For Day 4’s manual challenge, I approached it in two parts:
Part 1: News Interpretation
I evaluated each product news item based on two key dimensions:
1. Severity (impact magnitude)
2. Timeliness (urgency)
Detailed analysis:
Product 1: Gradual decline with rumors of big buyer
→ Probability: High chance of small drop, low chance of surge
→ Severity: Low, Timeliness: Weak
Product 2: Current price at 100, future bulk sales at 80
→ Clear 20% downside expected
Product 3: Preliminary link to Saturday accident
→ Severity: Medium, Timeliness: Medium
Product 4: New fishing rod tax law tomorrow
→ Severity: High, Timeliness: Strong
Product 5: Snowstorm 3 days ago caused pants shortage
→ Severity: Medium, Timeliness: Medium
Product 6: Sales up 32%, retention down 3%
→ Severity: Uncertain, Timeliness: Excellent
Product 7: Official report links iced tea to deaths
→ Severity: High, Timeliness: Good
Product 8: Unclear reference to last week’s event
→ Severity: Unknown, Timeliness: Poor
Product 9: CEO’s penguin tailcoat mandate
→ Non-actionable item
Part 2: Portfolio Optimization
With expected returns estimated from news analysis, I developed the optimization framework:
Transaction cost function (empirically determined):
F(x) = x²/625
Single investment profit function:
PnL(A) = (r% × x) – (x²/625)
Optimal position size (unconstrained):
x* = 625 × r% / 2
For multiple investments with total budget ≤ 750:
Total PnL = Σ(r_i × x_i – x_i²/625)
Solution approach:
1. Use individual optima if Σx* ≤ 750
2. Otherwise, solve constrained optimization via Monte Carlo simulation
