The Rickards 2016 Trump prediction is the call he keeps reaching back to. In October 2016, Jim Rickards went on television and told the world Donald Trump would be the next president. The consensus was overwhelming. Every major model gave Hillary Clinton a 99 percent chance of winning — Nate Silver’s FiveThirtyEight had her at 93, betting markets had her at 90, and the New York Times gave her 85. Rickards read the same data and saw the opposite result. That takes a specific kind of analytical confidence.
The Call
Rickards appeared on BBC, ABC Australia, CNN, and Fox Business in the weeks before the election. He said Trump would win. Flat out. Categorically.
At the time, that looked like a man who had lost the plot. Every model, every pollster, every analyst who had made a career on election forecasting was pointing in one direction. The consensus was so strong that people were not even arguing about the outcome. They were arguing about the margin of victory.
Rickards was the editor of Strategic Intelligence at Agora Financial, now Paradigm Press. He had just published his book “The Road to Ruin” in November 2016. He was not an election forecaster by trade. He was a macro economist who studied how complex systems behave when the consensus is wrong. And he saw something in the polling data that the models were missing.
The Polling Problem
Rickards’ argument rested on three observations.
First, social desirability bias. The idea is simple. When a pollster calls and asks who you are voting for, you might not tell the truth if answering honestly feels socially unacceptable. In 2016, supporting Trump carried a social stigma in many circles. People who planned to vote for Trump were telling pollsters they were undecided. The polls were not measuring support. They were measuring willingness to admit support.
This is a known problem in polling. It affected the 2015 UK general election, where polls underestimated the Conservative vote. It affected the Brexit referendum, where polls showed Remain ahead. Rickards had studied both. He saw the same pattern in the 2016 U.S. election.
Second, betting markets are not reliable predictors of political outcomes. The argument for betting markets is that they aggregate information through money. People put capital where their mouth is, and the price reflects collective wisdom. Rickards had spent his career inside financial markets. He knew that betting markets work well when the question is about a financial asset where the participants have real information. Political betting markets attract a different crowd. The participants are not informed traders. They are people with opinions, and opinions are not the same as analysis.
Think of it like a poker game where half the players are betting on feelings. The odds do not reflect the true probabilities. They reflect the crowd’s emotional state.
Third, the anecdotal evidence. Rickards did road trips through Spokane, Washington and the Ozark Mountains before the election. He talked to people in diners, at gas stations, in small towns. He saw Trump flags, Trump signs, Trump enthusiasm that the polls were not capturing. The mainstream media was covering the campaign through the lens of rallies, fundraisers, and horse-race polling. Rickards was looking at the ground.
The combination of these three factors told him the polls were wrong. Not slightly wrong. Structurally wrong.
The Asymmetric Bet
Rickards framed the election as a trading opportunity. He called it one of the greatest asymmetric trades of all time — even better than Brexit, which he had also predicted correctly earlier that year.
The logic was straightforward. The market was pricing in a Hillary victory with near certainty. If Hillary won, the market reaction would be muted. A 1 or 2 percent move. Nothing to trade. But if Trump won, the market would panic. Dow futures would drop 10 percent overnight. Gold would spike $100 an ounce. The volatility would create a massive opportunity for anyone positioned for it.
Rickards was not predicting a Trump victory because he wanted it. He was predicting it because the risk-reward calculus was so skewed that even a small probability of a Trump win created a trade worth taking. And he thought the probability was much higher than the market believed.
Election Night
The returns came in on November 8, 2016. State by state, the map turned red in places the models said were safe blue. Florida. Ohio. North Carolina. Pennsylvania. Wisconsin. Michigan.
The Dow futures dropped 800 points overnight. Gold spiked. The market did exactly what Rickards said it would.
Then something strange happened. The market reversed. By the next morning, futures were recovering. Within days, the Dow was hitting new highs. The Trump rally had begun.
Rickards was right about the direction of the initial move. He was wrong about the duration. The panic lasted hours, not weeks. The market decided that Trump’s economic agenda — tax cuts, deregulation, infrastructure spending — was good for business. The gold spike that Rickards predicted never materialized in the way he expected, because the dollar strengthened and risk appetite returned.
But the call itself was correct. The man who went on TV and said Trump would win while the consensus gave him no chance was vindicated.
The Method Behind the Call
The 2016 election prediction is a case study in how Rickards thinks. He builds models that identify when the consensus is wrong. That is a different skill than forecasting the future, and it is the skill that matters.
The framework comes from his work at Long-Term Capital Management, the most famous hedge fund collapse in history. Rickards was the principal negotiator of the LTCM rescue in 1998. He watched Nobel laureates discover that their models could not handle the real world. The lesson stayed with him: the most dangerous place in markets is inside a model that everyone believes.
LTCM taught him something about the relationship between models and reality. Models are simplifications. They work in normal conditions. But they fail at the edges, and the edges are where the money is made or lost. The 2016 election was a textbook case of a model edge. The polling models were built on assumptions that did not hold. Social desirability bias was a known problem, but the models treated it as a minor adjustment rather than a structural flaw. Rickards treated it as the main event.
He took that lesson to the CIA, where he advised the Director of National Intelligence on financial threats. He built predictive analytics systems that applied complex systems theory to market data. The systems caught signals that the rest of the world missed — the 2006 liquid bomb plot, the 2008 financial crisis, and, in 2016, the election surprise.
The common thread is pattern recognition. Rickards looks for situations where the model and the reality have diverged. He looks for hidden leverage, hidden information, hidden sentiment that the consensus is not accounting for. In 2016, the hidden variable was social desirability bias. The polls were measuring something that did not exist, and the models were amplifying the error.
What the Call Means
The 2016 prediction is one of four major macro calls that define Rickards’ track record. The 2008 crisis, which he warned the CIA about in 2006. Brexit, which he predicted when the consensus said Remain. The COVID crash, which he called in January 2020 with a note titled “CONTAGION.” And the Trump victories — 2016 and 2024.
Most professional forecasters do not get one call of that magnitude right. Rickards has four.
The full story of his career, including the calls that did not land, is in the Jim Rickards dossier. That dossier covers the LTCM rescue, the CIA work, the Currency Wars thesis, the gold predictions, and the misses that matter. The 2016 Trump prediction is the clearest example of his methodology applied to a specific, verifiable, public event. He went on television, made a prediction, and was right.
The lesson is about framework. Rickards’ method for identifying when the consensus is wrong works across different fields. Financial markets. Political elections. Geopolitical events. The mechanics are the same. The consensus is always comfortable. The consensus is always wrong about something. The question is whether you have the tools to see what it is missing.
Rickards had those tools in 2016. He used them. And he was right.
Flak Jacket Finance covers investment newsletters as an independent third party. We do not reveal paid picks, we do not trash the gurus or products we cover, and we do not sell the promos we cover.