Prediction Markets Like Kalshi: What's Really at Stake [2025]

Introduction: The Betting Wild West We Didn't See Coming

We live in an age where you can bet on almost anything. Want to wager on whether the Federal Reserve will cut interest rates by September? Done. Curious about the odds on a specific Supreme Court decision? There's a market for that. Feeling lucky about election outcomes? Multiple platforms are ready to take your money.

Prediction markets like Kalshi and Polymarket have exploded in popularity over the past few years, transforming from niche financial instruments into mainstream betting platforms. According to Front Office Sports, the sudden normalization of these markets raises serious questions that most people aren't asking yet.

What happens when ordinary people can profit from predicting political outcomes? How does endless betting on real-world events shape the information ecosystem itself? And perhaps most troubling: are we creating incentive structures that could undermine democratic processes, distort market signals, or encourage the spread of manipulated information?

These aren't hypothetical concerns. They're happening right now, as billions of dollars flow through prediction markets, and most of us are barely paying attention. The New York Times highlights that these markets aren't neutral tools. They're systems with winners, losers, and countless invisible consequences.

TL; DR

• Prediction markets exploded in 2024-2025 with platforms like Kalshi and Polymarket handling billions in bets on elections, political events, and economic outcomes, as noted by International Banker.
• These platforms create perverse incentives where financial profits reward spreading uncertainty, manipulation, and sometimes outright misinformation, according to War on the Rocks.
• Democracy takes a hit when election outcomes become pure betting events instead of civic processes, shifting focus from policy to odds, as discussed in The New York Times.
• Information distortion is real as people with money to spend can literally buy influence by moving market odds, as reported by Marketplace.
• Regulatory uncertainty remains with most prediction markets operating in gray legal zones, pending serious government oversight, as highlighted by Better Markets.

What Are Prediction Markets, Actually?

Let's start with the basics, because most people use the term "prediction market" without really understanding what one is.

A prediction market is simply a marketplace where people buy and sell contracts that pay out based on whether a specific event happens or not. That's it. The mechanism is dead simple: if you believe an event will occur, you buy shares. If you think it won't happen, you sell them (or buy shares of the opposite outcome). As more people trade, the price of those shares adjusts, and that price becomes a collective probability estimate of the event occurring.

Here's a concrete example: imagine a market on whether the FTC will approve a specific merger by December 31, 2025. If you think approval is likely, you might buy shares at

The underlying theory is elegant: by aggregating the "wisdom of crowds," these markets should produce better probability estimates than any individual expert or polling organization. Instead of asking one analyst, "What are the odds of this happening?" you're essentially asking thousands of people who've put real money behind their predictions.

But here's where it gets interesting. Prediction markets have existed in various forms since the 1600s. What's new in 2024 and 2025 is the scale and the focus. Most historical prediction markets were limited to sports betting or commodity futures. Today's platforms are betting on everything: elections, economic policy, geopolitical events, corporate decisions, even whether specific people will achieve certain milestones.

The reason for this explosion is partly regulatory. For years, prediction markets lived in legal gray zones. The CFTC (Commodity Futures Trading Commission) restricted them to very specific categories of events. But starting in the early 2020s, platforms like Kalshi began pushing boundaries, winning approval for broader categories of contracts. When Kalshi got regulatory approval to offer election-related contracts in 2023, it opened floodgates.

Suddenly, what was once a niche activity for finance nerds became accessible to anyone with a phone and $100 to spare.

The Rise of Betting on Everything: How We Got Here

Three decades ago, prediction markets were genuinely fringe. Almost nobody had heard of them. Even in 2010, they were still mostly ignored by mainstream media and the general public.

Then came 2016.

When traditional pollsters spectacularly failed to predict the US presidential election, people started looking for alternatives. Prediction markets, it turned out, had been slightly more accurate. Not by a huge margin, but enough to grab attention. Articles started appearing: "Why Prediction Markets Outperformed Polls." Academics began publishing research suggesting that these markets had value. Investors noticed.

