How to Use Correlation to Improve Your Investing
Today's article is inspired by my recent quest to find a way to improve my own investing strategy. You can read more about my strategy here.
First, allow me to define correlation.
Correlation - the degree to which two or more attributes or measurements on the same group of elements show a tendency to vary together.
If something is correlated with another thing, the two are likely to change similarly. If they are 100% correlated, they will each move the same amount at the same time. If they are -100% correlated they will move exactly the opposite of one another.
Correlation in Investing
Modern investing is strongly based on correlation of assets. Modern Portfolio Theory - the framework your financial adviser uses to suggest things you should buy - revolves around correlation.
In order to smooth out the wild ups and downs of the market, advisers will usually split your money between stocks and bonds, as they say. Why? Because when stocks down, bonds go up (or at least down less).
When assets are negatively correlated, one moves up and the other moves down. That would be something like a correlation of -30% (-0.3).
When assets are loosely correlated one moves up, the other moves up a little less. One moves down, the other moves down a little less. That would be something like a correlation of 60% (0.6).
If two assets move exactly the same, they are known to be 100% correlated.
If two assets move exactly opposite, they are known to be -100% correlated.
Assets with a correlation of exactly 0 have absolutely no statistical relationship to each others movements.
Stocks and bonds are an example of weak or negative correlation.
According to Portfolio Visualizer's Correlation Tool, the Vanguard Total Bond Index ETF is negatively correlated with the SPY S&P Index Fund. You can see the numbers here.
Looking at the chart above we can see that during the market turmoil of 2008, the correlation between the overall market as measured by the S&P when compared to bonds in general as measured by BND was about -60%.
So what does that mean for you?
That means in 2008 when the S&P fell $1.00, BND gained $0.60.
Unlocking the Potential of Correlation
Hopefully, it's beginning to dawn on you where I'm going with this.
In theory, if we can find something that is negatively correlated to an asset we are invested in, we can use it to make money when the rest of our portfolio falls.
If my portfolio is heavily invested in Apple stock, I might look for an asset that is negatively correlated with Apple. That way if and when Apple loses money, the other part of my portfolio gains money. That smooths out or entirely negates any swings in price.
This is a form of hedging.
Hedging - an act or means of preventing complete loss of a bet, an argument, an investment, or the like, with a partially counterbalancing or qualifying one.
More often than not we have assets which are loosely correlated, rather than negatively correlated.
If two assets are perfectly negatively correlated and we hold equal weights of both, our portfolio value will never change. If one rises $1 in value, the other will fall exactly $1 in value.
In most portfolios, we invest in unequal weightings of loosely correlated assets. 80% stocks, 20% bonds. This smooths out the ups and downs of the market but the bonds drag on our returns in good times, and the stocks destroy our portfolio value in bad times.
We end up getting a lukewarm porridge that's not too hot, not too cold, but it also tastes like crap.
A perfect investment portfolio would be invested in the best-performing asset when markets are going up, and then switch to the best performing asset as markets head down. That requires timing. There's one problem with timing...
So, we can all agree that in a perfect world we can bounce between negatively correlated assets with perfect timing and make tons of money - even when markets are falling.
We can also probably at least agree that timing is difficult if not impossible.
So my solution was to design an experiment that changes my portfolio between being invested in the S&P and being invested in gold based on a specific set of criteria. My theory being that gold will gain money when the market declines and I can simply let the market tell me when I should start buying gold.
Instead of trying to time it myself, I'll just let the market tell me when to change things!
This is also known as a Quantitative Strategy (Quant for short) and you can learn more about them here.
Quantitative Investing Strategy - use proprietary (mathematical) models to increase their ability to beat the market.
My strategy is quite simple, designed specifically to hedge against market downturns, and uses primarily index investing. It also beat the return of the markets as a whole by nearly 16% from 1998 - 2017 which includes the Tech Bubble and The Great Recession.
You can learn more about my strategy to outperform the market during economic downturns here in the full article.