Retail traders are beating big firms in guessing where U.S. interest rates will go next
Referenced Symbols
For the past three years, Brennan Robbins, a 32-year-old data scientist in the healthcare industry, has been a part-time trader in the prediction market. That means betting on everything from whether 2024 will be the hottest year in recorded human history to whether the coming movie “Megalopolis” will be hated by critics on Rotten Tomatoes.
In his world, thousands of dollars can be made or lost on a single bet about the outcome of future events, and when it comes to one of the most important issues facing financial markets, Robbins has been largely right this year. He anticipated that the Federal Reserve would be unable to cut interest rates by nearly as much as many had hoped.
Since the 2021 launch of the first federally-regulated exchange dedicated to trading on event outcomes — a platform known as Kalshi — the results are in: Investors like Robbins do a better collective job of anticipating where the Federal Reserve’s main interest-rate target will likely be, months ahead of time, than the U.S. central bank itself; fed-funds futures traders; and big-name firms like Bank of America, Goldman Sachs, J.P. Morgan Asset Management, State Street Global Advisors, and UBS Global Wealth Management.
Based on the Kalshi leaderboard as of Wednesday, Robbins has made $81,190 in profits since 2021 — with practically all of the gains coming this year, he said in an interview with MarketWatch. Robbins, who has positions on 36 different events, said he began trading last December on the number of Fed rate cuts in 2024.
He purchased a total of 65,000 shares, at 4 cents a share, on the likelihood of no 2024 rate cuts, when the odds were around 4%. Robbins has since made a total return of $11,511.94 from that single position after the odds kept climbing above 20%, based on snapshots of trades seen by MarketWatch. Separately, Robbins also put on positions in anticipation of just one or two rate cuts in 2024, which have had a combined total return of $10,426.12. The prediction market reflects an aggregate view of many people’s opinions, so there’s always a loser on the other side of his trades.
“The true underlying story about inflation is that it is only gradually moving lower, so I have been fading extreme market moves and expect the Fed to be easing less than expected,” said Robbins, who lives in New York.
A look at how individual traders have performed against their bigger counterparts underscores just how much predictive power lies in the hands of Main Street, which reaches its own set of conclusions based on the same data and remarks by central bankers that professional Wall Street traders sift through. Kalshi’s rise as a federally regulated exchange just happened to coincide with what turned out to be both the worst outbreak of U.S. inflation and the most aggressive rate-hike cycle in about 40 years.
Since the first half of 2023, retail traders have had a lower average error rate than much of Wall Street when it comes to guessing where the U.S. consumer-price index will be, based on data compiled by New York-based Kalshi. In the chart below, Kalshi defines the error rate as the average difference between the prediction and the actual headline CPI reading each month.
Average error rates on headline CPI readings for Kalshi’s top trader and all of its traders versus much of Wall Street. Data as of Wednesday. PHOTO: KALSHI
See also: Inflation rate slows in April, CPI shows, but prices pressures still haven’t gone away
Having a more accurate feel for the direction of inflation means also having a better sense of where interest rates might go.
Over much of 2023, Robbins said he often found himself on the losing side of bets on Kalshi when it came to guessing where a narrower measure of the consumer-price index, known as the core rate, would land on a monthly basis. In other words, his intuition about the likely path of inflation wasn’t very good. He estimates he lost about $4,700 by siding with the views held by major banks, until he concluded that other Kalshi traders were often more accurate.
Referring to the instant-messaging platform where many Kalshi traders communicate with each other, Robbins said “a lot of us on Discord were listening quite carefully to what the Fed was saying and knew there would be fewer 2024 rate cuts if inflation turned out to be sticky.”
An aspiring data scientist weighs in
When there is sufficient liquidity on Kalshi, `it is quite accurate’ and ‘even outdoes’ other platforms
— Blake Law, math major at Georgia Tech
A study of Kalshi versus the more widely followed CME FedWatch Tool was done by Blake Law, a 22-year-old aspiring data scientist and math major at Georgia Tech in Atlanta. Law blended his interests in data science, the Fed, and personal trading on Kalshi into an interest-rate analysis that concluded Kalshi, made up of mostly retail traders, provided a more accurate idea of where the fed-funds rate would be as of the central bank’s September, November, and December meetings last year.
Average error rates for Kalshi traders and fed-funds futures traders in predicting the outcome of the Fed’s September, November and December meetings last year. PHOTO: BLAKE LAW
This year, traders on Kalshi and fed-funds futures have moved closer to each other’s views following a shaky start in January.
At the start of the year, the latter group was pricing in a more than 60% chance of six or seven quarter-point rate cuts from the Fed by December — or roughly twice as many moves as policymakers themselves had indicated. Then, during the first two days of February, fed-funds futures traders saw an almost 60% chance that the first rate cut would arrive by May.
Kalshi traders had a somewhat different take on what would likely unfold. They expected Fed officials to cut interest rates in May by less than fed-funds futures traders expected.
The two sides ultimately aligned as the months rolled on, with the view that interest rates would probably remain at a more-than-20-year high of 5.25%-5.5% this month. The Federal Reserve ended up leaving borrowing costs unchanged on May 1, extending a pause that’s lasted for almost a year.
Expectations on Kalshi versus CME FedWatch Tool for where the Fed’s main interest-rate target would be as of May. The green line reflects a bout of illiquidity that occurred on Kalshi last November. PHOTO: KALSHI, CME FEDWATCH, BLAKE LAW
Law said that when Kalshi has sufficient liquidity, which tends to show up about 100 days prior to a Federal Open Market Committee meeting, “it is quite accurate” and “even outdoes” the CME FedWatch Tool. But during low-volume periods, Kalshi can be noticeably less accurate, he said.
