Nomination For A New Villain
It seems very strange to find myself suddenly on the more-optimistic side of opinion when it comes to the situation in Europe. That isn’t to say that I am optimistic in an absolute sense; I still think that the Euro will either fracture or dissolve, a number of banks will fail, be nationalized, or be zombied, and several nations will default or restructure debt significantly (although if the Euro disintegrates, then inflation in some countries may take the place of an overt default since countries will then control their own currencies and have that alternative). I think that still places me squarely in the long-term pessimistic camp. But I am more optimistic about the near-term trajectory of the crisis.
Maybe even “optimistic” is too strong. I suppose I just think that the huge ECB program is the best real attempt that the European policymakers have made so far to at least extend the game. Unlike all of the other measures that have been taken (or hinted at being taken, or promised and not taken), this is the first one that has at least a chance to push off the day of reckoning. Yes, as I wrote yesterday, there are lots of things that can still go wrong but there are at least incentives for the cartel to remain together for a while.
The solution so far looks to be just naked printing in everything but name. The take-up of the 3-year ECB lending facility today was an enormous €489bln, a number which will likely grow to be more enormous before it shrinks. And there is just no way that I can see that the ECB can sterilize the activity on such a scale. I could be wrong. I always could be wrong. But whatever this does for the ‘ol can, it certainly seems to represent another goose for the global reflation trade.
.
The other big news on Wednesday was the National Association of Realtors revisions to past Existing Home Sales and inventories. Sales from 2007-2011 were revised down 14-15% and inventories as well. The main ‘technical issue’ was that beginning in 2007 some houses and sales were double-counted when they showed up in multiple listings, and there were several related issues of that ilk (all of which, it seems, inflated sales). Changing the data doesn’t really change the depth of the crisis, but we now have a better idea of how deep the crisis actually was. But one Wall Street economist who shall remain nameless wrote this in his/her analysis:
“…the revision does NOT imply that the NAR has deliberately been cooking the books over the past few years, trying to paint an overly rosy picture of the housing market. There were genuine data issues to be overcome.”
Sure, and interestingly all of the data issues began in 2007. Hmmm, what else was happening in 2007? Oh, I remember – existing home sales were beginning to fall alarmingly (see Chart, source Bloomberg). The white line, which represents the original release for the pace of sales, actually rose in early 2007, right when the actual sales, based on the revisions, were falling off a cliff. The NAR at the time was assuring us that the market was healthy, and somehow their estimate of sales was suddenly off by a 1mm units/year pace and no one thought anything was funny with the data. Sure.
Bloomberg chart showing original release and new, revised level for Existing Home Sales.
The fact that home sales fell more steeply and more deeply than was previously thought would have been great information for policymakers to have had in, say, 2007 when the Federal Reserve was holding the Fed funds rate steady and considering hiking further. I wonder what would have happened if, data in hand, the FOMC had started easing 9-12 months earlier than they ultimately did. It’s an interesting thought experiment. Why is it only bankers’ heads that should be hoisted on a spit?
The “new” trajectory of sales has another implication. It makes it even more amazing that core inflation never declined. The housing crisis was significantly worse than we had thought, which makes the arguments of the classical economists even less defensible than they were. By all that is good and holy and Keynesian, the core price level – nearly half of which is housing – should have declined. In only two months out of the last 48 have prices declined on a month-over-month basis, and one of those was so marginal it rounded to 0.0. The only meaningfully-negative month, January 2010, saw a -0.13% decline in core inflation. The worst recession in 80 years had almost no effect on prices. Does that evidence count for anything?
Pardon the rant, but it continues to baffle me that this even remains a debate.
Tomorrow, I plan to write my next-to-last comment of the year, and look at one or two items from the Z1 data release that the Fed put out earlier in December. Ordinarily I get to it sooner, but the story has been constant for quite a while. There is one interesting little twist this quarter, and I want to bring it up before I forget it. The last comment of the year, which will come out sometime next week, will be my year-end portfolio-allocation exercise.
Leave a Reply
Michael, I really enjoy reading your posts. Keep up the good work. But for a while I’ve wondered about your allocation to TIIPS and other inflation indexed notes given that the inflation component is determined by the government CPI-U. This article brings to light the fact that the folks compiling the data have every incentive to keep “headline” inflation low, while real inflation is clearly raging; thus depriving the TIIP holder of the protection they hoped they had purchased when accepting negative real yields. Thoughts?
Hi Jeff. Thanks for the comment. This article concerns ‘errors’ committed by the National Association of Realtors, which has no affiliation to the government. That’s a very important point. The bureaucrats who produce the government’s numbers have jobs for life and generally have no incentive to monkey with the numbers even if the Congress or President would like them to! I have spent a lot of time with the inflation methodologies, and I am confident they are well-thought-out and measure what they are supposed to measure. I’ll also say that independent measures of inflation, such as the Billion Prices Project that I have mentioned before and Krugman (and this is probably the only time I will ever speak approvingly of Krugman) mentioned recently. http://nyti.ms/s1dY5x Real inflation isn’t raging, it’s in the 2-3% range. But there are reasons that people THINK it is raging, that have to do with patterns of cognitive perception – I wrote a paper (http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1661941) that will be published in the January edition of Business Economics on this topic.
