2008: Does Buffett do pairs trading?

AUDIENCE MEMBER: Hi. My name is Henry Pattener (PH). I’m hailing from Singapore, most recently.

In one of your older letters, you — your older partnership letters in 1964 — you introduced a fourth investment method called “Generals — Relatively Undervalued.”

In your description you say, “We have recently begun to implement a technique which gives promise of very substantially reducing the risk from an overall change in valuation standards.

“We buy something at 12 times earnings when comparables or poor-quality companies sell at 20 times earnings, but then a major revaluation takes place so that the latter only sell at ten times.”

Is this technique pair trading and, if so, how did you think about and calculate the ratio of longs to shorts?

WARREN BUFFETT: Yeah. I didn’t remember we started as early as ’64, but certainly in the ’60s we did some of what, in a very general way, would be called pair trading now, which is a technique that’s used by a number of hedge funds, and perhaps others, that go long one security and short another, and often they try to keep them in the same industry or something.

They say that British Petroleum is relatively attractive compared to Chevron or vice versa, so they long one and short the other.

And actually that technique was employed first by Ben Graham in the mid-1920s when he had a hedge fund, oftentimes — I read articles all the time that credit A.W. Jones with originating the hedge fund concept in the late ’40s, but Ben Graham had one in the mid-1920s — and he actually engaged in pairs trading.

And he found out it worked modestly — very modestly — well because he was right about four times out of five but the time he was wrong tended to kill him on the other four.

We did — we shorted out the general market for about five years in the partnership, to a degree. We borrowed stocks directly from some major universities. I think we were probably quite early in that.

We went to Columbia and Harvard and Chicago and different places and actually arranged for direct borrowing. They weren’t — it wasn’t as easy to facilitate in those days as it is now.

And so we would take their portfolios and we would just say, “Give us any of the stocks you want, and then we’ll return them to you after a while and we’ll pay you a little fee.”

And then we went long things that we thought were attractive. We did not go short things that we thought were unattractive; we just shorted out the market generally.

It was always kind of interesting to me, when I would visit the treasurer of Columbia or something like and I’d say, “We’d like to borrow your stocks to short,” and, you know, he thought his stocks were pretty good at that point.

And he’d say, “Which ones do you want?” And I said, “Just give me any of them — (laughs) — I’m happy to short your whole damn portfolio.” (Laughter)

I needed the Dale Carnegie course to get me through that kind of thing, you know.

We didn’t have any specific ratios in mind. We were always limited by the number of institutions that would give us the stocks to short.

So it was not a big deal, but we probably made some extra money on it in the ’60s. It’s not something that would fit our — what we do these days at all.

And, generally speaking, I think if you’ve got some very good ideas on businesses that are undervalued, it’s really unnecessary to do any shorting out of the market.

There’s a — for those of you who are in the field — I mean, there’s a — kind of a popular proposal — money managers always have some popular proposal that’s being sold to the potential investors — and now there’s something called 130-30, where you’re long 130 percent long, short 30 percent.

That stuff is all basically a bunch of stuff just to try and sell you the idea of the day. It doesn’t really have any great statistical merit.

But the fish bite, as Charlie says. Charlie can elaborate on that.

CHARLIE MUNGER: Yeah. We made our money by being long some wonderful businesses. We didn’t make it by a long-short strategy.