In a nutshell, “Kalshi gets to the answer quicker than CME FedWatch and it may be because there are more barriers to entry in becoming a fed-funds futures trader,” Law said.
All the ways to play rates
Fed-funds futures are the most widely known, but certainly not the only, way to play the interest-rate-guessing game.
Contracts on fed-funds futures have been around since 1988 and currently trade on what’s known as the CME — for the Chicago Mercantile Exchange — where banks, portfolio managers and others are looking to protect themselves from increases in short-term rates and to speculate on future moves. The contracts are designed to manage risks under all environments, and volume tends to increase during times of uncertainty.
Made up of some of Wall Street’s most sophisticated traders, the fed-funds futures market reflects a probable path for interest rates. According to CME Group
, its FedWatch Tool calculated the Federal Open Market Committee’s actual rate increases with an average probability of 88% going into each of the 11 meetings where officials hiked between March 2022 and July 2023.
Fed-funds futures “are one of the most widely used tools for hedging short-term U.S. interest-rate risk, providing deep liquidity for market participants to hedge and express views on upcoming changes to Fed monetary policy,” according to CME Group spokeswoman Laurie Bischel. While trading strategies vary across markets and client segments, CME provides a range of products that “serve a variety of needs for institutional and retail traders,” she said.
Newer instruments such as options on Secured Overnight Financing Rate futures, which also trade on the CME, are another way for global banks, hedge funds, asset managers, trading firms and others to hedge interest-rate risks. SOFR is the world’s leading short-term interest-rate derivatives market, and changes to that rate are closely linked to the fed-funds rate. The average daily volume on SOFR futures and options has reached 5.3 million contracts so far this year versus 384,000 contracts in fed-funds futures.
Both fed-funds futures and SOFR-linked options reflect a wider array of potential outcomes over different periods of time than Kalshi’s contracts do. A fund manager making money in a low-interest-rate environment, for example, may want to protect all or part of his or her portfolio by selling fed-funds futures contracts, SOFR futures contracts, or both in anticipation of higher borrowing costs.
Then there are prediction markets like Kalshi, where traders are playing for profit and describe themselves as largely incentivized to be accurate. These individuals trade on contracts in which the payoff is dependent on how a future event turns out. Kalshi caps trades on the Fed and CPI inflation data at $7 million — substantially higher than the $25,000 cap placed on most other markets, where it collects an average 0.5% trading fee. Kalshi’s competitors include the CME, PredictIt of New Zealand, Polymarket, and Augur, and it is currently barred from offering bets on U.S. congressional elections.
Xavier Sottile, who is in charge of markets for Kalshi, said retail traders are “more interested in the truth’’ and using an “intentionally simple tool’’ to take positions, whereas their counterparts in fed-funds futures are “hedging and offloading the risk of interest rates that can go up or down’’ in a more “opaque and hard-to-understand market.”
“The pros don’t know more than me or you. ‘Normal’ people are standing by their beliefs and putting money on what is essentially a strong opinion,’’ Sottile said.
In April, Kalshi announced the addition of its first institutional market maker, a unit of Susquehanna International Group — a move aimed at providing more market liquidity.
Historical origins
Prediction markets have been around in one form or another for more than 500 years, beginning with wagers on papal elections in the 1500s. The earliest form of prediction markets on Wall Street took shape around 1884, with betting on the outcome of presidential elections, according to authors Paul Rhode and Koleman Strumpf. In more recent decades, these markets have occasionally been confronted with ethical dilemmas over the propriety of betting on geopolitical outcomes, and described by some critics as “terrorism betting parlors.” Congressional criticism, for example, led to the shutdown of a prediction market created by an arm of the U.S. Defense Department in the early 2000s.
Despite the naysayers, experts said prediction markets can play a vital role in society, are reasonably efficient, and tend to perform well — despite occasionally being limited by low liquidity.
Eric Zitzewitz, an economics professor at Dartmouth College, said “these markets are better at getting liquidity on topics that people find interesting or enjoy betting on…What’s surprising, though, is how accurate the answers can be that come out of occasionally illiquid markets.”
The ability to forecast the most likely direction of interest rates, for example, can help businesses and consumers anticipate when and by how much they should be spending and investing.
How to play
Traders on Kalshi usually give a yes-or-no answer to an array of questions that have related contracts. Those contracts eventually settle at either $1 if certain event outcomes occur and zero if they do not, with odds expressed in terms of cents before they happen. Traders can buy or sell these contracts at any price between zero and $1 prior to the event. As of Wednesday, the question of how many times the Fed will reduce rates this year was trading at 32% odds of just one rate cut — a scenario priced at 32 cents and reflecting the highest-probability outcome.
‘What’s surprising, though, is how accurate the answers can be that come out of occasionally illiquid markets.’
— Eric Zitzewitz, economics professor at Dartmouth College
The performance of retail investors versus fed funds futures traders is similar to other little-guys-versus-big-guys patterns seen over the past few decades, including the time Wall Street Journal readers outperformed four investment professionals by losing less when it came to picking stocks during the early 2000s.
In 2021 and 2022, Robbins, the part-time Kalshi trader, remembers thinking how “crazy” it was that financial markets were “pricing in so little tightening, when during that time I thought the Fed would need to be tightening more.” The Fed had been “providing a narrative that a lot of this inflation was transitory, but you could look at how broad-based it was and see that wasn’t the case,” he added.
Fast forward to late 2023 and early this year, and the storyline “flipped” in Robbins’ mind — with fed-funds futures traders anticipating multiple quarter-point Fed rate cuts. Robbins said he recalls thinking “that was outlandish, given inflation was consistently higher than the Fed’s target and won’t go down nearly as fast as the market thinks.”
