What You Can Trade

This is the front door to the Proof exchange. It explains, in plain terms, the three kinds of instrument you can trade and how they fit together. It is conceptual — no code, no endpoints, no integration detail. Read it to get the mental model; the later docs go deep on market structure, settlement, margin, and pricing.

Proof has two layers. The first is ordinary perpetual futures on assets — the same instrument you already know from any derivatives venue. The second is what makes Proof different: conditional markets, which let you trade an asset conditioned on whether a real-world event happens, alongside a market on the event's probability itself. Everything here builds toward that second layer.

There are three instrument types in total:

1. Perpetuals — a future on an asset, with no expiry.
2. Prediction binaries — a $0–$1 instrument that pays $1 if an event happens and $0 if it doesn't.
3. Conditional perpetuals — a perpetual on an asset that only counts if an event resolves a certain way, and otherwise voids.

Each trades on its own order book. The last two come into existence together, as part of a conditional market.

Perpetuals in plain terms

A perpetual (or perp) is a future on an asset with no expiry date. You can go long (betting the price rises) or short (betting it falls), and you hold the position as long as you like. Throughout these docs we write a perpetual as asset-PERP, tracking some underlying asset.

Three things define how a perp behaves on Proof:

• Mark price. Your position is valued against a mark price derived from an external oracle feed — a trusted price source for the asset. The mark drives your unrealized profit and loss, your margin, and whether you get liquidated. It is deliberately tied to the oracle rather than to the perp's own order book, so it is hard to push around with a few trades.
• Leverage. A perp carries leverage. You post initial margin (IM) to open a position — a fraction of its notional value — and must keep at least the maintenance margin (MM) to hold it. If your collateral falls below MM, the position is liquidated.
• Funding. Because a perp never expires, a periodic funding payment between longs and shorts keeps its price tethered to the underlying. When the perp trades above the oracle, longs pay shorts; when it trades below, shorts pay longs. Funding is zero-sum and settles to cash. It applies to perpetuals only — never to any of the conditional instruments below.

We call the perp's fair price the forward, written F. For a conditional market, this one value anchors the whole structure.

If you only ever trade perps, that is a complete and standard product. The rest of this doc is about what you can build on top of them.

Prediction binaries (pay $1 or $0)

A prediction binary is the simplest instrument on Proof. It is a token priced in dollars between $0 and $1, tied to a yes/no event — for example, an event such as a scheduled economic decision, or a price-threshold question of the form "does the asset close above a level K?".

Its defining property: its price is the market-implied probability of its outcome. If the YES binary trades at $0.30, the market is implying a 30% chance the event resolves YES. At resolution it pays out simply:

• $1 if its outcome wins, $0 if its outcome loses.

So if you buy the YES binary at $0.30 (an illustrative price), you pay $0.30 now and receive $1.00 if YES wins, or $0.00 if it loses. (Read the price as the market-implied probability, not an objective one.)

Every event has a matching pair: a prediction binary (YES), whose price we write BY, and a prediction binary (NO), written BN. Because exactly one outcome wins and pays $1, the two prices should sum to about a dollar — BY + BN ≈ $1. This is the binary box: a fair YES probability and a fair NO probability add to one. The one exception is a Void (an event that does not produce a valid YES/NO outcome — whether because it is declared unresolvable, or simply because no valid resolution is produced in time): then both binaries pay $0, so the box sums to zero rather than a dollar. A Void is never derived from comparing a price to a strike; it is the absence of a valid YES/NO resolution. In other words, BY + BN ≈ $1 holds through a YES or NO outcome, but breaks to $0 through a Void.

A few practical traits:

• No funding. Binaries never carry funding.
• No separate position margin. A binary's worst case is fully known the moment you trade it — a buyer can lose at most what it paid, a seller at most a dollar minus what it received. Because that loss is already reflected in your account equity, the engine reserves no extra margin to hold a binary. (A resting binary order does reserve its potential loss until it fills or cancels.)
• A fine price grid. Binaries trade on a fixed, fine grid of small price and size increments, so the $0–$1 probability space stays granular regardless of how the underlying asset is priced.

A prediction binary by itself lets you take a clean view on whether an event happens. What it cannot tell you is how much the event moves the asset — every winning binary pays the same dollar no matter how dramatic the move. That magnitude is what the next instrument captures.

Conditional perpetuals (a perp that only counts if an event resolves a certain way)

A conditional perpetual is a perpetual on an asset that only fires if an event resolves its way. While the event is open, it trades like a normal perp — its own order book, its own conditional price, its own leverage. What makes it conditional is what happens when the event resolves:

• If its branch wins, it cash-settles to the underlying perpetual's mark — exactly as if you had held the underlying asset.
• If its branch loses (the other outcome wins, or the event turns out to be unresolvable), the position is voided: your profit-and-loss is zero, and your margin is returned (net of any funding and fees accrued while the leg was open).

That voiding behavior is the single most important property of conditional markets. A losing conditional perpetual is not a loss. It does not settle to $0. It simply disappears: nothing changes hands, and the collateral you posted is released. If you are long the YES conditional perpetual and the event resolves NO, you do not lose the position's value — you get your margin back and walk away flat on that leg.

Each event produces a matched pair: a conditional perpetual (YES), whose fair price we write CY ("the asset if the event is YES"), and a conditional perpetual (NO), written CN ("the asset if the event is NO").

Two consequences are worth stating up front:

• Conditional perpetuals carry leverage. Like the underlying perp, each has its own initial- and maintenance-margin requirements (both positive, with MM ≤ IM), and is margined as a perp in the branch where it fires.
• A conditional perp hedged with the underlying is only a hedge in the firing branch. If you are long a YES conditional perpetual and short the underlying perp to neutralize price risk, that holds while the event is open and if YES fires — but a NO resolution voids the conditional leg and leaves your underlying short naked. The rule that keeps you safe is to be solvent in the post-resolution world before the event resolves, by flattening or fully boxing across both branches; see Settlement, Resolution & Void for the full mechanic.

A conditional perpetual is the instrument that lets you express how much an asset moves because of a specific event, not just whether the event occurs — with no expiry to roll and no premium bleeding away while you wait.

How the pieces relate

Put one underlying asset and one binary event together and you get a conditional market: a linked set of five order books, created together, that we call a family.

• One underlying perpetual — the asset itself, priced at the forward F. It anchors the family (and keeps trading as a normal perp).
• Two conditional perpetuals — CY and CN, the asset's price in each of the two possible worlds (event YES, event NO).
• Two prediction binaries — BY and BN, the probability of each of those two worlds.

The intuition is clean. The underlying tells you what the asset is worth unconditionally, today. The two conditional perpetuals tell you what it would be worth in each possible world. The two prediction binaries tell you how likely each world is. Together they take a single asset's value and split it into a probability and a pair of outcome-conditional prices — exactly the structure you need to trade "what happens to this asset if that event goes one way."

These five books are linked, but not because the exchange forces their prices to agree — it matches each book independently. They are linked because their fair values relate to one another, and arbitrage keeps them roughly in line. Without writing any formulas: the underlying's price should sit between the two conditional prices, pulled toward whichever world is more likely; and the two binary prices should add up to about a dollar because one outcome always wins. The pricing doc makes these relationships precise. For now, the takeaway is that the family is internally consistent: the probability and the two conditional prices, taken together, recover the underlying — and where they don't, there is an inconsistency to trade.

Two more facts complete the mental model:

• Funding applies to perpetuals only. The underlying perp has funding; the entire conditional family — both conditional perpetuals and both prediction binaries — never accrues funding.
• Hedged positions are cheap to carry. Proof runs one shared collateral pool per account (cross-margin), and the margin engine checks that you stay solvent in the worst possible resolution outcome. A position that is fully hedged across both branches — a box — pays the same no matter how the event resolves, so it costs close to zero net margin to hold. This capital efficiency for hedged positions is a core reason conditional markets exist. The margin doc covers it.

A worked example of one conditional-market family

Take a generic asset trading at a forward of F = $100, and a scheduled binary event — say, an economic decision — that the family conditions on. (All numbers here are illustrative.)

Suppose the market believes:

• The event has about a 40% chance of resolving YES. So the YES prediction binary trades near BY = $0.40 and the NO binary near BN = $0.60 — together about a dollar.
• The asset is worth about $108 if YES happens and about $94.67 if NO happens. So the YES conditional perpetual trades near CY = $108 and the NO conditional perpetual near CN = $94.67.

Notice these hang together. Blend the two conditional prices by their probabilities:

0.40 × $108 + 0.60 × $94.67 ≈ $100

— which recovers the underlying forward F. That consistency is not enforced by the exchange; it is what arbitrage drives the books toward. The gap between the two conditional prices ($108 − $94.67 ≈ $13.33) is the impact of the event — how much the asset's expected price differs between the two worlds.

Now walk through what a few positions do.

You think the event is YES and the asset rallies on it. You buy the YES conditional perpetual at $108.

• If the event resolves YES and the asset is, say, $112 at that moment, your YES conditional perp settles to the underlying mark: you realize roughly $112 − $108 = $4 per unit, like closing a perp.
• If the event resolves NO, your YES conditional perpetual voids: profit-and-loss zero, margin returned (net of any funding and fees accrued while the leg was open). You are not down the $108 — you simply walk away flat on that leg.
• If the event is declared unresolvable (Void), same thing: the position voids, margin returned (net of any funding and fees accrued while the leg was open). (A Void also settles both prediction binaries to $0 — so on a Void the binary box sums to zero rather than a dollar.)

Contrast that with the YES prediction binary at $0.40. If you bought it instead, a YES outcome pays you $1.00 (a $0.60 gain) regardless of whether the asset moved to $101 or $130 — the binary pays for whether, not how much. The conditional perpetual pays for the magnitude of the asset's reaction; the binary pays for the probability of the event. That separation is the point of the family.

You want exposure to the asset only if YES, with no downside if NO. Buying the YES conditional perpetual gives you exactly that: full asset exposure in the YES world, and a clean void (margin back) in the NO world. There is no premium bleeding away day by day while you wait for the event, and nothing to roll.

One operating note from this example. A conditional perpetual hedged only with the underlying perp leaves that hedge naked in the branch where the conditional leg voids — so be solvent in the post-resolution world before the event resolves (flatten the conditional leg, or box it across both branches). It is the one thing most worth internalizing before trading a conditional market; see Settlement, Resolution & Void for the full mechanic.

Where to go next

• Market Structure & Instruments — the order book model, the five-book family, and the no-arbitrage relationships, in full.
• Settlement, Resolution & Void — exactly what each outcome pays each leg, how to exit before resolution, and the flatten-before-resolution rule.
• Margin — cross-margin, the worst-case scenario check, and why a hedged box is nearly free to carry.
• Pricing — how the forward, the probability, and the impact relate, and what drives the conditional spread.