Hallo!
Hier sind - leider auf englisch - einige interessante Gedanken zur Bewertung des US-Immobilienmarktes. Erstaunlich, dass P. Kasriel, der sonst eher dazu neigt die Ãœberbewertung zu sehen, zum Schluss kommt, dass im historischen Kontex davon keine Rede ist.
Sicherlich ist die Bewertung der Immobilienblase auch immer eine Frage des Messinstruments.
<font size="4">Is Owner-Occupied Housing Overvalued?</font>
April 22, 2002
There has been a lot written lately about US housing being the next bubble to burst. If housing is a waiting-to-burst bubble, then it must be overvalued. Is owner-occupied housing overvalued? For fear of losing my"Mr. Negative" credentials, I am compelled to answer"no," at least"no, in an historical context."
Ian Morris, chief U.S. economist at HSBC Securities in New York, has made much of the fact that the ratio of the value of owner-occupied real estate to disposable personal income, at 1.62 at the end of 2001, is the highest in postwar history. Mr. Morrison views this ratio as something analogous to a price/earnings ratio for equities. I disagree with this analogy. But more on that in the next paragraph. Although the value of household real estate may be at a record high in relation to disposable personal income, the NAR's Housing Affordability Index, which takes into consideration the median price of a house, median income, and mortgage financing rates, is near its record high, as can be seen in Chart 1. So, despite the runup in the value of residential real estate in recent years, it remains very affordable to the typical American family. Of course, a sharp increase in mortgage rates could change this.
Chart 1
<center>![[image]]([/img] </center><br />
<br />
<br />
If we are going to construct a P/E ratio for housing, let's use"earnings" data for housing rather than disposable personal income, which is not related to the earnings on housing. The Commerce Department provides us with data that are related to the"earnings" on owner-occupied housing. These"earnings" are a component of Personal Income - specifically NIPA Supplementary Table 8.12, the line item"rental income of persons with capital consumption adjustment." This measures the implicit value of housing services"produced" by owner-occupied houses less the costs to the homeowner for these services. These costs include mortgage payments. With these rental income data, we can now construct a price-to-earnings ratio for owner-occupied housing that is analogous to a P/E ratio for equities. Because asset prices incorporate expectations of future earnings, I have constructed a series of expected"earnings" on owner-occupied houses by multiplying each year's value of rental income of persons by 1 plus its rolling 10-year annualized growth rate. Such a housing P/E ratio is shown in Chart 2 along with the level of the Treasury 10-year note yield. Because the rental income data are not available past 2000, that is where the owner-occupied P/E data stop. Now, a P/E of 76, the 2000 value, would be considered very high for that on an individual equity or an equity index. But the average P/E for owner-occupied housing between 1972 and 2000 was 126. So, the 2000 value is low compared to the average over the 29 years ended 2000 and especially low in comparison to the P/Es in the second half of the 1980s, some of which were in excess of 200. All else the same, the lower are interest rates, the higher would be P/E ratios. But even with the relatively low Treasury 10-year yield in 2000, the P/E for owner-occupied housing still was low in an historical context. There are some 2001 data for rental income of persons. These data include more than the implicit income on owner-occupied housing. The story does not change using these 2001 data - i.e., the owner-occupied P/E remained low in an historical context. <br />
<br />
<i> Chart 2</i><br />
<br />
<center>[img]) </center>
In last week's commentary, I calculated a theoretical market capitalization series for nonfinancial corporations. Using that same methodology, I have calculated a theoretical market capitalization for owner-occupied housing. For expected profits or earnings, I used the same series as used in the P/E exercise above. Whether the Treasury 10-year yield or the 30-year fixed mortgage rate is used to capitalize housing's"earnings," the theoretical market cap was always under the actual market cap of housing - the value of residential real estate owned by households. I used the Treasury 10-year yield as the discount factor. Again, qualitatively, the results are not affected. Chart 3 shows the actual market capitalization of owner-occupied housing as a percent of the theoretical market cap. The 2000 data point is 458, indicating that housing was 458% overvalued. But, given that the actual market capitalization of owner-occupied housing always is more than 100% of the theoretical market cap from 1972 through 2000, it is clear that my theoretical market cap formulation is lacking. This formulation apparently cannot be used to judge whether housing in recent years is overvalued in absolute terms. But it might give us information as to whether housing is overvalued in relative terms - relative to the past 29 years. And the answer is"no." Chart 3 shows that the 2000 overvaluation figure is at the lower end of the distribution for the years 1972 through 2000.
Chart 3
<center>![[image]]([/img] </center><br />
<br />
<br />
<br />
If housing is overvalued, you might expect that developers would be overbuilding. But, as shown in Chart 4, although the year-over-year growth in the inventories of new homes is increasing, the increases are still low in an historical context. Moreover, relative to current sales, inventories also are low. <br />
<br />
<i> Chart 4</i><br />
<br />
<center>[img]) </center>
In sum, relative to the past 30 years or so, owner-occupied housing does not appear to be overvalued currently. This is not to say that a sharp rise in interest rates would not change the situation. And this is not to say that our tax laws encourage overinvestment in owner-occupied housing. But this is a different issue.
Paul Kasriel
Quelle
Gruss
Cosa
<center>
<HR>
</center> |
Thread-Ansicht
Mix-Ansicht
Reihenfolge
Thread gesperrt
