We’ve talked so much in these pages about why cross-asset correlations matter that our (digital) pens are dry and our (virtual) voices are hoarse.
There are four ways readers generally respond:
- Some folks get it.
- Some folks don’t.
- Some folks seem to think that when they hear the word “correlation,” they automatically need to recite the only thing they remember from high school science class (“correlation doesn’t equal causation”).
- Some folks seem to think that the words “cross”, “asset,” and “correlations” can only be used together in jet propulsion labs.
The thing is, this is really simple. If you understand why a 60/40 stock/bond portfolio works, then you understand cross-asset correlations – even though you might not realize it.
Well after the election, we saw the market’s faith in the reflation narrative (known to the investing public as “the Trump trade”) expressed in correlations. Have a look:
Note the green highlights. So on the left is the correlation between stock returns and bond returns and on the right is the correlation between S&P 500 returns and the performance of the trade-weighted dollar.
Those charts are not an accident.
These things are, to quote Ron Burgundy, “kind of a big deal.” Because see, when you get a massive equity drawdown, you’d better hope that the stock-bond return correlation is negative or, alternatively, that the rates-stock correlation is positive. Why? Well because you need bonds to provide a cushion. Have a look at this table from Goldman:
See the green? That’s your diversification. But ask yourself this: “how much lower can bond yields go?” That is, how can bonds rally during an equity drawdown in a very ZIRP-ish/NIRP-ish environment?
Good question, right?
Well anyway, watching the extent to which stocks, bonds, and the dollar move together is the best way (with the exception perhaps of simply watching USDJPY) of monitoring the market’s faith in the reflation narrative/Trump trade.
On Thursday, Bloomberg’s Luke Kawa is out with a great piece highlighting all of this and expounding on something that retail investors really, really need to get a better handle on. Namely that this is all about the “narrative.” When that narrative cracks – as it did last Friday – you need to pay attention to these correlations.
Will markets have a case of the Mondays or the Tuesdays?
That’s one of the the biggest questions facing investors as the relationship between U.S. equities, the dollar and Treasury bond yields becomes less correlated. After stocks deviated from the others in a move higher Monday, they traded in tandem again Tuesday in what has become known as the reflation trade since the election of Donald Trump.
“Once the U.S. presidential election was out of the way, global markets responded to the pick-up in global growth,” said Stuart Thomson, a Glasgow, Scotland-based fund manager at Ignis Asset Management. “As we started to move into 2017, the question became how sustainable is this.”
The deterioration comes as the administration struggles to gain traction for its legislative agenda, suggesting that traders are questioning whether an acceleration in U.S. economic activity will boost revenue growth, long elusive for publicly-traded companies throughout the bull market.
At the same time, the slipping relationships are welcome news to some analysts who have raised concern that continued gains in bond yields and the dollar will have an adverse effect on equities.
“The recovery will continue if we don’t have a big rise in the dollar and yields and oil,” said Christopher Sullivan, who oversees $2.3 billion as chief investment officer at United Nations Federal Credit Union in New York. “People kind of get skittish with yields when the 10-year gets to the December highs of around 2.6 percent or so.”
Note that last quote. Again: “People kind of get skittish with yields when the 10-year gets to the December highs of around 2.6 percent or so.”
He’s warning on “tantrum” risk.
On that note, we’ll close with one last chart which illustrates a very important point. Pre-crisis, the threshold for 10s beyond which the rates-stock correlation flipped negative (i.e. the stock-bond return correlation flipped positive) was ~5%. Post-crisis, it’s just ~3%. In other words: the “tantrum threshold” is materially lower.