Kalshi Contract Impermanent Loss: Why Liquidity Providers on Event Markets Face Hidden Volatility Decay

Market makers on event prediction platforms face a structural problem that separates them from traditional derivatives traders. When a contract on Kalshi moves toward resolution—whether it represents a Federal Reserve interest rate decision, employment data, or a binary policy outcome—the price converges to either $0 or $100. That convergence is mathematically certain, not probabilistic. A liquidity provider who has sold contracts at $45, expecting the event to remain uncertain, will eventually see those positions forced to either total loss or complete profit. Unlike equity markets where volatility is a two-way street, event contract volatility is a one-way compression that eats into maker profits through a form of impermanent loss specific to binary settlement.

This article examines why market makers accepting tight spreads on Kalshi face hidden decay in their expected returns, and how the platform’s current fee and matching structure may systematically undercompensate them for the risks they absorb. The mechanism is not difficult to understand once separated from the trading platform’s general appeal: predictive value and structured price discovery are genuine. But the compensation for providing the liquidity that enables that price discovery remains misaligned with the adverse selection and inventory risk that liquidity providers actually face as contracts age and information tightens the range of plausible outcomes.

How event contract dynamics differ from perpetual derivatives

On a perpetual futures platform, a market maker can hold inventory expecting mean reversion or gradual drift. If bitcoin falls from $45,000 to $42,000, a maker who sold at $45,000 faces losses, but the market could recover. Price volatility creates repeated opportunities. Kalshi contracts do not behave this way. An event contract settling on whether the unemployment rate will exceed 5% on a specific date has a known maturity. As that date approaches, information accumulates. Employment data, market expectations, and economic commentary all narrow the range of plausible outcomes.

A market maker who supplied liquidity at $50 (representing 50% implied probability) early in a contract’s life is betting that the event will remain uncertain. But if the event becomes increasingly likely or increasingly unlikely, the price will drift toward $0 or $100. The maker cannot achieve mean reversion because the mean itself is shifting toward the boundary. This is not volatility in the traditional sense; it is information arrival compressing the range of prices.

The asymmetry for market makers is severe. If a maker sells contracts at $50 and the true probability is later revealed to be 30%, the maker loses money. But the maker does not get to profit equally when the initial price underestimated the true probability. Orders get filled at the spread; takers capture the directional move. A maker’s profits depend on the spread width, the volume traded at that spread, and the accuracy of the initial prices quoted. If market participants are better informed or if new information arrives before the maker can adjust inventory, the maker’s long and short positions are not symmetric.

This dynamic becomes severe near resolution. When an event contract trades at $95, the implied probability is very high. A maker offering to sell at $96 is betting that the remaining 4% chance is real, not a mirage. If the maker is wrong, the contract settles at $100, and the short position loses the remaining $4 per contract. The maker’s fee income, typically measured in basis points, must cover not only the spread during uncertain periods, but also the tail risk when information crystallizes.

Adverse selection as information tightens

Adverse selection on prediction markets is not symmetric around the resolution date. Early in a contract’s life, there is genuine uncertainty. Takers on the platform may be arbitrageurs, sophisticated forecasters, casual bettors, or hedgers. Their motivations are mixed enough that a market maker can profit on volume without taking large directional risk. But as a contract approaches resolution, the remaining participants are increasingly informed. The takers left on the platform tend to be those with the strongest conviction that prices are wrong.

A maker on the Kalshi platform who quotes $96 to $97 on a contract near resolution is advertising vulnerability. Informed takers will aggressively buy at $96 if they believe the true probability exceeds 97%, or aggressively sell if they think it is lower. The maker’s edge—the spread—no longer compensates for the one-sided information. This is the textbook problem of adverse selection: the maker ends up trading with the best-informed takers and holding inventory that becomes increasingly mismatched to reality.

