By David Varadi and Corey Rittenhouse
The global economy and financial system have become increasingly interdependent. We live in a world where North American consumer demand drives sales of Chinese companies, and where a Greek pension crisis can cripple the U.S. stock market.
In early July the correlation between the S&P 500 index (SPX) and its component stocks climbed to its highest level since the October 1987 market crash. Moreover, correlations between the S&P 500 and other market indices have also recently hit new highs, and are quickly approaching the maximum value of +1.00, which implies perfectly synchronized moves (Figure 1). Some researchers believe correlation has spiked, in part, because large traders now focus more on stock-index exchange-traded funds (ETFs) than individual stocks.
High correlations affect traders and investors several ways. It makes diversification less effective at reducing portfolio risk, and it makes correctly sizing positions and managing risk increasingly important. Also, watching instruments that track the S&P 500 index (SPX) is often more important than watching other index instruments, because they represent the 500 biggest companies in the largest economy in the world.
The dynamics of intermarket correlation can also be used to improve trading strategies.Defining correlation
Let’s begin by defining correlation. The correlation coefficient “R” measures the linear relationship between two sets of data, ranges from -1 to +1, and is independent of the unit of measurement. R values near zero reflect little correlation between the data sets; values near +1 or -1 show a high level of correlation. When two data sets are positively correlated, an increase in one suggests a likely increase in the other. Negative correlation implies if one increases, the other tends to decrease.
While correlation does not imply causation, R values above 0.85 generally reflect strong statistical significance. The correlation calculation is included in virtually all statistical software, as well as spreadsheet programs like Microsoft Excel. (Visit www.activetradermag.com from Oct. 9 to Oct. 30 for examples of how to use it in a spreadsheet.)
The question is, when two markets are correlated, which one is causing the relationship? If the correlation between an individual stock or ETF and the S&P 500 tracking stock (SPY) is greater than 0.85, the broader market is likely pulling the smaller up or down, rather than vice versa. In a way, SPY functions as the “sun” in the solar system of stocks and ETFs, influencing their behavior significantly.
We borrow the concept of correlation from pairs trading, in which traders arbitrage between highly correlated instruments that are also conceptually related —for example, two stocks in the same industry or sector. Market indices are also good arbitrage candidates, especially stock-index ETFs such as the Nasdaq 100 (QQQQ), the MSCI European, Australasian, and Far East (EFA), the Russell 2000 (IWM), and Emerging Markets (EEM). To determine if trading strategies perform differently when SPY is either tightly or loosely correlated to other instruments, we will test a simple mean-reversion strategy that uses the two-day Relative Strength Index (RSI) on these stock-index ETFs. For the complete article, see the November 2010 issue of Active Trader magazine. Click here to subscribe.