How Far Down for Housing Values?
April 9, 2007
Now that selling prices for residential property are heading down pretty much nationwide, a sensible question to ask is: how far down can they go?
A simple answer is a metaphoric what-goes-up-must-come-down approach, and this is not such a bad way to look at things.
The must-see Real Estate Roller Coaster.
A slightly more sophisticated version of the same question is: at what point do houses get cheap enough that there should be a natural demand for them, that isn’t based on dreams of can’t-lose capital gains?
One way to look at it is cashflow. The 30-year fixed-rate fully-amortized mortgage is still the standard in the US. The rate for best-quality credits is about 6%, plus about 2% in amortization, for payments of about 8%. On top of that, we need to add about 4% of the purchase price of the house for maintenance, insurance, property tax, and so forth. That gets us to about 12% of the purchase price — or 1% per month. There is a tax deduction, so if the buyer’s marginal rate is 25%, the deduction is worth about 1.5% (25% of the 6% interest). That brings us back to about 10.5% of the sale price per annum. To be cashflow-positive with respect to equivalent rents, then the annual rent/price ratio must also be 10.5%. This brings us to 9.52x a year’s rent, or 114x a month’s rent. At that point, people would be motivated to buy houses not because “they only go up,” but because it is cheaper to do so on a short-term cashflow basis.
For the investment buyer, they would like to make a bit of positive cashflow as well. Typically this means a cap rate above the cash cost of the 30yr mortgage, which we saw was about 8% per annum. Typical figures for residential housing are gross nominal rent X 93% occupancy X 55% FFO margin = net cashflow of 51% of gross nominal rent. So, an 8% cap rate means 8%/51% = 15.7% gross nominal yield. That works out to 6.37x a year’s rent or 76x a month’s rent. (Costs are higher for investment properties because you have occupancy, property management fees for finding new tenants and evicting the bad ones, and professional repairs/maintenance. Most homeowners do most repairs/maintenance themselves and don’t consider this an expense.)
A friend of mine tells me that, during the last property bust in London (London!), in the early 1990s, he was buying apartments at a 12.5% cap rate. That seems to work out to about a 24% gross nominal yield, 4.1x annual rents, or 50x a month’s rent.
We looked at this in a little different way some months ago.
I’m talking about housing valuations here which is a little different than prices. Valuations are cashflow ratios, while prices are nominal figures. Rents, the basis of my valuation metrics, might rise dramatically in an inflationary environment. They might go up for reasons not directly inflation-related, especially in those areas that still have a strong local economy like New York. In an inflationary environment, mortgage rates would also rise so valuations would likely fall. A bad recession might have the tendency to drive rents down. We’ll just have to see how it works out.
The point is, we tend to end up with valuations in the neighborhood of 100x rents, if not a little lower for investment properties. This is not even a “crash” scenario, and given that I can see the possibility of fully 15% of all US homeowners going through effective foreclosure (unless some seriously nasty inflation bails them out), a period of distress, despair and depressed prices is certainly possible. The 12.5% cap rates of early 1990s London is a good benchmark for this.
So, if we have 50x a month’s rent as the floor of our crash-crater, where are we now? There’s a very nice site that helps us in that regard:
We are most interested in the price/rent ratios. This is generated as the median single-family home sales price divided by the median rent for a three-bedroom apartment. Maybe a little apples-to-oranges. What you would really want is the price of a property compared to the rent for directly-comparable properties. Anyway, here are some headline numbers as provided by this website, as of 4Q06:
|Charleston, South Carolina||224x|
|Las Vegas, Nevada||262x|
|Los Angeles, California||335x|
|New York, New York||323x|
|San Diego, California||343x|
|San Francisco, California||393x|
Just for comparison, there are some places where the price/rent ratios are closer to our hypothetical 100x benchmark.
|Elmira, New York||97x|
Also for comparison, let’s look at our above high-valued cities for the lowest price-rent ratios since 2001, when this website’s data begins.
|Charleston, South Carolina||224x||170x||1Q01|
|Las Vegas, Nevada||262x||143x||1Q01|
|Los Angeles, California||335x||208x||1Q01|
|New York, New York||323x||223x||1Q01|
|San Diego, California||343x||229x||1Q01|
|San Francisco, California||393x||181x||1Q03|
So, we can see that our 100x rents bogey is WAAAAY down there. And 50x rents? Armageddon! I’m not going to say that we’re going there for sure. There will be plenty of time for that later. For now, I think it is apparent that another 20% to the downside is very feasible. That would barely get us to 2001 levels, even with a bit of rent rise baked in. Even that, the softest of soft landings to midboom valuations, would have great implications for those many homeowners in cashflow difficulties, who would not be able to either make the payments, refi, or sell at a profit.
Put your helmet on!