The issue with forecasting: Improving the art and science of prediction

January 15, 2016 11:00 AM

This is a story about the future. About how each year, audience-hungry editors and TV pundits gather with sell-side analysts to speculate on what the world will look like a year from now. Even if a prominent forecaster projects that a U.S. state might break off and sink into the ocean, he or she will be invited back to the roundtable next year despite the outcome, and no one will bring up previous forecasts. 

December’s special issues of The Economist and Bloomberg’s Businessweek project the state of the world in 2016, while CNBC has been giving precious air time to perma-bears and perma-bulls. Eye-popping predictions push aside more grounded, probable outcomes. And the language is certainly hedged: “This could happen, this might happen, if “x” happens, then “y” could occur… maybe.” But it’s rare that someone offers a simple probability of events. The desire for an absolute prediction is part of human nature, and part of the state of financial journalism today.

Lost in click bait headlines and the sound-bite predictions is an important variable. Each annual prediction is just as important as the manner in which each conclusion was reached. 

As Modern Trader explores a wealth of projections and predictions on the year, this year’s inaugural forecasting issue also dissects the numbers behind the numbers; offering insights into how the art and science of forecasting has evolved in recent years. We also provide a framework for how traders and investors can rely more on their own intelligence and insight than on the talking heads and sell-side analysts pitching their version of our future. 

At the heart of this discussion is a revolutionary book: “Superforecasting,” by Wharton professor Philip Tetlock and best-selling author Dan Gardner. The recent bestseller has unlocked a better understanding of what makes a better forecaster (hint: practice) and the five common habits of what the authors have called “Superforecasters”—a special breed of prognosticator who take a rigid and humbling statistical approach that enables them to outperform their peers in predicting the outcomes of future events.

These five elements of better forecasting flow through this narrative. This enables a stronger understanding of how and why so many forecasts have been wrong over the years, and how we can become better at this skill, even if we lack a statistical or financial background.

Using a probabilistic approach, EidoSearch will help us make better sense of the past so that we can understand the future. 

We’ll look back at the 2015 predictions of leading research teams of the world’s top investment firms, and gauge their performance. Shifting our attention to 2016, EidoSearch’s proprietary platform helps give us a range of probabilities based on today’s and past market conditions, allowing us to understand the statistics behind these sell-side projections and the associated psychology behind them. 
In addition, a number of contributors offer their insights into some of the markets’ most important questions going forward. Given that such projections are less scrutinized by compliance departments and communications teams that have a knack for hedging such forecasts, we can have a more candid conversation with our contributors regarding the statistics and reasoning behind their projections for oil, gold, the S&P 500 and more.

Next, we explore the wisdom of crowds and explore how one of our Fintech Seven companies—Estimize—is evolving to create more performance-based tools that reward the best forecasters and examine their consensus outlooks for the S&P 500,Apple stock, GDP growth and more. 

Finally, Tim Melvin of Marketfy pulls back the curtain to exploit the more common and nefarious reasons why bad forecasts make their way into the mainstream. In an Alpha Pages exclusive, Melvin explains why “everyone is doing it”—making unreliable forecasts—for the hope of fortune and fame in today’s media environment.

In a world of click bait, outrageous projections, statistical ignorance and blind guessing, the inaugural Modern Trader  forecasting issue aims to improve understanding of the numbers behind the numbers and takes a high school math teacher’s demanding approach to all forecasts and projections looking forward: “Please, show us your work.”

About the Author

Garrett Baldwin is the Managing Editor of the Alpha Pages and the Features Editor of Modern Trader. An author and Baltimore native, he earned a BS in journalism from the Medill School at Northwestern University, an MA in Economic Policy (Security Studies) from The Johns Hopkins University, an MS in Agricultural Economics from Purdue University.