"Dr. Copper"

Almost every day it seems, we see a new article in the financial media about "Dr. Copper".

Copper has many industrial uses, including residential and commercial construction projects.  So the (apparently) universally accepted theory suggests that if Copper prices are surging, then that means demand is hot, the economy will go gangbusters, and stocks will rocket higher.

I've studied Copper in many different respects, and the theory above is, for the most part, pure bunk.  Especially in that last rung of association, future stock market movements.

The metal is getting a lot of attention now because it recently broke to a new 52-week high, and is very close to a multi-year high.  Copper has had a close correlation to the S&P 500 over the past two years, and has been a leading indicator before a couple of important turning points.

But one warning sign immediately stands out.  The chart below shows the last time "Dr. Copper" broke out to a new multi-year high.

One cherry-picked example can't tarnish the degree of "The Doctor", so let's just go back to the furthest I have data, August 1988, and see how you would have done by buying the S&P 500 every time Copper futures hit a new one-year high.

Your returns, especially in the shorter-term, would have been sub-par, and even negative in the slight majority of cases.

Across every time frame except one year later (and that just barely), you would have done better by just blindly buying and holding the S&P 500 on any random day than you would have by buying after "Dr. Copper" signaled the all-clear by trading at a new high.  In fact, you would have done a bit better by avoiding the S&P on those days.

It didn't make any difference if Copper was hitting a multi-year high, either.  Your future returns in the S&P would have been almost identical to the ones in the table above.

Let's look at this one other way.  If the S&P was trading at least 2% below its own 52-week high while Copper was hitting a new one-year high, how did the S&P fare then?

Well, in those cases it took the S&P a median of 52 trading days to reach a new one-year high.  Before getting there, it suffered a median drawdown of -3.2%.  There were several cases (1994, 2000, 2002, 2008) when Copper hit new highs yet the S&P endured large losses and long stretches of time (more than 9 months) before it hit its own one-year high.

There is a modestly positive relationship between how Copper does in one year and how worldwide GDP does the next year.  But how the economy does and how stock prices do are two very different things, and I would be exceptionally wary of anyone who suggests that we buy the S&P 500 just because Copper is at a new high.  That hasn't worked any better than random in the past.