Trading Case: EV & Sizing
Work a list of opportunities the trader's way: compute expected value and the Kelly stake, take only the real edges, and size them to the edge with fractional Kelly.
The scenario
A list of betting/trading opportunities lands on your desk — each with a win probability and odds. The job is the trader's core discipline: act on expected value rather than the likely outcome, take only the real edges, and size them to that edge with Kelly rather than to your conviction.
Where this shows up
'What's the EV?' rather than 'what's most likely?' is exactly the lens firms of this type (SIG especially) hire for.
Firms such as SIG, Optiver, Jane Street.
DeskPrep is not affiliated with, endorsed by, or sponsored by any named firm. Firm names are used for illustrative, educational purposes only and do not imply that these materials are official assessments of, or are connected with, those firms.
Task brief
# Trading Case: EV & Sizing **Role relevance:** A short trading-desk decision drill - expected value and Kelly sizing. **Estimated time:** 40-45 minutes · **Difficulty:** Intermediate · **Format:** Python (.py) + CSV ## What you are given - `opportunities.csv` - a list of bets (win prob p, net odds b) - `ev_sizing_starter.py` - with the EV and Kelly functions left to you ## What you must deliver 1. Expected value per opportunity 2. The Kelly stake per opportunity 3. Selection of only the positive-edge bets 4. Half-Kelly sizing and the total edge captured ## Constraints - Standard library only (csv). ## Submission note Implement `expected_value` and `kelly_fraction`, run it, then check the mark scheme and solution.
Your tasks
- 01Compute the expected value of each opportunity (p·b − (1−p)).
- 02Compute the Kelly stake for each (p − (1−p)/b), clamped to zero when there's no edge.
- 03Take only the positive-EV opportunities; pass the zero- and negative-edge ones.
- 04Size the winners at fractional (half) Kelly and report the total edge captured.
How you're assessed
The full points-based mark scheme is included with the pack.
What you'll learn
- Why you act on EV, not the modal outcome — a 15× payoff can still be a bad bet.
- How Kelly turns an edge into a stake, and why desks bet a fraction of it.
- Why folding (passing zero/negative edge) is itself a position.