By 2020, during the COVID-era lockdowns, several things happened simultaneously. First, Robinhood and other retail investment platforms had democratized stock trading, so millions of people were already comfortable with trading interfaces. Second, remote work meant more people were online, more of the time, with disposable income. Third, the pandemic itself created massive uncertainty about everything from infection rates to vaccine timelines to economic recovery, making prediction markets feel more relevant than ever.

But the real acceleration came in 2024. Why? Because major news organizations started treating prediction market odds as actual news. The New York Times began running articles with headlines like "Polymarket Shows Trump at 70% Odds." CNN contributors started discussing Kalshi odds as if they were expert consensus. Venture capitalists began funding prediction market platforms at massive valuations. What had been a quirky financial instrument became part of mainstream political discourse.

Several factors converged:

1. Regulatory loosening: Kalshi won approval for political and economic event contracts
2. Media amplification: Major outlets started treating market odds as news
3. Retail accessibility: Phone apps made entry trivial for non-expert users
4. Uncertainty asymmetry: People crave certainty during chaotic times, and markets feel like they provide it
5. Financial incentives: Early platforms attracted massive capital from investors betting on network effects

Today, Polymarket alone handles hundreds of millions of dollars in daily trading volume. Kalshi has a waiting list millions long. New platforms are launching constantly. The market ecosystem has become impossible to ignore.

And that's exactly when we should start worrying.

The Allure: Why People Are Actually Betting on This Stuff

Before we talk about what's wrong with prediction markets, let's be honest about why they're so appealing.

There's something deeply satisfying about turning uncertainty into a tradeable commodity. Markets feel scientific. They feel objective. When you see that the probability of a specific economic outcome has shifted from 42% to 58%, it feels like you're discovering a hidden truth about the world, like you're reading the actual odds rather than just guessing.

For some users, prediction markets offer genuine value. Professional forecasters and researchers use them to stress-test their models. Businesses use prediction market data to inform supply chain decisions. Policy researchers study them to understand how different information sets affect collective probability estimates. These are legitimate applications.

But for the millions of casual users now betting on platforms like Polymarket? The appeal is simpler and more primal.

First, there's the intellectual appeal. You get to feel smart. You're not just watching events unfold like a passive observer. You're actively predicting them. You're competing against other predictors. You're playing a game where knowledge, intuition, and information flow together into actual winnings. For political junkies, this is basically heroin.

Second, there's the financial angle. Even small bets can generate real returns in volatile markets. If you bought contract shares at

Third, and this is important: prediction markets feel more fair than traditional gambling. Sports betting is about luck and statistical inference. Prediction markets feel like they're about insight, information, and being smart. You're not betting against a house that's trying to profit from your losses. You're betting against other people. The house just takes a small cut. That psychological distinction matters enormously.

Fourth, for some segments of the population, these markets have become genuinely community spaces. Polymarket, in particular, has developed a culture around it. There are celebrities with public prediction portfolios. There are Discord communities strategizing about which bets offer the best risk-reward profiles. There's drama when someone makes a seemingly crazy bet that actually pays off, or when someone dumps money into a position that looks absurd.

It's not just trading. It's entertainment.

And that's actually where some of the problems start. When your information source becomes entertainment, when your probability estimates are coming from a community of strangers with conflicting financial incentives, when you're competing for money with people who have professional-grade research operations... things get messy.

The Problem: Creating Incentives for Manipulation and Misinformation

This is where we need to have an uncomfortable conversation about how prediction markets actually function as information ecosystems.

The central promise of prediction markets is that they aggregate distributed information efficiently. In theory, if you have thousands of people trading based on their private information, the resulting prices should reflect some kind of ground truth.

But here's the problem that economists have largely been ignoring: information markets can be gamed.

And not just gamed. They can be systematically distorted by people with sufficient capital and motivation.

Consider this scenario: You have $5 million and you want to influence the market odds on a specific political outcome. You don't need a majority of the market. You need enough to move prices. On Polymarket, you could accomplish this. Buy enough contracts at the current ask price to push the overall market position upward. Now the odds have shifted. Other traders, seeing the new odds, assume that someone with information has made that trade. They update their own beliefs. They start buying. The market moves further in that direction.

