Looking for ‘winners’ is the most common approach to active fund management. Some fund managers look for the cheapest assets in the belief that they will outperform, others look for the strongest growth stories and others, again, are oriented towards various kinds of winning themes or industries.

Not only do these fund managers come up with a list of assets they think will outperform, many are even investing more money in some assets on such list than in others. On the basis of what? That they are sure winners? That they are less risky? What is implied by such behaviour is a clear assumption of precision: not only are these fund managers confident that they can compile a list of winners but they also believe that they can grade the list from best to worst.

We^[1] see two main problems arising when looking for ‘winning’ assets. Firstly, there are so many people looking for winning stocks and stiff competition reduces the value of information. Secondly, particularly in complex systems, less is more, and ‘positive’ knowledge, that is knowledge of what is right or what works, is intrinsically a non-robust type of knowledge. Investors, therefore, should focus their analytical resources on information that is not easily competed away and that is robust to error.

Competition and the value of information

On the first issue, the effect of competition on the value of information, think of the following analogy. In an opaque jar I am holding several $1, $5, and $100 notes in equal proportions. Before me I have a group of 10 people to whom I auction the notes in the jar. As we begin, I pull out one of the notes and carefully shield it so none of the bidders can see its denomination. Because the expected value is $35.33, our group of bidders top out at say $35 or just below the value of the average bill. So far so good. When I then draw the next bill, which is a $100 note, I am less careful and two of the bidders, Mary and Paul, get to see it. However, the rest of the group does not know that Mary and Paul saw the bill and neither do Mary or Paul know about each other. In the ensuing auction, bids are not topping out before $99.99 as Mary and Paul are competitively bidding the price up to a marginal 1 cent profit. In the third auction, I’m sloppy again: this time around, Annie and Joe get to see I draw a $1 note and, consequently, they do not participate in the ensuing bidding which again goes to $35.

Note that Mary and Paul did not gain a real advantage from their secret information as the two of them bid the price up to a decimal point away from full value. On the other hand, Annie and Joe, avoid bidding $35 because they know it is only a $1 bill. With a 10% probability of holding the winning bid, the secret information is worth $3.50 to each of them. Thus exactly the same information, secretly knowing the value of the note, is of little use when selecting a winner ($100 note) but is very valuable when avoiding a loser ($1 note).

Something similar happens in the stock market. Because the general equity risk premium is positive and 99% of investors buy a stock before they sell it, it is much more common for a stock to be bid up than offered down. Short sellers are heavily exposed to ‘synchronization risk’ as they must get the timing right, and a short trade exhibits a concave risk profile with limited gains and unlimited losses. Thus the hurdle to sell a stock short is higher than to buy one long. This means that there are a lot more players eager to bid up the value of a stock to its intrinsic value than there are people eager to sell it down to its intrinsic value. Thus if a fund manager has information pertaining a potential winner it only takes a few fellow managers with similar information to bid the stock up to its fair value. On the other hand, if a fund manager has information pertaining a potential loser, there is not going to be the same competitive rivalry involved in offering the stock down. Competition reduces the value of positive information.

Robustness of knowledge

As for the second problem, a material difference in the quality and robustness of knowledge, depending on whether it is of a negative or positive kind, Nassim N. Taleb describes it thus: ‘The greatest—and most robust—contribution to knowledge consists in removing what we think is wrong—subtractive epistemology.’ He goes on to advocate: ‘…we know a lot more what is wrong than what is right…..’ and ‘….negative knowledge (what is wrong, what does not work) is more robust to error than positive knowledge (what is right, what works) ^[2].’ and ‘…..since one small observation can disprove a statement, while millions can hardly confirm it, disconfirmation is more rigorous than confirmation.’.

Taleb’s robustness of negative knowledge echoes Karl Popper’s life-long obsession with falsification, or that a hypothesis can never be proved right, only falsified: ’no matter how many instances of white swans we may have observed, this does not justify the conclusion that all swans are white.’^[3] In business, Steve Jobs was on to the same thing: ‘People think focus means saying yes to the thing you’ve got to focus on. But that’s not what it means at all. It means saying no to the hundred other good ideas that there are. You have to pick carefully. I’m actually as proud of the things we haven’t done as the things I have done. Innovation is saying no to 1,000 things.’^[4] In a similar vein, I remember well a piece of good advice received decades ago from the former head of IMD, dr. Peter Lorange, ‘business strategy is all about deciding what not to do!’.

