Going backwards to go forwards

Managing our medium-term risk exposure over the economic cycle is a key feature of our investment process. Our rule of thumb is simple: stay long risk and buy market dips from early to late in the cycle, which can help us avoid the fallout of the early recession phase and help prevent us from turning defensive too early in the cycle out of ‘reckless prudence’.

One of the key tasks for our economists is therefore to assess where the economy lies in its cycle. Today, we believe the US economy is early in the cycle but could be mid-cycle by the year’s end.

We nevertheless test our theses rigorously as part of our ‘prepare, don’t predict’ mantra. Our research team recently engaged in one such exercise, reverse macro mapping, which essentially involves running our usual process backwards. So instead of starting with the economic analysis and considering potential market implications, we put that aside and asked: “What are asset prices telling us about the economic cycle?” The answer to this question is found by observing dynamics in recent asset prices and considering when in the cycle such patterns have occurred in the past. This can provide a valuable cross-check on our economic assumptions.

So via a quantitative theory called Bayesian inference, we map the current performance of 10 asset classes against their historical performance – going all the way back to 1973 – in the respective stages of the cycle. Based on this, one can determine what asset prices are telling us about where we are in the cycle.

The answer is somewhat of a surprise: this analysis suggests there is a 33% chance we are indeed early cycle, a 47% chance we are already mid-cycle, a 19% chance we are late cycle, and a 1% chance of a recession.

Is it getting late early?

This challenges our fundamental ‘early cycle’ view, but there are a few caveats. Macro mapping looks at unemployment slack, inflationary pressures, and policy. All of these ordinarily move slowly during recoveries, so markets could just be recognising the faster speed of this cycle, looking three to six months ahead and pricing on that basis.

The strong recent performance of commodities is another interesting facet. It is usually more common late in the cycle, but reverse macro mapping may be biased due to the ‘China effect’ in commodity returns as the world’s second largest economy is already later in the cycle.

Finally, this is only the second cycle of the quantitative-easing era. Past recessions have often been caused by inflation, with inflation and bond yields typically falling in the recovery phase. With central bankers worried about getting stuck at the effective lower bound, and continued quantitative easing, the reaction of bonds in the recovery phase could be different in this cycle.

There is more work to do on reverse macro mapping, but we need to be aware this signal is challenging our fundamental research. We may be further in the cycle than people think.