The speculation shift from crypto to prediction markets

Hello friends, and welcome to The Unsophisticated Investor! Brought to you by Scott & Rob, the founders of Shuttle!

If you want to invest alongside the VC funds who've backed breakout companies like Revolut, Asana, JustEat, Bolt, Lets Get Checked, Loom, Runna, Charlotte Tilbury, Deel, Aircall, AngelList, Carta, TransferWise and many more, regardless of your knowledge, network or net worth, then sign up using the button below.

Now, let’s get into it 👇

Two news stories that caught our attention yesterday, that when read together are symptomatic of where speculative capital is shifting in 2026.

The first was Coinbase announcing it would cut around 700 jobs, roughly 14% of its global workforce, in what CEO Brian Armstrong framed as a combination of a crypto downturn and a pivot to an AI-native operating model. It was the latest in a sequence. Gemini (led by the Winklevoss twins) cut a quarter of its staff in February and pulled out of the UK, EU and Australia entirely, refocusing on the US and on prediction markets. Block, founded by Jack Dorsey, cut nearly half its workforce later that month. In March, Crypto.com cut 12% and Algorand cut a quarter of its team.

The second story was a Warsaw startup called Elastics that closed a $2m oversubscribed pre-seed round to build what it calls an AI operating system for prediction markets. The round was led by Paris-based Frst, with angels including the ElevenLabs co-founders and an a16z scout. The founders are an ex-Goldman quant and a mathematician. Their product scans Polymarket, Kalshi and Limitless for trading opportunities and executes trades around the clock on behalf of users.

Read together, the two describe the same shift. As crypto retreats from its original promise of becoming a parallel financial system, prediction markets are absorbing the speculative energy that used to live in altcoins. Same demographic, same risk appetite, same crypto-native onboarding. A different rail.

Crypto’s narrowing story

Crypto had a maximalist pitch and a much narrower outcome. Decentralised tokens were going to replace large parts of the financial system, programmable money was going to unlock new categories of consumer application, and the open internet was going to be rebuilt on blockchain rails. Most of that has not happened. Bitcoin has settled into a store of value, Ethereum is infrastructure for a much smaller universe of applications than its proponents anticipated, and most altcoins from the 2021 cycle have not delivered durable use cases.

What has worked is stablecoins. Fiat-backed, regulated tokens turned out to be a genuinely important piece of payment infrastructure. The EU's Markets in Crypto-Assets regulation, fully enforced from 2024, was the first comprehensive framework anywhere in the world to codify them as a category and set clear rules around reserves, redemption rights and authorised issuers. With that regulatory clarity in place, stablecoins have found product-market fit in two areas in particular: international remittances, where they undercut traditional corridors on speed and cost, and treasury and settlement flows between financial institutions. 

That is the shape of crypto’s enduring contribution so far: a regulated payments layer plus a speculative trading layer that has been steadily migrating toward whatever offers the next lottery ticket. Prediction markets are simply the latest iteration. Polymarket and Kalshi together processed around $24 billion in April; Kalshi closed at a $22 billion valuation in March, and Polymarket is in talks at $15 billion. This is no longer a niche category. It is one heavily backed by venture capital, and one that still sits outside any clear regulatory perimeter and is likely to face heightened scrutiny going forward.  

Two regulators, two answers

Prediction markets have grown partly because no regulator has been able to decide what they are. The same yes/no contract sits awkwardly between three regimes: gambling, derivatives, and crypto. Gambling laws were written for casinos. Derivatives rules were written for institutional traders. Crypto rules are still being drafted. The product has scaled faster than any could catch up. The same yes/no contract is now a regulated derivative in one jurisdiction and unlicensed gambling in another.

In the United States, CFTC chair Michael Selig delivered the clearest pro-prediction-market signal yet from a US regulator in January, withdrawing a Biden-era rule that would have banned political and sports event contracts. Federal policy is converging toward treating prediction contracts as commodity derivatives. But state attorneys general and tribal nations have filed more than nineteen federal lawsuits arguing the opposite. A Maryland court ruled that sports contracts are indistinguishable from wagering. Massachusetts has forced Kalshi to geofence the state. Sports represents roughly 72% of Kalshi’s volume, so a Supreme Court ruling against sports contracts, expected within twelve to twenty-four months, would gut the business. This week, the US Senate quietly banned its own members and staff from trading on prediction markets, citing insider trading risk.

In Europe, the picture is much harsher. There is no unified EU framework. Polymarket is banned as unlicensed gambling in more than a dozen jurisdictions, including France, Germany, Italy, the Netherlands and Poland. The British Gambling Commission has stated it does not believe prediction markets can classify themselves as anything other than gambling products.

