Credit risk analysis on commercial mortgages is a complex undertaking, handled differently by purpose.
With commercial mortgage-backed securities (CMBS), risk analysis is typically done at the pool level. It’s not easy, which is why some ratings agencies have declined to rate some pools.
And, initial underwriting risk analysis often involves cash flow analysis of rents versus cost, cap rate, and the like.
This, of course, does not account for risk on properties solely occupied by the owner. Cap rates also omit the cost of financing. Most unsettling, the typical methods assume that rents will remain stable or will increase, whereas the actual future income of a property is unknown. Rents may change, vacancies may rise or fall, or tenants may become insolvent.
We felt as though the most obvious piece of the puzzle was missing: a simple credit risk score for a commercial property, analogous to a credit score for a consumer loan.
Consumer credit scores rank-order risk for each person (now or later) on any loan. They are based largely on consumer payment histories.
In commercial real estate, many buildings are owned by shell companies set up merely to handle liability risk. As a result, data on owner payment histories can be thin or non-existent. So, we rely not only on owner data but also on property characteristics, geography, and loan characteristics to create scores.
These scores can be used by lenders to focus attention on existing portfolio loans that may be in some jeopardy. They can be used as well in underwriting — or in loan marketing. Why market loans to properties that have high risk? Or, why do so without adjusting price?
Property-casualty insurers can use our risk scores to indicate likely future claims experience. They are ideal for making pricing and underwriting decisions, and in marketing of insurance, where the rest of the database supplies other detailed information.
The scores also can be used in rating Commercial Mortgage Backed Security (CMBS) pools, and to segregate loans into A-piece or B-piece segments in CMBS pools. And they can be used to make CMBS investment decisions.
In consumer credit scoring, scores are set up almost in a logarithmic fashion. A score of 640 is not much lower than a score of 700, but there is a world of difference in results.
Our scores are easier to use and understand.
We set a score of 100 as the national average risk level. A score of 200 would represent roughly twice the normal risk level, and a score of 50 only half the national level of risk. In short, the scores are linear – and higher scores mean higher risk.
Most properties have low scores, because most properties are safe. But at the high end, scores can range above 10,000. We impose no artificial limits.
We have developed three scores:
Existing Mortgage Default Score
Underwriting Score
Insurance Claims Risk Score
Existing Mortgage Default Score
The Existing Mortgage Default Score is applied to more than 2 million records where commercial mortgages exist and where Notice of Default (NOD) history data are available. This score rank-orders loans for likelihood to “default,” meaning risk of delinquency over 90 days.
Underwriting Score
The Underwriting Score is a generalized credit score based on property characteristics and location. It is applied to each of more than 14 million records. The intent is to create a score useful for considering any new property buyer.
Insurance Claims Risk Score
The Insurance Claims Risk Score also is applied to each record in the database.
Years ago, property-casualty insurers noticed that on consumer policies, credit scores were important to know. Risky people, it turned out, made more claims. We believe the same dynamic is even more likely to take place in commercial property. Arson and fraud are clearly linked to financial desperation, therefore a financial risk score should be very useful.
Insurers have spent a great deal of effort to understand the physical risks of individual commercial properties — fire, flood and storms. A score for financial risk completes the equation.
Our Insurance Score utilizes owner data, such as a history of loan defaults, along with financial risk by use, rising or falling valuations, building size and age, tenant data, and other factors.
This score can be tested retroactively to prove it works.