The decentralized prediction market Polymarket has risen rapidly in recent years, becoming a focus of attention for the crypto community and macro traders. Most users have heard Polymarket's core statement: "YES + NO = 1", but this seemingly intuitive formula actually involves the pricing logic of prediction markets, the design of shared order books, and many misunderstood arbitrage strategies.
Researcher DFarm analyzes the mechanism behind Polymarket's requirement that the sum of the prices for YES and NO must equal 1, and how this design affects liquidity and trading behavior. The following is a translation by ABMedia.
YES and NO are not two different assets, but rather a split of a dollar.
Many beginners intuitively believe that YES and NO are like two independent stocks, each with its own price, leading to the idea that "YES 0.7 + NO 0.6 = 1.3 is fine." However, this understanding is incorrect in Polymarket's design.
Polymarkets don't trade lottery tickets; they trade a future redeemable $1 certificate. The market simply splits that $1 into two complementary outcomes:
- If the event occurs: YES = $1, NO = $0
- If the event does not occur: YES = 0, NO = 1 USD
Therefore, at settlement, regardless of the outcome, the sum of YES and NO will always equal 1. As long as you hold both YES and NO in the same market and under the same settlement conditions, it is equivalent to holding an asset that will definitely be redeemed for $1 upon maturity.
A multi-option market is essentially still a combination of YES and NO.
Not all markets on Polymarket are presented as simple YES/NO statements. For example, Bitcoin price ranges and the number of tweets from specific individuals often involve multiple options. However, from the perspective of the API or structure, each option is still an independent "yes or no" proposition. For example:
- Will Musk post 0–19 tweets within a certain period?
- Will 20–39 tweets be sent?
Each option corresponds to either YES or NO, still following the pricing logic of YES + NO = 1.
The same concept applies to the sports market. Taking the NBA as an example, since a game will always have a winner and a loser, the home team and the away team actually correspond to the YES and NO of the same event, respectively. In sports such as football, where a draw is possible, the result is split into three complementary sets of outcomes: home team, away team, and draw.
Shared order book: centralized liquidity, rather than two separate markets
Unlike traditional cryptocurrency exchanges, Polymarket does not maintain completely separate order books for YES and NO, but instead uses a shared order book design. The orders for YES and NO are displayed in a mirror image on the other side.
In practice, when a trader places a buy order (e.g., price 0.18, quantity 10), the system will automatically display a sell order (price 0.82, quantity 10) in the NO market. The buy and sell orders on both sides are perfectly symmetrical, and the price always satisfies the relationship 1 − x.
This is why users cannot truly short YES or NO. The sell orders that appear on the surface do not represent someone borrowing securities to sell; rather, they are price mapping results from a shared order book. The purpose of this design is to centralize liquidity, avoiding the fragmentation of funds across two independent markets, thereby improving price discovery efficiency.
Why is it impossible to arbitrage between "YES + NO < 1"?
A common arbitrage strategy circulating in online communities suggests that by simultaneously buying low-priced "YES" and low-priced "NO" in the same market, ensuring the sum of the two is less than 1, one can reliably profit from the price difference. However, under a shared order book mechanism, this scenario is theoretically impossible to display on the screen.
The reason is that when someone tries to sell YES at a low price, the system is equivalent to receiving an instruction to "buy NO at a high price"; if another trader happens to sell NO at an even lower price, the two orders will be matched immediately, and no unbalanced orders will be left in the market.
In other words, the shared order book acts as an automatic balancing mechanism. Whenever a potential opportunity arises where YES + NO < 1, the system will absorb it internally immediately, leaving no room for external traders to intervene.
What are some arbitrage opportunities that predict the market will still exist?
Although YES/NO arbitrage does not exist within a single market, other arbitrage opportunities still exist under certain conditions:
Multi-option arbitrage
In a market with a set of mutually exclusive options that cover the entire range, if the total cost of buying all YES options is less than $1, theoretically, profits can be locked in. However, such opportunities are extremely rare and are usually captured instantly by high-frequency trading bots.
Cross-event arbitrage
When two different markets describe events with nearly identical semantics but differ in pricing, a small arbitrage opportunity may exist. However, this strategy requires extremely high levels of semantic understanding and judgment of settlement rules.
Cross-platform arbitrage
The most common example is arbitrage between Polymarket and platforms like Kalshi and Opinion. This requires both platforms to describe the same event, have identical settlement criteria, and ensure that the total cost, after deducting transaction fees and the time cost of locking up funds, remains below 1.
This article, "Understanding Polymarket in One Article: What is a Mirror Order Book? Why Must YES + NO Equal 1?", originally appeared on ABMedia .
