Strategy backtests
Last updated
Last updated
We conducted a series of backtests to evaluate USDC vault performance under varying liquidity levels of $1 million, $3 million, and $10 million. Drawing on historical data from the Arbitrum network spanning August 5, 2024 (block #239724778) to October 15, 2024 (block #264048675), we tested three distinct strategies:
Simple Strategy – Allocates all liquidity to whichever protocol currently offers the highest APR.
PSO-optimized Strategy – Distributes liquidity across multiple protocols to maximize overall APR.
Benchmark – Allocates all liquidity to AAVE for comparative purposes.
🔵 The blue line — AAVE-only
🟠 The orange line — Simple strategy
🟢 The green line — PSO-optimized strategy
In all simulations we have observed that the optimized strategy sustainably outperforms all other strategies, only varying in how much it outperforms them depending on the vault liquidity.
Average APRs:
Aave-only: 3.31%
Simple strategy: 5.21%
Optimized strategy: 5.40%
Improvement vs Simple strategy: +3%
Average APRs:
Aave-only: 3.24%
Simple strategy: 3.69%
Optimized strategy: 4.32%
Improvement vs Simple strategy: +17%
Average APRs:
Aave-only: 3.03%
Simple strategy: 1.53%
Optimized strategy: 3.55%
Improvement vs Simple strategy: +132%
Below is an overview of how large liquidity injections affect lending/borrowing markets and why distributing liquidity across multiple protocols helps maintain higher APYs and speeds up yield recovery.
When a substantial deposit enters a single lending market, it temporarily pushes down borrowing rates, because additional capital is immediately available for borrowers. As a result, the APY drops for a short period, usually 12–48 hours, until the market rebalances and borrowing demand catches up to the new liquidity level.
Our simulations (based on November 2023 historical data and original APY formulas from integrated protocols) showed that allocating liquidity across several protocols mitigates this dip:
Less Impact on Individual Protocols
Spreading a deposit means no single protocol is overwhelmed by a sudden influx, so borrowing rates (and thus APYs) experience a slighter dip.
Faster Yield Recovery
By distributing funds, supply-demand imbalances correct themselves more quickly — often within 1–3 hours, whereas a single-protocol deposit can take much longer (12–48 hours) to normalize.
$100k
12.23%
12.83%
$250k
11.08%
12.35%
$500k
9.68%
11.89%
$1M
8.01%
11.67%
The numbers above represent momentary APY changes and their subsequent recovery. They are derived from multiple instances of similar liquidity-volume shifts in real DeFi protocols, tracked on-chain.
Large deposits suppress APY before rebalancing occurs.
Splitting liquidity across multiple protocols reduces the severity of the drop and speeds up yield recovery.
These figures illustrate typical market reactions, not guaranteed future yields. Actual APYs depend on real-time borrowing demand, protocol parameters, and overall market conditions.
Important! APRs in a simulation results above are not a projected market APY, but only a strategy efficiency demonstration, based on real onchain data.
After assessing the total APR for each strategy relative to the AAVE benchmark, our findings indicate that the Optimized Strategy consistently surpasses both the Simple Strategy and the Benchmark. Notably, its advantage grows significantly with higher liquidity levels. While the Optimized Strategy delivers only a 3% improvement at $1 million, this figure rises to 17% at $3 million, and surges to 132% at $10 million.
