Strategy 3: Discounted Benchmark
Overview
A conservative variant of Strategy 2 that uses a discounted benchmark for comparison. This makes the strategy more selective about when to accelerate, only buying aggressively at larger discounts.
Type: Active / Conservative-adaptive
Completion: Always 100%
Performance: +20-30 bps typical
Completion: Always 100%
Performance: +20-30 bps typical
Algorithm
Discounted Benchmark
Instead of comparing to raw TWAP, Strategy 3 uses a discounted version:
\[
\text{Adjusted Benchmark} = \text{TWAP} \times (1 - \text{discount})
\]
For a 50 bps discount: $$ \text{Adjusted} = \text{TWAP} \times 0.995 $$
Effect
- More conservative: Only speeds up when price is significantly below TWAP
- Higher bar: Requires ~0.5% discount before accelerating
- Reduced whipsaw: Less responsive to minor price fluctuations
Implementation
def strategy_3(prices, benchmarks, total_usd, min_dur, max_dur, target_dur,
discount_bps=50):
"""Adaptive with discounted benchmark."""
discount = discount_bps / 10000 # 50 bps = 0.005
for t in range(max_dur):
price = prices[t]
raw_benchmark = benchmarks[t]
# Apply discount to benchmark
adjusted_benchmark = raw_benchmark * (1 - discount)
# Compare price to discounted benchmark
deviation = (adjusted_benchmark - price) / adjusted_benchmark
# Same adaptive logic as Strategy 2...
Discount Selection
| Discount (bps) | Effect |
|---|---|
| 0 | Same as Strategy 2 |
| 25 | Slightly more conservative |
| 50 | Moderate selectivity |
| 100 | Very selective acceleration |
| 200 | Rarely accelerates |
Recommended
A discount of 25-75 bps typically provides the best risk-adjusted performance.
Characteristics
Pros
- More consistent performance
- Lower variance than Strategy 2
- Reduces "false positive" accelerations
- Better for range-bound markets
Cons
- May miss small opportunities
- Slightly lower mean performance in trending markets
- Still forced to complete by deadline
Performance Profile
| Metric | Typical Value |
|---|---|
| Mean Performance | +20 to +30 bps |
| Std Dev | ~55-65 bps |
| Min | -80 to -120 bps |
| Max | +180 to +250 bps |
The reduced variance makes Strategy 3 attractive for risk-sensitive mandates.
(c) Joerg Osterrieder 2025