| TWAP! TWAP! TWAP! Switchboard V2 now supports aggregator history
What is TWAP?
A time weighted average price (TWAP) is the average price of an asset over a specified time frame. Let’s say you’re a large trader and you want to enter a large SOL/USD position. If you execute a sudden market order, you could drain the order book and end up paying substantially more for the trade, plus bots will immediately start to arbitrage and your position will be underwater immediately. If you execute the same position using a TWAP, your orders will be spread throughout the given timeframe, allowing the market to absorb your liquidity, and giving you a more optimal price of the true market value.
TWAP use cases extend beyond just trading. Here are just a few examples of how a TWAP could be used in DeFi to reduce oracle price manipulations:
Funding Premiums: Using a TWAP over a random sample to avoid a single sample skewing the results
Futures Expiry: Futures expiration are correlated with higher volatility. Using a TWAP reduces the risk of manipulation or banging the close.
Lending Liquidations: A TWAP reduces scam wicks causing erroneous lending position liquidations.
On-Chain Trading Algos: An on-chain program could use a Switchboard TWAP oracle to execute positions on Serum based on a given criteria.
How to Use TwapTask?
To start using the new TwapTask with Switchboard V2, you will need to initialize a history buffer account for your aggregator. You can initialize a history buffer using the javascript API method setHistoryBuffer. A history buffer account stores a set number of accepted results for an aggregator, and given Solana’s maximum account size of 10MB, the maximum number of samples a single history buffer can support is 350,000 samples (Keep in mind rent cost, this action is currently irreversible). The screenshot below shows the SOL/USD feed’s price history in action on our explorer page.
Once a history buffer has been initialized for an aggregator, you can create a separate aggregator to track its TWAP. You can have multiple TwapTask tied to a single aggregator/history buffer. The TwapTask accepts the following values:
aggregatorPubkey: The public key of the aggregator that has a history buffer initialized
period: The time period, in seconds, to include in the TWAP calculation
weightByPropagationTime (Optional): Return the average, weighing each sample by the sample’s propagation time (currentTimestamp — previousTimestamp).
minSamples (Optional): The minimum number of samples that must be present in the history buffer, after filtering by time period, before accepting a result
The TwapTask first filters the history buffer and excludes any samples failing outside the set time period. If the minSamples field is set and the number of samples in the filtered list is less than the minSamples, the oracles will not report a result. The oracle will then calculate the final value by taking the weighted average between each sample’s interval. You can find an example of the TwapTask in our Job Directory.
Known Limitations
The TwapTask relies on accepted on-chain results being pushed to the history buffer. During periods of network degradations, aggregators may fail to post accepted results due to an insufficient number of oracle responses. This may cause TWAP calculations to have insufficient data to make a proper calculation. To mitigate this, you should specify a minSamples field in the job definition to only accept a TWAP result if enough samples are provided.
Check out our docs for a full list of tasks Switchboard supports: docs.switchboard.xyz/api/tasks
Or view our currently supported on-chain feeds: switchboard.xyz/explorer.