefit long term shareholders.
I'm also a little concerned about the idea that a share repurchase could destroy value. First of all, firms are awash with cash - cash that is just sitting in corporate coffers - and share repurchases are an efficient way of getting that cash back to the shareholders. Second, even if the firm buys back shares when the prices are high, this doesn't destroy value per se. All it does is transfers wealth from the shareholders not participating in the buy back to those who are participating.
So while it might be the case that share repurchases may not have been optimally timed in the past 2-3 years, the larger evidence doesn't support the article's contention.
The article does make an interesting point however, that managers may be tempted to boost EPS by buying back shares. They would do this because they are compensated on the level of EPS. Surprisingly, many managerial compensation contracts are often pretty vague and may not preclude manipulation of EPS via repurchases thus opening the door for such window dressing. However, if you buy stock in a company that uses such a poor performance metric to compensate the managers, then you probably deserve what you get!
For more on this topic - you can see an earlier post here and Damodaran's excellent post on the valuation effects of buybacks.