Backtest
See how these vaults would have done since 2019 before trusting them with anything.
What a depositor would have earned in Strike's weekly vaults on TSLA, NVDA, AMZN and SPY, week by week from January 2019, next to simply holding the stock.
A simulation on daily adjusted closes that follows the contract rules, not a track record. It assumes every option sells at the model price; see the caveats.
The protocol prices with a keeper-set volatility; the backtest stands in trailing 21-day realised volatility × 1.00 (no premium) or × 1.15 (15% above). Neither is market-implied. Target delta 0.20, premium at fair value, premium held.
- Annual return (CAGR)
- 23.1%
- Buy-and-hold 44.0%
- Volatility
- 33.7%
- Buy-and-hold 62.3%
- Sharpe ratio
- 0.78
- Buy-and-hold 0.89
- Max drawdown
- −46.3%
- Buy-and-hold −72.2%
- Avg weekly premium
- 0.75%
- Per option, of collateral
- Weeks assigned
- 18.4%
- Finished in the money, of 403
Volatility and Sharpe use weekly returns × √52 at a 0% risk-free rate; drawdown is measured at weekly closes.
Equity curve
TSLA covered call ended at 4.97× the money deposited, against 16.7× for holding TSLA.
Growth of 1.0 · TSLA covered call vs buy-and-hold
- Covered call vault
- Buy-and-hold TSLA
Weekly premium
An average of 0.75% of collateral a week; 18.4% of weeks finished in the money.
Weekly premium · 403 weeks
Stress periods
In the 2020 crash the vault returned −36.8%; holding TSLA returned −46.6%.
Stress windows · TSLA covered call
- Covered call vault
- Buy-and-hold TSLA
Delta sensitivity
Calls: a lower delta did better on every ticker. Over a period in which all four stocks rose strongly, the less upside the vault gave away, the better. A property of this sample, not a rule.
By target delta · VRP 1.15
- Covered call vault
- Buy-and-hold TSLA
Caveats
From docs/backtest.md, Assumptions and Limitations.
- The implied volatility is modelled. Every option is priced with trailing 21-day realised volatility × 1.00 or × 1.15, not market-implied volatility, so it misses skew, term structure and earnings spikes. For SPY the VIX ran about 30% above trailing realised volatility; the data says nothing about the ratio for TSLA, NVDA or AMZN. The two factors bracket a range; they do not estimate it.
- Full fills are assumed. The vault sells its full 80% of capacity every week at Monday's closing model price. On-chain, options sell only if buyers want them at the oracle-anchored price, and unsold size earns nothing. Demand is the largest gap in this backtest.
- Covered calls trail buy-and-hold in return but cut volatility. At 0.20 delta the covered-call vault lagged buy-and-hold on every ticker, with 25–46% less volatility and a smaller maximum drawdown. Its Sharpe ratio was below buy-and-hold's in every base case.
- One period, no costs. 2019–2026 was an unusually strong market for all four stocks, the kind in which covered calls look worst. USDG earns no interest here; gas, keeper costs and claim delays are ignored. The stock tokens did not exist before 2026, so this is the strategy run on the underlying stocks' adjusted closes.