Obviously, the British referendum is dominating the headlines. I don’t really have much to add to the narrative. What I want to do today is discuss the broader backdrop, which I expect is likely to play out eventually regardless of the referendum outcome, since direct economic linkages to non-EU areas suggest a limited role for contagion. (Though of course that risk is skewed to the downside)
I think the big questions are:
- Is the US employment slowdown real, and will it persist?
- Why isn’t US growth & inflation reacting more to low yields?
US employment has undeniably slowed from year ago levels. There is some debate as to why that is, and how much of the slowdown is noise. My take is that this is a result of the US growth slowdown over the past 6 months. (payrolls in white, ISM composite in orange as a proxy for growth)
Nevertheless, the last payrolls print very likely overstated the weakness in the labor market, even if payrolls growth slowed to a 1.4% annual rate (which would be the lowest since mid 2011) monthly payrolls should average 165k. In addition, there are some divergences in various labor market indicators. Layoffs, separations, and jobless claims remain low, but help wanted, new hires and new jobs growth has slowed. This corroborates a picture of slowing, but still positive growth. Historically, when that occurs in conjunction with tight monetary policy and financial conditions, the economy tips into a recession. But in this case, financial conditions has eased significantly since earlier this year. The implication is that unless the economy ‘tips’ soon, the easing of financial conditions is likely to support growth over the next couple quarters. A final point to support a low likelihood of a recession: the percentage of US adults with subprime credit scores is at the lowest level since 2005.
Another interesting facet of this environment is that growth does not seem to be responding to low yields. Secular stagnation has been mentioned a number of times. The CEA just published a report noting the consistent decline in male participation rate in the labor force since the 50’s, and with the decline especially sharp for men with a high school degrees or less. In addition, nearly 36% of primate age men not in the labor force lived in poverty. That certainly suggests some aspects of the stagnation hypothesis are at play.
In addition, the combination of low unemployment and low inflation is unique. The implied flattening of the Phillips curve is likely one major reason the Fed has downgraded its inflation forecasts in the dots. I think one driver there is that mortgage availability along with high house price to income ratios has sharply blunted the real estate transmission mechanism. With real estate prices a large portion of indirect costs in most services we consume, this has likely been a reason inflation has been contained – not just here, but globally.
A net result of that is that long term inflation expectations have fallen as fewer people expect inflation above 2.5%. This lends credence to the view that the inflation term premium may not return to substantively positive levels anytime soon. (Though it seems to be quite negative currently)
That segues to the topic of neutral rates, which has gotten a great deal of attention the past couple months. Yellen devoted a fair bit of time to it at the last FOMC presser. The vice president of the ECB also recently spoke about it. The Fed recently published a paper estimating the neutral rate for the EU, UK and Canada. In general, it seems that the consensus is that neutral rates are low globally as a result of various factors and do not show signs of going up anytime soon. My own take on it is that in addition to the often mentioned pressures of aging, which lowers growth, a likely additional factor is the very large non-petroleum trade imbalances in the world economy. Specifically, the very large Chinese, German and Japanese trade balances vs the US is effectively sapping domestic consumption and raising foreign saving. That’s a trend that seems to be ongoing, though highly unlikely to persist for very long. In the meantime, the savings glut is not being offset (as it should) by fiscal expenditures. The net result seems to be very low neutral real rates. (about 0.5%) Even relative to those estimates, however, current Fed policy is quite simulative. (2y real yields are roughly -0.9% currently) The offshoot of that, however, is that growth rates of 2.0% is likely above trend. A normalized rate, say 0.5% real implies 4 hikes from here, which is likely to take the 2y to 1.75%… which is above where the 10y is currently.
The paper above also estimated neutral rates for the EU. The estimation results imply that EU real rates are currently only neutral, rather than stimulative, despite the low nominal figure.
The paper did not produce any estimates for Japan, but the BoJ tried calculating it late last year. Like Europe, the estimated real neutral rate was negative, roughly -25bps at the 2y pt. This suggests that at current levels, JGB yields are only neutral rather than stimulative.
Let me put that differently. Neutral, or ‘natural’ rates in Japan is probably supposed to be negative. Which means the current policy setting may be too tight! Yes, I know all the arguments that negative rates causes a deflation mentality and so on. But if you think that over the long run, yields are supposed to reflect the economy’s inflation and growth rates, then why is it wrong that Japan’s falling population (-0.5% a year over the next 10y and -0.7% over the next 30) and minimal inflation should be matched with negative nominal yields? In fact, I think the idea that current BoJ policy is too tight is one of the reasons for the Yen strength we’ve observed this year.
