False Signals

“All models are wrong. But some are useful.”

George Box

Human nature is a little bit like water, in the sense that it always looks for the path of least resistance.

When considering some infinitely complex system like the global economy - it is in our nature to search for patterns and heuristics to use as mental shortcuts.

But there can be no such thing as a fully reliable indicator. Some indicators can, with the benefit of hindsight, be attributed to coincidence - a danger that only becomes evident after the fact.

Particularly painful, if in the meantime, we have built an overreliance on “when this happens, that evidently follows”. Maybe. Maybe not.

Campbell Harvey is currently an economics professor at Duke University, and by any measure, an incredibly smart bloke. Back in 1986, he published a landmark paper observing a relationship between US interest rates and recessions in the US economy.

In a “normal” economy, I should expect to receive a higher return for lending out my money for a longer period of time. This seems logical.

If we were to plot the return available for lending money against the length of time we are lending it for, then we might end up with something that looks like this.

This kind of chart is referred to as a “yield curve”, as what we can see are the rates of return (or yields) available for lending money over different periods of time.

When the rate of return available for lending money in the short term is in fact higher than that available for “locking money up” for longer - this is referred to as a yield curve inversion. The yield curve in this case looking like so.

What Harvey found in 1986, was that any time long term interest rates (as defined by the rate available for lending money to the US government for 10 years) fell below short term interest rates (specifically the rate available for lending money to the US government for 3 months) - a recession had always followed for the US economy within six to twenty four months. And this matters for investors because the biggest and longest past stock market drawdowns have occurred during recessions.

At the time Harvey wrote his paper, the yield curve inversion had correctly predicted four out of four US recessions.

By the time the yield curve once again inverted on 18th October 2022, when the Federal Reserve pushed up short term interest rates to combat post COVID inflation, the indicator had been right eight times out of eight. It was 8-0 in American sporting parlance.

Like all sensible academic theories, there are a number of potential explanations for why this one “works”.

• If short term interest rates are rising, then consumers and institutions have more incentive to save money. Consumers spend less and this slows the economy.

• If I can get a higher return on my money by lending it out for a short period of time, rather than a longer period of time - why would I, or any other institution, choose to finance a long term project? Investment in infrastructure slows, along with economic growth.

• Or maybe, a yield curve inversion is a sign of the bond market “sniffing out” a recession before it happens. When recessions occur, the typical modern day response from central banks is to cut interest rates, and when this happens investors in long term bonds make (potentially quite a lot of) money.

  
  

Whatever the logic for applying Harvey’s indicator the point is this - there are many sensible justifications to do so. It just looks right.

But lazily applying rules of thumb can occasionally get us into bother because there is a first time for everything.

If you would have invested since October 2022 based on the presumption that the US economy was on the cusp of recession, you would have been completely wrong.

The US economy has kept on rolling just fine thanks, and the S&P 500 is up by a casual 42.5% in Sterling terms since the date the yield curve inverted.

Source: US Bureau of Economic Analysis via St Louis Fed. Chart shows US Gross Domestic Product (the most common measure of economic growth) from 2020 to today.

I am a huge proponent of relying on what has worked in the past, primarily because we have no choice - what has happened before is the best guide that we can have for what might happen in future.

But sometimes it really is “different this time”, unprecedented things happen more regularly than they should and the market retains a constant capacity to behave in completely counterintuitive ways.

This being the case, acting based on certain mental shortcuts can be a dangerous game to play. Sometimes weird stuff just happens.

For the record, Campbell Harvey has stated on several occasions that he does not believe that what he has discovered will be perfect forever and there remains a possibility that it is flashing a false signal this time.

How the world could do with more of this kind of academic candour, and less ego. I suspect we would all be in a far better place.

Past performance is not indicative of future returns. None of the above is intended to represent advice to any individual. If you have any questions about your individual situation, please consult a regulated financial planner.