Your Dapp Is Being Held Back by Slow Oracles
Oracle latency undermines capital efficiency, security, and user experience in DeFi. Switchboard’s faster feeds offer a safer, smarter alternative.
Real-time data is the lifeblood of decentralized finance (DeFi). That’s where blockchain oracles like Switchboard come in. Yet, not all oracles are created equal, especially when it comes to speed. Slower oracles can introduce additional layers of risk to dapps and DeFi, leaving users exposed to avoidable risks.
When we say that some oracles are faster or slower than others, what we are really talking about is their latency. Latency essentially measures the time it takes data to travel from one point to another, or in this case, the time between when a dapp requests data and when it is delivered. The higher the latency, the slower the oracle. High latency can cause data staleness, making dapps use information that is outdated and might no longer be accurate.
What’s the Problem with Existing Oracles?
A data staleness of a few seconds might not seem significant, but in some cases, it creates inefficiencies which can lead to user and/or protocol losses, and even trigger wider systemic shocks. More specifically, slow oracles, or high latency, can not only cause issues with prediction markets and decentralized perpetual futures exchanges (perp DEXs), but can also lead to cascading liquidations which may affect broader markets beyond the dapp itself.
The main issue is data staleness creates opportunities for arbitrage. Oracle lag arbitrage opportunities have led to attacks commonly known as oracle latency exploits. In these exploits, attackers take advantage of the stale oracle price compared to a different or more up to date price from a different market, such as a CEX. Using the price difference, they make a deal (like a swap or a loan) to get a seemingly-false profit. Then, once the price catches up, they quickly reverse the trade, draining money from the dapp and hurting its liquidity providers (LPs). To combat this, instead of switching (no pun intended) to faster oracles, many projects end up sacrificing capital efficiency (whether inadvertently, apathetically, or even deliberately), with some protocols even opting to build cushions within their margins in preparation for such scenarios. This ultimately results in slower, inefficient dapps with more rigid parameters.
Oracle Latency Exploits in Perpetuals DEXs
Perp DEXs are arguably the type of dapp most susceptible to arbitrage opportunities created by stale data. In fact, ChainLink emphasizes that “any delay in data, even just a few seconds, can…expose” a derivatives dapp to front-running by arbitrageurs.
Derivatives markets, including perp DEXs, are an important category in DeFi, with over $350 billion in monthly volume and hundreds of thousands in daily revenue generated by the top dapps in the market. Jupiter leads the market with $1.7 billion in TVL and around $760 million in average daily volume during the month of April. Drift, a perp DEX on Solana secured by Switchboard oracles, rounds out the top 3 with over $800 million in perps TVL, over $67 billion in cumulative perps volume, and over $1 million average monthly revenue generated.
In late 2023, dYdX was exploited via an oracle manipulation attack on SUSHI and YFI resulting in ~$5M in gains before controls were raised. Most recently, KiloEx was drained of ~$7.5M after an attacker opened a position on stale data. These cases show that even well-known DEXs like dYdX can suffer multi-million‐dollar losses if oracle updates are slow. In other words, oracle lags create profit opportunities for sophisticated traders at the expense of the protocol’s LPs.
Modeling Exploitable Arbitrage Profit for Varying Oracle Latencies
Oracle latency becomes even more important when you become aware of the large disparity between different leading oracle solutions. More specifically, recent reports note Switchboard’s data arrives in 0.317 seconds on average, versus 1.2 seconds for Pyth and ~8.1 seconds for RedStone. In other words, Switchboard’s feeds are roughly four times faster than Pyth’s and 25 times faster than RedStone’s. Even modest price moves during those extra seconds could be lucrative to a front‐runner. To illustrate, let’s assume a token’s price is drifting upward at a constant rate. The exploitable arbitrage profit is roughly:
Profit = TVL x price-change rate x oracle delay
For example, at a steady 0.01% per-second price climb (about 8.6% per day), a 0.317 second delay yields only ~0.00317% price lag, whereas a 8.137 second delay can result in nearly a 0.1% price lag. On a $1.5B TVL DEX like Jupiter, these price lags can be the difference between $47.6K or $1.22M of exploitable profit respectively, depending on the oracle used (see table). In short, longer latencies translate linearly into larger arbitrage windows. In these cases, a faster oracle like Switchboard can save dapps, their LPs, and their traders millions during significant price-change scenarios. More specifically, during a 0.01% per second price change scenario on Jupiter, Switchboard’s low latency would provide Jupiter and its users with $1.17 million or $132,000 in savings compared to slower oracles like Redstone and Pyth, respectively.
While the table above is an oversimplified example which does not sort for specific markets, it helps highlight not only the opportunities that exist for arbitrageurs to exploit dapps with lagging oracles, but also the way developers, users and LPs can protect themselves with faster oracles.
What Does This Look Like in Practice?
Ultimately, the model is designed to illustrate how latency might affect real-world outcomes on existing dapps. However, in practice, dapps like Jupiter and Drift have already implemented safeguards like sanity checks, multi-oracle validation logic, and narrower confidence bandwidths to minimize reliance on stale prices and limit manipulation. For example, Drift uses sub-second price feeds with integrated confidence bands and time-weighted average pricing (TWAP) smoothing. Jupiter executes perpetual trades using a composite price taken from various oracle sources, resorting to the median in the case of a deviation.
Since many dapps choose multi-oracle architecture, it’s helpful for developers to know how different oracle providers stack up against each other in order to tailor a solution to their dapp’s needs. Expanding the metaphor of oracles as critical pieces of DeFi infrastructure, an oracle can be seen as a specific recipe for concrete, with a dapp’s multi-oracle architecture becoming the unique mix of concrete used to construct a particular building. In this scenario, it’s important to consider the features of each oracle solution in order to determine the ideal blend for a given project. Relevantly, one of Switchboard’s advantages is its Oracle Aggregator, a unique feature which allows developers to integrate different oracle providers into a single transaction.
Unleash the Power of Low Latency
In DeFi, speed isn’t just a talking point, it’s a requirement. As this example shows, even fractions of a second can create arbitrage windows that sophisticated actors can exploit at the expense of honest users and LPs.
That’s why fast oracles aren’t just better … they’re safer. Switchboard’s sub-second update speeds greatly limit exposure to front-running and latency-based exploits, cutting the impact of oracle arbitrage attacks by up to 95% compared to slower solutions.
Because it’s not just protocols that suffer. For everyday DeFi users, slow oracles can mean unfair liquidations, unexpected slippage, and getting front-run by faster, more resourceful players. And in the event of an exploit, LPs (often retail) take the biggest hit. That’s why leading perp DEXs like Jupiter and Drift invest heavily in risk mitigation to maximize oracle accuracy and freshness, and protect their traders and LPs.
Speed also unlocks capital efficiency and better risk modelling. Lower latency lets dapps run tighter risk parameters, such as higher loan-to-value ratios and quicker execution on liquidations and settlements, allowing users to do more with less capital.
When precision matters, low latency does too. Build with faster oracles like Switchboard.
Ready to Learn More?
Speed sets Switchboard apart from other, slower oracle solutions. But it’s not the only differentiator. Read our last article to see how Switchboard’s permission-less oracles tear down barriers to data:
To learn even more about Switchboard:
- Visit the Switchboard Website
- Explore the Developer Documentation
- Join the Discord
- Follow on X
- Read the Whitepaper
