Research
A backtest with a 2.8 Sharpe
Your new cross-sectional signal shows a 2.8 Sharpe over twelve years — after you tried 300 feature combinations on the same data.
All simulators
Signals, capacity and overfitting
Quantitative Investing simulator
Your seat: Systematic research seat, quant fund
You are paid for edge that survives out of sample, real costs and crowding. Every extra specification you try makes your best backtest a little more of a lie.
What you'll learn
- How overfitting and look-ahead bias manufacture Sharpe ratios that do not exist
- Why capacity limits mean alpha is a finite resource, not a scalable product
- What crowding does to a diversified book on the day everyone de-grosses
Who it's for
Systematic research, quant risk and data-driven investing roles.
It works as interview prep: you'll have made the calls an interviewer asks about, and seen what happens when you get them wrong.
Decisions from your first year
Three calls a year. Here's the kind of thing that lands on your desk before your first promotion.
Capacity
The signal works at $100m, not $900m
Your best strategy earns 14% net at $100m. Market impact analysis suggests it decays to roughly 4% at $900m. Sales has already promised capacity.
Risk
Your factor is everyone's factor
Prime broker data shows your top-decile longs are held by nine similar funds. The strategy has never been more profitable.
The ladder
Each rank has its own bar for returns and standing. Clear both to move up.
- Quant Researcher
- Senior Researcher
- Portfolio Manager
- Head of Research
- Head of Systematic
- Chief Investment Officer
Terms you'll use
- Market regime
- The prevailing pricing and risk environment.
- AUM
- Assets under management.
- Redemption
- Investors withdrawing capital from an open-ended fund.
- Concentration
- Too much of the book in one sector or one name.