PMTrader: Predicting the Future, Taylor Swift, Trading Advice and More | TG Podcast
PMTrader turned a one-year prediction-market side project into roughly $50,000 of part-time profit, then quit his five-year CPA career to trade and create content full-time. He now reports making about $2,500-$3,000 a week; at sign-off, he says he is at roughly $60,000 and is publicly pursuing $100,000 “within the next couple of months.” He distinguishes himself from seven-figure traders such as Domer and Phantom: the pitch is a transparent climb from a smaller bankroll, not guru mythology.
His structural thesis is that prediction markets offer unusually favorable terrain because they are peer-to-peer and lack the 17-20% take he encountered in horse racing. In his news-trading framing, markets are live 24/7: when an earthquake occurs or another batch of election votes arrives, “someone has to be there to adjust the prices.” Deeper liquidity has so far enlarged the opportunity because newcomers and proliferating market types have arrived faster than prices have become efficient.
Mention markets are research markets disguised as novelty bets. PMTrader studies earnings-call transcripts back to 2020 and more than 100 past NFL games, then combines historical word frequencies with “vibes” and fresh context—such as Middle East news changing the probability that Trump uses a phrase. The host’s “fourth and inches” intuition illustrates the exploitable mistake: what feels obvious may be rare, although PMTrader candidly admits he is only even or slightly down on his recurring “inch—No” trade.
The Brian Armstrong episode exposed the reflexive risk that subjects can observe prediction markets and potentially influence outcomes. PMTrader avoided Coinbase because “I didn’t feel like I had an edge,” but acknowledged Armstrong reading out the listed words was “probably not great for mention markets in the long run.” His qualified counterpoint is that low-probability candidates still sometimes win, limiting the extent to which displayed odds become self-fulfilling.
Two trades capture the opportunity set: specialized knowledge and nearly resolved “bonds.” A Taylor Swift superfan with no prediction-market experience went approximately 10-for-12, earning PMTrader about $3,000; an EA mention market remained open even after an acquisition meant no earnings call would occur, letting him deploy about $7,000 and earn roughly $2,000. Thread Guy estimated perhaps 5% deal-failure risk and characterized the trade as offering around 30% per dollar.
PMTrader expects competition to compress today’s edges, but not eliminate them, because new and smaller markets continually reset the research race. He moved from eight months of trading two correlated markets into mention analysis and believes opportunity should remain abundant for at least three years, provided traders keep adapting. His broader method is top-down: compare logically linked contracts—such as Biden dropping out versus remaining the Democratic nominee—and trade inconsistencies rather than forecast every event from scratch.
For newcomers, his prescription is “start small,” “pick a niche,” calculate a fair value, and explain exactly why the crowd has mispriced the contract. Read the resolution rules, learn the order book before risking meaningful capital, and distrust anything that appears effortless: “If something seems too good to be true or too easy on these platforms…it probably is.” His next bet is content—pairing markets with niche experts, from Swift fans to Powell veterans or linguists studying Trump.
1. Lower market friction turned a CPA’s side hobby into a full-time pursuit
PMTrader came from poker and horse racing, where he was only “semi-profitable”; he specifically said horse racing removed roughly 17-20% from every wager. Prediction markets changed the equation: they were peer-to-peer, had “really no house edge,” and offered enough distinct inefficiencies that he wondered, “Why would I waste time” on the older formats?
He traded part-time while working five years as a CPA, including 60-80-hour filing-season weeks, and made roughly $50,000 during his first year. Three or four weeks before the interview, he quit, moved to New York, and began reporting weekly profits of approximately $2,500-$3,000 while planning to increase position sizes as his portfolio grows.
His self-positioning is deliberately modest beside Domer and Phantom, whose disclosed P&Ls reach seven figures. PMTrader says he is at about $60,000 and is “fully confident” he can reach $100,000 this year or within several months, but the attraction of his public “road to 100K” is watching that progression rather than inheriting a finished success story.
2. Faster information and expanding liquidity are creating more mispricing, not less
PMTrader’s simplest strategy is “news trading.” If a market asks whether a magnitude-7.0 earthquake will occur, the first trader to act once it does can buy Yes while the price is adjusting; election returns offer the same mechanism when reporting moves from 20% to 30% and a new batch reveals the direction of the race.
He described the earthquake trade as very low risk, while Thread Guy characterized the election example as somewhat risk-free. Execution and interpretation still matter. What has changed is capacity: trades that might have supported a few thousand dollars a year earlier now sit inside markets where traders such as Domer and Phantom can generate seven-figure P&Ls.
