Negative equity datapoint of the day

By Felix Salmon
May 12, 2010

NegEqQ12010.JPGCalculated Risk has the latest First American CoreLogic negative-equity numbers, and a scary chart to boot:

The share of borrowers whose mortgage debt exceeds the property value by 25% or more fell slightly to 10.4% or 4.9 million borrowers, down from 10.6% or 5 million borrowers. The aggregate dollar value of negative equity for these deeply underwater borrowers was $656 billion.

Once your negative equity reaches 25%, chances are you’re going to default even if you don’t have to: walking away is becoming both more common and more socially acceptable.

I’d love to know whether any banks at all mark their mortgage holdings to market if the mortgages are performing. My guess is that they don’t — which means that there’s likely to be a lot more in the way of write-downs hitting banks over the next few years. Especially in the sand states.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see

yes – there will be more writedowns – which is precisely why the Fed is pleased that BAC/JPM/C/GS had no losing trading days last quarter – that’s exactly how it was designed – so that they could extend and pretend on the mortgage paper marks, while they tried to earn the future, yet-to-be-taken writedowns back in advance!

Posted by KidDynamite | Report as abusive

The only mark-to-market on home loans that I have ever seen was for super jumbo loans ($1 million and up).

Bank regulators generally allow banks to aggregate reasonably-sized home loans into “homogeneous risk pools.” And rather than mark an individual performing home loan to market, they might insist that the bank create a provision for homogeneous pool losses. Those provisions have tended to run anywhere from 0.25% to say 3.0% of the pool and are based on past loss history. Obviously looking in the rear view mirror these days doesn’t help very much. Moreover, that percentage provision in the sand states is typically less than the legal and brokerage costs of a foreclosure sale. So this provisioning works if:(i) only a few mortgages in the pool default, and (ii) if the historical underwriting Loan-to-Values(LTVs) aren’t breached by declining real estate values. But if the mortgages in these pools are highly correlated (which they are), then LTV-breaching declines in home values will create bank charge-offs that tend to look like a steep step function, e.g. a 3% historical charge-off rate goes right to say a 20%-30% OREO loss rate.

So you are correct. Many home loans may be paying for now but are significantly impaired. Even if the average home loan underwriting at inception was a relatively conservative 70-80% LTV (which was the exception as the market got exuberant), loans in markets that have experienced 40%+ price declines (sand states)have significant unrecognized impairment losses.

Posted by AABender1 | Report as abusive

Our real estate crash is very sooo yesterday. The great China real estate crash of 2010 is on, baby!

“Beijing Home Prices Plunge 31.4%” nset/@10063/post/1185337

(hat tip MISH)

Posted by DanHess | Report as abusive