Every portfolio has exposure to common risk factors, whether to an economic sector (like technology or materials) or to a trading style (like momentum or mean reversion). Quantopian's risk model will help you both identify and manage your portfolio’s exposure to common sources of risk.
As you review your portfolio, it is helpful to identify how much common risk factors impact your portfolio's returns. With that insight, you can proactively decide to manage away or permit some exposure to those factors, depending on your investment thesis.
Once the contribution of the common risk factors to the portfolio's returns is stripped away, the rest of the returns are explained by your model's ability to forecast future returns. Generally, exposure to common risk factors can be replicated by investors through ETFs, swaps and other cost-efficient means. However, the portion of returns that are attributed to your model's forecast are not easy to replicate. This is the return that is uniquely attributable to your investment idea (and sometimes called alpha).
Before you write your algorithm, you need an idea, expressed as a market factor, that you think might be profitable. You will probably use Quantopian’s Alphalens to evaluate a large number of factors before you find one that has promise. The risk model will help you further evaluate those factors by analyzing their common and specific risk exposures. It can help you understand whether a factor is truly novel or based on well-known risk exposures.
Quantopian offers a powerful Optimization API to drive portfolio constructions. The optimizer can use the risk model’s predictions to minimize common risks while maximizing the specific returns.
The risk model can be used to review your portfolio’s historical performance and to assess what fraction of the returns come from common risks and what come from specific risks. This can be an as-traded portfolio or commonly a simulated portfolio.
The risk model evaluates each holding in the portfolio individually. First, every company is categorized in one of 11 industrial sectors. Each position in the portfolio is checked for the level of exposure to the corresponding industry sector.
Fama and French first popularized style factors. A simple, well-known style factor is the company’s market cap. The idea behind this factor is that over time, companies with a smaller market cap outperform large cap companies. The Quantopian model looks at each position’s exposure to five style factors: momentum, market cap, value, mean reversion, and volatility.