Inventory of Unsold Homes Suggests that San Francisco Market Will Remain Weak

The inventory of unsold homes has historically been a good predicter of near term housing market performance. I addressed this topic in an earlier blog posting. For California as a whole, the historical relationship seems to have broken down recently. The breakdown probably resulted from banks dumping foreclosed homes onto the market. (I'll address this shortly, in another blog posting.) San Francisco has had relatively few foreclosures, however, so the historical relationship has held up pretty well.

Take a look at the chart, below, which compares the beginning-of-year inventory of unsold homes in San Francisco to the change in median price during that same year.

Inventory is stated as the number of months required to sell the current stock of for-sale housing, assuming that selling activity remains at current levels. I plotted it using an inverted scale in order to highlight the correlation with price changes.

Real estate practitioners often assert that the housing market is "in balance" when there is six months worth of inventory. The chart suggests, however, that the equilibrium level of inventory for San Francisco may be less than six months. Since 1997 (the beginning of my data series), the average inventory in San Francisco has been only 3.6 months, compared to 5.0 months for California as a whole. And although San Francisco began 2008 with just over six months worth of inventory, prices fell almost 20% during the year.

The San Francisco market had 5.8 months worth of inventory at the beginning 2009. That represents only a modest improvement over 2008. If I had to make a prediction based on this indicator alone, I'd say that 2009 is shaping up to be another weak year for the San Francisco housing market.

Unemployment Still Increasing Sharply

The national unemployment rate reached 8.5% in March. That's almost a full percentage point higher than the January reading of 7.6%. Ben Bernanke said at the time that unemployment would climb to 8.0% "for sure." Presumably, he felt that a higher figure would have sounded outlandish. In retrospect, he was being overly cautious.

The Bay Area unemployment rate has been climbing rapidly as well. It hit 9.9% in March, compared to 8.8% in January and only 5.1% in March of 2008.

It's hard to see home prices stabilizing when the job market is deteriorating so rapidly.

Some Perspective on Housing Vacancy Rates

The cover story of last weekend's edition of USA Today was entitled, Open House Anyone? One in Nine Homes Sit Empty. The nation does indeed have a lot of vacant homes, but it's not as bad as the title suggests. Even in 'normal' years, there are plenty of vacant homes. Take a look at the chart, below, which shows historical vacancy rates for the United States and the western region.

The blue line shows the vacancy rate for the US, and the purple line shows the vacancy rate for the western region. The numbers (and the definition of 'western region') come from the Census Bureau. The figures in the USA Today article excluded vacant homes that are used seasonally (such as beach houses) but the chart includes all vacant homes.

As the article suggests, the current vacancy rate is unusually large, at around 14.5% (which is actually closer to 1 in 7). But even in normal years, plenty of homes sit empty. From 1987 to 2000, the vacancy rate was fairly stable at around 11.5%. That's equivalent to 1 in 8.7, which is still higher than the ratio that USA Today was hyping in its article.

Setting hype aside, a three percentage-point increase in the vacancy rate means that almost 4 million units were added to the pool of vacant homes between 2000 and 2008. To put that in perspective, the national housing stock increased by roughly 11 million units over the same eight-year period. In other words, roughly 1/3 of the new homes that were built since 2000 are essentially superfluous.

Another way to get a handle on the vacancy rate is to consider the rate of new household formation. Lately, it's been running at around 1.5 million per year. Not all of the 4 million additional vacant homes are available for new households (roughly half of them are second homes or vacation homes), but there still appears to be more than a year's supply of excess vacant homes in the nation. Home prices will remain under pressure (at a national level) until those excess units have been absorbed.

While we're on the topic, I thought it would be interesting to plot the national vacancy rate on the same chart as the Case Shiller composite home price index.

It's no coincidence that the vacancy rate started rising at the same time as the housing bubble began inflating. Rapid price increases contributed directly to increasing demand for homes -- that's more or less the definition of a bubble. If the price increases had been driven by true demand for housing, we wouldn't have so many empty homes now.

Credit Spread Model Predicts Massive Job Losses

I find a lot of interesting ideas in the Wall Street Journal's "Heard on the Street" column, which is published daily on the back of the Money & Investing section. Today's lead article, Giving Corporate Credit Its Due, describes an economic model from a forthcoming research article by Simon Gilchrist, Vladimir Yankov, and Egon Zakrajsek. The model uses corporate credit spreads to forecast job market activity. Take a look at the chart, below, which compares nationwide hiring activity with the results of the authors' model.

The shaded green line shows the year-over-year percentage change in nonfarm payrolls. The blue line shows the corresponding prediction from the authors' model. Evidently, the model fits the historical data rather well.

There's a big difference between explaining the past and predicting the future. In other words, there's no guarantee that the model's predictions will be accurate. So much for the disclaimers. The current prediction is bleak. It calls for nonfarm payrolls to fall by 7.5% during calendar year 2009. Even if you assume that some of those job losses have already turned up in the official statistics, that would still take the unemployment rate well into double digits by the end of the year.

