Summary Again, the Fed is threatening to raise interest rates. The results of my statistical study show that increasing interest rates leads to an increase in overall market volatility. Whether liquidating or investing in an increasingly volatile market, you have several strategies that can give you an advantage. I have had several requests for statistical analyses on individual stocks, but recently I was asked to look into the correlation between interest rates and volatility. This request does not come as a surprise for two reasons. First, although Chairwoman Yellen recently passed on raising interest rates , others are stating that, regardless, we will see a rise in interest rates this year . Many are asking what will happen to the market once this happens. Second, the VIX and its associated ETF, the iPath S&P 500 VIX Short-Term Futures ETN (NYSEARCA: VXX ), are looking increasingly bumpy. This time of year tends to bring bumps in the VIX, with a dip and subsequent rally. Investors are wondering what will happen to the VIX (which can be thought of as an overall measurement of how risky the market is at the current time) if interest rates increase. An increase in interest rates could just be the catalyst to bring back market volatility. But does an increase in interest rates truly bring an increase in volatility? Though I could find a few articles online claiming this fact, I found no previous statistical analyses on the subject. Some images backing the claim of a correlation between interest rates and volatility are examples of exactly what you don’t want to rely on as an investor: curve-fitting. I’ve seen too many “analysts” run models over and over until they find a couple of curves that seem to line up. This is exactly what the following two images display: The first chart shows the T-bill yield and VIX apparently lining up in perfect accord. But there are three problems here: First, a logarithmic transform was applied to the VIX line. This changes the shape of the VIX line. I suspect this was done to make the VIX curve better resemble the yield. While logarithmic scales can be useful for looking at indexes or stocks – especially when comparing two stocks trading at drastically different ranges – logarithmic scales should not be used without reason. A proper statistical model first states that the logarithmic scale should be used and gives reason for using it. I suspect that this analyst simply found the logarithmic conversion to produce the curve he wanted, meaning he was playing with data to confirm his conclusion rather than performing a true analysis. Second, the yield was transposed two years. Again, this is likely an action with the motive of making the two curves match. If the yield was not transposed to the right, the graph would show the opposite of what the author wanted – i.e., the graph would imply that yield and the VIX have a negative correlation! It simply makes no sense to move one index two years forward in time. This is especially true when the T-bill used is only a 3-month T-bill! Is the analyst trying to say that the VIX today can predict the price of a 3-month T-bill two years from now?! The second chart is equally absurd. This time, the VIX is plotted with the 2-year and 10-year yield curve (i.e., the slope of the yield curve, measuring the difference between the yield of a 10-year bond and 2-year bond). The absurdities follow: First, this analyst does the same as the previous analyst; he moves the entire yield curve forward two years. Again, this would imply that the VIX today is predicting something precisely two years from now. This is another sign of curve-fitting. Second, the analyst inverts the yield curve. There is simply no reason to do so – unless, of course, your goal is to get a desired look to your chart so you can draw a conclusion, which is the exact definition of curve fitting. Interpreting the inverse of a function in words is a difficult task – so knowing that the VIX is correlated with the 2-year future inverse of the yield curve tells us nothing! As you have probably concluded, we need a more formal way of determining the relationship between interest rates and the VIX. In this study, I set up the following set of hypotheses and test them statistically: Set 1: H0: The VIX is uncorrelated with yield rates H1: The VIX is correlated with yield rates Set 2: H0: The VIX is uncorrelated with the yield curve H1: The VIX is correlated with the yield curve The Study As you can see, the test will be simple – no data transformations or curve-fitting. I collected the data from the VIX for each day, starting from 2004. I did the same for the bond market. Because the stock market and bond market have a few days per year in which one market is closed while the other is open, I removed such dates from the analysis. I did so to allow a one-to-one comparison for the VIX and yield each day in the market. Thus, daily movements in the VIX and movements in the bond market will be tracked. For the VIX data, I used the closing values. For the bond data, I used 2-year bonds, which is more or less the “middle ground” for bonds. For the yield curve, I used the difference between 1-month and 20-year bonds, giving the widest and most sensitive curve. If anyone has any qualms with these choices, please let me know in the comments section below and I can rerun the analysis with your chosen values (e.g., daily VIX highs vs. 20-year bonds). I used an alpha level of 0.05 as the comparison point for the p-value. Correlation tests for the hypotheses that reported p-values less than 0.05 would be considered evidence for the rejection of H0, giving strong evidence for H1. The Results The results follow: Yield Yield Curve Correlation with VIX -0.2680 0.3581 p-value for correlation depreciated dollar -> increased yield curve -> increased VIX But the yield curve is actually moderated by the supply and demand of capital. Decreases in the money supply (e.g., M2 money supply), increased government deficits, and less money flowing into savings can all increase the yield curve, thereby spiking market volatility. In addition, commodity prices affect the yield curve. Generally, decreasing prices steepen the yield