How much is too much to hold?

This memo is the latest one written by Howard Marks who is an American investor and writer. Marks is the co-founder and co-chairman of Oaktree Captial Management. The memo is of great value for all investors. The last section is as follows:

Certainly there are times when it’s right to sell one asset in favor of another based on the idea of relative selection. But we mustn’t do this in a mechanical manner. If we did, at the logical extreme, we would put all of our capital into the one investment we consider the best.

Virtually all investors – even the best – diversify their portfolios. We may have a sense for which holding is the absolute best, but I’ve never heard of an investor with a one-asset portfolio. They may overweight favorites to take advantage of what they think they know, but they still diversify to protect against what they don’t know.

That means they sub-optimize, potentially trading off some of their chance at a maximal return to increase the likelihood of a merely excellent one.

Here’s a related question from my reconstructed conversation with Andrew:

H: You run a concentrated portfolio. XYZ was a big position when you invested, and it’s even bigger today, given the appreciation. Intelligent investors concentrate portfolios and hold on to take advantage of what they know, but they diversify holdings and sell as things rise to limit the potential damage from what they don’t know. Hasn’t the growth in this position put our portfolio out of whack in that regard?

A: Perhaps that’s true, depending on your goals. But trimming would mean selling something I feel immense comfort with based on my bottom-up assessment and moving into something I feel less good about or know less well (or cash). To me, it’s far better to own a small number of things about which I feel strongly. I’ll only have a few good insights over my lifetime, so I have to maximize the few I have.

All professional investors want good investment performance for their clients, but they also want financial success for themselves. And amateurs have to invest within the limits of their risk tolerance. For these reasons, most investors – and certainly most investment managers’ clients – aren’t immune to apprehension regarding portfolio concentration and thus susceptibility to untoward developments.

These considerations introduce valid reasons for limiting the size of individual asset purchases and trimming positions as they appreciate.

Investors sometimes delegate the decision on how to weight assets in portfolios to a process called portfolio optimization. Inputs regarding asset classes’ return potential, risk and correlation are fed into a computer model, and out comes the portfolio with the optimal expected risk-adjusted return.

If an asset appreciates relative to the others, the model can be rerun, and it will tell you what to buy and sell. The main problem with these models lies in the fact that all the data we have regarding those three parameters relates to the past, but to arrive at the ideal portfolio, the model needs data that accurately describes the future.

Further, the models need a numerical input for risk, and I absolutely insist that no single number can fully describe an asset’s risk. Thus, optimization models can’t successfully dictate portfolio actions.

The bottom line:

we should base our investment decisions on our estimates of each asset’s potential,

we shouldn’t sell just because the price has risen and the position has swelled,

there can be legitimate reasons to limit the size of the positions we hold,

but there’s no way to scientifically calculate what those limits should be.

In other words, the decision to trim positions or to sell out entirely comes down to judgment . . . like everything else that matters in investing.