Housing inventory abundant at prices buyers can’t afford
Overall housing inventory is not low, but the inventory available at prices buyers can afford is artificially low.
Starting in late 2008, lenders began deferring foreclosures to stem the tide of REO flooding the housing market during a time when buyers were few and far between. This slowed the rate of decline in home prices, but it didn’t reverse the downward momentum.
Starting in 2011, lenders made significant changes to their loss mitigation procedures. They dramatically slowed the processing of foreclosures, aggressively modified delinquent mortgages, and they stopped approving short sales, particularly if the borrower had assets.
As a direct result of these policies MLS inventory plummeted, and remaining homes for sale weren’t the must-sell inventory that plagued the market from 2008-2011, the remaining homes were can’t sell cloud inventory, suspended at prices buyers can’t afford. Any cloud inventory that appeared on the MLS weren’t really for sale unless the buyer was willing and able to pay the price the seller needed to get out from under their bad bubble-era loan.
The hope among bankers and loanowners was that restricting MLS inventory would inflate prices by forcing the few buyers active in the market to compete with each other.
Lenders discovered that manipulating MLS inventory could reverse the problems of the housing bust. A cartel of too-big-too-fail banks in cahoots with government regulators set out to save the foolish and greedy bankers and borrowers from the housing bubble, and to my surprise, they pulled it off. Of course, this harms future homebuyers by forcing them to overpay for real estate that likely won’t appreciate, but nobody really cares about their problems.
- The average inventory level since 2006 is significantly higher than from 1999 to 2006.
- When the housing bubble popped, inventory levels tripled but homes sales dropped 61%.
- The data shows that since 1999, inventory levels do not drive the rate of sales.
One of the predominant storylines underlying the level of home sales for the past few years has been that a low level of inventory on the market has been constraining sales.
Low inventory has constrained sales, but not for any of the reasons commonly stated in the financial media. There is no major shortage of houses for sale on the MLS, but there is a shortage of houses for sale at prices buyers are willing and able to afford. That’s what cloud inventory does to a market.
Imagine a market like Orange County in 2012 where comparables for a certain property sell for $400,000. Now further imagine about 25% or more of the houses in this market have mortgages well in excess of $400,000, so they either don’t list their homes, or they list them for $50,000 to $100,000 over recent comparable sales. Are those homes listed well above comparable sales really for sale?
For those who truly understand how cloud inventory works, it should not be surprising that the MLS gets polluted with stale listings. When houses get priced based on outstanding loan balances rather than the market, it’s a recipe for stale listings. Thus we have a market where the few reasonably priced listings go quickly with multiple offers and the cloud inventory listings just sit there.
It may not be the new normal, but it’s a phenomenon we will see until fundamentals applied to mortgage rates elevate market prices above peak housing bubble levels all across the country, and in some markets that is another decade or more away.
However, in looking at the available data going back to the beginning of 1999, the explanation that relative inventory levels drives home sales is misleading and erroneous. The data shows that, if anything, since 1999 an inverse correlation exists between the rate of home sales and the supply of homes.
The housing inventory number that is used for analysis and discussion purposes by National Association of Homebuilders is the “months’ supply,” which calculates how many months it would take for all the current homes for sale on the market in a given month to sell based on the rate of sales calculated for that month. If supply remains constant, the months’ supply metric varies inversely with the rate of sales for that month. For any given month, the metric is calculated using the “seasonally adjusted, annualized rate” of home sales.
The basic problem with his analysis is that he uses the NAr’s months of supply, and the “Months of Supply” is worst indicator of housing market activity.
… the historical data shows that, if anything, there’s an inverse correlation between inventory levels and the rate of home sales:
The graph above is from the St. Louis Fed’s economic statistics database. It plots the months’ supply of homes vs. the SAAR monthly sales going back to January 1999.
First point of note is that between the beginning of 2006 through late 2009, month’s supply increased dramatically and sales declined dramatically. This would likely be expected given the economic and financial backdrop of that time period.
However, you’ll note that prior to 2006, the month’s supply of homes per the NAR calculation remained fairly constant between 3 1/2 and 4 months of supply (red circle to the left). Yet, starting in 2002, sales rose nearly continuously until late 2005. Also note on the graph above that from 1999 through 2006, compared to 2012 through now, the average level of sales was higher and the average level of inventory was lower. This fact negates the argument that the current “recovery” in the housing market would be stronger if there were more inventory on the market.
