Over the past few weeks the Investment Team has been reviewing our tactics for getting our portfolio back to a neutral stance and explaining them to our clients and other interested parties. However, the way we measure the positioning of our portfolios and define a neutral stance is itself worthy of discussion.
The two most common ways to measure the risk of a portfolio is by looking at its volatility and beta and comparing them to the volatility and beta of the portfolio’s benchmark. While volatility measures risk in absolute terms (how volatile the returns of the portfolio are), beta measures risk in relation to a benchmark (how much the portfolio is exposed to benchmark’s risk). The reason for looking at both is that part of the portfolio’s risk may be unrelated to the benchmark’s risk. The volatility and beta that portfolio managers typically report to clients are backward-looking: They are calculated over a trailing historical sample of portfolio returns and reflect the average risk of the portfolio over that time frame.
For tactically-managed portfolios like ours, whose composition can change relatively frequently, this is clearly not a good solution for estimating how the portfolios are positioned going forward. Instead, we use what we call pro-forma volatility and beta, which use the historical beta, volatility, and cross-correlations of the individual securities to estimate the volatility and beta of our current portfolios during a historical period. In other words, we answer the question: “Had we held the same portfolio we hold today over some trailing time frame, what would our volatility and beta have been?”
The answer is typically an accurate estimate of the kind of volatility and beta the same portfolio will experience going forward. The pro-forma allows us to estimate the risk profile of virtually any portfolio composition, which makes it a valuable tool for understanding the impact of potential trades on the risk of the portfolio before they are executed. The chart to the right plots volatility and beta of our latest pro-forma (trailing one year of daily returns), and shows that for the most part, we are within the margin of error of being neutral, or 100% of benchmark risk.