I've always been fond of Warren Buffett's schtick of being an adorable Cherry-Coke-drinking grandfather whose fuzzy sweaters hide some sharp elbows, talking up old-timey value investing while making his money on distressed sweetheart deals, and railing against derivatives while doing lots of shady ones. So I'm pleased that today he basically announced "we bought back a bunch of shares from one deceased buddy of mine, against our previously announced guidelines for how we'd buy stock, which we just amended to make this deal happen." That seems shady!
Felix Salmon covers the shadiness here but also says this:
Buybacks are considered a good thing, on the stock market, for three reasons. Firstly, they reduce the number of shares outstanding, which means that the value of the remaining shares goes up: the company is worth the same amount, so the value per share is higher. Secondly, they provide an extra bid in the market, which helps support and drive up the share price. And thirdly, they give shareholders the opportunity to sell their shares back to the company: if they want to sell where the company is buying, they have that option.
So I submit to you that there's a fourth reason buybacks could be considered a good thing: some stocks are a good investment, and if yours is a good investment, maybe you should buy it, instead of, like, "hoarding cash" or doing dumb M&A deals or whatever.
This sounds like a nice theory but is almost always wrong; typically, if company is all "the best investment we can find is our stock and it will only go up from here," their stock is about to crater. But Warren Buffett isn't a typical corporate CEO, stock-picking-abilities-wise. He's ... y'know, he's a guy whose whole thing is being good at picking stocks, plus the Coke/grandpa stuff. And so his theory about buybacks is not primarily about EPS accretion or providing a bid in the market: it's about buying stock at below "intrinsic value," whatever that is.
Also he's a guy who's good at getting good deals. So why'd he get a bad deal here? BRK/A closed at $130,831 last night, so Berkshire paid a, um, 0.13% premium over the last sale.1 On the other hand, you can't just go buy a billion dollars worth of stock - even Berkshire stock - at the closing price. The last 9,200 BRK/A shares that traded publicly did so at an average price of about $131,322 - and it took 18 trading days to get there.2 So Berkshire actually bought those shares at a, um, 0.25% discount to a reasonable trading level.
Those are small numbers! What would happen if Berkshire had instead attempted to buy $1.2 billion worth of stock in the market? I dunno, but it's hard to buy almost a month's volume worth of stock. Companies' open-market purchases are basically restricted to 25% of volume, so Berkshire would spend at least 72 trading days - almost four months - buying this much stock in the market. A lot can happen in four months; BRK/A is up almost 6% since the start of September. And buying 25% of the volume a day, every day, will push up the stock price a lot faster - beyond even Buffett's revised 120% of book value cap on purchases.
This is not a pretty trade, but it's how Buffett works: buying an asset he likes, at a price he likes, from a motivated seller. It looks bad - it opens you up to charges of conflict of interest, which are probably true, and to charges of hypocrisy about estate taxes, which are also probably true, and to charges of wasting shareholder money, which are probably not true. Buffett got to do a big buyback at a decent price. He bought 9,200 shares at $131,000, and they closed at $134,000 today. He pissed people off and was up $27.6 million on the day. I like it.
Berkshire buys $1.2 billion in stock from single investor [Reuters]
Berkshire’s weird buyback [Reuters / Felix Salmon]
Berkshire Hathaway News Release [BRK]
1.It's not 100% clear from Berkshire's announcement that they bought the stock last night, though that stands to reason; if it's not new news, why would you halt trading for the announcement this morning? Presumably it takes some time to negotiate a $1.2bn trade, though; if they got pretty much agreed on terms over the weekend then they got a discount to Friday's $131,090 close.
2.Not a particularly exact science but here you go: