You buy into a caper by sending XRD to its bonding curve from the swap panel on the caper’s own page. The curve mints you new tokens on the spot and sets the price itself, so there is no counterparty to find, no order to fill and no listing to wait for. The smallest buy the contract accepts is 0.001 XRD. What you get back is a token that carries three things a plain token does not: a claim on a treasury the group owns, weight in the votes that spend it, and a right to leave with your share.

What am I buying?

Tokens that did not exist until you bought them. A caper does not sell you inventory another holder is unloading – the curve mints against a fixed notional cap of 100,000,000,000 tokens, and your purchase moves circulation up that cap by however much your XRD pays for. The contract refuses any buy that would push circulation past the cap.

Holding those tokens is what makes you part of the caper rather than a spectator. The same balance is read three ways: as your stake in the treasury, as the multiplier on your voting weight, and as the amount you can redeem when you leave. There is no separate membership token to acquire and no application to be accepted.

How is the price set?

By a formula, not by a market. The price of the next token depends only on how much of the curve has already been sold. Writing s for the fraction in circulation – tokens sold divided by the 100bn cap – the marginal price is

P(s) = 0.001 × √s / (1 − s)

which starts near zero, rises as a square root through the early range, and steepens sharply as the curve approaches fully sold. Every buy advances s, so every buy raises the price for the buyer after you. That is the whole pricing mechanism: no order book, no market maker, no oracle, and nobody quoting a spread.

Because the price is a function of circulation, a large buy does not clear at one price. It walks up the curve as it fills, so the average you pay is worse than the price you saw before you pressed the button. That is slippage, and it is a property of the shape rather than a fee anyone charges. The swap panel does guard the size of it: the transaction it builds asserts a minimum output 2% below the quote it displayed, so a buy that would fill worse than that aborts rather than filling, and your wallet's Customize Guarantees can tighten the floor further. A headless signer building its own manifest gets neither and has to set the floor itself.

Where does my money actually go?

Your XRD is split three ways before any of it prices your tokens:

1. A trade fee off the gross. 0.5%, and it is not a lever anyone can pull. The rate is a field written once when the platform's logic component is instantiated; the deployed CaperMain exposes no setter for it, so changing it means deploying a fresh logic component and moving the registry's current_main to it through a governed UPGRADE proposal.
2. The founder’s collateralization skim. A slice of what is left, taken partly in XRD and partly as a share of the tokens minted. This is the founder’s reward for launching, and it is deliberately front-loaded.
3. The remainder into the reserve. Everything else lands in the curve vault, which is the pool that pays sellers and backs the token.

The skim is the part a buyer should understand before committing, because it is not flat. It is a straight-line taper: largest at s = 0, falling steadily, and exactly zero once 30% of the curve is in circulation. Past that point every buy is split between the trade fee and the reserve alone, and the founder takes nothing further from new buyers. The contract bounds the peak too, so the founder’s share of the mint can never exceed roughly 5% – that ceiling is peak × 1/3 × 30/31, and the remaining thirty-first of the token leg goes to the Commons treasury rather than the founder.

The practical reading: you pay the most to the founder for being early, and the earliest buyers pay it precisely because the reserve behind their tokens is thinnest.

Is my money safe? What backs the token?

Part of it is backed and part of it is not, and the split is a known number rather than a promise. At genesis a caper’s curve holds two thirds of its market capitalisation as XRD in the reserve vault. That is not a policy choice – it falls out of the curve’s shape. A curve priced as supply to the power n holds exactly 1/(n+1) of its capitalisation in reserve, and caper’s curve is a pure square root at the origin, so n = ½ and the fraction is 2/3.

The other third is a claim on buyers who have not arrived yet. If nobody else ever buys, that third is not there. This is the honest shape of any bonding curve and it is worth stating plainly rather than discovering later: the reserve can pay every holder the curve price at the current circulation, but it was never sized to pay every holder what the token is nominally worth.

What the reserve does guarantee is that the sell side always works. There is no scenario where you hold tokens and cannot find a buyer, because the curve is the buyer and the vault is already funded.

Can I sell, and how much do I get back?

Yes, at any time, and the amount is computed in closed form rather than negotiated. Selling walks the same curve backwards: the payout is the reserve required at your circulation before the sale minus the reserve required after it. There is one curve in both directions – no separate buy and sell price, no bid–ask spread – so the price you sell at is exactly the price the next buyer would have paid to reach that same point. A trade fee comes off the payout on the way out.

The consequence people underestimate: because you are selling back down a rising curve, selling into a caper that has grown since you bought returns more than you paid, and selling one that has shrunk returns less. Nothing about the mechanism protects you from that, and nothing about it depends on anyone else agreeing to trade with you.

Does buying give me a say?

Yes, and it gives you two of them. A caper decides in two phases, and buying puts you in both. In phase one you can cast a ranked ballot on a proposal – any holder of at least one token can, for a 100 XRD fee – and the ballots are folded into a Borda tally that passes an option only on a supermajority of 1.5 / option_count of the weight cast. In phase two, a proposal that won its ballot is triggered, which locks the trailing time-weighted average price as a baseline and opens the market window; it executes only if the TWAP over that window is at or above the baseline. Holding through the window is consent, buying is support, and selling the price under the baseline is the veto. So your say is your vote and your position both – and the position keeps talking after the vote is over.

