Some Thoughts on Cycle Length

First, a quick note. I haven’t posted many updates this year because frankly, I haven’t seen a whole lot of mispricings. My last post on in late Feb suggested getting long duration, which has worked out OK. I think rates are a bit rich currently, but the backdrop suggests more downside is likely over the rest of the summer. I actually think equities are pretty fairly priced here. Prices had ramped up ahead of fundamentals earlier this year, but it seems like the fundamentals have now caught up.

I’ve been thinking a bit about how this business cycle could play out. Historically, falling unemployment has generated rising inflationary pressures, which causes the Fed to hike enough to trigger a recession. What’s different this time is that falling unemployment has not been generating much inflation at all. The scenario being priced in by fixed income markets is that this will cause to Fed to slow or significantly shorten its hiking cycle. That’s certainly plausible to me. I don’t think anyone can rule out a scenario whereby very low unemployment rates coincide with low inflation. After all, we are already seeing it in many countries across the globe like Japan and Germany. And then there’s the chart from Haldane at the BoE that purports to go back to the 1500’s: that suggests no relationship before the industrial revolution:

But there’s another possibility, which is that unemployment will cause inflationary pressures only when unemployment is low, but then the inflation will accelerate. i.e. the slope of the Phillips curve is non-linear. Goldman recently published a study of the relationship between inflation and unemployment by metropolitan areas, which provides some support for this hypothesis.

In addition to that, there is data in US history that support this. The time periods in the early 50’s and the early 60’s exhibited both low unemployment (in blue and inverted below) and low inflation. (cpi in red) That period ultimately led to the inflationary spiral that dominated monetary developments for the subsequent decade.

Now, obviously this is not the 1960’s. We don’t have the baby boom wave coming, and rather than an inflationary oil shock, we are getting a deflationary oil shock. But the data as well as the rhetoric from the current Fed suggests that – as long as unemployment continues to fall, hikes are likely to be forth coming. In fact, that is exactly what the Fed did in the late 90’s, despite inflation that is only a bit higher than current levels. After all, the Fed doesn’t know how things will play out – whether unemployment and inflation will completely decouple, or whether some relationship still exists. So it will most like try to tread a path in the middle, that will allow it to accelerate or decelerate tightening as evidence for either case unfolds. Put differently, if unemployment is at 3.5% in a year’s time, the Fed is likely to be hiking still, even if inflation is unchanged.

The lowest unemployment reading in US history is 2.5%. If we extrapolate the trend decline in unemployment since it peaked, we could be there in 2019 – just 2 years away.

The Fed will want Fed Funds to be at least positive then on a real basis, which means 2.25% nominal. That’s another 4 hikes, or roughly 2 per year. But the risk is clearly tilted toward more hikes. The SEP median dot is near 3% for the end of 2019, with the unemployment dot basically unchanged from the last reading and core inflation at 2%. That’s pretty undemanding. And I think that because of the analysis above, the dots are sticky. Well, at least stickier than dots were in the past – simply because of the possibility of accelerating inflation.

In addition, my estimate is that 3.5% Fed Funds is highly likely to tip the economy into a recession. (a very very big) If things play out as listed here, that suggests 2020 as the potential start of the next recession.

Now – how is this useful, even if this scenario somehow plays out? Put differently, how could assets perform if this expansion extends another three years? There aren’t a lot of precedents, but here are a few thoughts:

· China’s forced rebalancing away from credit growth could be delayed until then. But the impact of a US recession on China could be yuuuuge….

· Corporate leverage ratios could make all time highs. Stock buybacks could increase until then

· I don’t think a huge equity bubble is particularly likely. The psychological pain from the past two recessions are probably going to stick around for a generation – I think if another equity bubble will happen this time, we would’ve begun the see signs already. Certainly we didn’t get an equity bubble in the 1950’s. A corporate credit bubble seems more likely to me given the massive demand for debt, the easing of covenants, and the fact that most corporate credit ex-financials did OK the last time around

· But by bubble, I mean equities may not get to levels that result in a crash. But they certainly could get to a very high plateau and stay there. A model that extrapolates 10y nominal price returns suggest that current prices embed a 2% return annually. There’s a pretty good chance that if the cycle goes to 2020, that figure goes down to zero. But that could be ~30% higher from here in price terms.

Advertisements

5 thoughts on “Some Thoughts on Cycle Length

  1. Excellent observations/thinking. Thank you. Your case for a corporate credit bubble is interesting. With the ECB likely to reduce QE due to issuer limits, the BoJ stealth tapering, and the Fed preparing the way for a balance sheet unwind announcement in September, I’m starting to think about CB balance sheet reduction trades. The effect of CB’s QE on long rates is unclear to me, but there does seem to be a fit with credit spreads (Citi recently published a piece showing the fit between CB securities purchases and US IG spreads to demonstrate the point). It makes sense to me, since the closest substitute for government bonds is IG bonds, so many private investors selling bonds into QE likely moved into IG. I’d expect then that as net supply of government bonds available to the private sector increases with diminshed QE, credit spreads would widen out. Also, I observe that credit spreads have hit their tights before stocks have hit their highs in previous cycles.

    Curious whether you have a view on CB balance sheets and their effect on long rates and credit spreads.

    1. Thanks johno. That’s a really good question. I’ve seen that Citi chart as well, and it certainly does suggest wider IG spreads. I am a bit hesitant about putting trades on to capitalize on it, however. One reason is that the slow speed of balance sheet reduction suggest that spread widening may not appreciably offset the negative carry of widener trades. And as rates rise, it seems likely that corporate issuance slows down. But as you alluded, spreads are certainly quite tight and the risk is skewed toward widening.

      I think that the big central banks are well aware of the effects their balance sheets have on credit spreads, and they are likely to only reduce at a pace that limits widening pressures. My guess is that they would be happy to see some widening of spreads here at this juncture, but the whole point of caps is to basically give the market comfort that the risk is skewed toward slower reduction. That also makes it harder to structure trades to capitalize on that, IMO.

      Please feel free to disagree – I am quite sure I don’t have the whole picture on this…

Comments are closed.