A recent article in the Economist suggests that managers with exposure to their firm’s share price would benefit by reducing dividends and instead buying back shares. From a strictly economic point of view, we don’t see the difference.
A Preoccupation with Share Price
The article points out that by repurchasing shares, the outstanding share price will increase, as the market capitalization of the firm is divided among fewer shares. This is true. The article further suggests that shareholder managers will prefer share repurchase because the value of their holdings will increase.
Let’s assume for the moment that Fred Futile’s (the manager) situation is identical to that of other shareholders: he enjoys the benefits of both share repurchase (higher share price) and dividends (immediate cash inflow). How is his situation changed if the company cuts its dividend and repurchases shares instead? Does it matter if he participates in the share repurchase, or waits to sell his holdings at the end of the analysis period?
To answer this question we’ll use the figures from the article (linked above):
“Fred is awarded a 10-year option over the company’s stock at the current price of $100 per share. There are initially 100m shares in issue and annual earnings are $1 billion. So he cancels the dividend and uses all the earnings (which remain flat throughout the 10 years) to buy back shares; earnings per share rise and the p.e remains the same. By the end of the 10-year period, the eps will have more than doubled from $10 to $25.8 and the share price will have risen to $258, earning Fred Futile $158m for himself.”
Calculating the Economics
Let’s examine the dividend case first. The firm pays $1B/100mm=$10 per share. If Fred buys (or, equivalently, is given by his firm) a share today at $100, and earnings are constant (and distributed in entirety), at the end of the 10 year window Fred has received ten $10 dividends and sells his share for $100. He has enjoyed a 10% IRR.
In the repurchase example, over the 10 years the firm’s outstanding shares fall from 100mm to 34.9mm; the share price rises from $100 to $258. If Fred doesn’t participate in ongoing repurchase, he realizes no intermediate cash flows, and sells his shares at $258 at the end of 10 years. Assuming he purchased them at $100 today, his IRR is 9.9%.
Fixing the Calculation
At this point you might be tempted to say Fred is in fact worse off under the repurchase policy: his return is 0.1% lower. It turns out this is only because there’s a mistake in the author’s original example: the share price is being incorrectly calculated. Note that just before the first year’s repurchase, the firm has earned $1B. Market cap isn’t just P/E times earnings (or $10B), it’s that amount plus the accumulated cash, or $11B total. Thus the share price is $110 at the time of the first repurchase, not $100.
With this correction, the firm buys back fewer shares (ending with 38.6mm outstanding after 10 years, not 34.9mm). The ending share price is $259 instead of $258, and Fred’s IRR is now 10.0%, exactly the same as under the dividend policy.
Note that the sum of Fred’s inflows is considerably different: $259 under share repurchase, but only $200 under the dividend policy. The difference of course is that the dividends are distributed earlier (and evenly) throughout the 10 year period, producing the equivalent IRR.
You can download the original and corrected calculations here: Distribution Form Irrelevance (3). That workbook also considers a third case: where the firm distributes earnings via share repurchase, and Fred participates proportionally (selling each period the same fraction of his shares that the firm is purchasing from all investors). In this third case Fred continues to enjoy a 10.0% IRR.
The Missing Incentive
This corrected example clearly indicates how, controlling for such factors as shareholder tax considerations, dividends are economically equivalent to share repurchases. If managers are shareholders, they will not benefit from increased share price if we account for the forgone (earlier) distributions. So the suggestion that Fred’s incentives will skew the firm’s distribution policy seem unfounded.
Is Fred’s situation different? Perhaps Fred doesn’t enjoy dividends, or for some other reason his economic interest is different from that of the typical shareholder. In this case it’s certainly possible that agency issues may push him away from serving the best interests of the firm’s investors. But that’s an issue for contract theory, not distribution policy.