(Bloomberg) -- Shortly after 10 p.m. on election night, bettors went all-in on Donald Trump.
Huge swaths of the country were suddenly turning Republican red on the electoral maps splashed on TVs and websites everywhere -- Florida was in the bag for Trump; North Carolina and Ohio were breaking that way too, and early results from Georgia, Michigan and Wisconsin were all heavily in his favor.
This, the gamblers told themselves, was a crystal clear sign that Trump had been badly underestimated by pundits and pollsters just like he had been in 2016 and that he would prove them wrong again and defeat Joe Biden. They wildly plunked down money on him on the British gambling site Betfair, pushing his odds of winning up from a mere 31% at the start of the night to 50% and, then, an hour later, 79%.
The frenzy quickly spread to much bigger and deeper financial markets. Investors from New York to Tokyo, seeing how gamblers were now nearly certain Trump would win, bid up prices on U.S. stocks and bonds and frantically dumped the Mexican peso. Futures contracts on the S&P 500 soared over 3% in little more than a half hour. The peso plummeted 4%.
There was just one problem with all of this: The gamblers didn’t know what they were doing.
They didn’t properly understand what they were watching unfold and they raced ahead to erroneous conclusions, pulling financial markets along for a wild ride with them. Yes, some of these early results pointed to a repeat of the kind of polling error that occurred in 2016 -- and that warranted a more modest re-calibrating of gambling odds -- but there was a key subtlety that was lost on the crowd that night: Because of the way mail-in votes were to be counted in several crucial battleground states, the early tallies were always going to disproportionately favor Trump.
They were not, as long-time election analysts repeated throughout the night, to be taken at face value.
But the animal spirits were up in markets and that message was ignored.
“I was looking at the betting markets and popping in and out of all of those sites,” Liz Ann Sonders, Charles Schwab’s chief investment strategist, said Wednesday morning. Once the main U.S. betting site, PredictIt, crashed, the market’s “focus was on Betfair. I looked at that one a bit more often than others just because it represents to some degree a market perspective.”
In an election unlike any other, the hazards of relying too much on often illiquid betting markets were once again laid bare.
Gamblers have “skin in the game” -- as investors like to point out -- and so their preferences, when taken collectively, can often be more accurate than traditional polls. Yet they are often betting in less liquid markets, which are prone to wild swings based on incomplete data.
In a faint echo of the colossal misread of Brexit four years ago, professional investors put too much confidence in punters’ snap judgments. That their mistake didn’t wind up costing many of them much -- stocks rallied the next day as Republicans looked to keep the Senate and Biden inched toward a victory that was finally declared on Saturday -- was a fortunate coincidence, but the lesson is the same.
Matt Glassman, a longtime amateur poker player whose day job is as a senior fellow at Georgetown University’s Government Affairs Institute, says betting markets, particularly less liquid ones, can send the wrong signal because they’re often full of casual gamblers who are willing to take fliers for entertainment and aren’t looking at the odds all that closely.
“At a racetrack, it’s well known that gamblers will over-bet on long shots because they like the rush or can’t tell the difference between 100-1 or 200-1 odds,” said Glassman, who made several small bets on Biden when the betting markets seemed way off. “You’re seeing that in political betting markets.”
On Betfair, the Nov. 3 vote attracted roughly $172 million of wagers from 7 p.m. on the East Coast to 7 p.m. the following evening. Only a fraction of that was needed to turn Trump from an underdog to an overwhelming favorite.
Complicating matters was the fact PredictIt crashed Tuesday night. That left many traders to take their cues from overseas punters, who may not have fully appreciated the forces behind the “blue shift” that even the major U.S. TV networks struggled at times to explain.
As the vote count got underway, you couldn’t be blamed for seeing the predictive power in wagering. S&P 500 futures slumped through most of the early evening on election night as Trump’s re-election was priced at less than a 50-50 proposition in betting markets.
But as gamblers enthusiastically boosted Trump’s chances on Biden’s disappointing tallies in Florida and the early count in states like Michigan and Pennsylvania, markets followed suit. Futures jumped around 10 p.m. as Trump’s odds peaked.
Christopher Harvey, Wells Fargo’s head of equity strategy, explained the surge in S&P 500 futures by simply saying “betting polls now favor Trump.” A note from JonesTrading flagged Betfair’s Trump odds moving to 60% as the reason 10-year Treasury yields were “coming in like a stone.”
But by sunrise in the U.S., those same gambling markets reversed themselves yet again, and were back to pricing Trump’s victory at less than 40%.
Of course, betting markets were just one of a multitude of factors influencing the financial markets. There were the real-time vote tallies, the ubiquitous electoral maps colored in red and blue, the various needles, and so on. And you can argue that Trump’s odds in the betting markets at the start of the night, at least, were more accurate than those implied by major polls.
But it wasn’t the first time political betting markets erred badly. Just over four years ago, gamblers in the U.K. heavily favored the “remain” vote during the Brexit referendum.
Back then, more money was placed on “remain,” skewing the odds and overshadowing the larger number of individual wagers that favored “leave.” Traders in the equity and currency markets who took those cues set themselves up for big losses once the votes were counted.
Setting aside conspiracy theories of well-heeled Londoners trying to influence sentiment with big wagers, if investors knew to ignore the top-line number and focused instead on the number of tickets for each outcome, they would have escaped the bruising sell-off.
During Brexit, bettors at least had the home-field advantage. Besides PredictIt, Americans are largely barred from gambling on elections. That meant in betting markets, at least, the outcome of the Trump-Biden contest was being determined primarily by punters in the U.K., Australia and elsewhere.
Granted, bettors did give traders a slight edge in anticipating short-term swings in various overnight financial markets. Nevertheless, Columbia Threadneedle’s Ed Al-Hussainy says it was a mistake for professional investors to get too caught up in the betting markets.
“I don’t need to see how bets were placed on an outcome in the U.S. election from someone in Australia,” said Al-Hussainy, who specializes in rates and currencies. “It tells me nothing about the future.”
The fluctuations in gambling odds and financial markets were emblematic of the whipsaw ride Americans themselves were on as they were glued to their TVs waiting for news of the results.
U.S. newscasters tried to explain and articulate each state’s idiosyncratic voting and vote-counting practices. Does Florida report its mail-in bailouts first or after in-person totals? What about Ohio, Michigan and Pennsylvania? While political junkies understood that red or blue mirages could disappear based on each state’s order of tabulation and urged patience, it was a message that was arguably lost in the heat of such a high-stakes moment.
Ultimately, gambling markets weren’t so much predicting the future as reflecting the present, and echoing many of the same distortions in the electoral maps Americans saw on their TV screens. It took until 9:20 a.m. the following day for Biden’s odds to finally match Trump’s Tuesday night peak on Betfair.
“Betting markets,” Al-Hussainy said, “are 100% noise, 0% signal.”
(Updates to reflect Biden was declared the winner by major networks in the 12th paragraph.)
For more articles like this, please visit us at bloomberg.com
Subscribe now to stay ahead with the most trusted business news source.
©2020 Bloomberg L.P.