1. China’s trade deficit.
Lots of people have linked to the most recent trade data from China, which show a deficit in March. One month may not be that important, as China is expected to swing back into surplus. But buried in the report is this almost mind-boggling statistic:
China’s exports totaled $112.11 billion in March, up 24.3 percent from a year earlier. Imports reached $119.35 billion, up 66 percent compared to the same period last year, the Customs Administration said in data posted on its Web site.
That’s right, those “mercantilist” Chinese just increased their imports by 66% in the midst of the worst worldwide recession since the 1930s. China generally has little effect on the US business cycle. But to the extent that it does have an influence, its impact since March 2009 has been strongly expansionary. March 2009 was when the big worldwide stock market rally began. And this tidbit from the FT suggests that China was under pressure to devalue the yuan back in the dark days of early 2009. They didn’t devalue.
Suppose you have many years’ worth of figures on a large number of economic indices, including inflation, employment and stock market prices. You look for cause-and-effect relationships between them. Bouchaud and his colleagues have shown that even if these variables are all fluctuating randomly, the largest observed correlation will be large enough to seem significant.
This is known as the “curse of dimensionality”. It means that while a large amount of information makes it easy to study everything, it also makes it easy to find meaningless patterns. That’s where the random-matrix approach comes in, to separate what is meaningful from what is nonsense.
In the late 1960s, Ukrainian mathematicians Vladimir Marcenko and Leonid Pastur derived a fundamental mathematical result describing the key properties of very large, random matrices. Their result allows you to calculate how much correlation between data sets you should expect to find simply by chance. This makes it possible to distinguish truly special cases from chance accidents. The strengths of these correlations are the equivalent of the nuclear energy levels in Wigner’s original work.
Bouchaud’s team has now shown how this idea throws doubt on the trustworthiness of many economic predictions, especially those claiming to look many months ahead. Such predictions are, of course, the bread and butter of economic institutions. But can we believe them?
To find out, Bouchaud and his colleagues looked at how well US inflation rates could be explained by a wide range of economic indicators, such as industrial production, retail sales, consumer and producer confidence, interest rates and oil prices.
Using figures from 1983 to 2005, they first calculated all the possible correlations among the data. They found what seem to be significant results – apparent patterns showing how changes in economic indicators at one moment lead to changes in inflation the next. To the unwary observer, this makes it look as if inflation can be predicted with confidence.
But when Bouchaud’s team applied Marcenko’s and Pastur’s mathematics, they got a surprise. They found that only a few of these apparent correlations can be considered real, in the sense that they really stood out from what would be expected by chance alone. Their results show that inflation is predictable only one month in advance. Look ahead two months and the mathematics shows no predictability at all. “Adding more data just doesn’t lead to more predictability as some economists would hope,” says Bouchaud.
In recent years, some economists have begun to express doubts over predictions made from huge volumes of data, but they are in the minority. Most embrace the idea that more measurements mean better predictive abilities. That might be an illusion, and random matrix theory could be the tool to separate what is real and what is not.
I’m in that minority of doubters.
4. Phoenix = Hong Kong
Thinking of the bubble as a Sunbelt phenomenon is a bad idea because it’s not correct, but also because it generates confusion over what characteristics were important in driving bubble inflation. So it’s important to note that outside of the Sunbelt, there were many other bubble markets, primarily on the east and west coasts—San Francisco, Portland, and Seattle, New York and Boston. What these markets all have in common, and have in common with Los Angeles and Washington, is that housing supply is relatively limited. So what emerged in these markets, initially, was a healthy price signal. This, incidentally, is how basically every bubble begins: with a healthy price signal. Demand for these coastal markets was high and rising, and housing supply was not keeping up. Therefore, prices rose. The bubble took shape thereafter, as rising prices combined with growing enthusiasm and rapid credit expansion, which fueled the growth of a bubble mentality.
Now, as prices rose, some housing demand shifted to other markets with strong local economies, including Phoenix, Atlanta, and Dallas. These markets tend to have very elastic housing supply, and so price increases translated into rapid construction, which prevented prices from rising and kept the bubble at bay.
Except that in Florida and the desert southwest, it didn’t. So has our housing supply model failed?
Not necessarily. As it turns out, you can “catch” a bubble from elsewhere. Migration to Las Vegas and Phoenix came overwhelmingly from Southern California. Residents of Los Angeles would cash out their homes and move east, buying one or two properties in cheaper markets, investing in those properties, and generally transmitting the bubble mentality that characterised the real estate markets of the California coast. Analysis of price movements has identified ripple effects from the Los Angeles property market to the Las Vegas property market, and thence on to the Phoenix property market. It seems likely that a similar phenomenon took place in Florida, which absorbed a great deal of migration from bubbly northeastern markets.