curve because they decrease short-term inflation expectations. This pulls the left side of the yield curve downward, making the curve steeper on the whole. In other words, when commodity prices drop, the yield curve steepens, and the VIX should see an increase. But our current market, in which commodity prices are at all-time lows, doesn’t seem to have an increased VIX, which is interesting from a theoretical standpoint. Overall, the picture is complicated: (click to enlarge) Investment Strategies for a Volatile Market For now, we can expect that the yield curve will steepen and prepare our portfolios for such an event. I don’t recommend buying the VXX outright because it’s a garbage imitation of the VIX and will cause you to lose money in the long run. However, a spike in VIX should result in a spike in the VXX, which could leave you with a handful of cash should you have call options on this ETF. But let’s look at some more realistic strategies (I hate the VXX). If volatility increases and you are a risk-averse investor, the easiest “safe” strategy is to exit the market – as reasonably as you can – before increased volatility hits. Of course, most people reading this are likely “buy-and-hold” investors, so such a method might be lost on you. One fundamental idea behind the buy-and-hold strategy is that you don’t want to miss those days with the most significant returns, which tend to happen during days of high market volatility. Of course, if you’re in the market all the time, you’re also gaining exposure to those days with the most significant losses. And a significant loss hurts a lot more than a significant gain. Going from 100 to 80 requires a 20% drop, but going from 80 to 100 requires a 25% gain. The uphill battle is harder. Perhaps the best selling strategy is a staged sale strategy. In this strategy, you sell predetermined chunks of your stocks and either hold cash or reinvest (see below). The staged sell is like the opposite of dollar cost averaging. If you don’t want to worry about market timing but want to liquidate, staged selling is your best bet. Nevertheless, for buy-and-hold investors, volatile markets can be gold mines. An increase in volatility in the general market will not hurt the fundamentals of a company. Thus, a volatile market will allow you access to sporadic dips on stocks with solid fundamentals. This is a good time to buy such stock. However, when buying, realize that some things are different in a volatile market. If you’re not in the habit of buying with limit orders, get into that habit now. Volatile markets move quickly and have high volume; your market order is likely much different from that what you expect. In addition, the bid-ask quotes you’re looking at now might be very different from the real bid-ask quotes. And then there’s increased delays and slippage… This is all general strategy. What about choosing individual stocks during a volatile market? As stated, a volatile market gives you access to a myriad of stocks that hit dips simply as a result of increased volatility on the stock. In the past, such a low would often be explained by the company’s fundamentals. But in a volatile market, the lows that looked large in the past will soon be considered the norm. As a fundamental investor, your best bet is to ignore the daily changes in stock price and instead set a buy limit order that you consider to be “too low.” Set the order as “good for the month” and get your stock at a discount. As for the types of stock to add to your portfolio, choose stock that are relatively safe and undervalued during periods of increased volatility. REITs make good choices. Switching out low-yield dividend stocks for high-yield dividend stocks makes sense, just as switching out growth stocks for value stocks makes sense. Depending on your portfolio, this might be a good time to step back and question the purpose of the portfolio. Are you focused on growth? Passive income via dividends? In the previous case, you should have an existing exit strategy. Perhaps now is the time to take your profits and look to restructure your portfolio with undervalued growth stocks. If your goal is passive income, holding on to your current dividend stocks and REITs makes sense in terms of your overall objective, and you might have no exit strategy at all. But at this time, a day’s worth of research into your current dividend stocks’ fundamentals can give you some clues as to whether dropping the stock for cash (or switching it out for a better option) is the right choice. Overall, for investors, getting defensive as the market has a seizure isn’t the right strategy because you should have been defensive in the first place. But let’s assume you need to get defensive all of the sudden. What are some immediate actions you can take? Switching out common stock for preferred stock is a good choice because preferred stock tends to have lower beta – i.e., it’s less correlated with general market moves. Dropping the beta of your overall portfolio can ensure that your portfolio contains companies that you believe are fundamentally strong and yet will not be hit hard by market corrections. Here’s a general common-to-preferred and visa versa strategy for volatile markets: Switch out common stock for preferred stock when the market appears to be overbought. You’ll have sold common stock at a high, switching them for preferred stock that are more protected against drops. If the market does drop for an extended time, drop the preferred stock, which protected value and brought you dividends, in favor of common stock, which you can now buy at a low. Overall, you want to drop your portfolio’s beta when you believe the volatility is coupled with a downward trend. You should still perform well during the good times at the same time you’re protecting your capital with a low-beta portfolio. The following are some low-beta stocks I recommend: Pfizer (NYSE: PFE ) Wal-Mart (NYSE: WMT ) Avista (NYSE: AVA ) Request a Statistical Study If you would like for me to run a statistical study on a specific aspect of a specific stock, commodity, or market, just request so in the comments section below. Alternatively, send me a message or email.