The decay is not immediate but intensifies exponentially as resolution approaches. A contract that trades with a 2% spread when its true probability is 50% might still trade with a similar spread when the true probability is 70%, because uncertainty still exists. But when the true probability is 95%, a 2% spread ($1.90 on a $100 contract) is absurdly large. Makers will tighten spreads to attract takers, or they will simply withdraw liquidity. The compression squeezes maker profits precisely when their inventory becomes most vulnerable.

Why order types and matching mechanics amplify maker risk

Kalshi’s exchange matching logic determines how quickly a market maker’s exposure can be adjusted. If the platform prioritizes price-time matching (oldest orders at the best price execute first), a maker who quoted early in the day is stuck holding inventory while newer takers hit the ask with better information. If the platform allows cancellations with latency, a maker may execute at stale prices while trying to adjust. The speed and cost of rebalancing directly determine impermanent loss.

In traditional derivatives trading, market makers use smart order routing, cancel-replace cycles, and algorithmic inventory management to stay hedged. On Kalshi, a maker’s tools are more basic: cancel existing orders and place new ones. If the market moves, the maker’s old quotes may execute against informed takers before cancellation can take effect. The latency between receiving new information and updating quotes creates a window where the maker faces adverse selection from takers who moved faster.

This structural limitation means Kalshi market makers must either build wider spreads to account for the risk, or withdraw liquidity during volatile periods. Wider spreads reduce trading volume and hurt the platform’s utility for takers trying to build or unwind positions. The fee structure does not explicitly compensate makers for latency risk. A fixed maker fee (such as 0.1% or 0.05% per side) is constant regardless of whether the maker is operating with stable prices or facing information-driven convergence. As a result, the compensation per unit of risk worsens as an event approaches resolution.

The resolution-day problem: pinning and forced settlement

On the day an event resolves, liquidity providers face acute pressure. If a contract is still trading at $75 hours before the outcome is announced, a market maker holding a short position is effectively betting that the event will not occur. If there is material doubt, takers will be aggressive buyers, hitting the maker’s ask repeatedly. The maker cannot accumulate more short inventory without taking losses. The only rational response is to raise the ask price closer to $100 or exit the position entirely.

But if the maker exits by buying back contracts, the maker realizes losses on any contracts that were undersold relative to the true probability. If the maker holds, the maker faces the full risk of the event occurring. This is the pinning problem: as settlement approaches, there is no profitable way for a maker to hold inventory that is mispriced. The maker must either absorb losses or abandon the market.

Forced settlement also creates a liquidity cliff. In the final hours before an event resolves, takers who need to square positions become price-takers rather than price-makers. They will cross wide spreads to exit. Market makers, seeing that liquidity is needed most, might quote even wider spreads, knowing that takers are desperate. This is rational from each maker’s perspective but leaves the platform with poor price discovery exactly when transparency matters most. A taker trying to offset a position hours before resolution may face spreads that are 5–10% wide, compared to 0.5% early in the contract’s life.

Fee structure misalignment with maker risk

Kalshi charges makers a percentage fee (typically 0.1% to 0.2% per filled side, though this can vary). This structure creates a hidden cost for makers operating in the adverse selection environment described above. A maker who earns a spread of 2% on a contract might generate 0.1% in fee income and 1.9% in gross margin. But if the maker’s inventory drifts because of adverse selection, the maker must later close at worse prices. The effective margin, after accounting for replacement costs and information decay, might be 0.5% or less.

Fee-per-side structures are particularly problematic for inventory management. If a maker buys contracts at $48 and sells at $52, the maker earns 4% spread on the round trip but pays 0.4% in fees (0.2% on each side). The net is 3.6%. But if adverse selection forces the maker to close the short position at $51 instead of waiting to sell at higher prices, the maker realizes only a 3% profit on that leg and a $1 loss on the long leg, netting $2 profit on the full round trip against a 2% fee cost. The compensation erodes as the maker is forced to rebalance.

