Tuesday, June 24, 2008

Median house price versus median income

(Calculated Risk) A research analyst at Harvard Joint Center for Housing Studies (JCHS) has been kind enough to send me (Thanks!) an update to the median house price to median income ratio I discussed yesterday (see: Ratio: Median Home Price to Median Income)

JCHS didn't post this data because they changed both the data source they use, and their methodology to calculated the ratio - and JCHS analysts felt this would be confusing because the ratios are different from previous releases. However the trends are the same - and therefore useful for us.

Here is some price-to-income data from the 2008 report (data through 2007). Once again I picked a few key cities and plotted the national average (dashed). Note: the new data only goes back to 1989.

House Price Income Ratio Click on graph for larger image in new window.

Different areas have different price to income ratios. There are several reasons for this (land restrictions, demographics), but on a national basis, the median price to median income ratio rose from around 3.5 in the 1990s, to 4.7 in 2005, and has started to decline since then. This would suggest that a combination of falling prices and rising incomes would need to adjust this ratio by about 25% from peak to trough.

For Los Angeles, it is reasonable to expect the price to income ratio to fall to below 5. This suggests prices at the peak were about twice as high as normal.

Hold on, you say ... the above graph shows the price-to-income ratio for Los Angeles only declined about 3% from 2006 to 2007, but Case-Shiller showed prices in Los Angeles declined 14% in 2007 ... what gives?

Although I don't know the exact methodology used by JCHS, the likely reason for the difference is the JCHS data is based on the annual average, whereas the Case-Shiller data is from the end of 2006 to the end of 2007.

Case-Shiller vs. House Price Income Ratio
The second graph illustrates this point for Los Angeles. The red line is the JCHS median price to median income ratio for LA.

UPDATE: For Case-Shiller Index, Jan 2000 = 100.

The blue line is the annual average of the mid tier Case-Shiller index (mid tier was used to approximate the median price). The difference between the red and blue lines is that nominal median incomes are increasing. This is exactly what we would expect.

The dashed line is the monthly Case-Shiller mid tier price index. This has fallen off a cliff. It is very likely that the median price to median income ratio (on a monthly basis if it was available) would now be around 7 or lower - well on the way to the historical norm of around 4.7.

No comments: