How to Short on Prediction Markets: The Mechanics Explained

If you're coming from stocks or crypto, you probably expect a "short" button on every market. Prediction markets don't have one — and they don't need one. The structure of binary contracts makes shorting either redundant or trivially expressible through the existing interface. This post walks through exactly how to "short" on Polymarket and Kalshi, when each approach makes sense, and the small handful of pitfalls that bite new traders.

---

---

Why Prediction Markets Don't Have a "Short" Button

In stock markets, shorting means selling shares you don't own — borrowed from someone else's account, with the obligation to buy them back later. The mechanism exists because stocks are open-ended (a stock can go to any price, theoretically), and there's only one side of the trade displayed in most order books.

Prediction markets work differently. Every market has two contracts listed:

• YES contract: pays $1.00 if the event resolves true.
• NO contract: pays $1.00 if the event resolves false.

The two prices always sum to roughly $1.00 (give or take the spread). If YES trades at 65¢, NO trades at ~35¢. The market is implying a 65% probability that YES wins, equivalently a 35% probability that NO wins.

To "short YES" on a prediction market, you just buy NO. That's it. There's no borrowing, no margin, no obligation to deliver shares back to a lender. You hold a contract that pays out if the event doesn't happen.

Mathematically, buying NO at 35¢ is identical to shorting YES at 65¢:

• If the event resolves YES (true): your NO contract pays $0.00. You lose your 35¢ premium.
• If the event resolves NO (false): your NO contract pays $1.00. You make 65¢ on the 35¢ you invested — a 186% return.

That's the same payoff profile as selling YES short at 65¢ in a hypothetical short-sell market. The exchange just structured it as a buy-NO transaction instead of a sell-YES transaction.

---

---

The Two Ways to Express a Short View

Even though the platform doesn't have a "short" button, there are actually two different actions you can take depending on whether you already own shares:

Method 1: Buy NO contracts (the most common)

You don't own any contracts yet, and you want to bet the event will resolve false. Click Buy NO and place an order at the price you're willing to pay.

Example. A market trades YES at 70¢, NO at 30¢. You think the event won't happen. You buy 1,000 NO contracts at 30¢:

• Cost: 1,000 × $0.30 = $300
• Max payout: 1,000 × $1.00 = $1,000 if event resolves NO
• Max loss: $300 if event resolves YES
• Risk/reward: pay $300 to potentially make $700

Method 2: Sell YES contracts you already own

You own 1,000 YES contracts that you bought at 50¢, currently trading at 70¢. You want to close the position because you no longer expect YES to win. Click Sell YES and place an order at the current bid (or higher).

Example. Current bid for YES is 69¢. You place a market sell on your 1,000 YES contracts:

• Receive: 1,000 × $0.69 = $690 (minus fees)
• Original cost: 1,000 × $0.50 = $500
• Realized profit: $190 (38% on cost)

This is "closing a long" rather than "opening a short," but the practical effect is the same: you no longer have exposure to the YES outcome, and if YES eventually loses, you don't care.

---

---

A Worked Example on Polymarket

Let's walk through a concrete trade. A Polymarket market asks "Will the Federal Reserve cut rates at the June meeting?" Current prices:

• YES: 42¢ (implying 42% probability of a cut)
• NO: 58¢ (implying 58% probability of no cut)

You think the market is underestimating the chance of no cut. You believe the true probability is closer to 70% NO, 30% YES.

Step 1: Identify your edge. Market price says NO is worth 58¢. Your model says NO is worth 70¢. Your edge is 12¢ per contract, or roughly 20% expected return on cost.

Step 2: Place a Buy NO order. Navigate to the market on Polymarket, click the NO tab (or the "Buy NO" button on mobile), and enter your order size. Let's say you want to risk $580 to potentially make $420 in profit:

• Order: Buy 1,000 NO at 58¢
• Cost: $580
• Max payout: $1,000 if NO wins
• Max loss: $580 if NO loses

Step 3: Monitor and exit (optional). If, two weeks later, NO has rallied to 75¢ because the market now agrees with your view, you can:

• Hold to resolution and collect $1.00 per share (max profit)
• Sell at 75¢ to lock in a $170 profit per 1,000 contracts (early exit, lower payoff but quicker realization)

Step 4: Resolution. When the Fed announces (or doesn't), the market resolves. If you're holding NO and the Fed doesn't cut, you receive $1.00 × 1,000 = $1,000. Your profit: $1,000 - $580 = $420 minus any trading fees.

---

---

A Worked Example on Kalshi

The same logic on Kalshi, with slightly different UX. Kalshi market: "Will US CPI YoY be above 3.0% in the next release?"

• YES: 38¢ (38% implied probability)
• NO: 62¢ (62% implied probability)

You think CPI is more likely to come in above 3.0% than the market does. You want to buy YES at 38¢, not short.

Wait — that's a long position, not a short. Right. Most short views express as a buy on the opposite contract.

Let's flip it: you think CPI will come in below 3.0%, so you want to short the "CPI above 3.0%" question. You buy NO at 62¢:

• Order: Buy 500 NO at 62¢
• Cost: $310
• Max payout: $500 if CPI comes in below 3.0%
• Max loss: $310 if CPI comes in at or above 3.0%

Kalshi's UI explicitly labels this as "Buy NO" — the same as Polymarket. The position will show in your portfolio as "500 NO shares of [market name]," with realized P&L calculated against the original premium paid.

---

---

The Things That Trip Up New Shorts

A few mechanics that catch new prediction-market traders off guard:

1. NO is sometimes harder to buy than YES

Order books aren't always symmetric. If a market has heavy YES demand, the YES side will have lots of resting orders and the NO side may be thin. You might see:

• YES bid: 67¢, YES ask: 68¢ (1¢ spread, thick book)
• NO bid: 31¢, NO ask: 34¢ (3¢ spread, thinner book)

To buy NO at a "fair" price, you may need to place a limit order at the midpoint and wait. Hitting the ask on a thin NO book costs you 2–3¢ that you wouldn't lose if you traded the more-liquid YES side. This is a real cost of shorting that doesn't exist when you go long the popular direction.

2. Implied price asymmetry near resolution

As a market approaches resolution, the prices can move asymmetrically. A YES that looks "locked in" at 95¢ doesn't mean NO trades at exactly 5¢ — it could be 3¢ or 7¢ depending on liquidity. If you're shorting YES at 95¢ thinking "this can't go above $1," you're right, but the upside is only 5¢ per contract — and the spread can eat half of it on exit.

3. Resolution risk in disputed markets

If a market has ambiguous resolution criteria, the dispute can take days or weeks to settle. During that window, you can't trade the position. For shorts who needed the capital to deploy elsewhere, that's a real cost.

4. Fees on small NO positions

Both Polymarket and Kalshi charge fees as a function of price. Fees are highest at 50¢ and taper toward the extremes. So buying NO at 40¢ has roughly the same fee per contract as buying NO at 60¢ — both close to the peak. Buying NO at 90¢ (deep ITM short) has very low fees per contract, but you're paying a high premium for a small expected return. Choose your entry price based on edge, not on what feels "safe."

---

---

When Shorting Makes Sense

Five scenarios where opening a short (i.e., buying NO) is the right move:

1. The market price exceeds your true-probability estimate

This is the bread-and-butter +EV trade. The market is at 65¢ YES, you think the true probability is 50%, so YES is mispriced. Buy NO at 35¢ — you're paying 35¢ for something you think is worth 50¢, an expected 15¢ edge.

2. Hedging a long position elsewhere

You bought YES on Polymarket at 50¢. The market has rallied to 70¢ on Kalshi. Rather than selling your Polymarket YES at 70¢ (which might have a wider spread or fee), you can hedge by buying NO on Polymarket at 30¢. Now you're flat: a YES win pays back $1.00 on the YES side and zero on the NO side (net YES). A NO win pays $1.00 on the NO side and zero on the YES side. Either way, you've locked in a net position worth $1.00.

3. Capturing volatility decay

For markets near resolution where the price is hovering around 50¢ with high uncertainty, sometimes the implied volatility is irrationally high. Selling premium (buying NO into a too-high YES, or vice versa) and waiting for resolution captures some of that volatility decay.

4. Arbitrage across platforms

YES on Polymarket trades at 64¢. NO on Kalshi trades at 35¢. Combined cost: 99¢. Combined payout: $1.00. Buying both legs gives you a guaranteed 1¢ per contract (minus fees and execution risk). See our arbitrage guide for execution details.

5. Expressing a contrarian conviction

Markets aggregate consensus. When the consensus is loud and you think it's wrong, shorting is the cleanest way to express that. The classic example: late-cycle election markets where one side becomes overwhelmingly favored (e.g., 92¢) and a contrarian thinks the actual probability is closer to 80%. Buying NO at 8¢ pays 12.5x if the upset happens.

---

---

When Shorting Is a Trap

A few situations where new traders short and regret it:

1. Shorting a market that just resolved (or is about to)

If a market is about to resolve and you're shorting because "it can't go higher than 95¢," remember: the contract can settle at exactly $1.00 in literally minutes. There's no time decay benefit, no theta to harvest, and your downside is the full 5¢ if you're wrong.

2. Shorting illiquid markets

Wide spreads turn even a successful short into a loss after fees. If NO is trading 28¢/35¢ (7¢ spread), you'd need to be right by more than 7¢ just to break even on entry and exit. Stick to markets with sub-2¢ spreads if you're not willing to hold to resolution.

3. "It's obviously going to crash"

The market is at 90¢ on something you think is "obviously" overpriced. The market has been at 90¢ for two weeks while you've been telling yourself it's a great short. Either you're wrong about the probability, or the market is slower to update than you are. Either way, your short conviction needs to be priced against the cost of waiting.

---

---

Bottom Line

There's no separate "short" mechanism on prediction markets because they don't need one. Every binary market already has both sides listed; buying NO is shorting YES, and the payoffs are mathematically identical. The cap on your loss (the premium you paid for NO) is actually safer than traditional shorting, where losses are theoretically unlimited.

The skills that matter for shorting are the same as for going long: identifying mispriced markets, managing position size, watching for execution risk in thin books. The mechanics are simple; the discipline is what's hard.

For real-time price comparisons across YES and NO contracts on Polymarket, Kalshi, and other platforms, Prediction Hunt's market dashboard makes it easy to spot when one side is mispriced relative to the other — including the cross-platform arbitrage cases where YES on one venue plus NO on another sums to less than $1.00.