Even though gold prices have failed to breach the 1347/1350 USD area, the precious metal has tremendously outperformed copper, whose fate remains highly correlated to China's growth and PMI. As a result, the copper/gold ratio has fallen back to a level unseen in the past year.
The chart shows several interesting features:
i. Prior to mid-2011, higher copper to gold prices would have suggested a buoyant global economy, and thus an over performance of emerging market (EM) stocks against the S&P 500. The ratio was still negatively correlated to the VIX or overall risk aversion, but this is no longer the case.
ii. The relationship has since inverted, hence the difficulty in assessing the signal sent by the recent disconnect between two series: should we play mean reversion and call for either an outperformance of EM stocks and/or a recovery in copper prices relative to gold. On the contrary, does the correlation break signal a return to the old regime and therefore that there is no mispricing? (Lower copper prices = lower global growth = lower risk appetite = outperformance of the S&P 500).
To answer this question, I use a cross asset relationship that has held up even between 2011 and 2013. As can be seen below, the outperformance of the S&P 500 to U.S. Treasuries is highly correlated to the price return differential between copper and gold.
From this chart we see:
i. Further proof than when you buy gold, you are actually long U.S. Treasuries;
ii. The relationship has held pretty well over the last five years. The recent disconnect would clearly call for a re-adjustment of copper (+) and gold (-) prices. Yet, if you believe that the truth comes from commodities, then short stocks and bet on U.S. Treasury yields close to 2.5% - something difficult to sustain after the March 19th FOMC.
iii. If history is a guide, over the last few episodes when both series diverged, the commodity ratio adjusted back to the level implied by the stock-to-bond relative return spread: this would favor a forthcoming hike of the copper/gold ratio.
On the FX side, another pattern should be stressed. The AUD/CAD used to be strongly linked to the copper/gold ratio. The relationship blew up and then inverted in a manner similar to that observed in the S&P 500/EM ratio.
As can be seen below, in spite of a recent hike in the AUD/CAD, the spread of 2year/3month swap curves suggests that there is no overshooting. I still target 1.036 for the AUD/CAD, helped by the dovish stance of BOC's governor Poloz. As for the stock chart above, there is still uncertainty on whether the current regime will last or not. If it does, a slightly stronger pair may not be a strong enough signal for a higher copper/gold ratio.
Bottom Line: The recent collapse of the copper/gold ratio may have called for a bearish view on U.S. stocks. This is what the first and second chart would suggest. Yet;
i. On the U.S./EM relative stock performance: the uncertainty on any regime switch remains, but over the next few months is that U.S. stocks will continue to outperform their EM counterparts, where many adjustment (weaker currencies in particular) are still necessary.
ii. On the U.S. stock/bond relationship with the copper/gold ratio: I would lean on historical precedent, an upward adjustment in the copper/gold ratio. The hawkish stance of the Fed would also confirm that view.
This analysis is on a relative value basis. It suggests being long U.S. stocks against EM stocks (SPY vs. GMM) and long copper vs gold (JJC vs. GLD). I would avoid any single long position, in particular for copper, where the price outlook remains particularly bearish (wait for a contango big enough to encourage financing trades before there can be any real support to spot prices).
Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.