Too Much Information

The conventional wisdom in investing is that more information is better than less. In the past Peter and I have sometimes questioned whether we have an informational deficit, but over time we’ve realized that it is not quantity of information that counts, but knowing which information is important and analyzing it correctly. In fact, to the extent that too much information leads to overconfidence and emotional bias, more information can become a stumbling block.

Last year we wrote an article in which we compared insurance to investing. The focus was around diversification: How insurers are in the business of selling diversification for a profit. We feel there is a lot that investors can learn from the way insurers go about their business, which in many ways is very similar to the business of investing.

This time our focus is around information, and how that relates to the selection side of the process.

Focus on key variables and aggregate outcomes

Insurance is about aggregate outcomes. You isolate a few key variables – age, gender, smoking, family history and other health markers (for life insurance) – which largely explain variations in aggregate claims (or base rates). These variables are readily determined at low cost and can be applied broadly. This is called underwriting, the insurance equivalent of selection.

Further variations in individual claims are treated as random, either because they are random, or because the cost of collecting and assessing further information that might explain these variations outweighs any potential benefit of doing so.

Your life insurer probably won’t send someone out to interview you and find out about your life. The conclusions would likely be unreliable, especially to the extent that they differ from what the key variables are suggesting. Apart from the added expense, your insurer wouldn’t disregard the base rates in favour of a compelling story when setting your premium.

Your life insurer also won’t waste too much time second-guessing its underwriting process for every random deviation from the average, as long as the aggregates are in line. Claims are expected.

What is important about this approach to selection is that it is logical, business-like, and reasonably free from emotional bias. And it works.

The informational deficit mirage

In contrast, the investment world is obsessed with specific outcomes and individual stories. There are huge stakes attached to correctly predicting big winners and losers, which drives the investment community to generate and consume as much information as possible – as though this would somehow enable us to see the future.

The idea is that more information leads to more accurate predictions. This seems intuitive. The reality is that more information leads to greater confidence in forecasting, but not greater accuracy. This information-driven overconfidence causes investors to take big bets on specific outcomes without the required accuracy.

 Stories appeal to our emotions

The obsession with individual stories creates another problem: Investors are far more likely to allow emotional biases to bypass logical reasoning when stories are involved. In their paper On the Psychology of Prediction, Kahneman and Tversky demonstrate how compelling, but ultimately irrelevant information, can cause people to disregard prior base rates when making predictions, leading to systematic prediction error.

This sort of thing happens all the time in investing. Overconfident from all the information we have about a company, we fall in love with the story and often lose sight of the few key variables that determine aggregate outcomes.

It is staggering to see how different the approaches of the insurance and investment industries are, considering how similar their businesses are.

What are the key variables in investing?

Perhaps one reason why investors consume so much information is because there isn’t broad understanding of what the key variables in investing are. Understanding the causality of investment returns helps to isolate which variables are important. There are only three broad variables that really matter:

Higher profitability is associated with higher reward and lower risk. Higher leverage is associated with higher reward and higher risk. Higher prices are associated with lower reward and higher risk.

These are the three key variables that drive aggregate outcomes in the stock market. Every bit of information that you consume should be directed at trying to understand these three aspects of a business. Every bit of information that draws your attention away from these variables is probably impairing your judgement. The investment community generates far more of the latter.

Let’s do an example, then add a story to make it interesting…

You have the opportunity to invest in a persistently loss-making business that is operating in one of the most competitive and capital-intensive industries in the world. The company is highly geared, issues a lot of new shares every year, and burns through cash. The industry is cyclical. If the company was as profitable as its competitors, it would be trading at a PE of 30x – 40x.

Loss-making, leveraged, expensive… What do you think the base rates of return look like for a group of companies with these characteristics? Would you invest? Maybe this one is an exception.

What if I told you the company is a fast-growing, high-flying, tech company that is disrupting one of the most established industries in the world. The CEO is a billionaire genius who has successfully grown several startups into multibillion-dollar businesses.

How about now? The story tells you little about profitability, leverage or price, but it is representative of a mental model of success. Would you disregard the prior base rates in favour of the story? The market does this every day.

Investing like an insurer

We’ve always preferred a more actuarial approach to investing. The idea is to drill down to the key variables, and to ignore marginal information when it is no longer useful, or when it is introducing emotional bias that might cause us to lose sight of the key variables.

Though we often talk about individual companies, we always think in terms of base rates and try to build a portfolio of highly profitable businesses, with prudent leverage at attractive prices. It makes it difficult to keep portfolio updates interesting when the reasons we bought company X are always the same.

Finally, in investing as in insurance, there are always claims. Some periods are worse than others. High base rates combined with small samples can be dangerous. This is why diversification and time horizon remain critical to achieving investment success.