Buying does not accrue the second, soulbound token. Only a cast ballot does, and it mints exactly one – VOTE_MINT = 1, fused into the vote call itself. Until the redeploy of 11 September 2026 buys and sells minted it too, at 0.01 per XRD of gross value, and that is the single change a returning reader most needs to know: one transaction could buy a position, mint the record that opens the exit, and redeem on both. The tokens cannot be transferred or sent to anyone – the depositor role is bound to the caper’s own component – and they can no longer be bought either, at any rate, by anyone. They are earned by turning up to a vote.

The formula those stake tokens feed is

w = (t × v) / (V × T)

where t is your governance-token balance, v your stake tokens, V the total stake supply, and T the tokens in circulation. It does double duty, and that is the point of it: it weights your ballot in phase one, and it prices your exit – what share of the treasury you take when you leave – using the identical helper. A large balance genuinely counts for more, and so does having voted rather than merely held, because both terms sit in the numerator and a zero in either produces zero.

What are my ways out?

Two, and they pay from different pots.

Sell back to the curve. The route above. You are paid out of the reserve vault at the curve price, and the tokens return to inventory for the next buyer. Fast, always available, and it pays you nothing from the treasury.

Exit. The right the whole model is built around: you surrender governance tokens and soulbound stake tokens together, and take a share of the treasury the caper has accumulated. Your share is (t × v) / (V × T), so the exit pays out on how much you hold and how many ballots you have cast – not on which way you voted. It also gates on stake: exit takes both amounts, asserts each is positive and withdraws them from your own account, so an account that has never voted has no exit to take rather than a zero-valued one. Plan for that before you need it. Until 11 September 2026 a buy opened the door by itself; since the redeploy removed the trade-side mint, the only way to acquire stake is to vote while a proposal is open, and there is no way to acquire it after the fact. One caveat worth knowing: the mint deposits with try_deposit_or_refund, so an account whose deposit rules reject unknown resources forfeits the accrual rather than aborting the ballot. Leaving a caper covers the mechanics in full.

The two are not alternatives to pick between at the moment you buy. They answer different questions: the curve answers “what is my token worth right now”, and the exit answers “what is my share of what we built”.

What buying does not give you

Stated plainly, because the gap between these and what a token sale usually implies is where people get hurt:

• No yield. Holding tokens pays nothing by itself. A caper’s treasury grows or shrinks by what its members pass proposals to do with it.
• No claim on the founder. The skim is theirs once taken, and no mechanism claws it back.
• No guaranteed floor. The reserve backs the curve price at the current circulation, not the price you paid.
• No veto you can cast for free. Blocking a proposal means selling the price under its baseline, which costs you the position. Objecting is meant to be expensive; that is the design, and it is the trade-off to weigh before you buy into a caper whose treasury you would not want spent.
• No say in whether the caper does anything. A treasury sits until somebody raises a proposal and the market lets it through; see Paying someone from a caper’s treasury.

Will the price go up?

Nobody can tell you, this page will not try, and you should treat anyone who does tell you as selling something. The curve fixes what a token costs at a given circulation; it fixes nothing about whether anyone else will buy. A caper’s price rises when people buy and falls when they sell, and there is no emission schedule, no yield and no mechanism anywhere in the contract that pushes it in either direction.

So the question “how do I invest in a caper” is worth reframing before you act on it. What you can actually assess in advance is not a price path but three things that are all readable on-chain right now: how much XRD sits in the reserve behind the token, how much sits in the treasury the exit would pay out of, and whether the caper has ever passed a proposal and spent any of it. A group that has never executed anything has not yet shown it can turn money into work – which is the only thing that would make its treasury worth more than the XRD already in it.

How do I buy into a caper?

1. Open the caper’s page from the roster on caper.network. Every caper has one.
2. Connect a Radix wallet holding XRD. You need XRD for the purchase itself and a little more for the network fee.
3. Enter an amount – at least 0.001 XRD – in the swap panel, which is set to XRD → the caper’s token by default. The panel shows the tokens you receive at the current circulation, and SWAP builds the transaction.
4. Sign. The tokens arrive in the same wallet and you are a member from that transaction onward – though not yet an exiting one: the soulbound stake the exit is gated on comes from casting a ballot, not from the buy.
5. Watch the proposals. There is nothing to opt into: from here your position is already counted every time a proposal resolves against the price.

Before you buy into any caper, read its treasury and its proposal history rather than its description. Both are on-chain and both are on the caper’s page – a caper that has never passed a proposal has never demonstrated it can spend what it has raised.

Where next

What is a caper covers the model itself, and Leaving a caper the exit right in full. For the pricing mathematics rather than the buyer’s view, see Bonding curve and Trading. If you are on the other side of this – launching rather than backing – start at Getting started. If you are reading this through an AI agent, Querying a caper from an AI agent documents the MCP servers that expose a caper’s live treasury and proposal history – free to read, with no key.