You've just used capital to manufacture the appearance of information. You haven't deceived anyone directly. You've just made a big trade.

This is called "market manipulation," and it's technically illegal in securities markets. But prediction markets exist in regulatory gray zones where the rules are... unclear.

More insidious is what happens when you combine prediction market incentives with information warfare. Here's the game theory:

Suppose you have a financial interest in a particular outcome (let's say a specific candidate winning an election). Your financial position pays off if the market assigns that outcome high probability. So what's your incentive? To increase the probability that people believe that outcome will happen.

You don't need to change reality. You need to change perceptions. And the most efficient way to change perceptions is to spread information, credible or not, that supports your position. If you can convince enough people to buy contracts predicting your preferred outcome, the market price will rise, and anyone holding that position profits.

Now here's where it gets dark: once the market price has risen, you can actually point to the market as evidence that your position is likely. "Look, the market is saying it's 68% likely. The market is always right." You've created a self-reinforcing loop where misinformation → market movement → apparent market validation of that information.

This isn't theoretical. Multiple researchers have documented this happening in real prediction markets. A Politico report examined prediction market behavior during the 2024 election cycle and found strong evidence of coordinated trading patterns that appeared designed to move specific markets, not based on new information, but based on desired outcomes. The traders involved made money not by accurately predicting events, but by successfully moving market prices.

Another concern: prediction markets specifically incentivize uncertainty and unpredictability. If an election outcome seems inevitable—say, 95% likely—the market becomes boring. There's no opportunity for profit. No drama. The market will naturally be smaller, with less activity.

But if the outcome seems close? If the market is 48-52? Now you've got maximum trading volume. Maximum engagement. Maximum profit opportunity. This means that, structurally, prediction markets have incentive to increase the perceived uncertainty around important events, even if that uncertainty isn't grounded in reality.

Consider the 2024 US election cycle. Before election day, multiple prediction markets showed the race at rough parity, with odds constantly shifting. This kept traders engaged, kept money flowing, kept the platforms in the headlines. But post-election analysis suggested that the outcome had been substantially more predictable than the markets indicated. The markets had, through their own incentive structure, contributed to overstating the uncertainty.

Was this intentional? Probably not in most cases. But it's a structural feature of these systems that we should acknowledge.

Democracy Under the Microscope: Elections as Betting Events

Let's talk about the elephant in the room: prediction markets are increasingly used to place large financial bets on democratic outcomes.

This is fundamentally different from anything we've done before in democratic societies. For centuries, elections have been treated as civic processes. You vote. Your vote counts equally with everyone else's. May the best candidate win.

But now elections are also investment vehicles. The electoral outcome isn't just about who governs. It's about making or losing money. It's about portfolio performance. It's about whether your prediction was correct.

What does this do to the information ecosystem? Several things, none of them good:

First, it creates financial incentives for information warfare. If you have a large position betting on a specific candidate, you have a direct financial incentive to spread information that helps that candidate win. Not just genuine information about that candidate's policies or qualifications. Any information that makes them seem more likely to win. This is distinct from normal political advocacy, where you might promote a candidate because you believe in them. Now you're promoting them because you're financially hedged on the outcome.

Second, it makes elections smaller and weirder to people who don't participate in markets. Media outlets, politicians, and analysts all start talking about "what the markets are saying" as if it's objective ground truth. But the markets are only reflecting the beliefs of the people actually trading on them. Which, statistically, are disproportionately male, financially literate, and wealthy. They're also disproportionately concentrated in certain geographic regions and demographic segments.

So you end up in a situation where the "wisdom of crowds" is actually the wisdom of a specific, narrow crowd. And everyone else hears about it and assumes it's universal wisdom.

Third, it can suppress voter turnout and participation. If you wake up on election morning and see that prediction markets are showing your preferred candidate at only 15% odds, what do you do? If you're a casual supporter, you might not bother voting. You might donate that money to a different cause instead. The market creates a forecasted reality that discourages people from participating.

