Today I focused on comparing my performance against other market participants in bananas trading to identify improvement opportunities. Here’s my analysis approach:
1. Profit Tracking Methodology:
⦁ For all market trades reported by exchange:
PnL = abs((trade_price – fair_price) * quantity)
⦁ For my own trades:
PnL = (fair_price – trade_price) * quantity
⦁ Plotted cumulative sums of both PnLs
Key Finding: My algorithm missed ~700 in potential profits based on the gap between these curves.
2. Market Trade Breakdown:
I categorized missed opportunities into two types:
⦁ Type 1: Trades executed outside my bid-ask spread (or exactly at my quotes)
Related to those mysterious “phantom trades” from previous research
Currently unsolvable until we crack that anomaly
⦁ Type 2: Trades executed within my spread when position limit wasn’t full
Caused by having too wide bid-ask spreads
3. Optimization Attempt:
Narrowed my bid-ask spreads to capture more Type 2 trades
Result:
Reduced Type 2 missed trades (success)
But overall profits decreased because other trades became less profitable
Similar pattern observed in pearls trading
Conclusion: There’s a delicate balance between:
✔ Capturing more trades by tightening spreads
✔ Maintaining profit margins on executed trades
The 700 gap suggests I need smarter dynamic spread adjustment rather than simple narrowing.