Another application of via negativa is that less-is-more. Applying complex solutions to complex problems is flawed because the parts of a complex system can interact in more ways that we can understand. A simple exercise of combining LEGO blocks demonstrates how quickly complexity spins out of control. One LEGO block can be combined only once. Two blocks can be combined 24 different ways and for three blocks there are 1,560 combinations. For decades, it was commonly accepted that for six blocks there were 103 million combinations. However, in 2014 two mathematicians revisited the problem and with the help of massive computer power they discovered 915 million combinations.^[5] When the smartest people on the planet have problems understanding how many ways six LEGO blocks can be combined, we can appreciate the extreme complexity of financial markets with all its moving parts and feedback loops. Dan Goldstein and Gerd Gigerenzer (among others) found that when dealing with complex situations ‘fast and frugal’ heuristics make better decisions^[6]. Less-is-more, is not perfect, it is not designed to be, but it proves robust to mistakes, and importantly, changes of mind. Complexity must be met by simplicity.

Focusing on negative information and less-is-more, or via negativa, as an approach to decision making goes back to ancient times and its efficacy is supported by empirical success in many quarters, from chess to investment management. Complexity and asymmetry between positive and negative information in the investment realm mean that investment research is much more valuable when focusing on the few but important things in the negative tail. That is, using simple rules to determine what not to do, what to avoid. Here are two examples from equity investing:

Earnings quality is a case in point where negative information is more valuable than positive. Whereas companies with the poorest quality of earnings tend to underperform, there is no pattern, of any statistical significance at least, showing that companies with the most conservative accounting do out-perform. Note that this does not mean that high earnings quality is not a good thing, rather it is an example of how competition reduces the value of positive information. There are so many investors that like state-of-the-art accounting that the stock price has already been bid up to reflect it. When we use information to avoid stocks with poor quality of earnings we get our money’s worth (Just as Annie and Joe did, by not bidding on the $1 note) as the bottom earnings quality quintile underperforms the market 75% of the time.

Another example is stocks with large short rebates or borrowing costs. The owner of Mayne Pharma Group stocks in Australia, for example, may currently lend this stock out for a fee of 7.75% p.a., placing the lending fee of Mayne among the 3% most expensive in Australia. This implies that if the stock were to perform in line with the general market, the owner would reap an excess return of 7.75%. The short seller would have to pay the lender this fee plus a compensation of any dividends paid by Mayne before he would turn a profit. So who do we think is going to win the bet? The shareholder, the short seller or – perhaps – someone else? On average, the short seller is winning by a solid margin. Otherwise the market would be grossly inefficient as excess return could be reaped simply by lending a portfolio of stocks. However, the real winner is neither the shareholder nor the short seller, but the investors who simply avoid buying Mayne, thus saving themselves the underperformance of the stock without having to pay for its borrowing costs. Avoiding stocks with the highest borrowing costs is another example of the value of negative information.

Diligently following an investment process, seeking to exploit information robust to error, in ways where its value has not been competed away, it’s possible to build a portfolio that, we believe, is superior to the market portfolio.  Just like Michelangelo, who said ‘Inside every rock there is a statue’, we begin with a large diversified universe of assets (our rock) and, through ‘chipping away’ what we don’t like, we end up with our desired portfolio (our statue).

Footnotes

• 1 Please note that Sector Asset Management is comprised of individual investment teams with their own distinct investment philosophies, that may, or may not, agree with the arguments in this essay. Therefore, the use of ‘we’, here, is solely related to the views of the author and his Omega team.
• 2 Nassim Nicholas Taleb, «Antifragile»
• 3 Sir Karl Raimund Popper, «The Logic of Scientific Discovery»
• 4 Goodreads, Steve Jobs quotes
• 5 Bergfinnur Durhuus and Søren Eilers, «On the Entropy of LEGO», Journal of Applied Mathematics and Computing
• 6 Gerd Gigerenzer and Daniel G. Goldstein, «Reasoning the Fast and Frugal Way: Models of bounded Rationality»