This is called the "prophecy effect." Markets don't just predict outcomes. They influence them.

Fourth, it creates new attack vectors for foreign interference. If a hostile nation-state wanted to influence a US election, one mechanism would be to place large bets in prediction markets designed to move the probabilities. They don't need to actually influence the election itself. They just need to influence the perceived probability, which influences media coverage, which influences voter behavior.

How much would it cost for Russia or China to move a prediction market meaningfully? Probably not as much as you'd think. Polymarket, despite its size, is still tiny compared to traditional financial markets. A relatively modest investment could meaningfully move prices.

And here's the thing: the trades would be visible, but the motivation wouldn't be obvious. Was that large bet based on genuine forecasting skill? Or was it foreign interference? How would we even know?

Finally, there's the corruption problem. If a politician knows that their decisions will be reflected in prediction market prices, what do they do? Do they make the best policy decision? Or do they make the decision most likely to result in a positive market movement?

More troubling: what if wealthy donors make bets and then insist that politicians support the outcome those bets require? "Vote my way, or my hedge position tanks." That's not quite a quid pro quo, but it's in that ballpark.

These aren't hypothetical concerns. Multiple journalists and researchers have documented actual instances of significant traders attempting to influence real-world outcomes to protect their positions. The New York Times reported that during the 2024 election cycle, a single trader on Polymarket made over $35 million by predicting the election outcome, then used their platform presence to make increasingly bold and granular predictions about voting patterns and outcomes. Whether those predictions influenced the information environment remains an open question.

Information Integrity: How Markets Distort What We Know

Beyond elections, prediction markets are creating problems for information integrity across the board.

Consider how prediction markets handle information about corporate mergers or acquisitions. A company announces a potential deal. A prediction market is created: will this deal close by Q3 2025? Normally, journalists, analysts, and regulators would evaluate the deal based on the merits. Is it good for consumers? Will it reduce competition? Are there antitrust concerns?

But now there's also a prediction market. And that market will attract people betting for and against the outcome. Some of those bettors might be insiders with actual information about whether the deal will close. Others might be spreading information (accurate or not) that supports their position.

Meanwhile, journalists are now covering the prediction market odds as part of their reporting. "Market now expects deal to have 62% chance of closing." They're treating the market as an independent information source, when really it's just reflecting the positions of people with financial skin in the game.

This creates a feedback loop where market prices become self-validating. A prediction market says something is likely. News coverage focuses on that likelihood. The market sees the news coverage, interprets it as new information, and moves further in that direction. Reality hasn't changed. Information ecosystems have just become weirder and more reflexive.

There's also the problem of unfounded rumors turning into market movements. On Polymarket, if a high-profile trader makes a large bet on something, it signals to others that there might be inside information or special insight. Sometimes there is. Often there isn't. But the bet itself drives price movement. The movement attracts media coverage. The coverage makes the rumor feel more credible. And before you know it, you've got a substantial market movement based on a bet that was based on... what? A hunch?

This is especially concerning in markets about company decisions, government actions, or personal achievements. Imagine a prediction market on whether a specific CEO will announce a resignation. Some trader, for whatever reason, buys up contracts predicting resignation. The market odds shift from 8% to 23%. Business media picks up on it. Now the CEO is getting calls asking whether they're planning to resign. The pressure increases. And maybe, just maybe, it becomes a self-fulfilling prophecy.

There's also the phenomenon of "toxic prices" in prediction markets. An outcome becomes priced so high or so low that the market is clearly divorced from reality. For instance, in late 2024, certain prediction markets were showing probabilities for specific outcomes in the 1-2% range despite significant evidence that those outcomes were more likely. Why? Because large traders had taken massive positions and no one wanted to trade against them. The prices became meaningless.

The Regulatory Mess: Who Actually Rules These Platforms?

Here's the uncomfortable truth about prediction market regulation: it's a complete mess.

Traditionally, financial markets in the United States are regulated by the SEC (stock exchanges), the CFTC (futures and derivatives), or state-level gambling commissions (sports betting).

