Share buybacks do not undermine long-term shareholder value.
Democratic Presidential candidate Hillary Clinton does not like share buybacks. Her criticism is based on the argument that share buybacks take cash out of the company that could have been used to invest in growth and employees. She is advised by William Lazonick, an academic who wrote a highly critical article on buybacks in the Harvard Business Review. The critique is that U.S. firms deliberately underinvest in order to increase earnings per share (EPS), which other critics say undermines long-term competitiveness and even generates negative excess returns in the long run.
Note that EPS can be increased in two ways: the hard way by increasing earnings or the easy way by reducing the number of shares outstanding.
Why would managers care about EPS? The first reason is because they want to make sure not to miss analyst forecasts. Missing analyst forecasts may hurt stock prices significantly. Buybacks are a short-term stock price manipulation scheme. The second reason is that managers’ bonuses are tied to EPS targets. The fallacy of earnings per share targets is nicely explained in a McKinsey article by Obi Ezekoye, Tim Koller and Ankit Mittal. If you borrow money to buy back stock, your expected EPS will rise but the increase in leverage will increase the risk, so that the present value of expected EPS remains the same.
Paying bonuses based on EPS targets is a bad idea. But that does not prove that buybacks are driven by EPS manipulation. A recent survey of Board members on buybacks by the Investor Responsibility Research Center (IRRC) dismisses this concern by arguing that EPS targets in compensation schemes are adjusted for buybacks (although they admit that this information is often not publicly available). The manipulation story also does not explain why firms that lose money – such as the firms in the oil and gas industry last year – buy back stock, as in this case a repurchase increases losses per share.
Nowhere else for the cash to go
The most serious challenge to the buyback revolution is that firms would undermine their long term competitiveness by buying back stock rather than investing in good projects. Note the emphasis on “good” projects. If a firm has no good (positive net present value) projects, it should give the cash back to shareholders who can then invest it in other companies that have good projects and need cash. It is not because you earned a lot of profits this year that you also have great investment opportunities. Some direct evidence that share buybacks don’t undermine economic growth is provided in another McKinsey study. Simply investing for the sake of job creation may well be consistent with stakeholder value maximisation ideology, but is a violation of fiduciary duties of board members, at least in the U.S. The claim that investment suffers because of buybacks is also dismissed in the IRRC survey mentioned earlier: You can do both, especially in a world where interest rates are low and intense competition is reducing the supply of positive net present value projects.
Here the prediction of the anti- buyback crowd is that, although a buyback generates positive abnormal returns in the short run, it will generate negative excess returns in the long run. Testing this hypothesis is challenging because in order to measure “abnormal” return you need a model of “normal” or expected returns. It is the return you could have expected if the firm had not done a buyback. During the last 30 years, the academic finance profession has moved away from the Capital Asset Pricing Model which assumes investors only want to be compensated for market risk (beta) to multifactor models. The most recent model is the Fama-French (2015) five-factor model which assumes that expected returns are driven not only by beta but also by firm size, market-to-book ratios, profitability and investment. This model explains expected returns better than any of the previous models used in the literature.
In a recent paper we use this five-factor model to test whether open market buyback authorisations announced in the U.S. between 1985 and 2015 are followed by negative long-term excess returns. We find the opposite results: On average, companies that announce open market buyback programmes earn significant positive excess returns of around 12 percent after four years. Not all buybacks are the same: Excess returns are larger for small stocks, value stocks and stocks that are beaten up in the previous six months, a result also reported in an earlier paper by Urs Peyer and Theo Vermaelen who combine these three indicators in an undervaluation index (U-index) that is a good predictor of excess returns. The interpretation is that managers, on average, are able to buy back shares when they are undervalued and they are better at market timing in high U-index firms.
In this recent paper we add two additional indicators that are positively related to long-term excess returns: volatility and company-specific (idiosyncratic) volatility. Highly volatile stocks are more likely to be mispriced, making the information advantage of the firm management potentially larger. Second, managers are more likely to have an information advantage when the information is company-specific, i.e. unrelated to general market movements. We combine these two indicators with the U-index in an extended undervaluation index (EU index). Figure 1 shows the long-term excess returns of six portfolios ranked on the basis of the EU index.
Figure 1: Cumulative Abnormal Returns from six months before until 48 months after the announcement of the buyback program for seven portfolios ranked on the basis of the Extended Undervaluation index (EU index)
The portfolio with the highest EU index earns excess returns of more than 70 percent after 48 months. No portfolio generates negative excess returns, in contrast to the claims made by opponents of buybacks in the non-academic press. The claims are typically “supported” by cherry picking specific companies that did a buyback and then underperform over a short time horizon. Of course, no one claims that all buybacks are followed by excess returns. You have to consider a portfolio, properly adjust for risk and use a long observation period. The critique that such a long observation period obscures the fact that in recent years the buyback anomaly has disappeared is inconsistent with the performance of the PV Buyback USA fund that was started in May 2011 by Peyer and Vermaelen. The fund invests in high U-index buyback stocks. The fund earned 80 percent since inception, received a 5-star Morningstar rating and is currently ranked among the top 5 percent of all 800 small-midcap U.S. funds looking back one, three and five years according to Citiwire Selector.
So why this hostility to this increasing common corporate finance activity? We believe it is not about concern for long-term versus short-term shareholder value, but because buybacks are the ultimate shareholder friendly event that may not be appreciated by other stakeholders. Managers, consulting firms and investment banks would make more money if the cash is used to make an acquisition, for example. Bondholders and workers don’t like the fact that a buyback increases risk. The government may not like the fact that buybacks financed with debt reduce corporate taxes. So Clinton may be right that share buybacks are good for shareholders but bad for others. But there is no evidence for the argument that repurchases undermine the long-term viability or competitiveness of U.S. firms or that buybacks sacrifice long-term shareholder value at the expense of short-term stock price and earnings per share manipulation.
Theos Evgeniou is a Professor of Decision Sciences and Technology Management at INSEAD and is the Academic Director of INSEAD eLab.
Theo Vermaelen is a Professor of Finance at INSEAD and the UBS Chair in Investment Banking, endowed in memoriam Henry Grunfeld.