Concentration or Diversification

Over the last couple of days, we’ve argued that it is a mistake for Mohnish Pabrai, or other successful investors, to expand the number of positions in their portfolio in order to lessen volatility. Yet, you’ve heard enough from me; what do the professionals have to say?

In You Can Be a Stock Market Genius (Simon & Schuster, 1997), Joel Greenblatt, whose annualized returns at Gotham Capital from 1985-1995 were 50%, states that a concentrated portfolio of five securities will only carry 3% more downside potential (within one standard deviation of the mean) than a portfolio of fifty securities. Said differently, a portfolio of fifty securities stands a two out of three chance of returning between -8% and 28%; a portfolio of five securities stands a two out of three chance of returning between -11% and 31%. The returns on ten securities should fall between -10% and 30%. In short, there isn’t a significant difference in variability if one owns five, ten, twenty, or five hundred securities.

Warren Buffett’s compounded annual partnership returns from 1957-1969 were 31.6%. Though it is not easy to determine the relative size of the partnership’s positions, it is well known that Buffett put 40% of the partnership’s assets into American Express during the Salad Oil Scandal of 1964. Martin and Puthenpurackal found that “Berkshire Hathaway’s portfolio is concentrated in relatively few stocks with the top five holdings averaging 73% of the portfolio value.” As Buffett himself has said, “if you really know businesses, you probably shouldn’t own more than six of them. If you can identify six wonderful businesses that is all the diversification you need… going into a seventh one, rather than putting money into your first one, has got to be a terrible mistake.” Buffett has even gone so far as to say that he would have been willing to allocate up to 75% of his portfolio in the distressed assets of Long Term Capital Management in 1998.

Charlie Munger, Buffett’s partner at Berkshire, has quipped that you could be adequately diversified if you owned the best office building in town, the best apartments in town, the dominant car dealership, and the highest grossing McDonald’s.

Now, to be clear, these three are speaking primarily to dedicated investors willing to devote time and intensity to studying businesses. Broad diversification achieved through index funds is likely the best strategy for the majority of market participants. But for those who make it their profession to beat the market’s returns, concentration is an important component of success.

All told, this series gives voice to my growing conviction that the majority of one’s investing returns will come from a few great ideas. Looking at Buffett’s returns, substantial gains came from relatively few holdings—National Indemnity, American Express (the first time), Geico, See’s Candies, Capital Cities, the Washington Post, and Coca Cola. Though countless reasons can be offered for diversifying, it is hard to argue against the success that has followed his concentration.

Ten or Twenty for Pabrai’s Portfolio?

Yesterday I argued that Mohnish Pabrai’s recent decision to invest Fund assets in twenty 5% positions rather than ten 10% positions runs contrary to his investors’ interest. And the more I think about this altered strategy, the more convicted I become that it’s wrong. Permit me then to grapple with some of the main lines of defense for Pabrai’s position:

Defense #1 (the paternalistic defense): Investors get scared when positions move substantially lower, and then redeem their funds. It is in the investors’ interest to avoid or assuage such fear, so twenty portfolio positions will alleviate violent downdrafts and encourage more rational investor behavior.

Response: If we assume that Pabrai’s best idea is substantially better than his twentieth-best idea, then future returns will likely be lessened by this paternalistic action. Though it is in the investors’ long-term interest to remain committed despite short-term losses, their irrational behavior shouldn’t bother the manager much (so long as he isn’t levered and forced to liquidate the Fund at market bottoms). Thus, we have an apparent trade-off; either the investor endures market volatility and gets higher returns, or the investor takes less volatility and lower returns. The investors’ long-term interests are clearly best served in the former approach, and the fund manager seems better served in the latter. Keeping invested funds high keeps management fees high through bear markets.

Defense #2 (the “no big deal” defense): Even if this strategy shift trades performance for temperance, future returns won’t sag much. The twentieth-best idea is still a worthy idea.

Response: In the early stages of a secular bull market, one’s twentieth best idea will likely do fine. Cast aside the swim trunks and jump in! But over a three to five year period, it is fair to say that number 20 will lag number 1 by a couple of percentage points per year. In isolation, losing a few points of return is no thing, but compounded over a decade, such a drag will significantly affect one’s overall performance. Also, any additional time spent working on ideas 11-20 will likely detract one’s attention from future opportunities.

Defense #3 (the “diversification lowers risk” defense): Giving up some performance is acceptable because greater diversification entails less risk.

Response: This strikes me as plainly false. The best investment idea is the one with the best prospects and the greatest margin of safety. To diversify capital from one’s best idea to one’s merely ‘good’ ideas means accepting a smaller margin of safety, and in no way can that strategy be less risky. The only way diversification is less risky is if one fears that his inner psyche will compel him to sell his best idea if it declines the overall portfolio too much. In that case, diversification serves more as a psychic crutch rather than a risk-management tool.

Defense #4: (the “loss mitigation” defense) Diversifying fund assets into twenty positions means any mistakes will impact the portfolio returns less.

Response: Though the point is accurate, the other side of the coin is that future successes will also impact portfolio returns less. For managers like Pabrai, future successes will far outnumber mistakes, especially given Pabrai’s method of betting when he can’t lose much. Even when faced with a handful of mistakes, one has to discern whether they signify that his approach is broken, or his fortune briefly turned. If only the latter, mistakes should not consume one’s virtue.