These “caught” bubbles were incredibly damaging, because they combined rapidly rising prices with rapidly rising inventory, leading to massive housing overhangs and price declines up to and greater than 50% from peak. But other Sunbelt metropolitan areas managed to avoid them, perhaps because they absorbed more workers from declining markets elsewhere in the south or northeast or midwest. Housing supply growth then prevented any big initial increase in prices which might have led to the enthusiastic growth in credit that triggered bubbles elsewhere.
I think this might be part of the story, but it doesn’t seem to fully explain the difference between Dallas and Houston on the one hand, and Phoenix and Vegas on the other. Consider this article from Demographia:
The Phoenix metropolitan area is sometimes erroneously characterized as having a responsive (traditional or liberal) land use market. In fact, the Phoenix market is highly prescriptive, as a result of the combination of strong land use regulations (“growth management”) and the large share of developable fringe land by the state of Arizona, which has been restricting sales to maximize revenues.
The state of Arizona owns a large share of the developable urban fringe land in the Phoenix urban area. The state has been auctioning land at a rate well below what the market could accommodate. This is illustrated by the large increase in prices per acre and in a comparison with agricultural land values.
In 2002, the average auction price of urban land was $32.600. By 2006, which was the peak of the Phoenix housing bubble, urban land sales reached an average auction price of $190,800.1 Rising land prices are the principal element of house price escalation in the Phoenix area over the period. As median house prices have declined in Phoenix (median house prices declined 39 percent in the year ended November 2008),2 average auction prices fell back to $68,600 in 2008.
Agricultural land in Maricopa County (the core county of the Phoenix metropolitan area) had a value per acre of approximately $8,500 according to the 2007 United States Census of Agriculture. Further, there was plenty of agricultural land, an amount in Maricopa County alone nearly equal to the entire urbanized land area of Phoenix in 2000. At the 2006 peak state auction prices, “raw”3 land was being sold at more than 20 times the value of agricultural land per acre. Moreover, the land ownership was highly decentralized, with nearly 1,800 farms. If “raw” agricultural land had been freely available for development, purchasers would not have paid such high prices for the land sold by the state.
. . .
Prescriptive Land Use Regulation and Price Volatility:
Not only does prescriptive land use regulation artificially increase house prices, but it also makes prices more volatile. Prescriptive land use regulation brings more chaotic “boom and bust” cycles to housing markets. They convert what would have otherwise been modest price bubbles into extreme price bubbles.
And it seems like a similar pattern occured in Las Vegas. Consider three types of housing markets.
1. Open markets like the central US, with lots of privately-owned, low price farmland that can be easily developed.
2. Markets with very little available land, due to high population and geographic barriers.
3. Markets with plenty of open land, but government development restrictions that cause the price of developable land to greatly exceed farmland prices.
Of course even in case 2 (say San Francisco and Manhattan) zoning rules prevent skyscapers from being built in many low-rise neighborhoods. But Phoenix and Vegas are especially interesting cases. It seems that zoning restrictions caused the bubble in land prices. If there is any other explanation for why farmland sold for $8500 while developable land went for $191,000, I’d like to hear it. If not, then it appears that Phoenix and Vegas have the same sort of housing market as Hong Kong, which also subject to unpredictable government land auctions and frequent real estate bubbles.
Even if lots of rich Californians and New Yorkers had moved to Texas, it is hard to see how housing prices could have risen anywhere nearly as sharply as in Phoenix. Unlike the western US, almost all the land in Texas is privately-owned. And zoning rules are pro-development. A big housing bubble can only occur if developable land soars in value. But how could that happen in Texas, where there are many thousands of privately-owned farms in close proximity to its major metro areas?
My hunch is that the migration patterns cited by Ryan Avent were a necessary condition for the extreme bubbles in Phoenix and Vegas, but not a sufficient condition. Development restrictions were also necessary. One weakness in my argument is that I know little about Florida. Obviously land near the ocean is somewhat limited, but less so in flat Florida than in mountainous California. So I am not sure what category Florida belongs in. Does anyone know what the Florida land market looked like during the bubble years?
Just thinking out loud, how much smaller would the housing,I mean land bubble have been if local officials had auctioned land fast enough to prevent prices from rising significantly above nearby farmland? How much smaller would the 2007 subprime crisis have been?