The Context of Risk In Mortgages

Written By:

Ingo Winzer, President

Local Market Monitor 

The traditional calculation of risk in real estate - the likelihood that a property will lose value or that a mortgage will default - is similar to the experience-based methods used by the insurance industry. If one percent of mortgages defaulted in the past, it's reasonable to assume that one percent will default in the future.

This approach works well as long as the context of the real estate market - the larger economic circumstance - doesn't change. But if the context changes, even the best experience-based calculations can be overwhelmed by other forces.

During the boom years of the mid-2000s, Fannie Mae, Freddie Mac, mortgage bankers, investment bankers and the rating agencies all relied on traditional historical data to tell them how much risk to assign to the large numbers of sub-prime mortgages being packaged and sold to investors. And they all were wrong, spectacularly so, because they failed to see that default data from periods of steadily rising home prices no longer applied when home prices had become too high; they didn't see that the context had changed.

The context of risk can change for many reasons. The jump in oil prices in the 1970s led to real estate bubbles in Texas and other energy states. The S&L crisis of the 1980s was largely caused by a doubling of government guarantees on deposits. The recent crisis might not have happened if the Fed hadn't kept interest rates artificially low, prompting investors to look for higher returns in sub-prime mortgages.

Risk Revealed in Home Prices

Even though a change in context can't always be seen for what it is, the risk itself often shows up as an unsustainable increase in home prices. This happens when the demand for homes - from just-hired oil workers, from speculators or from newly-qualified sub-prime buyers - increases faster than the supply.

It's easy to see in hind-sight when real estate markets were over-priced. It's easy to say afterwards that the home price increases were not sustainable. But can we see this beforehand, and therefore do something about it?

The steadiest anchor for home prices in the last 40 years has been income. Like home prices, income varies from locality to locality - and although the ratio of price to income in a California market might be different from the ratio in Iowa, the ratio in any local market is remarkably constant, changing only during those periods we're trying to identify: when prices are unsustainably high.

 

We calculate an Equilibrium Home Price for each of 316 markets in the US, based on the long-term ratio of local home price to income. We consider a market over-priced when the average local home price exceeds the Equilibrium Price by more than 20 percent. Seen this way, in real time, some markets in Florida, Arizona and California were already over-priced in 2004.

Combining Context with Traditional Risk

 

The context of risk doesn't directly translate into a mathematical formula of default (which is why it's so often ignored). Even though we may know there is a higher risk of default or of falling home values, we can't say exactly how much higher the risk is.

But we can use the context to take common sense management steps that reduce the risk.

The first step is to tighten underwriting standards. When business is brisk from strong demand, lenders easily get careless with their underwriting process. Just ensuring that no short-cuts are being taken is useful down the road.

1. We recommend that lenders review and tighten their underwriting process when the average home price in a local market exceeds the Equilibrium Price by more than 20 percent.

The next step is to reduce exposure by writing fewer high LTV loans or by requiring insurance on exposures above 80 percent LTV.

2. We recommend that lenders reduce their exposure down to 80 percent LTV when the average home price exceeds the Equilibrium Price by more than 40 percent.

The final step is to stop all exposure to new loans, either by not writing any new ones or by immediately reselling them.

3. We recommend that lenders stop their expose altogether when the average home price exceeds the Equilibrium Price by more than 60 percent.

This last is a rare occurrence but was apparent in many Florida and California markets by early 2006.