Long-term 'fair value' for housing extremely difficult to discern

---

FT Syndication Service

Nothing now creates more uncertainty for markets, and nothing has greater import for the future, than the course of house prices in the US.

US housing, of course, lit the fuse for the credit crisis. Among measures of housing, price is most important because of its link to the risk that owners abandon their homes, and thus to the risk to the value of the mortgage-backed socurities that sparked the crisis. The best indicator of the coming rise in subprime delinquencies was the slowing in house price rises two years ago.

House prices are also the greatest imponderable for those trying to model the future. There is no precedent for big falls, in nominal terms, in US house prices or for a situation where Americans' percentage of equity in their homes has dropped below 50 per cent. Both have happened recently.

It would appear, however, that if house prices fall and homeowners move into negative equity, the risk that homeowners will abandon their houses increases greatly. So moves in price to clear the housing backlog could have a greater impact than falls in construction (for which there are some precedents).

It is extremely difficult to discern a long-term "fair value" for housing. Research by Robert Shiller of Yale University suggests that prices are 90 per cent overvalued compared to their average ratio to rents over the past 80 years - but they are actually slightly undervalued compared to disposable incomes per capita.

So far, prices have followed construction activity down. Permits for new housing ticked up in April, sparking hopes that an end was in sight. But futures on the S&P Case-Shiller house price index suggest prices could fall 13 per cent over the next year, and 20 per cent by the end of 2012. That shows the depth of the uncertainty. How it is resolved is critical for markets far beyond US housing.