“When the facts change, I change my mind. What do you do, Sir?”

John Maynard Keynes had a reputation for flip-flopping. So when one of his detractors pointed out his ever-changing opinion, he turned the tables with the mocking retort above. It’s one of my favorite quotes because it lays bare the conflict between an unassailable logic and a deeply human sentiment. We prize constancy, probably because it suggests virtues like fortitude and durability. We like ‘unwavering’ and despise ‘flip–flopping.’ Unfortunately, personal virtues turn out to be liabilities to effective forecasting.

Professor Philip Tetlock of the University of Pennsylvania has been researching how people assess probabilities for 30 years. His book Superforecasting: The Art and Science of Prediction lays the groundwork for why most people fail to beat the proverbial dart-throwing chimpanzees―but also how it’s possible to succeed. A big part of the problem is that we are not always rational decision-makers. Apparently, the million or so years we spent evolving as hunter/gatherers hard-wired some sub-optimal decision rules into our brains. The fact that graduate students at Ivy League schools―trained in statistics―continually get trapped into making irrational choices should be a warning.

Tetlock and his colleagues conducted a massive experiment not to identify our weaknesses. Rather they endeavored to find the superstars among us and figure out how they worked differently. They got a boost from an unusual place, the U.S. Government. The National Intelligence Director sponsored a live contest with consortia of thousands of participants making millions of forecasts on events related to national security. To spoil the story somewhat, Tetlock’s group won by a large margin. They even beat the estimates produced by our intelligence agencies. For those who fear the singularity is here already, computer algorithms and deep learning played a secondary role. Most of the heavy lifting seemed to depend on the very human personality and mental characteristics of almost ordinary people. So there’s still hope for the rest of us.

One trait that distinguished the superforecasters from the normal humans was their approach to information. They were not domain experts, but rather gathered data from a multitude of eclectic sources. Tetlock denotes forecasters who view the world through a top-down lens as ‘hedgehogs,’ and draws the distinction with ‘foxes,’ who take an interdisciplinary approach to assembling data. This characteristic bears directly (I think) on how they revised their views. The superforecasters were humble about the limits of information. As Tetlock describes, they viewed their beliefs not as “treasures to be cherished, but as hypotheses to be tested.” The superforecasters constantly asked what it would take to convince them they were wrong. My interpretation: since they knew they were foxes, they were able to be dispassionate about their estimates.

Consider how different these people are from the classic hedgehog we encounter – the media pundit. When the media covers a story centered on China, we inevitably hear from a former China intelligence specialist. The expert states her views with eloquence and conviction. The effect is that the viewer is cowed into a kind of intellectual surrender. This lady analyzed classified information on China for 15 years inside the State Department, for crying out loud!

Despite their clear inferiority in domain knowledge, Tetlock’s foxes managed to outperform the best and brightest of our intelligence apparatus. Clearly, it was about temperament and disposition. Being open to new data-shifting probabilities rather than trying to re-interpret the new information into a narrative.

In this quarter’s webcast, we discussed some of the tectonic forces that are shifting the financial landscape.  The extraordinary experiment in large scale asset purchases by central banks is coming to an end. The Federal Reserve is raising borrowing rates into positive real numbers. The synchronous upswing in world economic activity is beginning to splinter, with growth rates diverging noticeably by country and region. In short, it is a world that is offering new and contrasting signals.

This past quarter, we made some modest changes to our strategic asset allocation in light of new information. We expect the cross currents to continue, and, with that, the opportunities to deploy capital. We will evaluate the changing landscape with a view to discerning what is genuinely impactful given the expectations discounted in asset prices. Many of these decisions will involve fine distinctions of probabilities. It’s unlikely that any single opportunity will inspire a table-pounding conviction. But Tetlock’s work suggests that it often pays to ignore the confident pontifications of the hedgehogs and heed the humble, evidence-driven, protean voices like Keynes’.

— T. Brad Conger, CFA, Deputy Chief Investment Officer

On a quarterly basis, Hirtle Callaghan publishes our perspective on the current market. If you would like to be added to our distribution list, please contact us.

Investment Perspective 2Q2018 Excerpt.