For novice investors I’ll start with a definition: At base, position sizing refers to a strategy for determining the appropriate amount of money to invest in any given trade within a portfolio. So in a nutshell, here’s the position sizing landscape:
Sizing positions according to a benchmark. This is a thing but I’m not a fan, so I won’t dwell on it and will move on to the next choice.
Sizing positions evenly. This approach works best when your stock selections can generate alpha, yet you simply can’t distinguish why the winners win. My main comment here is to focus on risk rather than attractiveness. That’s how you’ll separate the good from the great.
Sizing them in order of attractiveness. The most common approach among famous investors is to size positions by attractiveness of the investment on a risk/reward scale.
In the latter, your stock picking skills either a) generate alpha or b) are based on your ability to differentiate between potential for some alpha and heaps of alpha.
If your skills don’t match up with a) then spending time actively stock picking is a bad idea unless you’re convinced that through this experience you’ll absolutely learn how to generate alpha for the future.
If your skill don’t match up with b), you’re better off with a portfolio of even-sized positions.
However, if you possess both a) and b) skill sets, how much risk can you assume?
Let’s examine a mathematical equation I first encountered when I played poker professionally. You’d be surprised at the linkages. For each game, I needed to figure out the amount of my overall budget to bet.
That’s where the Kelly Criterion came into play. Whether it’s a question of sports bets or cards, keeping statistics on your bets will enable you to optimize your bets and increase your earnings. The same goes for investing, although it’s vastly more complex. Your portfolio is not a matter of a single bet in a short time frame. Also, in sports and poker, there’s no holding period. All the same, grasping the essence of the Kelly Criterion will lead to better investment decisions because it tells you much you need to diversify.
Be forewarned about the bumps inherent in applying the Kelly Criterion: In the short run, you’ll incur losses that will be hard to swallow but in the long run, you’ll optimize growth rate of the portfolio. It can a tumultuous experience leaving you wondering if your portfolio is winning or losing.
Even if you’re able to shrug off big losses, as a private investor you may end up in divorce court if your spouse doesn’t feel the same when the majority of your portfolio goes up in smoke. A professional who’s intent on optimizing portfolio growth rates may end up unemployed. Whether you’re in the market professionally or privately, smart investors will take these sorts of factors into considerations and reduce their “Kelly.”
Professional investors who generate the least alpha and face the most career risk tend to stick close to the benchmark. Other independently minded investors who have a big chunk of "permanent capital" and feel secure because it won’t be withdrawn are likely to come closest to Kelly.
This table shows the perentage of capital several guru traders invest in their top 10 holdings:
Source: Gurufocus / portfolio value in million $ / number of stocks / allocation in %
My list contains investors who hold non-equity assets and shorts or who diversify using other vehicles (like Blum). They’re actually more diversified than the table suggests. In spite of the considerable time and money it requires to actively engage the companies in their portfolios, note how many concentrated investors do it! It’s a sensible approach when you stand to generate alpha. Activists include Ackman, Lampert, Ubben and Winters.
Here’s another table that displays the opposite approach, with success:
Source: Gurufocus / portfolio value in million $ / number of stocks / allocation in %
Joel Greenblatt (Trades, Portfolio) and Jim Simons (Trades, Portfolio), renowned for their returns, rely more on statistics than qualitative research and they have the lowest concentration of assets in their top 10 picks. It’s clear that these two still bet heavily on their favorites.
How you generate alpha should determine whether you hold a few stocks or lots. For the former, deep research and engagement will work better for you. However, if your pattern is to generate alpha by identifying something “the market doesn't know” or by zeroing in on market inefficiencies, you’ll fare better with a little less concentration.
When you take your best ideas and plough them into an evenly weighted portfolio, that’s a mistake. If you are skilled enough to pick eight alpha-generating stocks, you need to go further and pick more, differentiating them based on the amount of alpha they offer.
Bottom line for position sizing
When I invest heavily in my own best ideas, I publish them through the newsletter I run, The Black Swan Portfolio. My practical approach is to diversify aggressively but to bet bigger on the best ideas. I’ll continue to hold lots of different stocks even when there are just five I love. I’ve experienced having high conviction and being wrong, company reporting being subpar given hindsight and deterioration of the underlying business that caught me completely off guard. I always buy stocks that improve the risks/rewards of my portfolio in comparison with the market or I don’t bother.
Disclosures: No positions.
