Training · QF-1 · Lesson 01 of 05
What quant funds do
The rule
A quant fund earns money on statistical regularities too small to see in one trade, harvested at scale with tight risk control. Because each bet is small, the edge has to survive costs and repeat across thousands of names.
Strategy families
Quant strategies differ mainly by horizon and by what they predict.
- ·High-frequency and market making: seconds or less, earning the spread.
- ·Statistical arbitrage: hours to days, trading mean reversion among related stocks.
- ·Factor and equity market neutral: weeks to months, value, momentum, quality.
- ·Managed futures / trend following: trends across futures markets.
- ·Systematic macro: rules-based views on rates, currencies and commodities.
The fundamental law
The information ratio of a strategy is approximately its skill per bet multiplied by the square root of the number of independent bets. This is why quants pursue breadth: a tiny edge applied thousands of times can produce an excellent risk-adjusted return.
- Information coefficient (skill)
- 0.03
- Independent bets per year
- 2,500
- √breadth
- 50
- Information ratio
- 1.5
- Same skill, 100 bets
- IR 0.3
Capacity and crowding
Every strategy has a capacity: the amount of capital it can run before its own trading erodes the edge. High-frequency strategies have high Sharpe but low capacity; slow factor strategies scale but have lower Sharpe. When many funds run similar signals, the trades become crowded and can unwind violently together.
Signal decay
Edges fade as they become known and competed away. A signal that earned 10% a year when published may earn half that a decade later. Research teams run continuously just to replace decaying signals.
Costs decide everything
A strategy that trades its whole book daily needs a large edge to beat costs. Researchers measure returns after commissions, spreads, market impact and financing. Many promising signals die at this stage, which is why cost modelling sits at the centre of the business.
Hold onto this
- ·Horizons range from microseconds to months; each has its own competition and costs.
- ·Alpha is small per name, so breadth across many uncorrelated bets is the whole business.
- ·Execution slippage, fees and borrow costs can erase a signal that looked real on paper.
Progress saved on this device
Education, not financial advice. Figures may be simplified for illustration.