Yesterday, our shares of Berkshire Hathaway (BRK.B) staged a 50:1 split, bringing our adjusted cost basis to $66.08. That means we’re up around 10% on the stock in a single week as of this writing.

I’ve been fielding a few questions about our Berkshire Hathaway play lately. The idea that Berkshire could rally just because of increased trading and index fund activity has caught the interest of quite a few investors and market commentators. The main stumbling block is the idea that our B shares of Berkshire are locked a fixed price relative to the A shares: How could class B shares move higher if the class A shares stay put?

Well, the way Warren Buffett set-up the class B shares is what’s important… He designed them so that they could never become worth more than 1/30 (or 1/1500 after the split) the value of the A shares for more than a few instants – when they do, arbitrage occurs, and specialists buy shares of BRK.A and convert them to BRK.B to collect the premium (the conversion only works one way). So many investors simply assumed that the value of BRK.A was a sort of “price ceiling” over the B shares. That’s not what happens though.

Instead of A shares being a fixed price ceiling, they’re really a moving ceiling that grows to accommodate the B shares.

When specialists buy shares of BRK.A to convert to B, they inflate the price of BRK.A shares, then when (and if) they convert them to BRK.B shares, they pour extra supply into the market, depressing the value of B shares until the two meet in the middle. In other words, B shares trim their gains a bit, and A shares increase more than they should.

But there are two other factors that make A shares move more than B shares do: the actual arbitrage process and liquidity.

At market speeds, arbitrageurs don’t usually physically convert A shares into B shares – they just buy A shares knowing that the market will logically move them to that magic 1/1500 ratio. How to we know?

We know because the conversion only works one way… Once turned into class B shares, the class A shares can’t be reconstituted. That means that over time, with all of this arbitrage going on to keep class B shares in check, the number of class A shares should be dwindling notwithstanding new share issues. They’re not.

That tells us that specialists are buying inflating the prices of A shares when B shares become overvalued, but they’re not increasing the supply of B shares in any meaningful way.

The second issue is liquidity – because A shares are much less liquid than B (even accounting for the conversion rate), it takes fewer shares to have a bigger price impact. That means that even when specialists do convert their A holdings into B, during the leveling-off process A shares will move up more than B shares will move down.

Either way, we’re left with a situation where trading in BRK.B controls the price of BRK.A.

Get Our Free Newsletter by Email

Tags: arbitrage