Contrary to the expectation that deeper liquidity automatically produces sharper pricing, PMTrader thinks the present phase may be easier. A single Powell-speech mention market can trade roughly $1.5 million, while Taylor Swift culture markets that once attracted $10,000 now reach millions; her album release generated tens of millions in volume.
The supply of contracts is widening just as quickly—from Trump and Powell speeches to NFL announcers, music, culture, tariffs, ceasefires, and local elections. PMTrader expects established markets to sharpen, but every newly invented format begins with a fresh problem: “Nobody knows what the true probabilities are.”
3. Mention-market edge combines transcript statistics, context and reflexivity
PMTrader rejects the idea that mentions are pure RNG. He cited one new trader turning $1,000 into $30,000 in a month, other traders documenting more than $300,000 primarily from mentions, and an eight-person “Mentions Cave” whose members each reportedly have P&Ls above $100,000—evidence he sees as stronger than his own hot streak.
The quantitative work starts with language history: earnings transcripts going back to 2020 and more than 100 past NFL games. The second input is “vibes”—current events that alter language probabilities. If Middle East news breaks before a Trump speech, for example, a phrase with a 50% historical frequency might deserve a 75-80% price.
Thread Guy pressed the much-memed “inch—No” position: surely “fourth and inches” or “third and inches” makes No a brutal trade. PMTrader’s answer was revealingly mixed. No initially cost about 20 cents and later rose toward 35-40 cents after he repeatedly put about $1,000 on it, yet he admitted he is probably even or slightly down rather than claiming a manufactured victory.
Reflexivity is the larger danger. Brian Armstrong reportedly looked at Coinbase’s mention contract and recited its listed words during the earnings call; PMTrader had skipped Coinbase because “I didn’t feel like I had an edge.” He called the event both viral and damaging—“documented in mention-market lore,” but also the perfect ammunition for critics—and expects it to blow over rather than delegitimize the category.
4. Taylor Swift and EA show two very different forms of informational advantage
For the Taylor Swift release, PMTrader invited a die-hard fan who had never traded prediction markets but understood music, concerts and Billboard Hot 100 mechanics. The experiment went approximately 10-for-12 and made PMTrader around $3,000; the fan missed two song-word long shots, including “Lover” near five-to-one, while correctly hitting several outcomes priced around 20%.
That result embodied his advice to seek domain expertise outside the normal trading crowd. He paid the guest $500 and argued that a genuine specialist can price a niche better than generalists even without market experience—a model he hopes to repeat with political, linguistic and central-bank experts.
The EA trade was closer to a “bond,” his term for high-probability shares overlooked because users prefer lottery-like payouts. After EA announced it was being acquired and taken private, PMTrader concluded there would be no earnings call; he bought roughly $7,000 of No shares on words that would not be mentioned, earning about $2,000.
He did not call the EA position absolutely risk-free. Thread Guy estimated roughly 5% deal-failure risk and said the trade offered around 30% per dollar. The mispricing persisted for two or three weeks even after PMTrader publicly explained it. Taylor Swift offered a similar anomaly: an outcome PMTrader viewed as “pretty much already 100% determined” still yielded 5-10%, partly because users misunderstood the dating language.
5. Durable edge requires selectivity, adaptation and an explicit theory of mispricing
PMTrader trades only around 10-15 markets in a typical week, concentrating where he believes he has an edge. Although mentions receive most attention, his original method was eight months of arbitrage across two correlated markets, exploiting prices that should align but do not.
He describes himself as top-down rather than model-heavy. Contracts on whether Biden would drop out and whether he would remain the Democratic nominee should have carried closely related probabilities; discrepancies offered a trade without requiring a ground-up political forecast. Thread Guy’s parallel to crypto was that more capital raises competition while simultaneously opening neglected verticals.
PMTrader agrees his current edges will erode, but believes smaller contracts, new formats and adaptation should preserve opportunity for at least three years. His own evolution—from arbitrage to transcript-driven mention analysis—is the template: “As long as you’re willing to adapt, you’ll always have edges.”
His beginner checklist is concrete: start small, learn the order book, choose a subject you already enjoy, assign fair value, explain why the public produced the wrong price, and read every resolution rule. Longer term, he sees a “content gold mine” in putting specialists beside live markets—from someone who has attended 50 Powell speeches to a linguist studying Trump or a political candidate confronting their own odds.