Understanding A Bank Loan’s Debt Service And Probability Of Default (2015)

Commercial Loan Probability of Default

Few banking school classes teach the finer points of loan structuring these days. This is a mistake as loan production is so competitive and spreads so thin that inexperienced lenders are at a distinct disadvantage. Last week, for example, we were discussing a loan here at CenterState and we determined the downside (stressed) case in a particular commercial real estate loan’s cash flows was a 0.8x debt service coverage ratio (DSCR). The question came up how much risk is that compared to a loan that cash flows at 1.25x stressed case? The question was important as you can’t mitigate risk through pricing, maturity, reserve funds and covenants unless you know what risk you are mitigating. Today, we lay the foundation and discuss the basics of cash flow and probability of default.

 

Going To The Data

 

When underwriting a commercial real estate loan, a bank’s probability of default is mainly driven by both the term debt service coverage ratio (DSCR) and the expected refinance DSCR.  The two are different as while both cover the term of the loan, the later takes into account the additional liquidity risk during any balloon structure. A fully amortizing loan, only needs a 1.0x coverage in its final year to be fully performing, while a loan with an amortization longer than maturity needs something greater than 1.2x not to run into problems with refinancing. We have quantified this risk in the past (HERE), but  for the sake of today, we will combine both risks in the same data set.

 

We have looked at some 8,000 loans from a multitude of different areas (community banks, large banks, securitizations and life companies) over the past ten years to construct the below probability of default curve for both fully amortizing and refinance debt service coverage. Using the chart, your bank can determine (on average) the risk difference between multiple types of debt service coverages. This is very germane when pricing a loan if you don’t have a loan pricing model and helpful to understand when structuring a loan.

 

Commercial Loan Probability of Default

 

In our previous example, the average loan with a 0.8x debt service coverage has a probability of default (POD) of 30% vs. a loan that has 1.25x coverage which has a probability of default of 8.4%.  Thus, the loan with 0.8x coverage is more than 3.5 times more likely to have a delinquency compared to the loan at 1.25x DSCR. Again, keep in mind that this data is for all loans for various size banks, structures and includes the recession.

 

The above chart is helpful to keep handy if you don’t have a loan pricing model as knowing the difference in the probability of default is the key to understanding why loans above 1.5x DSC were priced at an average of Libor + 2.19% last month, while loans at 1.25x DSC were priced at Libor + 2.28%.

 

Taking Into Account Loan Type and Borrower Into Risk

 

Of course, not all loans with 1.25x coverage are equal. A property’s cash flow has a certain variability to it and is largely driven by a loan’s property type, as well as other qualitative factors such as tenant lease expiration schedules, tenant quality, property quality and other such factors.

 

We measure cash flow quality by looking at the standard deviation of net operating income. Whereas a cooperative apartment might have a standard deviation of 16%, a limited service hotel has an average standard deviation in net operating income of approximately 48%. To compensate for this risk, banks need greater debt service coverage on properties with more volatile cash flow. Another way to look at this is if a bank wanted a 50 basis point probability of default (usually considered a high-quality borrower), what would the debt service coverage have to be. The chart below diagrams the required debt service coverage for various loan types. 

 

Commercial Real Estate Probability of Default

 

Conclusion

 

The last missing piece of this puzzle is taking into account probabilities of default for the type of borrower which we have covered in the past HERE or the lease structure for investor-owned property which we have covered HERE. Lenders that understand the probability of default driven by cash flows, loan type, structure and borrower risk can now have a quantified understanding of the major components of the loan. This allows lenders to quantify risk and price accordingly. Failing to do this and banks will find that they are either overpricing their risk and missing good loan opportunities in the marketplace, or worst, underpricing the risk and creating a ticking time bomb to go off at the next credit downturn.