I started doing this as a joke and did not expect to find a way to make money with Roubini Sentiment Indicator. Nouriel Roubini, a.k.a Dr. Doom has recently been stating that he is not Dr. Doom but he is Dr. Realist despite the fact that he is on TV mostly when things are going wrong, volatility spiking and stock market going down. You can clearly see this pattern from the graph below.
I gathered the Roubini sentiment index from weekly values provided by Google Trends. A simple correlation analysis shows that Roubini sentiment index has a negative correlation of (0.68) with the market and a positive correlation of 0.82 with the VIX. Some might dismiss the Roubini sentiment indicator because of its high correlation with the VIX, thinking that Roubini’s popularity is caused by the increase in VIX. Before jumping into any conclusions, we need to answer the following question: What is the exact nature of the relationship between VIX and Roubini Sentiment Indicator? If VIX is leading Roubini Sentiment Indicator, then you could rightly claim that Roubini is merely a parasite exploiting the increased fear in the marketplace. If the two indices are coinciding then Roubini is merely expressing what the market is thinking, hence he is truly the Dr. Realist or maybe Dr. Journalist? On the other hand, if Roubini Sentiment Indicator is leading the VIX, then it might be the case that Roubini is spreading the fear like the plague and causing the spikes in VIX and declines in the market. In this case, he is the dreaded Dr. Doom. Take a look at the graph of Roubini Index vs. VIX and see the high correlation yourself.
How do we test our hypothesis about VIX and Dr. Doom, I mean Roubini? Granger Causality Test is the perfect tool for this job. We use VIX as the dependent variable and lagged values of VIX and lagged values of Roubini index as independent variables. So the idea is if lagged values of Roubini index can explain the movements in VIX in the presence of lagged values of VIX, then we can say that Roubini index is Granger causing the VIX (we run some other tests to make this claim, I don’t want to bore you with the technical details). We also test whether VIX Granger causes the Roubini index. Our test results clearly show that Roubini Sentiment Indicator Granger causes the VIX up to two weeks in advance with a very high degree of statistical significance. Naturally VIX does not Granger cause Roubini index. Roubini is indeed the Dr. Doom and maybe he should be put away for the sake of our economic recovery and millions of unemployed workers.
Maybe not. Since Roubini index is a leading indicator of an imminent spike in VIX and usually spikes in VIX come with a decline in stock market, we can develop a profitable trading strategy around this finding. Hey, I have a full-time job and did this quickly for fun, so the strategy I will present now may not be refined enough. However, it is good enough for a modest blog post. Here is our simple trading strategy: Go long the VIX when Roubini Sentiment Index goes %25 above its four-week moving average. Following this strategy over the 2007-2010 period will earn us a weekly average return of an amazing 2.4% (If we had simply gone long VIX at all times, the average weekly return would have been 1.6%). If we hold on to our long VIX position for three weeks, the average return over this time period is 8.25%. Dr. Doom has been sitting on a huge pile of gold all this time, yet he does not trade (Would it be insider trading if he traded based on what he is going to say on TV the next day?). Google guys could also utilize Roubini index in real-time and trade profitably :)
Another trading strategy is shorting the SPY when Roubini Sentiment Index is 25% above its 4-week average. In this strategy when we keep our short position for three weeks, SPY declines an average of (1.6)% over three weeks. Normally the average weekly decline in SPY is only (.06)% from Jan-2007 to Aug-2010. It is really interesting that we can make money by using Roubini Sentiment Indicator.
What does the Roubini Sentiment Indicator tell us lately? The latest value for Roubini Index is 2.2 and it is more than 25% above its 4-week average. The latest reading was from Aug 22nd. So this means that the stock market is expected to decline by an average of 1.6% between Aug 22nd and Sep 12th. Of course stock market losses would be much higher if Roubini shows up on CNBC tomorrow and scares the bejesus out of investors :)
Disguised Advertising: This article is brought to you by insidermonkey, which is going to be your source for free real-time insider trading data. At least, we hope so. We also aim to educate ordinary investors and give them the tools to help them make informed decision. Again, we hope so. Come back in a few days, and I will present another interesting and entertaining article. Btw, I have a Ph.D. in financial economics with a focus in insider trading. Maybe you will come back for serious articles on insider trading.