Thursday, October 5, 2017

Stock-Picking Skill

When researchers talk about someone having skill at stock-picking, they are using the word “skill” differently than we’re used to. I might be impressed that a stock-picker seems very smart and knows far more than I do, but this is far from having skill in the technical sense.

To illustrate what we typically mean by “skill,” let’s consider golf. Over the years, I’ve golfed with many people whose abilities impressed me. There are dozens I’ve played with who I’d say have golf skills. Worldwide, there are millions of people who I would judge to be skillful at golf if I saw them play.

But what if we define “golf skill” differently? What if we decided that only those who have an expectation to earn more in prize money and endorsements than they spend on travel and equipment count as being skilled at golf? By this definition, I’ve never golfed with a skillful player. Worldwide, there are perhaps a few thousand players who have skill in this sense.

Does this strict definition of golf skill make sense? I suppose it makes sense for someone considering making a living at golf. But the vast majority of golfers play for fun, and there is nothing wrong with deciding that a player is skillful, even if they have no chance of making a living at it.

Getting back to stock-picking, does it make sense to call someone skilled just because I’m impressed with their knowledge and insight about stocks? The answer is no because the purpose of stock-picking is profits. I’ve never met anyone who admits that their stock-picking efforts tend to lose money but they do it anyway because it’s fun. A few such people may exist, but the vast majority of stock-pickers believe their efforts are likely to be profitable; they wouldn’t pick their own stocks otherwise.

Roughly speaking, the definition of stock-picking skill is the expectation of earning higher compound returns (after factoring in costs) than an appropriate benchmark. The word “compound” is important here because it implies a rational level of risk aversion. This is similar to the concept of risk-adjusted returns. (The way the math works out, the expected compound return is approximately equal to the median return rather than the arithmetic average return.)

A major challenge with this definition of stock-picking skill is that there is so much luck involved. Even Legg Mason Value Trust’s record of beating the S&P 500 every year for 15 years seemed certain to be skill before 3 years of substantially underperforming the market. It’s very difficult to say with any confidence that a particular stock picker has skill. Another challenge is that stock pickers can shop around for a benchmark they look good against.

A paradox of skill at stock picking is that the better everyone gets at it, the less skill there is in the world. The definition of skill at stock picking involves beating the market, which implies beating the average returns of other investors. So, you can only have skill if you’re enough better than other stock pickers.

As more and more talented stock pickers enter the field, the talent threshold for having skill rises. With each passing decade, market professionals increase their domination of trading in stock markets. Retail stock pickers are no longer much of a factor. To have skill, you need to be substantially better than the typical professional stock picker, a tall order.

There is even some doubt in the case of the great Warren Buffett. I’m satisfied that Buffett demonstrated skill over his long and hugely successful career. But does he have stock picking skill today? He has two huge forces working against him. The first is that his competition has been getting better over the decades. The second is that he’s investing so much money that he’s forced to focus on the stocks of only the largest businesses that already attract a huge amount of attention. He may still have skill, but it’s hard to be certain.

I’m satisfied that stock-picking skill must be very rare. Among professional stock pickers, skill is, at best, uncommon. Among retail investors, skill is so rare that you can safely assume that anyone you meet almost certainly doesn’t have skill, regardless of any claims they make.


  1. I'm re-reading a biography of Warren Buffett. What struck me this time was that a lot of his investments were not based on buying stocks in the open market. In the earlier years he would often buy a block of shares directly from someone else, and negotiate down the price to get a small advantage. Now of course he will buy whole companies but even when he had a lot less to work with he would try to influence the management of the stocks he was buying.

    It's hard to be sure but it seems that most of what he did involved going far beyond entering an order with a broker, though he does also seem to have a pretty good record at the times when he did buy shares at market prices and didn't have any influence over the company. Then again when he started out there were a large number of companies selling for less than their cash on hand. The opportunities today are likely more limited.

    It seems that his skill is much more than simply picking stocks and involved a lot more work than that.

    1. @Richard: I agree that Buffett has skills beyond stock picking. His humble persona masks a great memory and powerful intellect. But even if we isolate just his stock-picking, I'm satisfied that he demonstrated skill over the years.

  2. To me when I hear skill, I keep thinking that person has an ability that makes them better at something, however, now with the growing flim-flammery of FINTECH, now I am being told that "Algorithms" will make me rich. Is an algorithm a skill ? Currently it seems to be?

    1. @Alan: I'd say that if an algorithm has the expectation of producing a market-beating compound return, and you created the algorithm, then you've got skill.

    2. If you mean by "skill" magical foresight, sure, that sounds about right. Just the right amount of fairy dust as well.