A more sophisticated fee model might offer rebates for makers who provide liquidity near resolution, or variable fees that reflect the maker’s time-to-settlement risk. Instead, Kalshi’s current structure treats all market-making equally. A maker quoting tight spreads on a contract with one day to resolution faces the same fee percentage as a maker quoting on a contract with three months to resolution, yet the risk profile is entirely different. This structural misalignment is why impermanent loss persists even on well-functioning exchanges: the fee model does not capture the full economic cost of the risk.

Inventory decay and the illusion of tight spreads

Tight spreads—say, 1% or 2%—on Kalshi attract traders. They signal a liquid market and encourage participation. But a tight spread is only profitable for a maker if the underlying probabilities remain stable or move slowly. If a contract’s true probability shifts from 50% to 70% in a single day, a maker who quoted 49%–51% spread has already lost before earning any fee income.

This creates a classic market-making paradox: the most competitive makers, who offer the tightest spreads, are the first to suffer impermanent loss as information arrives. Conservative makers, who quote wider spreads, protect themselves better but reduce platform liquidity. The aggregate effect is that competition for taker volume during certain periods (early contract life) creates conditions for losses during other periods (late contract life).

The illusion of tight spreads persists because early-contract volumes are high and competition is fierce. By the time a maker experiences the full impact of adverse selection, the contract is approaching resolution and the maker is trying to exit. New makers arrive and repeat the pattern. The platform appears to have good liquidity, but the sustained profitability of individual makers is lower than gross spread numbers suggest.

Strategies makers use to mitigate impermanent loss (and why they are incomplete)

Experienced market makers on prediction platforms employ several tactics to reduce exposure. One approach is delta hedging: if a maker is short a contract at $50, the maker might short the same exposure on another platform or trade through a different account to maintain neutral exposure. This works only if correlated contracts exist and if transaction costs are low enough to make hedging profitable. On Kalshi, hedging options are limited, and the cost of establishing offsetting positions elsewhere often exceeds the spread earned.

Another tactic is inventory rebalancing: a maker actively closes unprofitable positions rather than waiting for mean reversion. But rebalancing forces the maker to realize losses early, which is only rational if the alternative is larger losses later. A maker must constantly estimate the true probability and compare it to the current market price. If the maker is wrong, rebalancing converts unrealized losses into realized ones.

A third approach is time-weighted quoting: a maker quotes tighter spreads early in a contract’s life and widens them sharply as resolution approaches. This signals to takers that liquidity will deteriorate, which may discourage them from waiting to trade. But it also concentrates the maker’s risk in the most uncertain period, which is when adverse selection is least predictable. The net effect depends on the timing and magnitude of information arrival, which no maker can control.

None of these strategies fully eliminates impermanent loss. They manage it, shift when it occurs, or trade it for other risks. The fundamental issue remains: event contracts converge to $0 or $100, and makers who quote in the middle do not have symmetric outcomes. Smart makers accept this and price it into their spreads. Naive makers discover it after suffering losses.

What platform design would better align incentives

A more maker-friendly fee structure might include rebates for orders that rest longer near resolution, or variable fees that decrease as time-to-settlement shrinks. This would reward makers for holding inventory during the most information-sensitive periods, directly offsetting the adverse selection risk they face. Alternatively, the platform could offer market maker bonds, where makers commit to quoting minimum spread widths and receive fee reductions in exchange. This creates transparency about the cost of liquidity.

Another improvement would be better matching transparency. If Kalshi published real-time data on who is trading what, when, and at what prices, makers could more quickly update their probability estimates and reduce information lag. Currently, information asymmetry advantages takers with better external data sources. Reducing that asymmetry would shrink the profitable opportunities for informed takers and reduce the adverse selection problem.

Longer contract lives could also help. If Kalshi offered contracts that settle months or years after event resolution, makers could hold inventory longer without facing the acute pinning problem near settlement. Event outcomes are most uncertain when the event is far away; providing liquidity during that period should be more profitable and less risky than current designs allow.