To be clear, the BoJ may not be able to ease policy much here, even if it thinks policy is too tight. In fact, it is interesting to me that so many analysts expect additional easing from the BoJ here, but none have proposed any idea that seems likely to work. Monetizing debt will not do much, as the NY Fed has observed.
With the ECB and BoJ not able or willing to do much more, currency movements will be increasingly driven by the Fed and economic fundamentals. Since both the EU and Japan have current account surpluses vs the US, that means the Fed needs to tighten policy just to keep the crosses in place.
With respect to China, I find it interesting that the recent jump in credit/money creation has had a minimal impact on growth. The chart below shows CNY M1 growth in white vs real GDP growth in orange, and the ratio of the two in the lower panel. M1 growth has been at a rate not seen since the great recession – and the lack of an uptick in growth the past couple quarters is striking.
With growth no longer responding to credit creation, an eventual slowdown seems inevitable.
I noted last August that China is likely to try to depreciate its trade weighted currency by about 20% over a few years. The CFETS index has already depreciated by ~7% on a nominal basis since it was published in November, or roughly 12% annualized. With the growth backdrop, a continuation of trend seems likely, but the probability of the weaker currency supporting domestic growth is low given both China’s size as well as the limited appetite for more trade deals by the G7 electorate.
I listed a number of reasons that argue for lower yields above. But there may be one more reason for yields to be low, which is that risk free assets have become increasingly negatively correlated to equities. It makes sense that insurance costs money, and on that logic, the better treasuries function as insurance for an equity book, the richer the price is likely to be.
Despite that, and despite the other fundamental reasons I listed above, I still think yields are too low. I think current yields have already discounted many of these effects, and in any case it’s not clear how those factors will shift over time. I.e. will the negative correlation actually persist in the US? Or are current levels an artifact of the oil sell off? Could correlation move towards zero as inflation picks up?
What is clear, however, is that inflation forwards seem unlikely to be realized. 1y Inflation swaps are currently just 2.0%, and the 1y1y forward is just 1.5%. That’s with core CPI at 2.2% currently. In fact, the lowest core CPI print over the past 5 years is 1.6%. A sub 1.5% outcome only seems likely in the event of a recession.
Furthermore, a sharp further decline in yields here is likely to be self-defeating. Currently, 30y mortgages are ~15bps above the 2012 lows. If mortgage rates were to make new lows on the back of Brexit, Fed, overseas buying, whatever, and holds there for a bit, there will likely be a decent response from the housing sector in the form of a refi surge / housing price increase.
Finally, spec positioning in futures suggest long bond futures positioning is quite stretched to the long side.
IMO, the backdrop I presented above is a good backdrop to get long equities after or potentially just before the British referendum. Now obviously, a Brexit vote is likely to coincide with a risk off. And obviously there are headline risks around the US election. But if we look through that, I think a brighter picture emerges.
- As I noted earlier, given the easing in financial conditions earlier this year, growth is likely to turn up from here.
- Higher year over year inflation is already baked in
- Earnings expectations is likely to pick up as well. In fact, they have already started moving up – the 3m chg have hit the highest levels since last summer
- Current valuations vs fixed income are the most attractive they’ve been in a year. S&P dividend yield at 2.2%, vs 30y swap at 2.0%.
- Sentiment & positioning is still bearish. But it is beginning to turn. A break to new highs is likely to drive quite a squeeze. The chart below shows the results of the most recent BAML survey. Equities weighting is at levels last seen in 2011 and 2012.
- On account of the easing in financial conditions and the passing of inflation & currency shocks, recession risk over the next 12 months is reduced. If the economy was fragile enough to tip into a recession, the shocks over the past 12 months should’ve done it.
I’m aware of the various argument for an upcoming recession and/or bear market. Little available slack in the labor market, profit margins have peaked, Fed is tightening, etc. The fact is, however, that the reason why recessions have historically followed those factors is due to tightening fed policy / financial conditions. The Fed’s reaction function this time is quite different. Which means the lag between those indicators and the next recession is likely longer than usual also.
Finally, there are several similarities to late 90’s here, after the Fed was dovish after oil & EM shocks. Imagine the narrative when the S&P makes new highs, and managers are underweight equities, and 10y is still below 1.75%. What do you think they will do? Bonds are arguably already in bubble territory. Why can’t equities be also? (Yes, the Fed did note that forward p/e ratios are well above their 30y median. But then Greenspan did speak about ‘irrational exuberance’ in stock prices… in 1996. And if you have other concerns that risk taking is blunted – apparently someone is hovering up leverage loans)