How’s that for thoughts!? 🙂 Keep reading, and commenting! Thanks.
I’m looking forward to your asset allocation commentary.
After you mentioned USCI last week, I took a look at its performance versus GCC, a collateralized commodity fund which replicates the 17-member, equally-weighted CCI (Continuous Commodity Index).
From USCI’s inception date of Aug. 10, 2010 through Dec. 20, 2011, it turned in a 10.12% annually compounded return, versus 8.23% for GCC. Monthly volatilities were almost identical (20.66% and 20.86% respectively), so USCI’s higher return delivered a higher Sharpe ratio as well.
USCI outperformed in 11 of 17 months, so its consistency looks encouraging, though it’s probably too early to assert statistical significance. My calculations indicate that from June 1959 to March 2004 (the period covered in Geert Rouwenhorst’s paper), a collateralized CCI commodity portfolio would have returned 8.12% — well short of the 11.0% ‘equity-like return’ Rouwenhorst found for his equal-weighted portfolio.
As best I can see, Rouwenhorst’s arithmetic averaging of monthly returns introduced a positive bias, as compared to the CCI’s geometric averaging. Roger Ibbotson, in ‘Stocks, Bonds, Bills and Inflation,’ states that ‘One key concern with an equal-weighted index is the effect of bid-ask bounce … equally weighting these returns can produce a substantial upward bias.’
If a real-world portfolio can capture that positive bias, then all is well. I remain skeptical of Rouwenhorst’s claim that collateralized commodities offer an ‘equity-like return,’ as his results are on the high side of most studies. But if USCI’s selection methodology (which picks the 7 most-backwardated and the 7 highest 12-month momentum contracts from a larger candidate group) can deliver the consistent outperformance shown in historical backtesting, then the promise of ‘equity-like returns’ from collateralized commodities might actually be realized.
That would be quite a sensational result, since equity-like returns are hard to come by — particularly UNCORRELATED ones! If inflation and bond yields are tracing a great secular trough now, then collateralized commodities will be an essential asset class in the coming rising-yield, rising-inflation environment.
Great comment Jim! I haven’t independently confirmed Rouwenhorst’s figures, mainly because the time series of discontinued futures contracts are hard to come by. I will say that other studies have found similar ‘re-weighting’ returns but criticism of Rouwenhorst on this matter seem to focus on the fact that some of the rebalancing return he claims isn’t technically a rebalancing return. I don’t understand all of the intricacies of this argument but there was an article in a recent Financial Analysts’ Journal about it. Unfortunately I don’t remember the title of the article but it probably had ‘rebalancing’ or ‘volatility’ in it somewhere. 🙂
Oh, one other thing I ought to add – we should be equally careful of the term ‘equity-like returns,’ since most careful observers say the numbers bandied about for those returns omit a large survivorship bias and have other significant problems.
What I like about CCIs, in addition to the nominal risks and returns, is that they tend to have very good returns and risk characteristics in inflation-surprise events, which is when most other asset classes suck (to use a technical term).
Jiiiiimm…you didn’t buy 1 million shares of USCI today, did you? Looks like someone did. Actually, it looks like they SOLD 1 million shares, which is more fascinating.
Not me. But maybe some anonymous billionaire reads your blog and shadows you. 😉
Nor have I confirmed Rouwenhorst’s analysis with its individual components. Of the 34 contracts for which he listed returns, 17 are in the CCI and 17 are not. His returns for CCI-component contracts show an arithmetic mean of 11.48% CAGR, while the average for non-CCI contracts is only 11.16%. Thus, to a rough first-pass approximation, adding LME contracts (most of which, other than nickel, showed weak returns) and discontinued contracts did not boost the portfolio return, as compared to simply analyzing the 17-member CCI.
I’d like to run the full, detailed analysis, but it’s a low priority for now. And futures data is messy to work with, owing to rollovers. For now, I’ll stick with my homebrew collateralized CCI index, which is at least conservative compared to Rouwenhorst’s results.
Thomson Reuters publishes their own collateralized CCI index starting in 1982 — it’s the benchmark for GCC. But ‘CCI-TR’ appears to have a severe downward bias (to the tune of about minus 2.5% annually) from what I interpret as double-counting of negative roll return from contango. Consequently, it returned only 5.35% (CAGR) from 1982 to 2010 — hardly better than T-bills! To combine your technical term with vernacular English, it sucked real bad.
Mike, I am shocked that you think NAR knew better then the ’07 published Home-Sales-Gate numbers. Shocked; especially because everyone that I know in that business seems to be managing the fallout so well and honorably. You know, full disclosure on the shadow inventory, short sale transactions..etc… I think they are just misunderstood.
Ha! Yes, there really should be a “Realtor Recognition Day.” 🙂