Numerous classic cognitive errors are on display at once in these markets. We have “overconfidence,” with large bets being made on the basis of strongly-believed models and forecasts…but these are forecasts of the dynamics of a system whose configuration is distinctly unlike anything we have seen before, even remotely. What does a “taper” do to rates? How can we know, since we have never even had QE, much less a taper, before? How aggressively does it make sense to bet on the outcome of such a transition period, given rational-sized error bars on the estimates?
We also see naïve extrapolation of trends. TIPS go down every day, it seems, for no better reason than that “core inflation is low, and the Fed is no longer going to be maintaining as loose a policy.” Ten-year TIPS yields have risen 83bps since April 25th (5y TIPS, +107bps since April 4th). Ten-year breakevens have fallen from 2.59%, within 15bps of an all-time high, on March 14th to 2.03% – the lowest since January 2012 – now. What has changed? Our model identified TIPS as cheap to Treasuries (that is, breakevens too low for the level of nominal rates) and went nearly max-long when breakevens were still at 2.30%. It is some solace that this position has fared better than a long position in TIPS, but when markets simply follow recent momentum mindlessly it can be painful.
Year-ahead core inflation is priced in the market at roughly 1.50%, despite the fact that current core inflation of 1.7% is only at this level because of persistently soggy core goods prices (and core goods are much more volatile than core services prices). Meanwhile, although core services prices remain buoyant, housing rents have not even begun to respond to the sudden boom in housing prices. To realize the core inflation priced into the 1-year inflation swap, core goods prices need to remain low and rends would need to decelerate while a shortage of owner-occupied housing drives the prices of existing homes skyward. It is possible, but it would be a very unusual economic occurrence. As I have previously written, we are maintaining our forecast for core inflation in 2012 at 2.6%-3.0%; although we may tweak that lowers if next week’s CPI is disappointing, we will not be changing it dramatically. Based on both top-down and bottom-up forecasts, we think the inflation market right now is very wrong. However, in accordance with paragraph 1, above, our 80% confidence interval for that estimate would be quite wide. Still, we feel that most errors looking out at least one year are going to be in the direction of higher inflation, not lower inflation.
Now, our forecast relies significantly on the behavior of the housing market, since shelter is the largest share of the budget for most of us. There has been a lot written recently about how the rise in rates could shatter the housing recovery. But let me explain why I don’t think that will happen.
I remember reading many years ago in “The Money Game” by Adam Smith (a pseudonym) that “you make more money with good investing decisions than with good financing decisions.” At least, I think that’s where I read it. In any event, it is true: if you are creating the next Microsoft, it makes very little difference if you finance it at 2% or at 15%, because the investment performance will completely obliterate the cost of financing. And this is why higher rates, even significantly higher rates, will not derail the housing market while prices are rising at 10%+ per annum. A home buyer is clearly happier to borrow at 3% than at 5% (tax-deductible), but if the home price is appreciating at 10% per annum (tax free, for much of it, and tax-deferred in any event) then it is a home run for the buyer either way. What hurts the housing market is when the expectation of future home price changes goes from go-go to stop-stop. And, with most consumers concerned with inflation and recent price trends in the home market, this isn’t going to change soon.
Here is an illustration of the real-world response of housing to rates. This first chart is the Mortgage Banking Association’s Refinancing index, plotted against 10-year Treasury rates (inverted). You can see that the recent rise in rates is having a significant impact on refinancing activity.
And this next chart is a chart of the MBA Purchase Index, showing activity on mortgages related to new purchases of homes. Again, the 10-year Treasury rate is inverted. You can see that there is no meaningful correlation here; if anything, purchase activity has been rising over the past year while rates have also been rising.
So, rest easy: higher interest rates are not going to meaningfully impact the housing market, unless they go much higher. Indeed, homebuyers might reasonably believe here that there is a “Bernanke put” on home prices in the same way that investors (correctly) believed there was a “Greenspan put” on stock prices. The Fed (and for that matter the state and federal governments) clearly have responded and can reasonably be expected to respond robustly to a future home price bust. So why not be long real estate here, if your downside is protected…and in any case, is limited to your home equity?
And if home prices do not decline, then rents are not going to decline, and in fact need to accelerate to keep up with the previously-seen rise in home prices. That is going to cause core inflation to rise going forward.
One final note: over the next month or two I hope to put out a few more articles like the one I wrote on June 8th about equity returns and inflation, but focusing on other asset classes such as real estate, infrastructure, commodity indices, etcetera. But in the meantime, I wanted to point out one security to keep an eye on. It is one of only two inflation-linked bonds that is traded on an exchange with a daily price and reasonable bid/offer spreads. The symbol is OSM (for Bloomberg users: you need the <CORP> key), and it is a floating-rate inflation-linked bond issued by SLM Corp with a March 2017 maturity. The problem with this security is that it is very hard to figure out what its true yield is unless you have an inflation derivatives curve, and even harder to figure out whether the issue is priced correctly given that you own SLM credit. The recent selloff has driven the real yield of this issue to (approximately) 3.40%, which is obviously much higher than is available for TIPS. The bad news is that the bond is still fairly expensive given the spread that should exist for a SLM bond, but in terms of raw real yield to maturity there are not many inflation-linked bonds out there with that yield. I am not recommending this security, but mention it as a point of information for investors who may want to check it out on their own.