Wednesday, October 18, 2017

Howard Mark's recent G & D interview quotes

I enjoy reading the Graham-Doddsville newsletter.

Some pertinent quotes, relating to the issue of indexing and quantitative/AI factor.

Howard Marks:

“If people take their money out of active management, then active managers would fire all their analysts, and then the market would not stay efficient. Then the necessary condition is satisfied for active to work. The point is, I don't think this move is permanent, I think it's rotational.”

“Because every dollar that goes into a truly passive fund is invested on autopilot, the fund must buy the stocks that satisfy its criteria, and that’s without regard to value. That suggests to me that prices can go farther in diverging from value before they get corrected. Think about what would happen if 95% of the money went into index ETFs or index funds. Who would be setting prices? There’s something called price discovery, and it’s done by thoughtful buyers and sellers. The price of a security in the marketplace is set by buyers and sellers coming together, and seeing if they can find a place to transact where the buyer thinks it has good upside, and the seller thinks it doesn’t. Who provides that function if all the buying are on autopilot? People put their money in index funds, with the presumption that they’re minimizing error, but how much of your money do you want to have managed in a fund where nobody’s thinking about the price of the stocks or the weightings within the portfolio? The thing about investing is that the efficient market hypothesis says that price equals value. Active management is about the assumption that price sometimes deviates from value, finding those deviations, and then taking advantage of them. It seems to me that the fewer the people who are looking at value, the higher the likelihood that price can diverge from value. But that’s just a hypothesis.”

“None of this stuff is easy. The greatest quote in my book is from Charlie Munger, who said, “None of this is meant to be easy, and anybody who thinks it’s easy is stupid.” All this stuff is really complex. It’s easy to talk about, but it’s hard to implement. How do you tell the ones who are good but unlucky, from the ones that are bad? It’s not easy. It takes judgment. That’s why I believe that this whole thing can never be completely computerized, because I think exceptional investment success requires judgment, and I don’t know if AI can be taught to make those judgments.”

Dear readers, any thoughts?

Yours One-Legged