Earlier this week, we discussed why community banks need more granularity in their credit grades in order to compete with more sophisticated banks (HERE). Our example focused on just the credit and loan loss allowances between a set of loans that were all rated with a credit grade four by one bank as we showed what happens over the course of a quarter when they have to compete. We stopped short on pricing, which is our topic today. In this article, we will look at some real time examples that are before our bank right now where we will adjust the reserves to match the risk and see how it impacts pricing and competitiveness.
In April, we have three owner-occupied businesses that all want five-year loans with 20-years of amortization in the amount of $650,000. We have average credit that most community banks would grade as a “three” on a scale of one to eight, the first five grades of which are all “pass” credits. However, these average credits are in the following industries:
Customer A: Trucking – they want to finance a distribution/warehouse center
Customer B: Private school that wants to refurbish their gym
Customer C: A manufacturer of non-patented medical supplies
For the sake of simplicity, let’s say these loans are all priced at Libor + 2.55% since you have them all a grade of three. It does not matter how these are structured, fixed, floating or adjustable; we will just assume that they are all priced such to take interest rate risk into account so that the average spread over the life of the loan is Libor + 2.55%. Finally, we will assume that the bank is average regarding their cost structure so that we will assume a 75% efficiency ratio and a 60 basis points cost of funds. You run each loan through your risk-adjusted pricing model, and you smile because you get an 11% return on equity. However, that is not reality because your effective single credit grade lacks accuracy. Over the life of the loan, this is what you earn:
The difference, given a single credit grade is vast between all three credits. Trucking, it turns out, is much more competitive, and much more volatile in terms of earnings. As such, the probability of default is more than four times as great as the private school. This bank might have booked this loan thinking they were going to get an 11% return, but they would have received a 9%. The private school has less than a 0.90% probability of going into default any given year while the manufacturing company has about a 2.15% expected default rate.
Put another way, if you hold the return constant and adjust for risk in the form of different reserve levels, then you get the following pricing differences to equate to the 11% target return on equity:
If this bank had faced competition these deals, it likely would have won both Customer and Customer C, while Customer B, with the highest return on equity, would have been taken away at a lower price. The bank would have then ended up with an underpriced portfolio and one that is skewed towards higher future defaults. Meanwhile, competing banks would have won the Customer B, the customer that not only had the highest return but the one that would have performed best in an economic downturn.
Like our previous view of this problem, community banks need to get more granular in their credit grades to prevent this adverse selection. Having a loan pricing model that takes into account the cost of production and risk is mandatory given our place in the credit cycle.
Submitted by Chris Nichols on May 11, 2017