Prediction markets don't fit neatly into any of these categories. They're not quite securities. They're not quite futures contracts. They're not quite gambling.

So what happened? Platforms basically created their own regulatory frameworks and dared the government to stop them.

Kalshi, for instance, operates under an exemption granted by the CFTC that allows it to offer contracts on certain political and economic events. But this exemption is narrow and contested. The CFTC has expressed concerns about whether it should have granted this exemption at all. There's an ongoing push from various regulators to clamp down or establish clearer rules.

Polymarket operates in a much hazier gray zone. Technically, Polymarket is a peer-to-peer exchange based offshore, and US residents are technically prohibited from using it. But the prohibition is largely unenforced, and millions of Americans use it regularly. The platform has spent years trying to navigate US regulatory requirements while maintaining plausible deniability about whether they're actually operating under US jurisdiction.

This regulatory ambiguity is actually a feature, not a bug, from the platform's perspective. If the rules were crystal clear, many current platforms would likely become non-compliant. The uncertainty is what allows them to operate.

What happens if (or when) regulators decide to crack down? Several scenarios are possible:

Scenario 1: New restrictive rules. The CFTC or Congress could establish clear rules that dramatically limit what kinds of contracts can be offered or who can participate. This would be bad for prediction market growth but potentially good for information integrity.

Scenario 2: Legalization with safeguards. Regulators could legalize and standardize prediction markets but with requirements around transparency, preventing manipulation, and limiting participation by certain groups. This has been proposed by some researchers.

Scenario 3: Status quo continues. Regulators remain unclear on jurisdiction, rules stay murky, platforms keep operating in gray zones, and we just accept that we're conducting a massive experiment on information integrity without proper oversight.

The third scenario is where we are right now. And it's not sustainable.

The Economics: How Money Really Flows Through These Markets

To understand prediction markets, you need to understand the economics.

Platforms make money through fees. On Kalshi, you typically pay a small fee (around 2%) on winning trades. On Polymarket, the fee structure is more complex but roughly similar. These fees add up. A platform processing hundreds of millions in daily trading volume is generating tens of millions in annual revenue.

Where does that money come from? From users. Prediction market participants, on average, lose money. Not all of them—skilled or lucky traders make bank. But the aggregate flow of funds is from users to platforms and to sophisticated traders who know how to exploit retail participants.

This is important because it means prediction markets are, fundamentally, a wealth transfer mechanism. Money flows from casual participants to platforms and professionals. And that money is real—it's not being created from thin air. It's opportunity cost. That's money people could have used for other things.

How much? Estimates are hard, but consider: if prediction markets are currently handling billions annually, and the average participant has a negative return (which they do), then we're talking about potentially hundreds of millions of dollars annually flowing from retail participants to professionals and platforms.

Now, is that inherently problematic? Casinos operate on the same model. Sports betting operates on the same model. Any marketplace where there are winners and losers will have this property.

But prediction markets present themselves as intellectual endeavors, not gambling. They market themselves to people as a way to make money by being smart, by having good forecasting skills. In reality, most participants are losing money, and the profits are going to people with more capital, more information, and more sophisticated trading strategies.

There's also the venture capital dimension. Prediction market platforms have raised hundreds of millions in venture funding. This creates pressure to grow user bases, increase trading volumes, and expand into new types of predictions. All of this is driven by the venture capital model, which demands exponential growth.

So you get platforms that are incentivized to attract as many casual users as possible, encourage high trading volumes, and expand into increasingly niche or concerning prediction categories. Not because it's good for the information ecosystem, but because it drives growth metrics that venture investors care about.

Accuracy and Reality: Do These Markets Actually Predict Anything?

Let's talk about the core claim: are prediction markets actually accurate?

The answer is complicated. In controlled settings, with clear resolution criteria and sufficient liquidity, prediction markets do tend to aggregate information reasonably well. If you're predicting whether a coin will land heads or tails, a prediction market will figure it out.

But for the kinds of predictions that matter—political elections, corporate decisions, geopolitical events—accuracy is more mixed.