How regulatory gaps are opening platforms up to insider trading

In April, French police opened an investigation into suspected tampering with a Météo-France weather sensor at Charles de Gaulle Airport. On at least two occasions in mid-April, the sensor reading spiked sharply at moments that aligned with large Polymarket bets on the day’s high in Paris. One trader netted around $21,000 from a $119 wager, and no other weather station in the area recorded the spike. The leading theory among local meteorologists is that someone used a battery-powered hair dryer at the publicly accessible sensor; the investigation remains open and no arrests have been made. Polymarket quietly switched its data source to a different airport.

Three months earlier, US Army Special Forces master sergeant Gannon Van Dyke was arrested over Polymarket bets placed in the days before the January raid that captured Venezuelan president Nicolás Maduro. Van Dyke had helped plan the operation. He turned a roughly $33,000 stake into more than $400,000 in profit by wagering on Maduro being out of office by the end of the month, then allegedly tried to delete his account once scrutiny landed. It is the first US prosecution of insider trading on a prediction market.

In February, Israeli authorities indicted an Air Force reserve major and a civilian for using classified information to bet on the timing of Israel’s June 2025 strikes on Iran. Prosecutors say the pair earned around $162,000 from a single successful wager.

Prediction markets settle financial outcomes against real-world data, and people with access to that data have strong incentives to nudge it. This is a structural risk regulators have not yet figured out how to police. The patterns are familiar from crypto’s pre-regulation years; what is new is that the event being manipulated is sometimes tied to the physical world.

The crowd is mostly machines

The retail demographic showing up at Polymarket and Kalshi is increasingly trading against machines. According to LayerHub, around 30% of wallets on Polymarket already use AI agents. Third-party data suggests only between 7% and 13% of human traders on prediction markets achieve positive performance. The rest lose money.

Elastics’s pitch is that this is a problem worth solving for retail: build the agent infrastructure that institutional traders already have and let everyone else compete on equal footing. It is reasonable, but if the entire premise of prediction markets is the wisdom of crowds, what happens when the crowd is mostly machines pricing in news flow faster than humans can react? The answer is probably more efficient markets, and probably worse outcomes for the retail traders the migration narrative depends on.

The private markets read, and the caveats

This is a category being aggressively backed by US venture capital. 5(c) Capital, a new dedicated fund with both the Kalshi and Polymarket CEOs as backers alongside Marc Andreessen, raised $35 million for its first fund. Coinbase, Robinhood and Crypto.com have formed a Washington lobbying coalition to push for federal pre-emption. Kalshi quadrupled its valuation from $5 billion to $22 billion in six months. The shape is the one crypto traced between 2020 and 2022: rapid valuation expansion, category funds, institutional onboarding, then a sharp reset that wiped out a generation of retail entrants.

The risks that could trigger something similar are real and stackable. A Supreme Court ruling against sports event contracts would remove most of Kalshi’s revenue overnight. Insider trading is now the defining vulnerability.

The bottom line

Crypto promised to rewire finance and ended up rewiring two specific things: a parallel store of value in bitcoin, and a programmable dollar in stablecoins. Both have settled into something durable. Everything else, the trading layer that produced most of the speculative excitement of the last cycle, has been looking for a new venue. Prediction markets are now that venue.

That does not automatically make them a fad. The wisdom-of-crowds proposition is genuine, the institutional research coverage is real, and the trillion-dollar volume forecast is not implausible if the regulatory questions resolve cleanly. It also does not automatically make them durable. The same product is a regulated derivative in one country and unlicensed gambling in another, the manipulation problem is unresolved, and the user base is the same retail cohort that lost money on altcoins, now trading against AI agents that already account for nearly a third of activity on Polymarket.

Prediction markets are an interesting segment of financial markets right now, but it’s clear there are structural risks and similarities to boom and bust business models we’ve seen before.

What we’ve been working on at Shuttle

• Continuing to build commits to our investment round 💰

• Working on our implementation and launch for the UK market 📝

• Submitting our application to the FCA’s Supercharged Sandbox to apply AI in financial services 🏦

The Unsophisticated Investor is brought to you by Scott & Rob, the founders of Shuttle. We’re both sick of private markets being a playground exclusive to the ultra-wealthy so we started a company to challenge the status-quo. Shuttle’s singular focus is to unlock private markets for Millennial and Gen Z tech professionals and help them build wealth through the highest performing private market opportunities.

Scott & Rob
Shuttle Co-Founders