Book Review by Adam Pascarella: Superforecasting

Editor’s Note: This is the fourth book review we are publishing as part of our Get Paid to Read contest. Vijar Kohli’s excellent review of The Intelligent Investor was published last week in sharp juxtaposition to the temporary meme rally in stocks like GameStop and AMC following the return of Roaring Kitty on Twitter.

There are few books that stay with you for years after you read them and Superforecasting by Dr. Philip Tetlock and Dan Gardner was one of them for me. I was excited to read the following review by Adam Pascarella. Adam is the founder and CEO of Second Order Capital Management. He is a graduate of the University of Michigan and the University of Pennsylvania Law School.

As investors, our job is to make predictions about the future. This is true whether we are value investors estimating the free cash flow for a microcap industrial company or macro strategists considering the next Fed decision. The job requires us to use our best judgment to navigate an increasingly uncertain world. 

It’s almost like we are all surfers trying to catch waves. We get up on our surfboards, make some predictions, and, ideally, ride waves as they crest. We don’t always get it right, but we make the best decisions with the information that we have. 

But what if we could improve the quality of our predictions? What if we could leverage publicly available information and key analytical principles to more accurately forecast the future? 

These are the enticing possibilities that Superforecasting raises. 

Originally published in 2015, Superforecasting is a guidebook for helping readers mimic so-called “Superforecasters,” who are known for making extremely accurate predictions about global and financial events. While some may think that these Superforecasters have access to the latest information or have the best teams behind them, authors Philip Tetlock and Dan Gardner reveal that they are often solitary, “normal” individuals who have a passion for forecasting. It isn’t necessarily their backgrounds or careers that make them Superforecasters. Rather, it is their curiosity, mental flexibility, and process-oriented analytical style that separates them from the rest.  

Forecasting a Murky Future

The book covers a significant amount of ground, ranging from overarching principles about forecasting to the specific steps that Superforecasters take when making predictions. You’re probably already familiar with some of the higher-level concepts in this book, as they have been shared among many equity and special situation analysts (particularly in the value investing community). 

For instance, many Superforecasters rely on base rates when making their forecasts. As Michael Mauboussin often states, base rates are critical when projecting things like sales growth rates and future cash flows. Tetlock calls it the “outside view,” which traces back to Daniel Kahneman’s monumental book Thinking Fast and Slow. Essentially, when forecasting, we can significantly increase our accuracy by using base rates as a starting point and then tweaking our estimates based on the “inside view,” which is the specific background and details of the thing we are trying to project. 

Another familiar concept may be the problems of public forecasting. From activist investors managing billions of dollars to retail investors posting on Twitter, it becomes that much more difficult to adjust our forecasts when we publicly commit to a certain stance. As Tetlock argues, public commitment to a certain belief makes it that much more difficult to change, which increases the likelihood that our forecasts—which constantly evolve—are off the mark. 

A secret sauce of a Superforecaster is the ability to quickly update their views when facts change. It’s why investors like Stanley Druckenmiller are so great (and so rare). They truly adopt the maxim “strong opinions, loosely held” and can make decisive decisions in an increasingly dynamic world. 

Much of Superforecasting comes down to getting out of our heads and imagining if the other side is correct. It echoes the late Charlie Munger’s well-known maxim: “invert, always invert.” By questioning our assumptions, we can deliver forecasts that are less biased and more probable. 

Tetlock supports his argument with one fascinating study from researchers at the University of Basel. The researchers asked participants to make a judgment about a particular prompt. Later, they asked those same participants to assume that the first judgment was wrong and to produce another estimate. The researchers found that combining those participants’ first and second judgments produced a more accurate response versus the first judgment alone. The same was true if those participants made a judgment and then waited several weeks before making the second judgment. Called dialectical bootstrapping, this process lets individuals (including Superforecasters) leverage the wisdom of the crowds within themselves. 

Yet another key principle that Superforecasters leverage is so-called “Fermization.” Tetlock references Italian American physicist Enrico Fermi and his propensity to break down difficult questions into their component parts. Fermi recognized that some parts of the future are knowable and some are unknowable. Sure, we may need to make some black-box guesses on some seemingly unknowable elements of a forecast. However, if we focus more of our efforts on trying to predict the knowable, we’ll deliver more accurate estimates. 

There are plenty of other Superforecasting principles that are likely familiar to you. Great Superforecasters don’t hesitate to engage in postmortems to gauge how their forecasts could have been improved. They also engage in scuttlebutt research to supplement their forecasts. Simultaneously, they recognize that not all information is created equal. Especially in today’s world where we are flooded with data and information, it is even more critical to focus on what truly matters. It isn’t an easy task, but Superforecasters can more readily identify the key drivers, questions, and assumptions that will tip a forecast in one direction or another. 

With all of that said, scope is key. Not everything can (or should) be forecasted. Accurately forecasting the winner of the 2036 presidential election, for instance, is probably a fool’s errand. Some forecasts belong in the “too hard pile” and Tetlock recommends that you focus on questions in the “Goldilocks zone of difficulty.” It requires a humble mind to recognize that your time is best spent on questions and forecasts that are actually forecastable. 

Practice, Practice, Practice

At the end of the book, Tetlock provides ten commandments for aspiring Superforecasters. It is a helpful guide as you work to implement the advice and lessons from the book. The bottom line, however, is that the best way to get better at forecasting is through practice. It’s about making inaccurate forecasts, learning from your mistakes, and using that feedback to improve your future forecasts. It is a mentally demanding practice, but it is a practice that can set you apart from other forecasters. 

Superforecasting is an entertaining read. I’m sure you’ll get something out of it, whether you are a brand new investor or have been investing for decades. Fortunately, Tetlock provides some simple, actionable advice that you can leverage as you navigate any type of market environment. While it may take some time, implementing the lessons in this book may be one of the best decisions that you make this year.