The market is currently pricing just ~0.35% of hikes over the next TWO years. The infamous Fed dots imply ~1.75% of hikes. Even with the knowledge that the Fed dots have been way too high the past several years, that gap is very, very large. The consensus view seems to be that the market has lost faith that inflation will return to anywhere close to 2%.
But that doesn’t really seem to be the case. According to Bloomberg, the median private sector economist projection for core PCE at the end of 2018 is 1.8% vs the Fed at 2.0%. 1y1y inflation forwards are pricing in a headline CPI of roughly 2.0%. So I have a very hard time believing that the consensus actually think a lack of inflation likely.
In fact, betting on more hikes appears to be a popular or even crowded trade now, with Bloomberg actually running an article this morning entitled: “Bond Market’s Hot Trade is Betting a 2017 Fed Hike Still in Play.”
So if the market doesn’t believe that the Fed will stop hiking, and may in fact be betting for more hikes, the big question is: why is bond market pricing so different?
There isn’t a smoking gun type answer to this one, and probably never will be. But my hypothesis is that there are two primary factors: the shortage of risk-free bonds, and their insurance properties. Neither hypothesis are all that novel, but perhaps some of the conclusions will be of interest.
The bond shortage argument has been around for a long time. Many banks have tables showing annual net issuance of risk-free bonds, after accounting for QE programs, relative to demand from private sector investors like pension plans and the like. That gap has been very wide, and in addition has recently exacerbated by the increase in USD FX reserves. (Foreign central banks tend to park their USD reserves in US treasuries, mainly at the front end of the curve. The recent rally in EMFX has allowed many of them to replenish their reserves, which are now back to mid 2016 levels)
In addition, since the 1990’s, risk free bonds are essentially the only asset class that is negatively correlated to risky assets. In other words, they provide insurance for the rest of a portfolio. An in a world where asset prices are universally deemed expensive, the price of insurance should naturally rise.
In fact, insurance, as we all know, typically costs money. So if owning treasuries are akin to owning insurance, should we be shocked that owning treasuries… ultimately costs money, relative to cash?
To some extent, the various term premium models out there showing negative readings across the curve already says that. Consensus views seem to remain that term premiums will revert to above zero over the long run, after QE programs are unwound, etc. But if QE becomes a recurring part of the central bank tool kit, as most central bankers assume, it may well be that negative term premiums are here to stay.
To take the insurance analogy a bit further – historically insurance pricing fluctuates such that during quiet periods, insurance companies compete to drive down the price of insurance to uneconomical levels, until a natural disaster or other shock causes the insurance market to reprice as the weaker players are forced to step away from the market. Perhaps we could see a similar dynamic in the treasury market – long periods of notably negative term premiums, until a shock (like an a Fed hike) causes a temporary repricing to more justifiable levels.
For the record, I think the consensus view on inflation is right. My own inflation model, which only puts a minimal weight on the slope of the Philips curve and has done a decent job of forecasting inflation out of sample, suggests Fed and consensus forecasts over the next several quarters look pretty reasonable, and if anything could be on the low side.
In my last note in mid June, I noted that rates looked rich but it wasn’t time to fade it yet, which turned out OK over the longer run. Unfortunately that was immediately before the Sinatra conference, which caused a temporary spike in yields, which has since been more than unwound. At this juncture, I think there is enough uncertainty at the Fed about inflation that they will want to see more than one strong inflation print to hike in Dec. There are only 3 more core PCE prints to be released between now and the Dec FOMC, and 4 CPI prints. So my guess is that the inflation data will need to start strengthening very soon for a Dec hike to be realistic. In addition, since the recession there’s been a seasonally bearish bias to the MoM core PCE prints in the last few months of the year. So my personal take is that current market pricing for Dec isn’t all that crazy.
Otherwise, macro pricing doesn’t look terribly mispriced at this juncture –
· Credit spreads look a bit tight, but there’s nothing that suggests a reversal is likely. I’ve been hearing that underwriting for leveraged loans have been weak given strong demand, but that could persist for some time.
· Spec positioning on the dollar is starting to get stretched, but I don’t think it’s a binding factor – yet. I don’t think the USD is terribly mispriced – fundamentally, the trend turned in 2014, IMO.
· Equity prices still look pretty reasonable here. I don’t think a Fed hike or QE unwind will cause a major correction. A major reason is that the long end of the curve doesn’t look too out of whack.