A synthetic asset is a token or position that tracks the price of something else, such as a currency, a commodity, a stock index or another crypto asset, without holding the underlying. Building one means solving three problems well: how the exposure is collateralized, how the price is known on-chain, and what happens when collateral is no longer enough.
The main designs
Overcollateralized debt positions
A user locks collateral worth more than the synthetic they mint, say 150% or more, and receives a token pegged to a reference price. If the collateral value falls below a threshold, the position is liquidated. This is how MakerDAO's DAI started, and it generalizes to any price feed. It is simple to reason about but capital-inefficient, and the peg depends on arbitrage being profitable.
Shared debt pool
Synthetix popularized a design where stakers of the protocol token collectively back every synthetic in circulation. Each staker owns a share of the total system debt, so if traders profit on one synthetic, every staker's debt rises. Traders get zero-slippage exchanges between synthetics at oracle prices. The cost is that stakers carry an unhedged, hard-to-understand exposure, and the system is extremely sensitive to oracle latency: if a price can be front-run, traders extract value from stakers. Synthetix itself shifted its focus toward perpetual futures over time, and its sUSD stablecoin has had periods trading below its peg, which illustrates how fragile debt-pool pegs can be.
Perpetual futures as synthetics
Many teams now deliver synthetic exposure as perpetual swaps rather than tokens. A funding rate pays longs or shorts to keep the contract near an index price. This avoids minting a token that must hold a peg, but brings margin engines, liquidations and much higher regulatory scrutiny.
Delta-neutral synthetic dollars
A newer pattern backs a dollar-like token with a spot crypto position hedged by a short perpetual on exchanges. The token earns funding and staking yield. It depends on counterparties and custodians off-chain, and on funding rates staying favorable; when they turn negative for long periods, the backing erodes.
Optimistic and dispute-based synthetics
UMA's approach uses an optimistic oracle: a price is proposed, and anyone can dispute it within a window, with disputes escalated to token-holder voting. This works for assets without a reliable push oracle, at the cost of settlement delay.
Comparing the approaches
| Design | Capital efficiency | Who carries the risk | Critical dependency |
|---|---|---|---|
| Overcollateralized vaults | Low | Each minter | Liquidation incentives, oracle |
| Shared debt pool | Medium | All stakers collectively | Low-latency oracle, staker incentives |
| Perpetuals | High (leverage) | Counterparties, insurance fund | Liquidation engine, funding design |
| Delta-neutral dollar | High | Token holders | Exchanges, custodians, funding rates |
| Optimistic oracle | Varies | Minters and disputers | Active disputers, voter honesty |
Oracles decide whether it survives
Every synthetic is only as good as its price feed. Use decentralized oracle networks for liquid assets, combine sources, and check staleness and deviation on every read. For assets that trade only during market hours, such as equities, decide what happens on weekends and holidays: freeze minting, widen collateral requirements, or allow trading with a stale price and accept the risk. Mirror Protocol on Terra offered synthetic US stocks and collapsed with the Terra ecosystem in 2022, a reminder that a synthetic inherits the failure modes of its collateral and its host chain.
Building one, step by step
- Choose the exposure and the user. Hedgers, traders and yield seekers want different products.
- Pick the collateral model and stress-test it against a 40–60% collateral drawdown in a day, which crypto markets have produced.
- Design oracles and circuit breakers: maximum price change per update, staleness limits, pause rules.
- Build liquidation: who can liquidate, the bonus, partial vs full liquidation, and how bad debt is absorbed.
- Write the contracts in Solidity with invariant tests, then audit; see the DeFi smart contract guide for process.
- Launch with mint caps and raise them as liquidity and monitoring prove out.
The collateral and liquidation mechanics overlap heavily with decentralized stablecoin design and with lending platforms, so it is worth studying both.
Legal reality
Synthetic exposure to stocks or commodities is generally treated as a security-based swap, a derivative or a security in major jurisdictions, including the US and the EU. Offering it to retail users without authorization is a serious risk, and geofencing a front end is not a complete answer. Synthetic crypto-to-crypto exposure is less clear-cut but still draws scrutiny when leverage is involved.
Frequently asked questions
Is a synthetic asset the same as a tokenized asset?
No. A tokenized asset represents a claim on a real underlying held by a custodian. A synthetic only tracks the price and is backed by collateral of a different kind.
Why do synthetics depeg?
Usually because collateral falls faster than liquidations can keep up, oracles lag, or demand to exit exceeds the arbitrage capacity that restores the peg.
Can synthetics track real-world stocks legally?
Only within a regulated framework in most jurisdictions. Treat it as a regulated derivative product from the start.