Friday, June 25, 2004

HSBC, Contrary to Fed, sees U.S. housing bubble

Reuters is reporting (link) that the world's largest bank, HSBC, disagrees with the US Federal Reserve ("fed"). The fed recently published a reported that concluded that there was no significant evidence of a housing bubble in the United States. In a 47-page report, HSBC concluded that the US was in the midst of a severe housing bubble. Prices were 10 to 20% too high. "We think the party stops by mid-2005." A gradually fall over a number of years was likely, with some overshoot on the way down before prices began rising again. The impact of a collapse in housing prices was likely to be felt much more strongly than the recent bursting of the stock market, HSBC feared, due to the greater "wealth effect" associated with housing prices. HSBC thus fears a hard landing in the economy.

I can only cite my own experiences out here in Southern California as evidence that a housing bubble is real. Some condominiums (aka "condos" or town houses) can be valuated like stocks. The same complex may have identical units that are both for sale and for rent.

One can think of the sale price of a condo unit as the price of a stock. The annual rental gross rental income for the same unit, minus costs (maintenance fees to the homeowner's association, insurance, property and income taxes, &c) can thought of like a stock's earnings.

Classic Capital Asset Pricing Model (CAPM) theory relates the fundamental price of a stock (or condo unit) to its future earnings and the current interest rate. If condos (or stocks) were a risk free investment, the expected multiple of price over earnings (P/E) is simply the reciprocal of the risk-free interest rate, normally around 15x. (Interest rates are low right now, so the optimal P/E is a little higher, but not much higher. This is one of the reason stocks tend to fall to lower valuations when the Fed raises interest rates.) If a stock or condo trades at higher P/Es, then it is essentially a "growth stock" or "growth condo", meaning the buyers expect the stock or condo's annual earnings to rise with time. During the Internet bubble, CAPM predicted that some Internet stocks would need to earn more than the entire US GDP within a few years to justify their valuations, so CAPM valuation can be a powerful tool to detect bubble pricing. (We are ignoring risk in the CAPM model. With due-diligence on tenants, a properly insured, leased condo's risk is only somewhat greater than that of a risk free bond. Increased risk implies condos should have a lower P/E than a government bond with comparable earnings, since the condo must earn greater income to compensate for higher uncertainty.) CAPM theory predicts how much a stock or condo's earnings must rise to meet the expectations implicit in the P/E.

The article states that Washington DC is thought to be the most overvalued region, with California the second. I'm aware of some 2-bedroom condos here being rented for about $22K, and sold for around $430K (rents in some parts of California are among the highest in the world.) Depending on how much one allows for maintenance, taxes, and insurance, this is a P/E of between 25 and 35, comparable more to a pricey tech growth stock during the stock bubble than real estate valuations. Now, real estate prices in coastal states have historically risen faster than the rest of the nation (there's only so much ocean to go around, you know, while there's plenty of undeveloped land in the interior), but these multiples seem high even for California. In less overpriced areas, one sees P/Es closer to the usual risk-free 15 multiplier determined by interest rates.

There is other evidence. Real estate agents, like stock brokers during the bubble, tell their clients that prices can only continue to go up. (Real estate prices in California dropped by 30% about 15 years ago during the last housing correction, but everyone seems to have forgotten this.) This is in spite of high vacancies and other signs of slack demand in the rental market (a sign of overcapacity), despite a rush to buy.

Real estate agents out here urgent clients to take on the maximum house they can afford, arguing they will make a killing as housing prices continue to rise 10, 20, 30% each year as they have in the last few years. They suggest "interest-only mortgages", where the seller only pays the interest payments on the mortgage, as a way of affording an even larger house. The theory of an "interest-only mortgage", like the buying of stocks on margin that ruined many a stock trader, is that the value of the home will continue to inflate, allowing the mortgage to be paid off at considerable profit when the home is ultimately sold.

HSBC has some company out here in California in its belief that a housing bubble may exist.
However, these ideas are not popular with real estate agents out here (who, like their stock broker predecessors of a few years ago, argue passionately that prices will just continue to go up.).

As I've argued in these pages in the past, the Fed will be reluctant to raise interest rates prior to the election in November. (Raising interest rates a little now would help cool off the overheated housing market in the hopes of a economic soft landing.) This is because former President Bush reported blamed the Fed and Alan Greenspan's rate hike prior to the election in 1992 for Bush's loss to Clinton, and the Fed reportedly hasn't forgotten. Readers may recall how the Fed resorted to an emergency rate hike right after the 2000 election, in the midst of the political instability caused the Florida vote recount crisis.

However, the Fed has made it clear an interest rate hike is coming "soon" due to rising inflation in the economy. (The inflation is probably caused by high oil prices and the continued U.S. currency devaluation triggered by the yawning U.S. trade deficit. Inflation is bad because it causes managers to try to get out of pricing agreements with their customers, slowing economy growth.) Let's hope the fed's rate hike doesn't come too late to produce the "soft landing."