Key components of a quantitative trading strategy
Probably the simplest ever structure of quantitative trading strategy was proposed in a book “Inside the Black Box: A Simple Guide to Quantitative and High Frequency Trading” by Rishi K. Narang:
So, according to this model, basically each strategy consists of the following components:
- Alpha model.
Alpha model consists of some sort of rules or predictive factors that derive the active value of the strategy.
- Risk (constraints) model.
Risk model has some sort of constraints, for example, how much we will buy or sell, that keep the risk of the strategy under control.
- Transaction costs model.
Transaction costs rules manage everything associated with trading fees, slippages, spreads and market impact.
- Portfolio (optimization) model.
Portfolio model may have rules for assets, their selection and weights we would trade.
- Execution model.
Execution model will manage everything associated with placing, closing, canceling trades and how they would be executed.
To create great strategies, all those models should be in place. Although this is exceptionally time consuming, but it makes, as an example, Goldman Sachs to be profitable 93.6% of the time:
Many (profitable) traders, who even come to algorithmic trading don’t discern from where their profits are coming, and as such, miss on weaker returns and/or higher risks.
For example, a trader can trade on noise (his alpha factor isn’t predictive, let’s say it directly, — missing whatsoever) and still become profitable if he employs strict risk or some sort of smart execution rules. Pension funds, as another example,may just hold portfolios, but have no alpha or risk models in place, thus making their clients’ money susceptible to system-wide risks.
