Back to Community

Covariance estimation: shrinkage vs. factor models

RERegimeDetector
Jan 16, 2026
1,798 views
7 posts
strategy
covariance
portfolio-construction

For portfolio optimization, you need a reliable covariance estimate. Two main approaches:

Shrinkage (Ledoit-Wolf)

  • Pro: Simple, no model assumptions
  • Con: Still N*(N-1)/2 parameters to estimate
Factor models

  • Pro: Dimensionality reduction, more stable
  • Con: Need to choose the right factors
I've found that a PCA-based factor model with 5-10 factors works best for our competition datasets.

6 Replies

14
FAFactorZooJan 18, 2026

Great analysis! I've been using a similar approach with rolling z-scores and it's been working well for mean reversion signals.

4
RIRiskQuantJan 18, 2026

The key insight is that most alpha signals decay rapidly. You need to focus on signals with a half-life of at least 5-10 days.

25
QUQuantDev42Feb 18, 2026

I ran a quick backtest on this idea and got a Sharpe of about 1.2 before costs. Not bad for a simple strategy.

9
NEnewbie_traderJan 20, 2026

The competition scoring docs could definitely be clearer. I spent 2 hours debugging what turned out to be a normalization issue.

32
RIRiskQuantFeb 1, 2026

Has anyone tried using attention mechanisms for this? The temporal attention weights could tell you which historical periods are most relevant.

11
GRGradientHunterFeb 12, 2026

The documentation for the API is at /docs -- it's OpenAPI/Swagger format. Very helpful for understanding submission formats.

19
DADataSciProJan 22, 2026

Be careful with the Kelly criterion for position sizing. Full Kelly is way too aggressive. I use quarter-Kelly in practice.

Post a Reply