The Outlook For Inflation
Retail Sales was softer-than-expected (+0.6%, +0.5% ex-auto, plus downward revisions, versus 0.8%/0.7% expected), Industrial Production stronger-than-expected (+0.8% versus +0.5%), and CPI a smidge above expectations (maintaining 0.8% y/y on core, and rising to 1.5% y/y on headline). More on CPI later.
"Close enough!" cried the equity traders, who subsequently put up prices 0.7% on the day, to 28-month highs in the S&P. Bond traders also felt the balance favored a stronger economy and faster price increases, but moved yields only a few basis points higher with the 10y note to 3.33%. Inflation swaps curiously softened 2-5bps despite the reasonably sunny outlook for carry; some traders and investors feel the inflation market is a bit frothy right now - which it is, but supply is tight and I am not sure I'd be very aggressive about shorting inflation-linked bonds even at these valuations.
With Friday's trading, the Jan-2011 TIPS have matured. The yield of the bond over its lifetime is a picture of the economy of the 2000s (see Chart, source MorganMarkets). The recession of the early part of the decade shows up clearly, as does the expansion from 2004-07; the dip in late 2007 as the developing recession became apparent was followed by the spike as Lehman collapsed and balance sheets became allergic to anything except TBills. It was clearly the buying opportunity of the last few decades in fixed-income, as the ongoing crisis was more consistent with sub-zero real yields, and the issue subsequently rallied nearly 700 basis points in 15 months from December 2008 to March 2010.
Jan 11s, we'll miss ye.
So now, let's talk about CPI.
Core CPI was +0.092% month-on-month; the annual rate of change rose slightly to +0.804%. I'd mentioned yesterday the potential for a surprise higher in the monthly change because of the reversal of last month's seasonal adjustment quirk (which, in November, held the monthly change down). This didn't happen, and my suspicion is that the main part of the effect will actually be seen in this month: January 2010's seasonally-adjusted change in core CPI was a rather surprising -0.14%, which accounts for most of the difference in the seasonally-adjusted and non-seasonally-adjusted year-on-year series. We will have to wait a month to see.
But that is just sharp-pencil trivia for the bow-tied set. The bottom line is that the year-on-year change in core CPI is now rising. The headline figure did surprise on the upside, printing +0.505% to put the year-on-year rise at +1.496%. This was accomplished mostly through the rise in commodities (gasoline contributed 0.37% to the headline number, so with core+gasoline you have almost all of the month's change), and more of that is to come over the next few months.
Perhaps more surprising is that the second-largest contribution came from housing, which added 0.08% to the overall figure and therefore accounts for just about all of the rise in core CPI. This is remarkable - it was the largest such contribution in years. This is probably a reasonable time for a step back and a re-think about the overall outlook for inflation.
The Outlook For Inflation
The basic pricing conditions in the United States at present are:
- Housing continues to drag on core inflation, but far less than it had been. The y/y change in Shelter is 0.447%, compared to 0.804% for the overall core. The main two components of Shelter are "Rent of Primary Residence" (that is, if you are a renter rather than an owner) and "Owners' Equivalent Rent of Residences" (that is, what is happening to the consumption value of your home, as distinct from its investment value). These two series are illustrated in the chart below. While OER is only up 0.282% y/y, Rent of Primary Residence has risen 0.798%: essentially in-line with core. The chart (source: BLS) illustrates that at least the near-term pricing dynamic has changed.
Surprisingly, housing is now contributing to inflation.
- Offsetting this surprising rebound in housing (Shelter is around 40% of core inflation), just about everything else has seen price inflation decelerating. The core-ex-housing number, which was cause for much alarm last year when it spiked near 3%, has plunged to just over 1% (see Chart below). Apparel (about 4% of core) has plunged from +2% to -1%, Recreation (about 8% of core) is contracting at -0.8%, Education inflation (about 4% of core) has shaved from 5% to 4%, and Communication inflation (about 4%) has gone from 0% to -1%. About the only major core element (other than housing) that hasn't decelerated very much is Medical Care (8.5% of core), but it also isn't accelerating.
To my surprise - although anticipated by my models - core ex-housing has regressed to the core number rather than the other way around.
- This is not wholly surprising. Two models of core inflation that I follow have predicted this declining trajectory of core inflation; both models predict a turn higher now. I wouldn't have expected core inflation to evolve this way - with housing inflation bouncing higher while core-ex-housing provided a drag - but the outcome is approximately in line with my models, which broadly suggested core inflation declining to around 1% until Q3 or Q4 and then remaining at approximately that level through mid-2011. In the event, core inflation has overshot on the downside somewhat, but not egregiously so.
- Commodities will continue to place upward pressure on headline inflation relative to core inflation at least for the near-term, and I think probably for the medium-term as well.
- Because of the general decline in the volatility of price changes, and in the dispersion of price changes (that is, most items are experiencing broadly similar inflation), inflation feels lower than it has for some time previously. With reported CPI around 1.5%, I estimate the "real feel" inflation at around 2.7%, which is nearly as low as it has been at any time in the last forty years except during the financial crisis (You can read more about the "real feel" inflation temperature here). Consumers are perceiving a broadly stable pricing environment, which is probably one reason that the stock market is doing so well and confidence is rising. Gasoline, however, could change all that: the real-feel jumped from 2.4% in November.
Aggregated models of core inflation and one which separately forecasts housing inflation both project rising inflation going forward. The average of the models is 1.5% for 2011 core inflation. Interestingly, the model that separates out housing inflation projects 1.8%, but then converges in 2012.
