Correlation: Now You See It, Now You Don’t?

By: Russell Sands

The following is an excerpt from Russell Sands' Turtle Secrets

Correlation is a tough subject to get a handle on.  Especially for me, since I always used to get confused when people started talking in higher mathematics, using funny looking symbols instead of regular old numbers.  Nonetheless, correlation is an interesting subject, and one that is important to traders, so it’s worth devoting a little time and effort to understand how the markets interact with each other.  However, keep in mind that in the final analysis, we should not be overly concerned with trying to find reasons why markets are acting (or not acting) the way they should, we just have to go with the flow.

Probably the easiest example to start with will be Stocks and Bonds.  What do you think, are they positively correlated, negatively correlated, or is there no correlation?  If you said positively correlated, you are right; most of the time, except when they are not.  See, we’re confused already.  Most of the time, stocks will follow bonds.  They move together, with the movement usually being a reaction to the state of our national economy.  Since the fixed income market is so much larger than the equity market, bonds are usually the leader and stocks are usually the follower.  How many of you have seen a rally in the bond market and picked up the phone to buy S&Ps?

If you want to do this type of trading, you have to look at the underlying reasons for the move.  This gets dangerously close to ‘fundamental analysis’, for which the great master Turtle had a large disdain.  But even if you don’t go this far with it, you still must be aware of why the relationships exist as they do, so you can be alert to when they might change or even break down completely.  For example, normally if bonds (which open first) rally in response to a (bullish) economic report, the S&P will tend to gap open higher.  But, what if bonds rally because inflation was reported as lower than expected during a recessionary phase of the economic cycle, when business and industry are actually looking for a boost in productivity?  Well, bonds might rally because interest rates can remain low, but stocks will sell off in disappointment that the economy (i.e. the major public industrial companies in this country) are not staging their hope for recovery.  The point is, whether stocks follow bonds or go the other way may depend on where we are in the business cycle.  The trader who blindly buys stocks because he sees bonds rally, and knows that stocks ‘usually’ follow bonds (due to positive correlation), may be in for a disappointment.

And of course, let’s not forget the negative correlation that occurs when stocks sell off too swiftly and the fixed income markets rally in a ‘flight to quality’ (remember October 1987).  Not only do the markets move in opposite directions, but now it is the bonds who are reacting to (following) the stock market, instead of assuming their usual role as the leader.

Normally, bonds and currencies are positively correlated, with the conventional reasoning going something like this: In a global marketplace, ‘money’ flows to where it can get the highest return.  If bond prices move higher, that means interest rates are moving lower, making the dollar less attractive as a vehicle in which to invest global capital.  Thus, ‘other’ foreign currencies become ‘relatively’ more attractive to large and sophisticated global investors.  So, as bonds rally, currencies rally also.

However, during a two week period in 1997, I noticed that every time the bonds rallied, the currencies sold off.  And every time the bonds sold off, the currencies rallied.  Why should this be happening?  The answer lies in the commodity markets.  The bond market had been selling off as a reaction to the strong upside breakouts in grains, metals and foods.  The bonds were afraid of inflation!  So every time Beans or Silver or Coffee took off, the bonds retreated.

Knowing that the bonds were reacting to inflation pressures, we understood why the currencies were negatively correlated to bonds at that time.  Under a strong inflationary scenario, although interest rates move higher as bond prices fall, global capital is no longer attracted to the higher rates (as they normally would be) because they were afraid of investing in a currency which might become worthless if the inflation got too much out of hand.  In this case, whether bonds and currencies are positively or negatively correlated depends on the perception of global investors and other market participants as to whether or not the inflation is controllable, or too much out of control to be worth the risk of getting paid a higher rate of return on their money.  And this perception (and thus this intermarket relationship) can change overnight.