Inside the Black Box: The Simple Truth About Quantitative Trading

4.1. LIMITING THE AMOUNT OF RISK

Size limiting is an important form of risk management. It is easy to imagine having a tremendously good trading idea, seemingly a "sure thing," but without some sense of risk management, there can be a temptation to put all one's capital into this single trade. This is almost always a bad idea. Why? Because, empirically, a sure thing rarely exists, so the correct way to size a trade in general is certainly not to put all your chips on it. Otherwise it is likely that in the process of "going all in," at some point the trader will go bankrupt. In other words, it is prudent to take just as much exposure to a trade as is warranted by the considerations of the opportunity (alpha) and the downside (risk). Quantitative risk models focused on limiting the size of bets are common, and many are quite simple. The following sections explain how they work.

There are several kinds of quantitative risk models that limit size, and they vary in three primary ways:

1. The manner in which size is limited

2. How risk is measured

3. What is having its size limited

4.1.1. Limiting by Constraint or Penalty

Approaches to the size limits come in two main forms: hard constraints and penalties. Hard constraints are set to "draw a line" in terms of risk. For instance, imagine a position limit that dictates that no position will be larger than 3 percent of the portfolio, no matter how strong the signal. However, this hard limit may be somewhat arbitrary (e.g., imagine a 3.00 percent ...