The paper “Forecast-to-Fill: Benchmark-Neutral Alpha and Billion -Dollar Capacity in Gold Futures (2015-2025)” shows how a NASA physicist and MIT engineers would think about building a trend model. It is very similar to what we see with traditional trend-followers. The results are very good, but the process matters more. We highlight their key concepts.
We conclude that forecast-to-fill engineering—linking transparent signals to executable trades with explicit risk, cost, and impact control—can turn modest predictability into allocator-grade, billion-dollar-scalable alpha.
We build a process that consistently transforms small, noisy signals into statistically stable profits through structure:
1. Smoothing extracts persistent direction from noisy returns without overreacting to random fluctuations.
2. Confidence mapping converts that slope into a bounded, interpretable probability that scales exposure naturally.
3. Volatility targeting keeps realized risk constant over time, preventing uncontrolled leverage in quiet markets.
4. Friction-adjusted sizing ensures that every position has positive expected log-growth after trading costs.
5. Disciplined exits enforce an asymmetric payoff: small losses, large and infrequent gains.
These principles can be applied to any quant model.
No comments:
Post a Comment