What Polymarket Says About Oil, SpaceX/Tesla, and the AI IPO Race
By datatrekresearch in Blog
Prediction markets give traders a real-time edge on moves in oil, tech stocks, and AI – and most investors aren't paying attention yet. In our latest video, DataTrek co-founder Jessica Rabe walks through current Polymarket odds on Strait of Hormuz/oil price risk, the hidden merger premium in Tesla's stock, and what the Anthropic vs OpenAI IPO race means for your AI trade positioning.
Watch it here on our YouTube channel! Please hit like/subscribe and share this video if you find it useful. Sign up on datatrekresearch.com to sign up for a 2-week free trial to our daily investment newsletter!
Transcript
Hi, I’m Jessica Rabe, co-founder of DataTrek Research, and today we will explain why prediction markets matter for every investor and trader, including those who will never place a single bet on one. Prediction markets used to be novelty, a place to wager on elections or obscure geopolitical events. Polymarket, Kalshi and similar platforms have fundamentally changed that. Today, you can trade contracts on whether a major virtual currency rises or falls over the next five minutes, or on where individual stocks close the day’s trading session. There's also an entire Earnings section on Polymarket dedicated to whether companies beat or miss estimates, updated in real time, and priced by people with actual skin in the game. Prediction markets are no longer fringe venues you can dismiss as irrelevant. Rather, they are becoming important, market-based sentiment indicators underpinned by the same dynamics that drive traditional capital markets.
So why should a traditional investor care? Two reasons.
The first is the growing connection between prediction markets and the markets that set the daily prices of your investments. As platforms like Polymarket add users and contracts, mispricings between prediction markets and spot or derivatives markets become opportunities to make money. Institutional players are already doing this by identifying gaps between what a prediction contract implies and where the underlying asset is trading, then closing that gap for a potential profit. Over time, that tightens the linkage between the two ecosystems. Every time capital flows in to exploit an inefficiency, prediction markets become a little more connected to the markets you already trade. You may not be on Polymarket — but Polymarket is increasingly in your market.
The second reason is the data itself. Prediction market pricing is fundamentally different from a poll or a survey.When someone fills out a survey, they're sharing an opinion. When someone trades a prediction contract, they're putting real money behind that opinion. That's a completely different and more valuable signal. Hedge funds have figured this out. For example, they're using prediction market contract pricing as real-time reads on retail investor positioning. Quant shops are feeding API data directly into models that forecast asset prices. Mispricings between prediction contracts and macro outcomes are being used as potential leading indicators. Prediction market data isn't just interesting; it's becoming part of the actual trading infrastructure. Bloomberg reported that Susquehanna and Jump Trading are already providing liquidity on these platforms. And in January, Dow Jones signed an exclusive deal to embed Polymarket data directly into the Wall Street Journal, Barron's, and MarketWatch — treating prediction market odds as a financial data feed, not a novelty.
The bottom line is that we have both a new information source and a fast-evolving parallel market, neither of which really existed even five years ago. As participation grows and liquidity deepens, prediction markets are becoming an incremental driver of volatility, price discovery, and positioning across many important public markets.
The traditional tools — polls, analyst surveys, sentiment indexes — have a fundamental flaw: they measure what people say. Prediction markets measure what people are willing to bet. Now, that distinction only matters if the bets are informed ones. Most people who go to Vegas lose. Money behind an opinion doesn't automatically make it a better opinion.
What makes prediction markets different is who is betting and what they stand to gain from being right. Some people on these platforms, and maybe most, are casual gamblers, just as many stock market day traders essentially are. But, as with equity markets, there are also many prediction market users who do in-depth research and bet their own hard earned cash on their views being more accurate than the less-informed crowd. There’s a lot of money to be made doing that well, and it is that potential reward that both makes liquid prediction markets contracts efficient and means their price signals are much more valuable and timelier than surveys or other measures of popular opinion.
With that framing in place, let's put it to use by walking through Polymarket odds across 3 geopolitical and capital markets outcomes based on prices from today, June 23rd 2026.
First, let's start with the reopening of the Strait of Hormuz. Polymarket currently puts the odds of shipping traffic returning to normal by year-end at 87 percent, and that figure has moved up 14 percentage points within the last 2 weeks. That's a significant repricing in a short period of time, and it tells us something important.
When a prediction market moves that fast, it means new information is getting priced in quickly, and in this case, it suggests that a substantial portion of the anticipated decline in oil prices due to renewed global access to Mideast crude has already been discounted by commodity markets. The Strait of Hormuz is one of the world's most critical oil chokepoints, with roughly 20 percent of global petroleum supply passing through it daily. If the recent bottleneck due to the US/Iran conflict is permanently resolved, oil prices will increasingly reflect their underlying supply and demand fundamentals rather than a geopolitical risk premium. In other words, the good news may already be in the price.
The oil futures curve backs that up. Front-month WTI contracts are trading around $72.7 a barrel, and that price declines steadily to $69.79 by January 2027. A downward-sloping futures curve like this tells you that the market has already priced in a meaningful normalization of supply. Again, these are prices as of June 23rd.
