Still, there may be benefits to empirical risk models. If the theoretical risk models are any good at being right, an empirical model should capture these effects without having to know the names of the factors beforehand. If market risk is indeed a big driver of stock prices, an empirical model should pick this up from the data. If the data don’t bear it out, what good is the theory? Furthermore, the competing objectives of statistical significance and adaptiveness can be dealt with in part by using intraday data. For example, if a quant uses one‐minute intraday snapshots of price activities instead of simply a single closing price for each day, he is able to extract almost 400 data points for each day in his sample, which allows him to use far fewer days to achieve the same statistical significance as another quant using a single data point for each day (the closing price).