The main influences on the 2011 inflation outlook are the late-2008 spike in money supply, the decline in the dollar over 2009, and the level of core inflation in late 2010. It is important to realize that there are long lags in the pass-through of monetary policy to inflation. Most of the broad currents of 2011 inflation have long since been formed. Policymakers are right now working on policies that will influence 2012 and 2013 inflation, but not much will drastically change the outlook for core inflation in 2011.
Risks to the Outlook:
However, the model is not "fully specified." That is, not every possible influence on inflation is included in the model - mostly because there are all kinds of influences that I can't imagine, or because some influences operate with variable lags. For example, my models do not include consumer expectations of inflation. Partly, this is because I don't think they have a lot to do with inflation, but the Federal Reserve thinks they are very important and if they're right, my model will miss inflation zigzags that are caused by changes in consumer expectations and not captured in other variables. I also do not model changes in money velocity directly, but this can be as important or even more important than the level of the money stock itself.
The risks to the outlook have been generally to the upside over the last year or so, but seem to be somewhat more balanced now that core ex-housing has decelerated. Of course, when we are near the lows we should expect that inflation feels saggy and that the risks seem more balanced. Here are what I see as the main risks to the outlook.
My models do not have a role for the output gap. In my econometric work, I haven't found that the output gap adds any explanatory power to a model that already includes monetary variables (and I'm not the first person to notice this: Fama in 1982 is the first person I'm aware of to show it). But this may have happened if I looked for a linear influence, while the influence is actually nonlinear - that is, maybe small output gaps don't matter but large output gaps, about which there is limited experience in the data sets, matter. If this is the case, then the large output gap we currently have will dampen inflation more than I expect it to.
In my models, leverage plays an important role. High levels of private leverage tends to dampen inflation; moreover, since leverage is related to the velocity of money, stable leverage tends to imply stable money velocity and thus a stable quantity of money. After signs in 2008-09 that an important deleveraging trend was underway, that trend slowed in late 2010 as the Federal Reserve has worked hard to maintain the degree of systemic leverage. I think this is a really bad long-term idea, but in the short-term there's no question that a collapse of leverage could lead to extremely bad growth outcomes. From an inflation perspective, though, high levels of private debt relative to public debt tends to dampen inflation, so by encouraging leverage the Fed is somewhat inhibiting medium-term inflation. If the deleveraging trend turns back into releveraging, it will tend to truncate some of the more bullish inflation scenarios.
Money supply growth has remained tepid although it is accelerating, but the Federal Reserve is clearly trying to increase money growth. As I noted before, normal money supply growth passes into inflation with a long lag but if there was sharp money growth then some of that increase would likely pass through into inflation in a shorter time-frame (quantitatively, my models assumed fixed lags but in reality the lags ought to be distributed...that is, the pass-through happens over time. When the changes are small this is not terribly important but for large changes we could feel some effects sooner). Now, the Fed doesn't want a sharp increase in transactional money (e.g., M2) even though it is pumping massive amounts of liquidity onto bank balance sheets. As I wrote several months ago (see comment here), we don't know how or if that money multiplier will return to normal. If it doesn't happen until the Fed withdraws the M0, that produces one outcome; if it suddenly snaps back then that could cause a massive, uncontrolled rise in M2. That isn't my null hypothesis, but we really don't know how this will work (no matter how confident the Chairman is about it) and the direction of the risk is quite clear.
The Fed, despite protestations to the contrary, has no way to unwind liquidity provisions in a rapid, yet orderly way and likely would be slow to do so even if inflation began to rise. This doesn't affect 2011 inflation, but it implies that the medium-term trajectory for inflation probably needs to take into account the fact that the Fed will not be leaning against the wind in its normal fashion.
Majority political preferences all favor higher inflation because it seems to solve some problems (at least at moderate levels of inflation) at a low cost. Therefore, most political decisions are likely to be made without particular concern for the effect on the pricing dynamic.
The dollar is unlikely to rise sharply (although it may rise), but conceivably could fall sharply if public debt grows too onerous. See Iceland, whose currency fell 50% against the dollar in 2008. There is little doubt that Greece's and Ireland's currencies would have been similarly penalized had they not been on the Euro (incidentally, this is also the reason that Iceland will recover long before Greece or Ireland - the latter countries lack one crucial automatic stabilizer). It isn't likely to happen here, but it could.
I mentioned above the risk that private leverage reasserts itself, but the flip side of that risk is the risk that public debt simply explodes. What really matters to inflation is actually less the amount of debt but whether that debt is largely private (which is disinflationary) or public (which tends to be inflationary). For a long period of time, the ratio of Federal, State, and Local debt as a proportion of total debt has been fairly stable between 22% and 37%, but the huge deficits of the last couple years combined with whatever private deleveraging there was has pushed that ratio to the upper end of that range (as of Q3. When the Q4 numbers come out it will be shocking if we aren't at new multi-generational highs of that ratio). Even if deleveraging stops, the increase of the public debt creates ever-larger incentives to inflate it away. This wouldn't likely manifest as a 2011 risk except through a currency collapse as mentioned above, but it is a very large problem for the medium term. I've run charts before but the Financial Times had an excellent article on the debt problem on Thursday. Below you can see the key illustration from that article.
Source: Financial Times, Jan 13, 2011
I'll end there. The bottom line is that the 2011 trajectory will be to roughly a doubling of core inflation with a good chance of headline figures outpacing the rise in core (and perhaps significantly). The less-stable pricing environment will also likely produce a larger jump in perceived inflation. There are both upward and downward risks to this forecast, but the upward risks in my view predominate - especially in the medium- and long-term outlooks.
The bond market is closed on Monday for the Martin Luther King Jr. Day holiday, but on Tuesday the Empire Manufacturing (Consensus: 13.00 from 10.57) report is due out. The next regular installment of this comment will be on Tuesday evening.