Case Study
651.56% backtested return
from a 72H Z-score Bitcoin strategy
Overview
We asked our AI to identify datasets correlated with BTC’s future price.
It found several, with one showed a 0.67 correlation. From that dataset, our AI
engineered features and ran a backtest with defined cost assumptions and
basic risk controls to evaluate trading performance.
Trading Strategy
The base trading strategy testing was designed around entry and exit rules
using Z-score signals specific lookback window, with a short holding period
to manage exposure and cost.
Assets
BTC
Signal Filtering
Only certain % of Z-score signals were used, selected by our AI Data Science signals.
Transaction Cost
0.025%
Performance Summary
Backtest Period
Jan 3, 2018 — May 24, 2025 (2,697 days)
Why This Matters
Our strategy outperformed many BTC benchmarks using clear entry and
exit rules, short holding periods, and cost assumptions. The backtest highlights consistent performance in terms of returns, drawdowns, and trade behavior.
We added Monte Carlo simulations to understand the range of possible outcomes rather than a single equity curve. By simulating thousands of P&L paths,
we could estimate the probability of adverse outcomes.
Walk-forward testing showed that the model holds up across different
market regimes including bull, bear, and sideways periods. By retraining and validating on rolling windows, we confirmed the strategy's adaptability over time.
Want a tailored strategy?
Let’s Talk
Your Data Science Team
Automated for Alpha
This information is for informational purposes only
and should not be considered investment advice.
Past performance does not guarantee future results.
Investing involves risks, and results may vary.
Strategies and performance provided are
generated by AI-driven algorithms.
© CL1CK 2025