We’ve all been required to shock our balance sheet. We’ve all done it and for the most part, it’s 100% wrong! The biggest problem is that our shock is to our own institution as if it only effects us, and we operate in a vacuum. Neither is true. Let’s look at how reality strikes down most of our “assumptions”.
CD’s are repriced at maturity – Really? Let’s look at a current $100,000, 1% CD for one-year. It earns $1,000 and the 90-day early withdrawal option costs approx. $250. IF rates really did go up by 300pb, the institution would be offering the same CD at 4% effective tomorrow. How long will it take members to figure out that a $250 penalty is insignificant when the increased annual income on the account is now raised to $4,000? You could double the penalty to $500 (180 days) or even charge the entire interest as a penalty, and the member would still be ahead by taking the option. So, let me ask you, would you pay $1,000 to receive $4,000? You bet you would!!! The Cost of Funds analysis for fixed term shares is completely off-base in an instantaneous shock.
How much can the institution afford to increase rates? Even the most liquid institution can’t put on enough loans at the higher rates to offset the increased cost of funds (CDs and non-maturing shares). Again, assuming all rates went up. You have an effective 100% repricing of your deposits by 300bp. Who has that much NII to absorb that massive increased amount of expense? This is the same for each and every institution in the world. Institutions simple cannot afford to raise rates that high that fast. Even in good times, the ROA was only around 100bp. A -300bp hit to ROA is not a good number! There will be a world-wide hesitancy for all FI’s to raise rates. They won’t need to collude to keep rates low. No one can afford to reprice all their cost of deposits by 300bp. Even if the shock is instant, the response will take an extended amount of time. Assuming a 300bp instant increase in cost of funds is unreasonable and unattainable by FI’s.
There is the negative convexity of exiting loans to deal with. If rates go up, how much longer will your loans be on the books? If I have a 2% car loan and the rates are now 5% for car loans, will I pay it faster or slower? Even more so with home loans. What is the incentive to prepay my loan? There is no incentive to prepay, and every incentive to put those funds somewhere else that’s earning more income than I’m paying on my loan. This further impedes the institution from reinvesting the paydowns into higher yielding loans. Negative convexity is not considered in most evaluations of a shock scenario. Look at it this way: When rates went down, members were fast to refinance at the lower rate. They didn’t stay on the books very long. So in a downward rate environment, having a 6% loan in a 3% market is just begging for a refinance to 3%. When rates go up lending stops and a 7-year average loan just became a loan until maturity or death of the borrower. Either way the FI doesn’t get what they had hoped for. No institution will realize long term high rates in a down market, and they will face a longer loan when rates are up when the FI would like it off their books. Constant repayment assumptions only work in a flat rate environment.
Models don’t include member behavior. If you were just cruising along and knew the current rates were low and now you just heard that rates are more than double for a loan, would you rush out and make a new loan? Of course you wouldn’t. You would postpone, repair, re-anything other than make a new loan. Fix that car. Repair that house. Put off that vacation – “Not at those rates!” A 300bp shock does not account for total and complete lack of loan demand. Who is the institution listed above, with all the liquidity, going to make a loan to? The answer is no one. The economy and individuals will respond negatively towards a quick rate increase. Assuming brisk future loan volume after a large raise of rates is counter to human behavior.
The current economy and demographic situation. We have to realize we are in a very weak recovery. The Fed has stimulated growth for the last 4 years with no results. The demographics of 10,000 workers retiring every day for the next 10 years will have an impact. Borrowers are skittish at best and already over-indebted. They are also just now feeling the effect of the ACA. This increase tax has a further chilling effect on future purchases. We are not in an expanding economy. Qualified loans are very hard to come by. Assuming that a raise in rates will return us to the heady growth of the past is not an economically based assumption.
An instantaneous 300bp shock is a financial nuclear bomb! Everyone loses. So, why do we use it? Because that’s the best model we had. That, and the regulators were married to the idea. Except now we know the shortcomings of the model outweigh the usefulness of what is being tested. Why should we be concerned about something that can’t happen? Either rates move up slowly and the economy adjusts, or rates move quickly and the economy contracts significantly.
We need to drop the instantaneous 300bp as a shock test! We have better technology and ways of factoring in more variables. We have been dipping our finger in the soup to judge the temperature when we have a digital thermometer right next to us.
It’s way past the time to enhance our modeling to better predict outcomes. Our regulator has held us to this flawed standard test, and of actually making decisions based upon these crude and ineffective models. We all need to put a stop to that and push for better modeling standards.
We have been analyzing the rate shock as if the credit union operated in a vacuum. CU’s operate in a much broader environment of members, economies, and with other FI’s. It’s time we reconsidered the implications of a rate increase within a wider scope and with more realistic assumptions. The time for a 300bp shock is well past in almost every part of the model; time, response by members, demographics, interconnectedness of the financial community, and current economy.
Certainly we can now see that drawing conclusions by our current instantaneous method will provide poor, if not contrary conclusions with regards to reality, which could lead to disastrous results if we take actions based upon our analysis.
The old adage is very true here: Garbage in, garbage out. We need to change the input to our instantaneous shock scenario. And, the sooner the better!