The Wisdom of Crowds

This article by Barry Pasikov is excerpted from a letter of Hazelton Capital Partners.

In the days leading up to the 1948 Presidential Election, all of the polls, political analyst, and media predicted that Republican Governor, Thomas Dewey, would easily defeat incumbent Democratic President, Harry S. Truman. In fact, the Chicago Tribune was so confident in the outcome, it printed its first edition (150,000 copies) with the now famous headline: “Dewey Defeats Truman.” As history has taught us, never pop the champagne bottle until the final votes have been tallied. Nearly 70 years later, Americans were treated to another “Dewey” moment when Donald Trump defeated Hillary Clinton, recording the biggest political upset in our nation’s history.

Say what you will about 2016, it was anything but boring, especially when it came to U.S. politics. Virtually no media outlet, political analyst/insider, or statistician could create a scenario in which Clinton would not become our nation’s 45th President. Even Nate Silver, the statistician who correctly predicted the presidential outcome in 49 out of 50 and 50 out of 50 states in 2008 and 2012 respectively, had the probability of a Clinton win at a conservative 72% the day of the election. This conventional wisdom was also reflected in the stock market. The final two days before the election, after Clinton’s email investigation was once again closed, the S&P 500 index rallied over 2.5%. Tuesday night, as the election results began to filter in and a Clinton Presidency began to fade, the S&P 500 index collapsed, the futures indicating a 6% decline by the open on Wednesday morning. That, too, proved inaccurate, as the S&P 500 had not only fully recovered, but even posted a 1% gain by the close of trading on Wednesday afternoon.

Now, more than ever, we live in a world where data is constantly collected and analyzed. When you get into your car, your phone, more often than not, predicts the time to your expected destination based on previous driving habits and your calendar. Open your email and Amazon has sent you a list of products you “need.” And then there is the famous story of an infuriated father screaming at the manager of his local Target store because his teenage daughter was receiving coupons for pregnancy and baby items. His daughter received these coupons because her purchasing patterns were similar to women who were or became pregnant. It turns out that Target knew the man’s daughter was pregnant before she admitted it to her father. So, how is it that Target can know that a random teenage girl is pregnant but the pollsters, political forecasters, and the media can get the election results so terribly wrong?

In James Surowiecki’s book, The Wisdom of Crowds, he explains how a group’s predictive ability can be far superior to nearly all of the collective individuals’ estimates. The power of crowdsourcing (harnessing a collective’s input) has been successfully demonstrated in marketing campaigns, beta testing, funding creative projects, and even predicting the outcome of elections. However, what Surowiecki also discovered was that as powerful and accurate a tool crowdsourcing can be, it is also very susceptible to biases. One of the biggest biases impacting collective predictions is independence. Studies have shown that when a collective’s estimates are made public, it will influence future estimates and negatively impact the group’s predictive ability. We are all familiar with peer pressure and have at some point succumbed to its influence. Our primordial brains have been hardwired to seek out the protection of the pack as a survival mechanism; distancing yourself from the collective is not only dangerous but socially alienating.

Leading up to the Presidential Election, I was part of the general consensus that believed that a Hillary Clinton Presidency was all but a certainty. I only knew of two people that would openly admit that they were voting for Trump. Of course, I live in Illinois where it has been rumored (but never proven) that the state legislature is considering reducing election costs by only listing the Democratic candidates on the ballots. But even friends and colleagues outside Illinois did not show any support for Donald Trump. Of course, my informal poll was not diverse or expansive enough to represent the zeitgeist of the nation. Polls and surveys are biased and inherently unreliable. They are dependent on a number of factors in order to be accurate: Sampling diversity, non-conformity, accurate reporting, and honesty. It is believed that the reason the 2016 election polls and results were so divergent was mostly because people would not admit to be voting for Trump because of the social stigma attached to their decision. Without accurate data, polls are worthless.

Over the past 15-20 years, business news networks have created a niche segment within the news industry by providing real-time market quotes, business news stories, and investing strategies. Unlike conventional news, which reports on the events that are occurring at the moment, business news focus on the future expectations of a company’s share price. To help in their reporting, the networks invite “experts” to share investing insights. Leading up to the election, nearly every market pundit who took to the airwaves warned of the dire market outcome if Donald Trump were to miraculously win the election. Because these “market experts” shared the same general consensus, the investing public took these insight to be highly probable.

In the short-run, markets are surveys. They are the aggregate predictions of investors. It is widely believed that markets are efficient because they distill all news and information available. But even though mathematics is involved in determining the price of a stock, investing is not a science. Just like election polling, markets are very susceptible to biases, including investor emotions of fear and greed. On election night, investors disbelief turned to terror as a Donald Trump became the presumptive winner. The market’s short-term sell-off was an emotional and Pavlovian response to an impossible outcome, and therefore, the market swoon was believed to be a self-fulfilling prophecy. But in the long-run, markets, and more importantly, a company’s share price, will ultimately reflect the fundamental strength or weakness of a company. For the long-term, patient investor who was feeling the same political distress, the market swoon was an investing opportunity.

Opportunity always comes gift wrapped in a fog of uncertainty. And given the “Trump Tweets” that we have witnessed so far, the next four years promise to be very unpredictable. There are numerous unknowns about a Donald Trump Presidency and the future of the economy. Many people believe that Trump’s desire to lower corporate taxes will be a boon to corporate profits. Or that his promise to revamp Dodd Frank will lead to less regulation and benefit the financial sector. And just as many people are fearful that a Trump Presidency will lead to tariffs, trade wars, alienating countries, and political unrest. But mostly, we have no idea what he is going to do or “Tweet” about next. Just as the market got it wrong on election night, the year end rally could easily have been an overreaction leading to a “Trump Dump,” or the beginning of another extension to the ongoing bull market leading to a “Trump Bump.” Those who pretend to know are either very lucky or trying to sell you something.

Hazelton Capital Partners is not in the prediction business. We are not trying to outsmart the investing collective. We focus on our competitive advantage: Ignoring the “sound and fury, signifying nothing,” and instead, continuously search for our next investment opportunity where the price paid for an asset is significantly lower than its expected future value, ultimately leading to long-term success.