There are a few empirical studies of prediction market accuracy. A 2023 analysis of prediction markets during the 2020 US election found that the markets were, on average, slightly more accurate than polling. But the margin was small—within the margin of error. In some cases, they were worse.

During the 2024 cycle, prediction markets in multiple countries showed significantly more uncertainty than turned out to be warranted. This could be interpreted two ways: either the markets were appropriately accounting for genuine uncertainty (and the outcomes turned out to be more predictable than expected), or the markets were distorting reality by increasing perceived uncertainty.

The problem with studying accuracy is that we rarely have clean tests. By the time a prediction resolves, the world has changed in ways that make it hard to evaluate whether the prediction was truly accurate or just lucky.

Also important: prediction market accuracy varies wildly depending on the type of prediction. Markets predicting near-term, objective outcomes (will this FDA decision happen by June 30?) tend to be pretty accurate. Markets predicting subjective, long-term, or value-based outcomes (will this person be remembered as a great president?) are often not.

What we do know is that prediction markets have become more expensive to move with capital. Early prediction markets (2015-2018) could be significantly influenced by large trades from individuals. Modern prediction markets are more efficient, which means they're harder to manipulate. But they're still far from perfectly efficient.

The other issue is that prediction market accuracy isn't distributed evenly. Some categories of markets are predictable and accurate. Others are chaotic and essentially random. Users often don't know which category they're trading in until they lose money.

Behavioral Economics: Why Smart People Make Dumb Bets

One dimension that rarely gets discussed is the behavioral economics of prediction markets.

When you give humans money and ask them to forecast uncertain outcomes, interesting things happen. People don't act like rational agents weighing probabilities. They act like psychological creatures with all sorts of biases and incentives.

Overconfidence is huge. Studies show that about 70% of prediction market participants are overconfident in their forecasting ability. They think they're better at predicting than they actually are. This keeps them trading, despite negative returns.

There's also anchoring bias. The initial price of a prediction contract, even if essentially arbitrary, heavily influences subsequent trading. If a contract starts trading at $0.40, people assume there's a reason. They assume there's information embedded in that price. Even if the price was determined randomly, it will influence where the market settles.

Social proof is powerful. If you see a bunch of people buying a specific contract, you assume they know something. You follow them. This can create bubbles where prices disconnect from fundamental value.

There's also the "narrative" problem. Humans think in stories. We prefer to believe that events happen for reasons, that the future will rhyme with the past, that our intuitions are correct. Prediction markets that align with compelling narratives attract money, even if they shouldn't.

Consider: during periods of high political polarization, prediction markets often skew toward the predictions that align with the user base's preferences. If Polymarket users skew Democratic, you'll see Democratic-favorable prediction markets get bid up. The market isn't predicting accurately. It's reflecting the preferences of the traders.

There's also availability bias. People overweight information they've recently encountered. If you've been reading stories about recession risks, you'll overestimate the probability of recession in prediction markets.

Finally, there's the "gambler's fallacy" and hot hand fallacy. People believe that if something hasn't happened in a while, it's "due" to happen. Or conversely, if something just happened, it's more likely to happen again. These beliefs influence prediction markets.

The Information Warfare Angle: Can Bad Actors Exploit These Systems?

Let's be direct: prediction markets are vulnerable to information warfare.

Here's how a sophisticated actor might exploit them:

Step 1: Identify a market with moderate liquidity but not so much that individual trades are meaningless. Maybe it's a market on whether a specific country will impose tariffs on US imports. Trading volume is decent but not enormous.

Step 2: Accumulate a large position. Buy up contracts predicting the outcome you want. This requires capital, but not necessarily enormous amounts. Millions of dollars, not billions.

Step 3: Spread complementary information. Use social media, bot networks, or friendly media outlets to spread information that supports your position. The information can be true, false, or misleading—doesn't matter much. The goal is to move market sentiment.

Step 4: Watch the market move. As others see the information and the market movement, they start buying in the same direction. The price rises. Your position becomes more valuable.

Step 5: Exit the position. Once the price has risen, sell. You've made money. And you've also influenced how people perceive the likelihood of the outcome.

