One of the most important challenges for the individual investor is when (and how much) to worry about the Fed. The swirl of controversy about interest rates and the implications for the economy is roiling markets. Time for some clarity. I’ll describe the sound conclusions we should draw, and in my next installment turn to some of the less reliable opinions making the rounds.
Fed Decisions. The decision and Chairman Powell’s news conference provide reassurance on several key points---the likely path of future rate hikes, timing for the balance sheet runoff, and confirmation that the Fed is paying attention. Top Fed watcher Tim Duy, describes this change, noting that the traditional Fed model of employment and inflation is not working. He writes as follows about the importance of this change:
It is hard to understate the importance of this shift. The Fed’s models haven’t worked this way in the past. In previous iterations of the forecasts, the expectation of unemployment remaining below its natural rate would trigger inflationary pressures. To stave off those pressures, the Fed perceived the need to raise rates above neutral to slow the economy enough to nudge unemployment upwards. Now the Fed believes it can let unemployment hold persistently below the natural rate without triggering inflation and without Fed policy becoming restrictive.
These announcements are bullish for stocks.
Yield Curve Research and Statistics. Let me begin with the most important point.
There is no way to do a sound statistical analysis of recessions.
The problem is the number of recessions is too small – way too small. If you reach for more history, you are just collecting less relevant data. It is a great subject for study but does not lend itself to normal statistical approaches. The problem is the statistical concept of degrees of freedom. Simply put, the number of possible independent variables is too large compared to the number of cases. You can achieve perfect explanation, because you can always add another variable or adjust the criteria.
Even when you stick to the yield curve, there are actually scores of variables.
- You can choose the slope between any two points on the curve – about fifteen choices.
- You can focus on rate of change rather than level, so double the fifteen.
- You can ask whether an inverted yield curve is a bull flattener or bear flattener, which doubles the 30 cases.
And we are just getting started.
Is there an answer?
Yes, but it is not purely statistical. Taking the statistical approach allows for retrospective perfect results, but there is little predictive value. Recognizing the importance of this question, and suspicious of the widely accepted message, I embarked on a quest to find the best approach. I describe this as well as many other important facts about recessions at my blog in Recession Forecasting and Misinformation. The key aspect of my search was that I was not seeking perfection in results – a sign of overfitting. I wanted excellent results with these characteristics:
- Openness — with the potential for peer review
- Small number of input variables. Most people do not understand that “small is good.” If you have a lot of variables, it is easy to do back-fitting on a few cases. Beware.
- Real-time performance. This means that you do not go back in history doing any data-mining. You create an indicator and live with it through time. (While the ECRI predictions are a matter of record, no one knows what changes they have made in their indicators).
As a conclusion to the research I described the work of Dr. Robert F. Dieli and implemented it in my own analysis. Dr. Dieli uses only four variables – each important and all combined in a way that makes intellectual sense. Two variables represent the yield curve – Fed Funds to Ten-year note. Two represent the economy and inflationary pressure—unemployment and inflation.
If someone described why these were important variables, it would make intellectual sense. It was a hypothesis created and now used in real time for thirty years or so.
And crucially, Bob treats the results of his analysis, a constant, nine-month look ahead, as like a weather alert. It is a signal to examine potentially confirming data.
Is Anything Different This Time?
I know, we are never supposed to admit this possibility. There are three important differences, stated briefly at this point.
- Unemployment is not giving the traditional signal of labor tightness. Whether it is because of marginal workers returning, or boomers working longer, there apparently is still some slack in the labor market. This reduces inflationary pressure and the need for pre-emptive Fed action.
- The carry trade, for the first time in my long experience, is using US bonds as the “funding currency.” This means that traders can borrow at low or negative rates and lend by buying treasuries. Since this is a pairs trade, it can be done with significant leverage. There are risks, of course, and I’ll take those up at some point. Right now the biggest risk is an adverse currency more, which can be hedged.
- QE in the aggregate reduced the ten-year rate by about 1%. As it is unwound, there is some upward rate pressure.
The most important conclusion is that the stock market is reacting to a questionable signal from a questionable indicator. Traders reacted last week to the Fed’s assurance of rate stability by buying the two-year and five-year sections of the curve. This made sense as a fixed-income trade, but it has no message for stocks.
A secondary conclusion is that long rates are being held down by the arbitrage with Europe. This makes them less useful as a US economic signal, whether in their own right or as a part of the yield curve. One could argue that it is signaling European weakness.
The Investment Message
Investors have an opportunity to enjoy the last part of the economic cycle – often a real surge. You can’t do that from the sidelines or by owning utility stocks. The best stocks for the next two years were unloved last year. It is a classic repeating market cycle.
This means a focus on non-FAANG tech, homebuilders, and some financials. Since index funds are market-cap based, the overpriced stocks are over-represented. Why not focus on those that are historically cheap?