Who gets the money spent on share buybacks?
(Hint: It’s not only shareholders.)
Andrew Ross Sorkin recently discovered the secret to how companies like IBM keep their per-share earnings rising steadily upwards: share buybacks. IBM’s revenues have gone basically nowhere over the past six years or so, he explains, but its earnings (or at least its earnings per share, which is all that shareholders really care about) have been doing great — because the number of shares outstanding has been shrinking.
The cost of IBM’s buyback program has been truly enormous: $108 billion, since 2000. And the effect of the buyback program has been pretty impressive. Here’s a chart:
What you see here is the total number of IBM shares outstanding, on a fully diluted basis, falling from 1.8 billion at the end of 1999 to just 1 billion at the end of the second quarter this year. To be precise, the share count has fallen by 797,604,000 shares. That’s a lot of shares to repurchase!
But here’s the thing: IBM actually bought back significantly more than 797,604,000 shares. I don’t have good data going back to 2000 on how many shares it bought back each year. But in its most recent annual report, IBM does say that it repurchased “approximately 73 million shares” in 2013. Which is interesting, because it ended 2012 with 1.13 billion shares, and ended 2013 with 1.08 billion shares. That’s a difference of just 50 million shares.
What happened here is that IBM didn’t just destroy 73 million shares by buying them back. It also created some 23 million new shares, largely as a result of compensation deals — restricted stock, options, that kind of thing. (IBM also sometimes uses its stock as an acquisition currency, but I don’t think it did that to any great extent in 2013, and in any case insofar as such deals are often acqui-hires, the newly-issued stock still ends up going in large part to employees.)
There are two ways of looking at this. The first is that IBM spent $13.859 billion buying back those 73 million shares — an average of $190 per share. But in fact the share count only fell by 50 million shares, which means that IBM actually paid $277 per share by which the share count was reduced. That’s a crazy sum of money, far greater than IBM’s all-time-high share price. (The current price, by the way, is $162.)
The other way of looking at this is that the $13.859 billion didn’t really go entirely to shareholders, as Sorkin implies. Rather, some 50/73 of it went to shareholders — that’s about $9.5 billion — while the rest of it, about $4.4 billion, went, in one way or another, to employees. And especially to senior employees.
Sorkin quotes Warren Buffett’s 2011 annual letter, where the Sage of Omaha explains why he likes stock buybacks. He explains that if IBM starts with 1.16 billion shares, and buys back 250 million shares over five years, then at the end of that period there would be just 910 million shares outstanding. On the other hand, he says, if IBM bought back 167 million shares, there would be 990 million shares outstanding.
But that’s simply not true: Warren Buffett should spend more time reading blogs like Ultimi Barbarorum. While shareholders love to think that share buybacks are a way to give money to the owners of capital, a lot of the time they’re a back-door way of giving money to high-priced producers of labor, a/k/a senior management. The Ultimi Barbarorum post mentions Cisco, for instance, which managed to spend $600 million on its own stock in one quarter, without the share count falling at all.
The fact is that share buybacks are complicated things. In principle, I think that Sorkin is wrong. He says that share buybacks are unlikely to be effective when a company is in decline: that’s simply not true. Indeed, share buybacks are pretty much the only way that a company in decline can ensure a steadily rising share price. And there’s no shame in such a strategy. There’s no reason why all companies should always want to get bigger: often, all attempts to do so simply end up destroying value rather than creating it. If you’re a company throwing off a large amount of money, but those cashflows aren’t likely to last indefinitely, then there are a lot of good reasons to use those profits to buy back your own stock.
What’s more, it’s also OK to pay employees well, and to pay them in stock rather than in cash. Shareholders often like that because it tends to align incentives. And those of us who fret about the way in which capital is winning out over labor can maybe take a little bit of pleasure in this clever way that companies’ employees have found of using sluices designed by financial engineers to divert billions of dollars into their own pockets.
And yet. All too often stock buybacks are deceptive things, which create a sugar high in the share price, a nice little windfall for management, and pretty much nothing in the way of actual value creation.
There’s also some basis for the widespread feeling that buybacks are a sign of desperation: the triumph of financial engineering over genuine value creation. It’s probably no coincidence that Apple became the biggest repurchaser of stock in the world at exactly the same time as it was going through a fallow period in terms of innovation; now that Tim Cook is getting his mojo back, I wouldn’t be surprised to see him ease off on the buybacks a bit.
All companies should think about buying back their own stock. When a company like Amazon reinvests profits from relatively mature business in newer businesses with higher growth, it’s doing exactly what innovative companies are meant to do. On the other hand, it also needs to set some kind of baseline for when such investments make sense. And the obvious baseline is buybacks: is there a decent chance that shareholders’ return on new investment will exceed the return on buybacks?
Journalists love to talk about heroic management decisions as the things which drive companies to greatness, and there’s even some truth in such stories, sometimes. Investors, too, like to latch on to a narrative which can make them lots of money. But the fact remains that share buybacks are an easier and much more reliable means of generating shareholder returns than strokes of corporate genius. So it’s hardly surprising that shareholders tend to love them.
That said, of all the things that executives should be paid millions of dollars for, monster stock buyback schemes are surely at the bottom of the list. Any schmuck can do that. And even if shareholders clamor them, it behooves boards to keep a very close eye on them. Because a lot of the time it’s extremely unclear where the money is actually going.
Felix Salmon is a senior editor at Fusion.