Step 6: (Optional) Continue spreading information. Even after you've exited your position, you might continue spreading information if it suits other goals. Maybe you're a government trying to influence public perception. Maybe you're a rival business trying to hurt a competitor. The point is, you've decoupled the market from reality.

How easy is this? Easier than you'd think. The barrier to entry is financial capital and willingness to be deceptive. Both are readily available in the world.

Would it work? In smaller markets, absolutely. In larger, more liquid markets, it would be harder but still possible, especially with sufficient capital.

Would it be detectable? Not necessarily. Market manipulation is hard to prove. Was that large trade based on genuine information or was it manipulation? Regulators might never know.

The Platform Problem: Polymarket, Kalshi, and the New Generation

Different prediction market platforms have different structures, and that matters.

Kalshi operates as a US-regulated platform under a CFTC exemption. This means it has legal obligations and regulatory oversight. It's more compliant but more constrained. Kalshi can't offer certain types of contracts because regulators said so. Kalshi has to report things to regulators. Kalshi can be held accountable if things go wrong.

But Kalshi also has to turn down profitable opportunities because they don't fit within its regulatory exemption. This means Kalshi is smaller and less feature-rich than unregulated competitors.

Polymarket operates as a peer-to-peer exchange without US regulation. The theory is that it's a platform, not an exchange. Users trade directly with each other. Polymarket just provides the infrastructure.

In practice, Polymarket is obviously an exchange. But the legal structure creates plausible deniability. Polymarket can offer any type of contract it wants because it's not operating under US regulation. Polymarket can have users from anywhere in the world. Polymarket can be more feature-rich and profitable because it's not constrained by regulation.

But Polymarket also operates in legal ambiguity. If the US government decided to enforce laws against prediction markets, Polymarket could be shut down or forced to cut off US users.

There are other platforms too: Manifest Markets (academic focus), Hypermind (older platform), INFER (acquired by Metaculus, which focuses on longer-term forecasts), and new platforms launching constantly.

Each platform has different fee structures, different markets, different user bases. This fragmentation means that the same event might be priced differently on different platforms. If you knew this, you could potentially profit through arbitrage.

But it also means the "prediction market consensus" isn't as clear-cut as people think. There's not one prediction market saying that something is 60% likely. There are multiple markets, with different prices, and you have to aggregate them somehow.

Addiction and Behavioral Health: The Hidden Cost Nobody Talks About

Here's something that rarely comes up: prediction markets can be addictive.

They have many of the hallmarks of gambling or social media addiction. Immediate feedback. Variable rewards (sometimes you win big, sometimes you lose). Sense of control (you can make informed bets). Community involvement. Frequent opportunities to engage.

For people with gambling addictions or similar impulse control issues, prediction markets are particularly dangerous. The platforms don't typically have meaningful addiction safeguards. There's no automated spending limit. There's no warning that tells you you've lost X amount this month. There's no cooling-off period for large bets.

The platforms are designed to encourage engagement, which means engagement toward addiction, at least for vulnerable populations.

How many people have experienced significant financial harm as a result of prediction market engagement? We don't have good data. The platforms don't publicize negative externalities. But given the size of the market and the lack of safeguards, it's probably not zero.

There's also the phenomenon of people checking prediction market prices constantly. It becomes a compulsive behavior. Like constantly checking email or social media.

More broadly, prediction markets can distort your relationship with uncertainty. Normal uncertainty is uncomfortable, and we develop coping mechanisms. But if you can always find a market that gives you a precise probability, you might start using that market as a substitute for actual thinking. You might stop making your own judgments and just accept the market price.

This is cognitive outsourcing. And while markets can be helpful sources of information, replacing your own judgment entirely with market prices is probably not good for your own reasoning abilities.

Possible Solutions: Can We Make Prediction Markets Less Problematic?

Given that prediction markets are growing and not going away, what could make them less harmful?

Several possibilities:

Better regulation. Clear rules about what types of contracts can be offered, who can participate, how to prevent manipulation, and how to minimize social harms. This would kill some of the innovation but might be worth it.