Reflecting on one of my earlier blog postings about jobs and home prices, this new model suggests that home prices will remain soft at least through the end of the year.

(By the way, a 'credit spread' is the difference between two interest rates, i.e., the rate that a corporate borrower pays and the rate that the government pays. Government debt is assumed to be free of default risk, while corporate debt exposes the holder to the possibility of not being repaid. The difference between the two interest rates encapsulates the bond market's estimation of the likelihood that the corporate borrower will default.)

Benchmarking the Financial Crisis - Part 2

In December of last year, Carmen Reinhart and Kenneth Rogoff published an article comparing the current financial crisis to historical financial crises from around the world. I summarized their findings in a previous blog entry. Yesterday, Barry Eichengreen and Kevin O'Rourke published a similar article comparing the current crisis to the Great Depression. Such comparisons have become commonplace. They've generally found that the current crisis has been mild compared to the Great Depression...in the United States. Eichengreen and O'Rourke depart from the usual analysis by comparing the two crises at a global level. Their findings are sobering.

Using the peak in global industrial production (i.e., April 2008) as a starting point, Eichengreen and O'Rourke found that:

• Global industrial production has fallen by over 10%. Thus far, it has more or less followed the same trajectory as in the Great Depression.
• Global stock prices have fallen by 50%. In contrast, stock prices had fallen by only about 10% in the first twelve months following the 1929 peak in industrial production.
• Global trade has fallen by over 15%. Again, the rate of contraction is much greater than in the first twelve months of the Great Depression, when trade fell by only about 5%.

The authors point out that global policy responses have thus far been much more aggressive (and appropriate) than in the Great Depression. Evidently policy makers have learned from past mistakes. Let's hope that we get better results than the last time.

Bay Area Home Prices Back in Fair Territory

Bay Area home prices have fallen more than 40% from their peak. Judging from the heavy demand for bank-owned homes, many investors seem to think that prices have bottomed out. There may in fact be good deals available in the foreclosure market, but recent price trends are not encouraging for the broader market. What do the fundamentals have to say about the issue?

One way to assess the value of an asset is to compare it to the cash flows that it generates. Homes generate rents, so many analysts focus on the ratio of home prices to rents.

(In the past, I've compared home prices to incomes. I've also pointed out that rising incomes lead to rising rents, so the two series typically move in tandem with each other. Comparing prices to incomes is therefore equivalent to comparing prices to rents. Indices of historical rents are easier to find, which is why so many people use them.)

The Case Shiller indices are widely used measures of home prices. Their main strength is that they are based on repeat sales of the same homes, and thus are not distorted by unusually heavy activity in low-end neighborhoods. Unfortunately, these indices don't cover the period before 1987, so I combined them with price indices from OFHEO in order to cover the pre-1987 period. That shouldn't invalidate the present analysis, because we're mainly interested in the post-1987 period anyway. (The two indices were in close agreement until around 2003.)

As for rents, the most readily available measures are the Owner's Equivalent Rent of Primary Residence indices, which are provided by the Bureau of Labor Statistics.

Take a look at the chart, below, which shows Bay Area home prices, rents, and median household incomes for the period since 1983 (when the BLS rent series began).

The solid blue line shows the Case Shiller home price index (with the pre-1987 period represented by the OFHEO index). The dashed blue line shows the BLS rent index. The purple diamonds show an index of median household income, assembled from data provided by the Census Bureau. (The absolute levels of the price and rent indices are meaningless, so I re-based every index to have a value of 1.0 in 1999.)

It's risky to make sweeping assertions on the basis of aggregate statistics, especially when the statistics are compiled from multiple, unrelated sources. But based on this snapshot, I'd say that Bay Area home prices are back in fair territory.

Note: I don't have long-term historical price or rent data for the City of San Francisco, so it's harder to do the same kind of valuation analysis for the City. I'll tackle this in a future blog entry.

Update: The most recent figures from Case Shiller are for December. I should have mentioned that I brought them forward to the end of February using numbers from Dataquick. That approximation is probably no more brutal than any of the other approximations that go into this kind of analysis. (Case Shiller will release its figures for January on Tuesday, March 31st.)

How Significant is Fed's Plan to Support Mortgage Market?

The Fed announced a plan last week to buy another $750 billion of mortgage-backed securities, as well as another $100 billion of bonds issued by mortgage giants Fannie Mae and Freddie Mac. Combined with earlier purchases, the new plan will bring the Fed's total purchases of mortgages and agency debt to $1.45 trillion in 2009.

That sounds like a big number, but everyone knows that the mortgage market is huge. How significant is $1.45 trillion?

The Mortgage Bankers Association announced yesterday that it expects mortgage originations to total $2.78 trillion in 2009. That means that the Fed will be providing more than half of the capital used to create new mortgages or to refinance existing mortgages this year.

Without the Fed's intervention, the housing market would be in much, much deeper trouble.