Is Dr. Copper diagnosing a stock market crash?
The following is a reprint of the market commentary from the September 2019 edition of The Option Advisor, published on August 23. For more information, or to subscribe to The Option Advisor -- featuring 10 new option trades each month -- visit our online store.
Conventional investing wisdom has long held that copper, the humble base metal, has predictive powers when it comes to the economy. Due to its wide use in the industrial, energy, and technology sectors, the theory suggests that strong copper demand (and, by extension, strong copper prices) is correlated with economic growth, while weaker copper demand and prices are suggestive of an imminent downturn.
But if that theory holds any water, it's probably true that we should all be reading this commentary from our underground bunkers right now, with a stockpile of canned goods and gold bullion close at hand. Copper prices have been in a long-term downtrend for about eight years now, and futures have shed roughly 20% from their June 2018 highs around $3.27 per pound. Zeroing in on the most recent drop, copper has already slumped 12.5% from its peak just shy of the $3.00 level in late March.
Framed another way, the ratio of the S&P 500 Index (SPX) to copper just crested above 11 to reach its highest point since November 2003. In other words, copper hasn't underperformed the broad U.S. equities market to this degree in nearly 16 years. And -- not inconsequentially -- the last time the ratio was this high, it was on the downswing, as opposed to the current uptrend.
What's more, at least one group is braced for a steeper slide, based on the weekly Commitments of Traders (CoT) data. Large speculators are fresh off a record net short position on copper futures, with the pinnacle reached as recently as Aug. 6, at net negative 60,135 contracts. Schaeffer's Quantitative Analyst Chris Prybal notes that this amount represents 28% of open interest -- the most elevated reading since February 2015 -- while total copper futures positions among large speculators are at their highest since January 2018.
The recent market panic over the Treasury yield curve inversion has effectively propelled recession hysteria into the spotlight, even though this bond market anomaly has typically pre-dated past economic downturns by more than a year and a quarter, on average. So, for those investors who tend to hit the panic button and sell stocks at the first tick of the 10-year yield below the 2-year, there's perhaps a surprising degree of opportunity loss over the next year or two.
With this in mind, we crunched the numbers to see whether the pronounced weakness in copper futures of late might similarly be a "dumb money sell signal," or whether there may be genuine cause to flee toward "safe havens" when this industrial metal really starts to tank.
To drill down on previous instances where copper's performance has negatively diverged from the S&P's by a similar magnitude, Schaeffer's Senior Quantitative Analyst Rocky White looked back to 1971 to find times the S&P was up 10% or more on a year-to-date basis by Aug. 19, with copper simultaneously in negative territory for the calendar year. There have been only eight such occurrences, starting in 1980.
Bulls can certainly find a lot to like about the results following those previous signals, per the tables below. Looking at S&P average and median returns one month and three months after a signal, the index outperforms its typical "anytime" returns for those time frames at every checkpoint. Most notably, the S&P averages a 2.60% gain three months after a signal, easily blowing past its typical 0.98% advance for this time frame.
Meanwhile, the rest-of-year returns are likewise remarkable. The S&P's average return from mid-August through year-end is 6.44% following one of these copper underperformance signals, compared to 2.85% anytime. The percentage of positive returns also improves to 88% from 71% -- while the standard deviation of returns is cut by more than half, collapsing from 11.59% "anytime" to 5.67% after a signal.
Bearing in mind the usual caveats that accompany this type of analysis -- this is only one study, looking at one data point, with a single-digit sample size -- the results here do suggest that a low-volatility S&P rally could be in the cards through the end of 2019, which would certainly be a welcome development after the fourth-quarter 2018 meltdown in the equity markets.
For a more detailed look at how each individual signal has played out on a year-by-year basis -- including the most recent instance, in 2013 -- please see the table below.
As for copper itself, the commodity tends to enjoy a short-lived burst of outperformance following a signal... then woefully underperforms through the end of the year. After racking up an average one-month return of 5.58%, with 88% positive (compared to its anytime 0.74% rise, 61% positive), copper goes on to tumble an average of 4.79% through the rest of the year. By contrast, in an average "rest of year" period, copper would gain 4.39%.
One wild-card factor that could bolster the case for short-term outperformance in copper is that multi-year extreme net short position on the part of CoT large speculators. Generally speaking, a more "average extreme" (so to speak) would be a net short position of around 30,000 contracts, and large speculators have hit this bearish benchmark nine times since 2008 -- including on June 11, 2019, en route to the current accumulation of 60,000-plus contracts.
Interestingly, data from Prybal shows that copper tends to exceed its average anytime returns as far out as the six-month mark following these signals, with outperformance peaking at three months. And this time around, that three-month CoT signal checkpoint just so happens to fall in mid-September -- right within the same time frame of the expected one-month return from the S&P underperformance signal highlighted above.