Last Updated: 2026-10-25
Risk: Prediction-market positions can lose the entire stake. The examples below are synthetic accounting exercises, not financial advice or a study of actual wallets.
The earlier “92% of traders lose” headline is withdrawn here. Its stated groups were 52% large losses, 27% smaller losses and 13% within plus or minus $10. The first two sum to 79%; the third includes possible gains and cannot all be counted as losses. The article did not supply the dataset, query, observation window or cost-basis reconciliation needed to establish a population loss rate. This does not establish that 79% is the real rate either.
First decide what the statistic measures
A wallet is not necessarily one person. A person may use several wallets; one wallet may represent multiple participants. Deposits are not profits, withdrawals are not losses, and an open position is not automatically worthless. A leaderboard may omit abandoned wallets or exclude users who never traded during the chosen window.
Before quoting a loss rate, ask for:
- The included markets, dates, wallet-selection rule and reproducible query.
- Cash-flow reconciliation, including transfers between related wallets when identifiable.
- Realized results, open-position valuation method and unpriced positions shown separately.
- Actual fees and whether results include rewards, rebates or unrelated activity.
- A declared break-even interval, missing-data treatment and denominator.
If these are absent, report the statistic as unverified rather than using it to sell a course or explain why any particular reader lost money.
A cost example without charging the spread twice
Suppose 1,000 YES shares are bought at an actual fill price of $0.71. Entry cash is $710. They subsequently settle at $1, producing $1,000. Assume all additional transaction costs total $1.50.
Scroll sideways to see every column. Keyboard: focus the table region and use the arrow keys.
| Item | Cash amount |
|---|---|
| Settlement proceeds | $1,000.00 |
| Actual entry payment | −$710.00 |
| Additional costs | −$1.50 |
| Net cash profit | $288.50 |
The $0.71 fill already reflects what was paid. If the earlier midpoint was $0.69, the $20 difference can explain execution quality, but subtracting it again from this cash profit would double-count it. If the contract instead settles at zero, the loss is $711.50 under the same cost assumption. Verify real fee records; $1.50 is only an input to this example.
An alternative-use benchmark belongs on a separate line. At an assumed 5% simple annual rate over exactly three months, $710 × 0.05 × 3/12 = $8.875, or about $8.88. Subtracting that hypothetical benchmark gives about $279.63 of excess result, not a new cash balance or an inflation-adjusted return. An APY convention would require compounding instead; a real alternative also has its own risk, terms and availability.
Holding, selling and drawdown require the right denominator
After a share rises to an executable bid of $0.85, holding it to binary settlement exposes roughly $0.85 of sale value for another $0.15 of upside, before costs. Its current price alone cannot establish whether holding has positive or negative expected value. That depends on an evidence-based probability estimate and the available exit.
A 40% drawdown leaves 60% of the starting balance. Returning to the original balance requires 40 / 60 = 66.67% growth. Recovery is possible mathematically; it is neither assured nor achieved simply by “being disciplined.” Avoid doubling exposure just to recover a past loss.
Review a loss without inventing its cause
Use observable categories: a forecast was wrong; rules were misunderstood; a quote changed; a partial fill left exposure; a cost was omitted; or the available record is incomplete. A losing trade does not prove panic, lack of discipline, insider advantage or a bad strategy by itself. A winning trade does not prove skill.
Keep the original thesis and expected failure cases, then reconcile the actual fills and outcome. Group similar cases only after defining the group. If only a few completed trades are available, say so rather than drawing a general conclusion. The probability worksheet separates forecast quality from trading results.
Read the specific contract and the provider's resolution rules. Check current fees and access before any activity; nothing here endorses bypassing restrictions. AITR may earn commissions from marked referral links elsewhere. This page makes no claim about an observed population loss rate.