That asymmetry is where the risk comes in. When both prediction markets and futures markets are pricing near-certainty of a permanent resolution, the market calculus shifts. The upside from things going right is already reflected in prices. But, if US-Iran relations deteriorate and there is a renewed military escalation, markets would have to rapidly reprice in the other direction.
This is exactly where prediction markets earn their keep. Everyone knows oil prices would spike if the Strait closes again — that's not the insight. The insight is that prediction markets give us a hard number, an 87 percent probability, on a very specific fulcrum point that would otherwise be purely a judgment call. Instead of debating how much geopolitical risk premium is embedded in oil prices, you can look at a single contract and ask: do I agree with the market's 87 percent confidence that the US/Iran conflict is resolved? If you think that number is too high, you have a trade. If you think it's right, you know the risk premium has largely been priced out. That's what prediction markets add — not the scenario analysis, but the ability to put a precise probability on the specific issue that matters most.
Next up is a topic that's generated a lot of buzz lately, namely a potential merger between Tesla and SpaceX. Polymarket currently puts the odds of an official announcement by year-end at 39 percent, and the September 30 contract sits at 26 percent. So, the market is pricing this transaction as possible, but not likely — and we'd actually argue that these odds are much too high.
Here's why. When SpaceX filed its S-1, there was no mention of any merger discussions with Tesla ever having taken place. That's not a small detail to omit if a combination is genuinely on the table. Now, you might rightly say that Elon Musk doesn’t care if he gets sued for materially omissions to an SEC filing, but every investment bank on the cover of the S-1 certainly would. If there were even a remote chance of a merger in the year after the SpaceX IPO, they would have insisted on its inclusion in the S-1.
And from a financial standpoint, the math is difficult — SpaceX has an extraordinarily capital-intensive product development roadmap, and Tesla doesn't yet generate the free cash flow needed to fund it. A merger doesn't solve SpaceX's capital needs, and it potentially burdens the combined company with a conglomerate discount.
But here's what makes this interesting from a markets perspective. If around a third of Tesla's current valuation reflects some modest but still significant embedded merger premium — which is roughly consistent with where Polymarket odds sit at 26 to 39 percent— then that premium is at risk. Not necessarily because the merger is impossible, but because as conviction fades and those odds drift lower, that portion of Tesla’s valuation will decline. Markets that reprice a low-probability event toward zero can move faster than people expect. That's the hidden risk embedded in Tesla's stock that prediction markets are actually helping us quantify.
Our third and final example is one that matters enormously for anyone with exposure to the AI trade, and it's actually two stories in one.
Let's start with Anthropic. Polymarket currently puts the odds of an Anthropic IPO by year-end at 75 percent, and by October 31st at 37 percent. The more important signal is in the valuation distribution. Markets are pricing a 79 percent probability that Anthropic lists above $1.5 trillion, a 43 percent chance it clears $2 trillion, and a 20 percent chance it reaches $3 trillion. That implies a median expected valuation somewhere between $1.75 and $2 trillion — which puts it in the same neighborhood as where SpaceX went public.
What's notable is how sharply these odds have moved in recent weeks. That repricing coincides with SpaceX's IPO, suggesting its successful public debut materially lifted sentiment around Anthropic's own prospects. But it also looks fundamental. Polymarket simultaneously gives Anthropic a 64 percent probability of fielding the best AI model by year-end, well clear of Google at 14 percent, OpenAI at 11 percent, and xAI at 9 percent.
That combination matters because it tells you the valuation isn't just hype. A high IPO valuation with weak model leadership would make little sense. But when the same crowd that is betting on Anthropic's valuation is also betting on its technical superiority, those two signals are reinforcing each other. Model leadership is ultimately what drives revenue, pricing power, and market share in AI — so if prediction markets are saying Anthropic wins on both dimensions simultaneously, that's not noise. It's a coherent thesis, expressed in two separate contracts by people putting real money behind it.
The OpenAI picture is considerably more uncertain, and the contrast is striking. Polymarket assigns only a 54 percent probability to an OpenAI IPO happening by year-end, meaningfully lower than Anthropic's 75 percent. The valuation distribution tells an even more sobering story: the single most likely outcome is no IPO by December 31, sitting at 48 percent odds today. Conditional on a listing actually happening, the most likely outcome is a valuation between $1 and $1.25 trillion, well below what markets are implying for Anthropic. That gap almost certainly reflects OpenAI's deteriorating model leadership. When prediction market traders are giving you only an 11 percent chance of having the best AI model by year-end, that affects how equity market investors price your IPO.
There's one more takeaway here that matters for all public equity market investors. The IPO overhang risk that has weighed on sentiment, namely the fear of a flood of new AI supply hitting public markets simultaneously, may prove less severe than the bear case assumes. If OpenAI's listing slips into 2027 while Anthropic comes to market first, the pipeline is more manageable than it looks today. That's a meaningful distinction for how you think about positioning in AI names heading into the second half of the year.
With that, thank you very much for taking the time to watch this video, please hit like and subscribe if you enjoyed it. And if you'd like to see our daily research, you can start a 2-week free trial on datatrekresearch.com. Thanks again for watching and we hope you have a great day!
