⏮️Backtest report
Backtesting that speaks for itself
Last updated
Backtesting that speaks for itself
Last updated
Backtesting allows us to evaluate the past performance of a given strategy, taking into account predetermined parameters and assumptions. Thanks to backtesting, we can get a very detailed idea of how much the strategy would have earned if we had used it in the past. Although past results are no guarantee of future performance, enabling us to fine-tune the strategy as much as possible. Backtesting is therefore a key stage in assigning a sufficient level of confidence that the strategy will work in the future.
For optimal backtesting, we need all the historical market data and a high-performance backtesting software. Lobster has one of the most sophisticated backtesting software on the market in terms of delta-neutral strategy overlaying concentrated liquidity pools. Our quant and engineering experts worked together to create what would become our greatest strength in the ecosystem of active tailor-made strategy creation in decentralized finance. Lobster's job is to optimize all the parameters of the strategy in order to maximize its profitability and minimize risk through stress testing.
For our first strategy, for example, we carried out backtests on around ten Uniswap v3 pools, on networks such as Polygon, Arbitrum, and Optimism, in different market configurations. These tests generated hundreds of millions of data points, allowing us to improve the mathematical models that guide our algorithm while training it for various scenarios.
Here are the detailed results of our backtesting analysis after parameters optimisation.
This specific strategy aims to outperform Bitcoin, with the BTC/ETH pool on the Arbitrum blockchain and Uniswap v3 protocol as the underlying.
In green, the APY corresponds to the strategy's outperformance against Bitcoin. In purple, the almost negligible slippage costs And in gray, the blockchain costs ($100/year), which are not even visible because the backtesting was carried out with more than $1 million under management.
On this graph, we can see the backtested profitability over 17 months. This corresponds to an annual return of 15.1%, with a drawdown of just 4.2%. if you are not familiar with finance, it is quite low compared to the existing solutions.
In blue,the pool's quote (BTC price divided by ETH price) In yellow, the outperformance of our strategy In grey, the concentration range chosen on Uniswap v3 In this representation, we can see the algorithm deciding to change range approximately every 48 hours. This corresponds to 78 trades per month. The average concentration of these backtest decisions on Uniswap v3 is around x9.
In purple, the strategy allocation on the Uniswap protocol In blue strategy allocation on the AAVE protocol Here we see the algorithm's allocation decisions on the underlying protocols. We can see that throughout the arbitrages, a small part of the allocation is dedicated to hedging on AAVE, and a large part is used to generate returns on Uniswap.
In blue the resulting current exposure to Bitcoin In red, the liquidation threshold on AAVE In orange, the last borrow ratio applied during an arbitrage In gray, the current borrow ratio on AAVE
On this last chart, we first look at the strategy's current borrow ratio over time. When it touches the red curve, a liquidation takes place and causes the overall strategy to lose around 1% of total yield, which is not much and, above all, never happens over the 1.5 years of backtesting. We also see a blue curve representing the current hedging ratio of all positions in the strategy.
