Common risk factors

Commercial real estate risks vary greatly by location and property type, and many idiosyncratic factors determine building-specific risks: if it is on the waterfront, amenities and updates, management, etc. But there are some common factors.

1. LTV \(LTV=loan/property value\) The higher the LTV, the higher the risk. Once LTV is bigger than 1, it becomes very risky.

2. DSCR \(Net operating income/debt service coverage ratio\)

   The higher the DSCR, the better the deal, and smaller the default risk.

   Sometimes even when LTV is larger than 1, but NOI is higher than interest payments, borrowers are willing to hold the property and continue the loan because it is profitable in terms of cash flow.

3. Loan contractual terms Recourse loans will have much lower expected loss severity than non-recourse loans.

4. Interest rates Floating rate loans will suffer a lot more in a rising inflationary enviroment.

   IO (interest-only) loans may have higher risks as well. IO loans do not pay down principal overtime. And the final payment is a balloon payment. If borrower cannot refinance at affordable terms then there is a strong default risk. Loans with large balloon payments can be distressed when maturity dates are near.

5. Lease tems While hotel property type has an immediate response to macroeconomic shocks and disasters (e.g., 9/11 terror attack, Covid pandemic), offices and industrial property types often have long term leases, some are well over 10 years, and will show a lagged response to stress.

   Private negotiations over lease terms may not be reflected in public data.

Inflationary enviroment and risks in CRE

Inflation enviroment could be a booming period for CRE if inflation is related to rapid economic growth.

However, when inflation is caused by what we are having right now (too much stimulus, supply chain issue, not enough production, lower productivity), then inflation will negatively affect CRE sooner or later.

As eonomoy contracts and businesses cut down on spending, occupancy will drop.

Combined with higher construction costs and wages, NOI will drop as well.

Further, with rising rates, DSCR will decrease. Decreasing DSCR will push down property value, causing LTV to rise. Overall default risks are higher.

While the above relationships are true at a high level, it does impact different property types very differently.

Hotel

Inflation may actually be positive right now when the lockdown-induced pent-up demand can make hotel business very good. Rooms at good hotels may be almost fully booked even if hotels charge higher fees in tandem with inflation. So incomes are rising.

Whereas the cost of financing is still low (yes, the 10 Year Treasury is only slightly above or below 3%), hotels are expected to benefit from inflation in near term as long as economy is not in severe recession.

Industrial

Industrial is not expected to suffer either. Demands for warehouse and distribution centers are still strong while cost of financing still low.

Office

Office is a very different story, which is not so much inflation related.

After 2+ years of staying home, office works are coming back to office. But some never go back. And many are working in hybrid mode. As a result, occupancy rates are staying lower than prepandemic levels. If the remote + hybrid mode stays, and those who have not come back to office never returns, then this sector may have serious problems.

With the crime rate increasing in some major metropolitan cities, office workers are even less inclined to come back to office 5 days a week.

Using myself as an example, before Covid, I used to be in the office nearly 7 days a week. I liked to study and work on my book when the office was quiet during the weeekend. Now, with the NYC crime rate, I try not to get out of my place if I can help it.

CRE risk models

CRE models are commonly segmented by property type, and sometimes also geographic locations.
Often they take a two-step approach: first model the CRE fundamentals like the NOI, vacancy rates,

1. Some models are segmented by property type and geographic locations. Probability of default and rating migration models: First simulate NOI and value paths, along the paths determine DSCR and LTV; then the PD and TM models are functions of the LTV and DSCR. The LGD model is a function of LTV, which is estimated by using an adjustment for accrued interest, cost and distress property value for each simulated path.
2. Some models may also be segmented by property type and geographic location and have similar drivers except not based on simulation.
3. Some others may first models NOI growth rate and price (or value) growth rate using econometic approach using rental rates, occupancy rates, and MEVs. FOr PD, besides DSCR and LTV, other factors are: recent changes to the property value, property type, vintage, delinquency, market tier, loan type, and remaining term.
4. In addition to default rate, some may use survival modeling approach and have time to liquidation as target, which tries to capture the effects of accrued interest, deterioration of property and fees/costs.