Portfolio Rule Seven

Rebalance the portfolio whenever a stock gets more than 20% away from its target weight. Run a largely equal-weighted portfolio because it is genuinely difficult to tell what idea is the best. Keep about 30-40 names for diversification purposes.

Let me begin with the last part of the rule.  I’ve gotten different reactions over the years from holding 30-something names my portfolio.  From some friends who run concentrated hedge funds they say, “Why so diversified?”  From those that run mutual funds they say, “Why so concentrated?”

Since I concentrate industries, owning 30 to 40 names my portfolio is relatively undiversified, compared to an index fund.  My view is that industry performance is the main driver of stock performance.  I also think that owning industries in proportion to the index is a recipe for mediocre performance.  If you don’t break free from the industry weights of the index, you will never achieve index beating performance.

Now, some may criticize the idea that I don’t know what my best ideas are.  Good, go ahead and criticize.  My experience has been that ideas that I thought were marginal often did quite well and ideas that were highly promising sometimes did marginally.  On the whole I think there was some rough positive correlation between how well I thought it would do, and how it actually did, but not enough to make me change this rule.

I do occasionally make a name for my portfolio a double-weight, or once even a triple-weight, but those are rare.  To have a high weight in my portfolio means that it must be exceptionally cheap, and be exceptionally safe.  In other words, it must be very, very, misunderstood.

At present I have 34 names my portfolio.  Ordinarily I like to have 35.  But, I don’t change the number rapidly.  I have a greater tendency to raise the number when the market is cheap, and I add equity exposure.  That said, during the bear market from 2000 to 2002, I decreased the number of names as the selloff got worse, concentrating into the names that got hit the hardest that still deserved to be invested in.  As the market began to rise, I added to the number of names that I invested in.

Keeping the number of names relatively fixed in the short run helps to facilitate swap transactions.  Swap transactions are a more intelligent way of managing an equity portfolio, because people are reasonably good at doing binary decisions, and not at doing decisions that have a lot of degrees of freedom.

I can’t tell you at any given point in time that I have the best set of stocks in the world.  But, it is relatively easy for me to look at the stocks that I am buying versus the stocks that I am selling, and conclude that I am coming up with a better portfolio as a result.  More on that when I discuss portfolio eight.

=-=–==-=–=-=-=-=-=-=-==-=–==-

Now, on to rebalancing.  I have gotten some criticism over the years, arguing that my rebalancing discipline leads me to pump money into losers that lose yet more.  Part of the misunderstanding is that when a stock falls by 20%, it doesn’t trigger an automatic buy, but an automatic review.  If my review comes the conclusion that the stock is fundamentally okay, then I rebalance up to the target weight.  I don’t double the position; that would be lunacy.  I only add to bring it up to target weight.  This is a moderate and disciplined way to buy weakness.

If my review finds that I made a fundamental mistake, I sell the position out.  Now, all that said, my worst losses typically came from cases where I rebalanced down and rebalanced down, and never caught the fundamental error.  Hey, I’m human.  But I can say that my gains more than paid for my losses.  Also, most of the companies that hurt me had balance sheets that were less strong then I thought.

When I was on the other side of the table, choosing investment managers, I ran across three managers with excellent track records that used this strategy.  They felt that it added on 1-3%/year.  Surprisingly, one was a growth manager, one was a core manager, and one was a value manager.

My sense is this: markets oscillate, and performance by industry and company oscillates.  This simple strategy catches some of those oscillations, buy low and selling high.

It also helps portfolio management from a psychological standpoint, because you realize that by realizing gains partially, and buying lower selectively, you don’t care so much about day-to-day fluctuations.  Rather, you realize that on average, fluctuations are working for you, and so, you focus on fundamentals, not the noise.

At least, that is what I have experienced.  And that is why I rebalance.

2 Comments

  • David, I was actually going to question the rebalancing on the other side – when a stock rises. you wrote:

    “Run a largely equal-weighted portfolio because it is genuinely difficult to tell what idea is the best.”

    Sometimes the stock prices themselves answer that for you! But I’m guessing/hoping that if a stock rallies such that it’s more than 20% above its target weight you don’t automatically sell it just like you don’t automatically buy those that fall…

    That’s just a long way of saying that sometimes (Especially in markets like this where you have so many of these stocks with 100% straight line up moves) you have to let your winners run.

  • I let 80% of my winner run. Given the choppiness of markets, and that I don’t buy pure growth stocks, I rarely get parabolic moves. I leave something on the table for the momentum guy some of the time when I sell after a 20% move up, but often the stock will move down, and I buy more, before it moves up again.

    In short, most of the the time I automatically trade some away to come back down to target weight. For stocks where the cash flows are relatively determinate, risk rises as prices rise. Risk falls as prices fall.

    That’s why this seems to work for me, and, some other large institutional investors that I know.

3 Trackbacks