HostNexFlow › Library › Prediction markets vs perps
Prediction markets vs perps: same conviction, very different payoff
Both let you act on a view. A prediction market pays you a fixed amount if a specific thing happens by a specific time. A perpetual pays you continuously as price moves — and takes from you continuously when it does not. Most people who pick badly are not wrong about the world; they picked an instrument whose payoff shape did not match the sentence in their head.
Start from the sentence, not the product
Write down your view as one sentence. If it has the shape "X will happen by date Y" — a token lists, a proposal passes, a number prints above a threshold — that is a binary claim, and a prediction market is the instrument built for it. You buy YES or NO shares; each share resolves to a fixed value if you are right and nothing if you are wrong. The price of the share is the market's implied probability, and your maximum loss is what you paid.
If the sentence has the shape "price is going up from here" or "this is overextended", with no date and no threshold, that is a directional claim about a continuous variable, and a perpetual is built for it. You open a long or short with leverage, your gain or loss tracks price tick by tick, and there is no expiry — the position lives until you close it or the exchange closes it for you at your liquidation price.
The mistake is forcing one shape into the other. Trading a binary event with a perp means carrying funding and liquidation risk through every wobble on the way to an outcome that only cares about the final state. Trading a directional view with a prediction market means picking an arbitrary threshold and date that your view never actually contained.
Side by side, with sources
Each row is read from the NexFlow feature page named in the source column on the date given. Fees quoted are the platform's own stated figures; they are not measurements of what any given trade will cost after slippage and network fees.
| Dimension | Prediction market | Perpetual | As of · source |
|---|---|---|---|
| Payoff shape | Binary — share resolves to fixed value or zero | Linear in price, multiplied by leverage | as of 2026-09-27 · /features/predictions/ and /features/perps/ |
| Maximum loss | Premium paid for the shares | Margin posted, reached at the liquidation price | as of 2026-09-27 · /features/perps/ |
| Expiry | Resolves at the event / resolution date | None — open until closed or liquidated | as of 2026-09-27 · /features/perps/ |
| Carry cost while open | None beyond the entry | Funding rate, paid or received periodically | as of 2026-09-27 · /features/perps/ |
| Platform fee (stated) | Flat 1% on entry | Flat 1% platform fee | as of 2026-09-27 · /features/predictions/ and /features/perps/ |
| Settlement asset | SOL-native front end; positions on-chain | On-chain positions | as of 2026-09-27 · /features/predictions/ and /features/perps/ |
| Key document to read first | The market's resolution rules | Funding schedule and liquidation price | as of 2026-09-27 · /features/predictions/ and /features/perps/ |
| Jurisdiction restrictions | not published | US, UK, Canada and sanctioned jurisdictions excluded | as of 2026-09-27 · /features/perps/ |
The "not published" cell is deliberate. The predictions feature page did not state a jurisdiction list on the date we read it; check the app before you assume it is unrestricted.
Where each one hides its risk
A prediction market's risk is almost entirely in the fine print. The share pays out based on how the market resolves, and resolution is a rule written by a human, not the thing you think you are betting on. "Token lists on a major exchange by October" has to define "lists", "major" and which timezone October ends in. Markets with sloppy rules resolve in ways that surprise one side, and the surprised side is usually the one that did not read the rules. The second hidden cost is liquidity: a thin market lets you buy YES at 40 and then shows you a bid at 25 when you want out early. Your maximum loss is bounded, but your ability to exit before resolution is not guaranteed.
A perpetual's risk is in the path, not the endpoint. You can be right about the direction over a week and still be liquidated on Tuesday because leverage turned a normal wick into a margin call. Funding is the second slow leak: when the crowd is on your side of the book you pay them to hold your position, every period, for as long as the position is open. A perp that is "right eventually" can lose money the whole way there. The perps feature page tells you to know your liquidation price before you open; treat that as the minimum, not the whole checklist.
Rule of thumb: a prediction market can only hurt you as much as you put in, but it can hurt you for reasons that were in the rules. A perp can hurt you for reasons that were in the chart — and it can do it before you were wrong.
Fees look the same and are not
Both feature pages state a flat 1% platform fee as of 2026-09-27. Read that as "the same sticker", not "the same cost". A prediction position is entered once and settles once; the platform fee is a one-off against a payoff that is fixed in advance. A perp is a position you might scale, partially close, re-enter and hold across many funding periods, and the stated fee applies to the trades, with funding on top. Over a long hold the funding can matter more than the fee, and it is a number that changes with market positioning rather than one anyone publishes as a constant. We are not going to give you an expected funding figure, because there is no honest one.
On both, the on-chain leg has its own costs: network fees to open and close, and on the prediction side the spread in a thin order book. None of these are NexFlow's fee and none of them are fixed.
Who should pick prediction markets
Pick predictions if your view has a date and a yes-or-no answer, and you want your worst case written down before you click. It suits people who are good at reading rules and bad at watching charts; people who want exposure to an event without being shaken out by the noise between now and then; and people who want to be able to say "I risked this much, no more" and have it be literally true. It also suits small, frequent, uncorrelated bets: many independent binary views are a very different portfolio from one leveraged directional position.
Before you buy, read the resolution rules twice and look at the depth on the other side of the book. If you cannot say who resolves the market and on what evidence, do not buy the share.
Who should pick perps
Pick perps if your view is about direction and magnitude with no natural deadline, and you are willing to manage a position rather than place a bet. It suits people who already size trades in terms of liquidation distance, who understand that leverage is a choice about how much path they can tolerate, and who check funding before they open rather than after they notice the drift. Short exposure is the other real reason: a prediction market on "price will be lower" is an awkward construct, while a short perp is the direct instrument.
Do not pick perps because the leverage number is large. Pick the smallest leverage that expresses the view and put the liquidation price somewhere you genuinely believe price will not visit. If you are in a restricted jurisdiction per the feature page, the answer is that you do not pick perps at all.
Using both without doubling the same bet
The sensible pairing is a prediction share for the event and a perp for the drift around it, with the two sized so that being wrong on both is survivable. The trap is stacking them on the same view: long the perp and long YES on "price above X by date" is one bet wearing two hats, and it will lose twice on the same news. If you find yourself opening both on the same afternoon about the same thing, stop and reread the sentence you wrote at the top. One instrument matches it. Use that one, and let the other go.
Frequently asked
What is the core difference between a prediction market and a perp?
A prediction market pays a fixed amount if a defined event resolves in your favour by a set date — your maximum loss is what you paid for the shares. A perpetual tracks price continuously with leverage, has no expiry, charges or pays funding while open, and can liquidate you before your view plays out.
What does NexFlow charge on each?
As of 2026-09-27 the predictions feature page states a flat 1% entry fee and the perps feature page states a flat 1% platform fee. Network fees, spread, and on perps the funding rate, are separate and vary; we do not publish an expected figure for them.
Can I lose more than I put in?
On a prediction share, no — the premium is the maximum loss. On a perp your loss is bounded by the margin you posted, reached at the liquidation price, which can happen on a temporary move even if the direction later proves right.
Which one should I use to short something?
A perp. Shorting is its native use: open a short with the leverage you choose. Expressing a bearish view through a prediction market requires picking an arbitrary threshold and date that your view may not actually contain.
Are there jurisdiction restrictions?
The perps feature page lists the US, UK, Canada and sanctioned jurisdictions as excluded, as of 2026-09-27. The predictions feature page did not state a jurisdiction list on the date we read it, so check the app rather than assuming.
What should I read before entering either?
For a prediction market: the resolution rules and the depth on the opposite side of the book. For a perp: your liquidation price and the current funding rate. Both feature pages tell you to do this; it is the minimum, not the whole checklist.