Transparency requirements. Require platforms to disclose who's trading what, at least to regulators. This would make manipulation easier to detect. It would also reduce the information advantage of insiders.

Addiction safeguards. Implement spending limits, cooling-off periods, warnings about negative expected value, and links to gambling addiction resources. Basically, treat prediction markets like gambling products, because they are.

Restrict certain categories of predictions. Ban markets on elections, or at least restrict who can trade in them. Some things shouldn't be financialized. Democracy might be one of them.

Improve market design. Some designs (like market makers with skin in the game) are less prone to manipulation than others. Support better-designed markets and discourage poorly-designed ones.

Consumer education. Teach people that prediction markets are not oracles. They're betting venues. They contain no special truth. They're expressing the positions of people with financial incentives, not objective reality.

Mandatory disclaimers. Require platforms to be clear about the lack of accuracy, the presence of manipulation risk, and the typical negative returns.

Competitor regulation. Allow multiple platforms to operate but regulate them clearly and enforce the regulations consistently. This prevents monopolistic problems.

Will any of these things happen? Probably some of them, eventually. Prediction markets are growing fast enough that they'll attract regulatory attention. The question is whether that attention comes proactively or reactively, i.e., after a major scandal.

The Future: Where Are Prediction Markets Headed?

Prediction markets are growing. That much is clear. But how much will they grow? And into what?

Several trajectories are possible:

Scenario A: Mainstream adoption. Prediction markets become as normal as sports betting. You can place bets on basically anything through mainstream platforms. Millions of people participate regularly. The markets become important inputs into business decisions, policy decisions, and public perception.

In this scenario, the risks I've outlined become much more significant. Information warfare becomes more attractive. Democracy becomes more vulnerable. Wealth concentration increases as casual users lose money to professionals.

Scenario B: Regulatory crackdown. Governments crack down on prediction markets, either banning them outright or imposing such strict regulations that they become much less profitable and thus less appealing. Most current platforms would need to drastically change their business models.

In this scenario, the worst harms are mitigated, but you also lose some of the potential benefits (better forecasting, aggregated information, etc.).

Scenario C: Niche evolution. Prediction markets remain popular but remain somewhat niche. They grow, but not to sports betting levels. They become more sophisticated, more professional, and more useful for certain purposes (corporate research, policy analysis) but less central to mainstream consciousness.

This might actually be the best outcome, but it requires prediction market platforms to stop aggressively pursuing growth, which is counter to how venture capital works.

Scenario D: Crash and burn. A major scandal involving prediction market manipulation or misinformation causes public backlash. Platforms lose credibility. Users leave. The whole thing shrinks back to a niche.

Historically, lots of financial innovations have followed this trajectory. They grow explosively, cause problems, scandals hit, and then they either reform or collapse.

My guess? We're heading toward some combination of Scenario C and B. Prediction markets will face increasing regulatory pressure. Some platforms will adapt and survive. Most will probably disappear. The industry will settle into something smaller and more constrained than current trajectories suggest.

But I've been wrong about these things before.

The Broader Conversation: What Are We Trading For?

Ultimately, prediction markets raise a deeper question: what are we willing to sacrifice for better forecasting?

They promise something valuable. Aggregated information. Better probability estimates. A way to see hidden truths about the future.

But the cost of getting those things is that we're converting uncertainty into tradeable commodities. We're creating financial incentives for people to spread information about important events. We're making democracy, business outcomes, and public understanding subject to manipulation by people with capital and motivation.

That's not a crazy trade-off. It might be worth it. But it's worth being honest that there is a trade-off. Prediction markets aren't costless.

They're efficient information aggregation mechanisms with significant negative externalities. Like many financial innovations.

The question is whether the benefits outweigh the costs. And that's a question we should be asking more carefully before we let this industry grow unchecked.

Because right now, in early 2025, we're conducting a massive, uncontrolled experiment on information integrity and democratic processes. We're betting—literally betting—on the outcome. And most of us don't even know the game is happening.

That's the real cost of prediction markets.