Thursday, May 18, 2006

Quantifying Downside Risk

The market action was atrociously weak today. It open higher with light trading. By mid afternoon it has begun to roll over. The downward volume accelerated to extend the losing streak to a consecutive 8th day. My portfolio continues to bleed, albeit at a slower pace than the general market due to my hedges. Month-to-date, the entire portfolio is down 2.7%.

Certain housing stocks finally showed signs of life, although their gains also started to fade with the broader market. Some of them managed to finish in the green. If the rebound continues, I will consider a small nibble on the short side in view of the CME housing futures to debut next Monday. Of course, if the home builders get massacred tomorrow, I will likely cover my BZH shorts.

I was asked about my current hedging amounts and I gave a rough figure of 15% net long in the actively managed portfolio. Wanting to be more precise, I went through a more rigorous exercise today assuming a 10% drop in the S&P (example only, not my prediction), and calculating the downside risk using the beta of each security. The beta of a security describes how it moves (on average) relative to a standard index, usually the S&P. The movement of the security is the movement of the index times beta. For example, if the S&P declines 10%, a security with a beta of 1 will also decline by 10%, while a security with a beta of -0.5 will increase by 5%. Adding up the responses of each individual security in a portfolio, one gets an estimate for the entire portfolio. The table below shows the results for my portfolio under the assumption of a 10% decline in the S&P.


  • The beta values were obtained from Yahoo! Financial unless otherwise noted. They can be found on the “Key Statistics” page for most stocks and the “Risks” page for most ETFs and mutual funds. I made estimates where they were not available.
  • The beta for the asset allocation portfolio was calculated from the monthly changes YTD. Although there were only a few points, the fit was reasonable.
  • I omitted the bond funds in the actively managed portfolio.
  • The low betas for CEF and MGN were probably due to poor correlation between precious metals and stocks rather than a sign of their sluggishness. In that sense, the risks were underestimated.
  • ZPWMB is IWM Jan 80 puts (I own 10). EEMXB is EEM Dec 110 puts (I own 3). They are deep in-the-money, so I assumed a 1 to 1 relationship with the underlying ETF.
  • The column "Effective Control" is equal to the value of the securities except in the case of the puts where the value of the underlying security under control (100 sh per contract) is used.
I was a little surprised to see that my actively managed portfolio has been fully hedged. Of course, the beta values were computed from long term trends, whereas during the recent plunge, the commodity stocks dropped much further than the betas would have you believe. That’s why I came up with the 15% net long estimation before.

As a result of this work, I now have a much better grasp of what’s needed to fully hedge my total portfolio or switching to a net short stance. This exercise also pointed out some non-core, high-beta stocks, such as GVA that I should consider selling. So even if one has no plan to go against the market, checking the beta is still a good way to identify pruning candidates and manage portfolio risk.