The Engines · Ch. 13
Oracle-Priced Perps
You buy ten ETH, or ten thousand, and you fill at the exact same price. Nothing you do moves the number. The market stopped asking what you are willing to pay and just told you the price.
The Idea
Intuition
The last two engines had one thing in common: they discovered the price by trading. The orderbook set it wherever the resting orders sat, the vAMM set it wherever the curve sat, and in both, your own trade pushed it. That push was the slippage.
An oracle perp drops that idea entirely. It does not discover the price. It reads it, from an outside feed, the same index price from Chapter 7. Chainlink, an exchange median, a pull-based oracle, whatever the source, the protocol treats that number as the price and fills you there.
So your trade moves the price by nothing. Buy a little or buy a lot, you fill at the mark, with zero slippage from your own order. There is no book to walk and no curve to climb. In effect there is infinite depth at one price: the oracle’s.
The bill comes due on trust. Your fill is only ever as honest as the feed. A stale oracle prices your trade on the past. A manipulated one, like the low-liquidity market from Chapter 7, poisons every fill at once. And because the price never moves to discourage a crowd, someone still has to take the other side when everyone leans the same way. That someone is the next chapter. Send an order and watch the fill stay pinned to the mark.
The Math
How It’s Calculated
In plain terms: the fill price is just the oracle price, full stop. Size does not enter the formula, so the slippage term that defined the last two chapters is simply zero.
Whatever you trade, your fill is the oracle price :
There is no size term, so the price impact that defined the last two engines is gone by construction. What replaces it as the real cost is not in this equation: it is the risk that is wrong, stale, or gamed, and the funding you pay when the crowd piles onto one side of a price that will not push back.
The bigger picture
That an oracle reads the price instead of discovering it is most powerful with real-world assets. A stock, an ounce of gold, a currency pair: these trade off-chain, have no native onchain market makers, and would strand a vAMM’s funding exactly the way the last chapter warned. There is no crowd to bootstrap and no book to fill. But if a trustworthy feed already publishes their price every second, an oracle perp can quote a market on any of them from day one, with no makers and no deep reserves to stand up first. Discovering the price onchain was the hard part; reading it from where the asset already trades quietly removes it. That one move is what lets perps reach past crypto into the far larger world of real-world assets.
Note
Reading the price instead of discovering it moves the hard problem, it does not delete it. Oracle latency invites front-running: a trader who sees a price move before the feed updates can trade against the stale mark. And with no price impact to absorb a lopsided crowd, the imbalance lands somewhere real. It lands on whoever is quoting the other side: a funded pool. Who that is, what they earn, and the risk they carry is the next chapter.