How Far Might Housing Prices Fall?

Our original thesis back in May 2005 was that Home prices could retrace as much as 25%-35% from the peak to re-establish a normalized pricing.

Now, a new study shows exactly how and why that might occur: Home Price to Rent Ratio:

"U.S. house prices "likely would have to fall considerably" to return to a normal relationship with rents, says a study by one former and two current Federal Reserve economists.

The study, which doesn't necessarily reflect the views of Fed policy makers, suggests prices would have to fall 15% over five years, assuming rents rose 4% a year. House prices would have to fall further if the adjustment took place more quickly.

The study tracks rents and home prices back to 1960 and found annual rents fluctuated at around 5% to 5.25% of home prices until 1995. At the end of that year, the average monthly rent was about $553 (or about $6,600 a year) and the average home price was about $134,000.

But starting in 1996, home prices started to grow much more rapidly than rents. By the end of 2006, they had more than doubled to an average of $282,000, while the average rent had risen 48% to $818. That drove the annual rent/price ratio down to 3.48%.

That means the rent/price ratio is about a third below its long-term average. To return to normal would require some combination of falling prices and rising rents. The paper suggests house prices would need to fall about 3% a year, if rents grew in line with their 4% average annual growth this decade."

That's a pretty hefty pull back over time, especially on an inflation adjusted drop.

~~~
Note: The study was authored by Morris Davis, an economist at the University of Wisconsin-Madison and until 2006 a staff economist at the Fed; and Andreas Lehnert and Robert F. Martin, staff economists at the Fed.

The PDF of the study can be found here: The Rent-Price Ratio for the Aggregate Stock of Owner-Occupied Housing.

Sources:
Home Prices Must Fall Far To Be In Sync With Rents
GREG IP
WSJ, January 3, 2008; Page A2
http://online.wsj.com/article/SB119931831334463571.html

The Rent-Price Ratio for the Aggregate Stock of Owner-Occupied Housing
Morris A. Davis, Andreas Lehnert, and Robert F. Martin
Department of Real Estate and Urban Land Economics, University of Wisconsin-Madison
Federal Reserve Board of Governors
December, 2007
http://morris.marginalq.com/DLM_fullpaper.pdf

Comments

[RichShinnick]

Price to rent ratio makes sense because real people (as opposed to analysts and CDO engineers) make decisons based on such things. I remember hearing somewhere that in a given area, when the average family could afford the average house with 1/3 of their income (as opposed to their "stated" income) that prices were basically in equillibrium.

Prices falling to where the average family is able to afford the average home is, of course, to be avoided at all costs.

[Paul Meisel]

Wait just a minute. The vast majority of rental properties in most of the US (other than resort areas) are utilitarian, smaller houses, the equivalent of starter homes. The people who rent them are largely either income or cash constrained from purchasing, are intentionally in a short term situation (e.g. divorce, relocation), or just don't want to be burdened by home ownership (think singles).

The average family home, on the other hand, has grown in size over the last 15 years -- a 2500 sf house used to be a mansion and now is just "nice".

This comparison is logical but statistically biased when looking at the universe. There needs to be something to sort apples from oranges.

[Malkiel]

The problem I have with most discussions of "how far home prices will fall" is that the discusser (but not our esteemed correspondent Barry, who I admire as one of the smart guys) seems not to have a handle on the "median income" issue and how home prices are reverting to a mean, their ratio to median income being the mean. Percentage decreases in all housing prices represent quite a variety of percentiles, from starter homes and condos that might not drop at all to bubble Mcmansions that may drop 60%. What are we knowing from an overall percentage number?

[WAKEUP]

All these stats are from previous real estate situations. This situation ain't been seen before, by anyone under 80 years of age. So far, almost everybody is afraid to even glance at the probable eventual drop in house prices. It's going to be a 50% drop, by the end of 2010. Even the most entrenched denial-freak will have gotten the message, by then.

[JohnnyBitchdog]

Let's establish a few things:

1. The value of any asset is the present value of it's cash flows

2. Rents are the only sensible way to estimate the cash flows of a residential property

3. Adjustments must be made for the tax benefit of the depreciation deduction and some premium for control

4. It is clear that in most urban areas the current price of residential real estate far exceeds the theoretical present value of rent derived cash flows....far in excess of what could be explained by the tax benefit and control premium

The problem with so many of the so-called analyses of how much houses should come down is because they are based on such flawed data......all this crap from the BLS and the MLS.

None of this data properly accounts for capital expenditures, size of house, kick backs from builders, kick backs from owner/sellers, etc.

The thing we can say with some confidence is this:

House prices went too high due to a corrupt and faulty system of granting credit. The corruption was widespread and included the arrogance and blindness of Greenspan, to the greed and corruption of real estate agents, mortgage brokers, appraisers, lenders, and Wall Street securities packagers. The whole thing stunk to high heaven and now it's unravelling.

If you want to keep it simple - estimate the monthly rental of a property, assume it goes forever (50 years is fine), and take the present value at a discount rate which includes inflation (6-8%, run it for both numbers). Anyone that can operate a financial calculator can do this. That gives you a good starting point for the value of the property.

Now add some premium for control and tax benefit (10-30%).

If after doing this the asking price of the property exceeds the high range (6% discount rate, 30% premium), tell the sellers to take a hike.

If the property is at the low end of the range (8% discount rate, 10% premium), BUY.

There.

JBD.

[RichShinnick]

JBD

Are you serious? You mean I don't just compare the neg-am payment to the rent I can charge and then calculate that the property will go up by 20% per year forever? The mortgage guy said I could refinance and pull out a Ferrari next year!!!

Financial calculator? Net present value? Tax benefit (I don't pay no stinking taxes)?

I just don't get you man!

2005 lives forever!!!!!!

[MAS]

JBD,

You shouldn't include inflation in your discount rate because it will presumably affect house prices and rents in the same way (unless you're betting you know which one will grow faster).

I think the appropriate discount rate for rent is closer to the after-tax yield on treasuries, or even better, the yield on tax-free munis. Using munis to discount avoids having to make a rough tax adjustment. So this leads to a discount rate of about 3.75%, which gets you to a much higher reasonable range for a home. But what's missing? You can't forget maintenance and taxes. These depend on the location, but will surely decrease the fair value of a home. The security premium is idiosyncratic, but I think it's about 20 bps per year for me.

[RichShinnick]

Mas,

I think JBD is right about including inflation because you are discounting FUTURE cash flows to arrive at PRESENT value. Factoring inflation in to the discount model gives you the proper price of the asset. If you do not factor in inflation, the PRESENT value of the asset will not include FUTURE inflation.

I agree that you need to factor in maint and taxes (just pull them from the rent cash flows).

Please tell me if I am missing something.

[JohnnyBitchdog]

You either have to include inflation in the discount rate (as I suggest) or you have to estimate future cash flows with inflation added. The former is much easier since it's an annuity calculation.

And instead of playing with premiums for control and tax benefit, you can adjust the rent cash flow as suggested. It probably is better to adjust the rent number, especially for taxes which are reasonably estimated. The premium for control, which is pretty subjective, can be added to the valuation.