In reflecting on 2011 and the great risk on/risk off dialogue, we began to consider our own contributions to the conversation. The isolation of a day as being risk on or risk off in our publications had been far from scientific. Deriving inspiration from Oliver Wendell Holmes, we never attempted to define what made a day risk on or risk off, because we simply knew it when we saw it by the streaks of neon red or green across the trading screen.
However, as we put together our piece on 2011 volatility earlier this week, we realized that, without the pretty colors flying in front of our face, recognizing a day as risk on or risk off was a little more difficult. Was it based on what percentage of the markets were headed in a certain direction (i.e. 80% of stock, commodity, and foreign currency markets up), or on how much those markets moved on average? (i.e. the average gain across stock, commodity, and foreign currency markets was x% today) Or both, a majority of markets and a high amount of movement?
After sorting through a decade of data relative to claims of a “risk on” or “risk off” atmosphere, we found the following framework most useful in defining what exactly a risk on/risk off day looks like:
- First, we decided on the groupings of markets which tend to move in tandem on risk on, risk off days. For the purpose of this post, risk on markets include Gold, Silver, Platinum, Palladium, Copper, Crude Oil, RBOB Gasoline, Heating Oil, Corn, Wheat, Soybeans, Cocoa, Cotton, Kansas City Wheat, Sugar, S&P 500, Nasdaq, Dow Jones Industrial Average, the Euro and the British Pound. Risk off markets include U.S. 30 Year Bonds, U.S. Ten Year Notes, and the U.S. Dollar.
- Next, we found it made more sense to use just the “risk on” markets to define both risk on and risk off. In this manner, risk off is not necessarily when those risk off markets are rallying, but, instead, when the risk on markets are falling. Think of “risk off” as “risk on is off.“
- Finally, we looked for the average daily move of the risk on market grouping between 2000 and 2011 and defined a day as “risk on” when the average gain across all of those markets was over 1%. While 1% may seem somewhat small to be a full bore “risk on” day, that is the average across all those markets, and we found it to be best aligned with the reporting of risk on/risk off days in both the mainstream media and our own commentary. A “risk off” day was the inverse- a day where the risk on market grouping was down over 1%.
Having established a framework for identifying risk on and risk off days of trading, we started to look at the aggregate data for any interesting trends that we might find, especially in light of 2011’s risk on/risk off laments. As it turns out, those frustrated with the environment may have had good reason. While 2011 was no 2008 in terms of risk-oriented behavior, it was still witness to far more risk on and off trading sessions- both in quantity and severity- than 2010. Perhaps even more frustrating was that the moves led to more severe drops in most markets on risk off days last year than in the previous ten combined, without a corresponding extension of gains on risk on days.
Will this year continue along the same lines? If RBOB Gasoline is really the risk on bellweather in purports itself to be, that may well be the case, with 3 of the first 7 days of 2012 trading witness price swings that fall squarely into those same risk-oriented patterns. Should this be the case, long-term trend followers may face continued struggles in a land of fickle trendlines, while short-term trend followers may get another shot to shine. At this point, it’s a game of wait and see.