The point here is that, going back to 1999, which is when the St. Louis Fed database begins to track existing home sales, the data shows that in times of relative economic stability that there is no correlation between the housing supply and the rate of home sales.
For an indicator to be useful, at least occasionally, it must foreshadow some economic event. Months of supply is supposed to indicate busts and rallies, but its track record is dismal. From 1960 to 1982, house prices rose steadily, but the months of supply indicator predicted 5 busts, including two major ones that didn’t materialize. Then in 1982-1984 when house prices really did decline in California, the months of supply indicator was signalling a rally. The same happened from 1991-1994 when the months of supply also indicated a rally that corresponded to falling prices in California.
So this indicator falsely signaled eight busts and incorrectly signaled two rallies during the only two previous instances of house prices declining anywhere in the United States. How much more wrong can this indicator be?
From 1987 to 1990, California inflated an epic house price bubble, yet the months of supply remained consistently above 6.
From 1991 to 1997, house prices steadily declined in California, yet the months of supply fell below six and stayed there for most of the 1992 to 1997 period.
Quite honestly, based on the above analysis, it’s hard to see where realtors came up with the idea that the months of supply indicates anything.
Alternatively, I would suggest that the “affordability” of the down payment and the monthly mortgage payment has driven home sales since the housing bubble popped:
I presented this same observation this way:
To his point that affordability drove home sales since the housing bubble popped, that was the whole reason mortgage rates were pushed down from 6.5% to 3.5%. During the bust, bankers became far more concerned with return of capital than they were with return on capital.
…Ultimately, I believe that the “pool” of potential home buyers who can afford home ownership has largely been “saturated.” This is reflected in the rapid decline of the rate of home ownership in the U.S. since 2005:
The buyer pool is exhausted, but as the economy improves, more people will get good paying jobs, form new households, and the home ownership rate should stabilize as new demand enters the market.
In my view, that big drop in the home ownership rate reflects the increasing unaffordability of buying and owning a home for the average household unit. This is despite policy measures implemented by the Fed and the Government designed to stimulate housing sales.
In my view, the big drop in home ownership rate reflects several million foreclosures and a shortage of potential buyers with good jobs, good credit, and sufficient down payment savings to close the deal. Plus, perhaps a little rational fear of ownership due to the bust and the market manipulations that followed.
With the above analysis that’s based on the housing data from the Federal Reserve, I have shown that the “low inventory” narrative does not explain the slow “recovery” in the housing market, as historically there is no correlation between the rate of home sales and the relative level of inventory since 1999.
Actually, this analysis does not show that at all. He has shown the “months of supply” indicator is worthless — which it is — but that’s all he’s established. The “recovery,” which is more accurately described as a reflation of the old bubble on better financing terms, the recovery is slow because fundamentals are still weak, though they are improving now. Further, this study does not refute the idea that prices inflated because low inventory of for-sale inventory at price points buyers can afford created an artificial shortage.
I also believe that, unless the Fed and the government can devise and implement policies which will re-stimulate home sales by further lowering the both the down payment and the monthly cost of financing, home sales will re-enter the housing bear market that occurred when the big housing bubble popped in 2005/2006.
While I agree that pressure to lower down payment requirements and the cost of ownership will mount, when Dodd-Frank limitations prevent reckless behavior among lenders, the housing market won’t see a repeat of the last housing bust. The next housing bust will be different.
Lenders learned is that no matter how foolishly irresponsible their lending gets, they will get bailed out by government cash and federal reserve interest-rate policy, and they can avoid mortgage default losses by loan modification can-kicking until prices rebound. As long as they don’t foreclose and resell for a loss, they can amend-extend-pretend their way out of any lending disaster.
If we do see another housing bust, I predict we will see sales volumes fall to lows never before measured, even lower than the worst sales years of the most recent housing bust.
Why will sales volumes fall so low? If sellers are trapped in cloud inventory, unwilling and unable to lower price, and if buyers are abruptly limited by lower borrowing power, financed buyers simply won’t be able to transact, no matter how much they substitute down in quality. If financed buyers can’t buy, sales volumes will crater — prices won’t go down much, which is what lenders are after — but sales volumes will necessarily suffer. That’s how I believe the next housing bust will turn out.