CryptoCrispy – Annual Performance Analysis
Performance overview generated on 30 May 2025
Executive Summary
Over the last twelve months the AI Assistant analysed 225 predictions across 9 major coins, achieving an average ROI of 2.27 % at 24 % directional accuracy. The model’s best-performing asset was BTC, returning an average 4.04 % per trade.
Total Predictions Analysed
225
Average Directional Accuracy
24 %
Average ROI (per trade)
2.27 %
Best-Performing Coin
BTC
Top ROI Coins
| Rank | Coin | Avg ROI (%) | Avg Accuracy (%) |
|---|---|---|---|
| 1 | BTC | 4.04 | 52.0 |
| 2 | TRX | 3.88 | 40.0 |
| 3 | SOL | 2.79 | 16.0 |
| 4 | ETH | 2.09 | 32.0 |
| 5 | DOGE | 2.06 | 12.0 |
Recommendations & Next Steps
- Allocate a higher weight to BTC and TRX while monitoring momentum shifts monthly.
- Investigate low-accuracy assets (e.g., ADA, LTC) to refine feature engineering or exclude them from live portfolios.
- Track Sharpe-ratio stability as ROI scales to verify sustainable risk-adjusted returns.
- Deploy retraining triggers when rolling-window accuracy drops below 20 %.
Methodology Snapshot
All performance figures are calculated on a trade-by-trade basis, net of fees. Accuracy measures the percentage of correct directional predictions; ROI is the mean percentage return per closed trade. The dataset spans 30 May 2024 → 29 May 2025; raw trade logs and statistical breakdowns are provided in the accompanying JSON file.