Probabilities & Base Rates

The great and powerful Michael Mauboussin wrote some of our favorite books on investing and thinking. One of his lesser-known "books" is The Base Rate Book. Since reading it, we have put thinking in terms of probabilities and base rates front and center in our mental model tool kit.

For us, there is no sense in trying to be precise about things that are impossible to be precise about. A better approach is using base rates and probabilities. Start with the base rate - the past probability rate of a certain event - and then adjust the probability as new information comes in.

Let's take the COVID-19 pandemic as a current example. Of course, as a disclaimer, you should know we are not virologists or epidemiologists, only curious thinkers that try to put numbers into some type of context for proper interpretation. An extreme extension of this is attempting to figure out the probability someone will die from the virus. What is the kill-rate of this thing? Tough to know, but we can use history as a decent starting place. In 2009, the H1N1 pandemic (yes, it was a pandemic like COVID-19, yes it was a new virus-like COVID-19, yes it was very deadly…but no one seems to remember that for some reason) blindsided the health of the world. Here is how the Center for Disease Control & Prevention (CDC) describes it:

In the spring of 2009, a novel influenza A (H1N1) virus emerged. It was detected first in the United States and spread quickly across the United States and the world. This new H1N1 virus contained a unique combination of influenza genes not previously identified in animals or people…From April 12, 2009 to April 10, 2010, CDC estimated there were 60.8 million cases (range: 43.3-89.3 million), 274,304 hospitalizations (range: 195,086-402,719), and 12,469 deaths (range: 8,868-18,306) in the United States due to the (H1N1)pdm09 virus. Additionally, CDC estimated that 151,700-575,400 people worldwide died from (H1N1)pdm09 virus infection during the first year the virus circulated. Globally, 80 percent of (H1N1)pdm09 virus-related deaths were estimated to have occurred in people younger than 65 years of age. This differs greatly from typical seasonal influenza epidemics, during which about 70 percent to 90 percent of deaths are estimated to occur in people 65 years and older.

In trying to get an answer to our question, we first must estimate the base rate of infection for COVID-19. As mentioned above, the CDC said approximately 61 million people (42-89 million estimate range) were infected, and that was without any of the social distancing, quarantining, border closings, flight restrictions, restaurant & bar closures, shelter-in-place, mass hysteria going on like there is today. That seems like a better guess than all these crazy prognostications the fear-mongering media is projecting. To be conservative, let's use the 89 million people estimate, which is roughly 27% of the 330 million U.S. population. If 27% of the population gets COVID-19, then we need to figure out the mortality rate of those infected. Again, the fear-mongering media suggests it is 2% (our guess is that it will be lower, but higher than the rate of H1N1, which was 0.02%). To be conservative we will use 2%, and putting it all together, you get a 0.5% mortality rate. That seems pretty low to us given the mass hysteria that resulted, especially given our 2% mortality rate feels aggressive. Others don’t seem to think so. We will put that in the good to know category.

Here is another application, this time from a macroeconomic perspective. Once again, another disclaimer: we are not macroeconomists nor do we spend time trying to divine the future direction of economic activity, interest rates, currency moves, etc. That said, it is helpful to think about the potential impact of the pandemic. It is still very early in trying to estimate the negative hit to economic activity from COVID-19, especially as the uncertainty of when the economy will "reopen" and the daunting 2nd order effects of the whole debacle. But it will probably be bad. Very bad.

In our post-war economy, the worst year-over-year decline in U.S. economic activity (measure by gross domestic product) was 12% back in 1946 when our economy was demobilizing after WWII. That is a good base rate as any to start with given what is going on today and it is a huge decline. Our economy is roughly $22 trillion, and a 12% decline equates to a loss of $2.6 trillion. All in, it will be a big decline whatever happens.

Now let's extrapolate that to the U.S. stock market, which is valued on economic activity and corporate earnings. At its peak in February, the total market capitalization of all the public companies in the U.S. was around $36 trillion. At the nadir in March, it was $23 trillion, a loss of $13 trillion. Today (April 15th) it is $28 trillion so we lost $8 trillion in market value due to the virus fallout. You read that right - $8 trillion! To us, that seems a bit extreme compared to the approximate $2.6 trillion of potential loss to economic activity. But that is what the stock market does: shoots first and asks questions much later. Of course, the counter-argument to that set of facts is, perhaps, the market was "overvalued" in February at its peak. Maybe so. Probably.

Bringing it back to base rates for the market, unfortunately we don't know if the market will continue shooting for a while or let up and start asking questions. But we do know the mean and the median peak-to-trough declines for the U.S. equity market, as measured by the S&P 500 Index, going into and through recessions are 25% and 35%, respectively. So far, we were down a bit more than 35% at the bottom in March, about spot on the median decline. However, this recession does not seem to be "average" in any way, shape, or form. Rather, it seems there is a high probability that it will be more severe than average which argues for, maybe, lower price levels from here. Who knows? But for us, putting the numbers in context is helpful.

Again, there is no sense in trying to be precise about things that are impossible to be precise about. However, we do find comfort in thinking in terms of probabilities and base rates to help us put macroeconomic and capital market events in context. This framework helps filter out the sometimes outlandish, baseless noise of the media's irrational prognostications and grounds our interpretations of events and, ultimately, our decision making in something meaningful and useful.