Originally, I was hoping to start this post with a link to some research a colleague and I just completed that discusses how lenders may be overestimating property values prior to foreclosure. But it has not made it through formatting and on to the web yet, so I will instead share the findings with you.
In this research we find that lenders may be overestimating property values prior to foreclosure in weak housing submarkets. (By “lender” I mean banks servicing their own loans or securitized loans.) We find evidence of overestimating values by looking at the difference between the sale price at foreclosure auction (in this case the lender’s reserve/minimum bid) and the subsequent sale price of the home out of REO in submarkets in Cuyahoga County, OH (home to Cleveland). As the housing market gets weaker, the gap between those two sale prices grows. We also find that lenders’ value estimates may be dramatically improved by incorporating a few simple factors such as the age of the home and the poverty level in the home’s census tract. So we would expect lenders to pick up on this at some point and adjust their models accordingly. But we don’t see that happening. There are three possible explanations I can think of, though I welcome others.
First, lenders may not be overestimating the value at all. The price they pay for property may represent bidding in accord with an Ohio law that automatically sets the minimum bid at the first foreclosure auction, rather than waiting for subsequent auctions when the minimum bid can be adjusted. The way Cuyahoga County interprets this law, prior to foreclosure the County pays for a drive-by or walk-around appraisal. The initial minimum bid is set at two thirds of that appraisal. (If anyone can think of a good reason for this law, please share in the comments.) If no one bids at the first auction, the lender can lower the minimum bid at subsequent auctions. Anecdotally, bankers report credit-bidding their judgment to meet the minimum bid to obtain control of, and begin marketing, the property.
Automatically placing the minimum bid may be routine for bankers, but it probably does not always payoff: we find that the worst 25% of REO property sells for less than half of its minimum bid, if it sells in the quarter it is taken into REO. If it stays in REO for four quarters, it sells for less than 10% of its minimum bid. If the property’s minimum bid was $50,000 (remember, this is the worst 25% of property taken into REO), the lender recovers $5,000 before the broker’s commission, maintenance, taxes, and transfer costs. It is unclear why lenders would be in such a rush to obtain such low-quality properties if they were valuing them correctly.
The next two explanations differ from the first, because they assume that lenders are actually bidding at or close to their estimated value of the property. The second explanation may be that the methods used to value property just don’t work well in weak submarkets, and lenders’ valuation models are not correcting for that. It is not hard to imagine that a walk-around appraisal is a reasonably accurate way to value most property in most markets. If brokers want to find non-foreclosure sales to use as comparables, they have to reach back further in time in weak markets than they do in others, so the prices they use are more likely to be stale. Walk-around appraisals may also miss interior damage(stripped copper pipe and wire, appliances, etc.) that properties are more likely to have suffered in weak markets.
The third possible explanation is that lenders are shifting accounting losses from loan portfolios to REO portfolios. This could be accomplished by using the inflated estimated value to prevent recognizing losses on the loan, and instead writing down the value in the REO portfolio. There are two potential benefits to this. The first is that capital markets tend to pay more attention to loan portfolio performance than REO portfolio performance. The second benefit is that most solvency tests for banks focus on loan portfolio performance metrics, and pay little or no attention to REO portfolio performance. So shifting these losses could potentially make lenders look healthier and more attractive than they actually are.
Any way you slice it large REO portfolios are bad for banks and communities. One way to reduce the size of these portfolios is to lower foreclosure auction reserves, increasing the chance that others will purchase the property at auction instead of it becoming REO. If there is no market for the property, then donation to a land bank or similar entity may be the answer.
TrackBack URL for this entry:
Links to weblogs that reference Are Lenders Overestimating Property Values Prior to Foreclosure? And if so, why?: