BID® Daily Newsletter
Mar 2, 2011

BID® Daily Newsletter

Mar 2, 2011

PREDICTING PROBLEMS AND FDIC-ASSISTED DEALS


It is hard to decide who is crazier between embattled dictator Muammar Gaddafi and embattled actor Charlie Sheen. Both egocentric, sex-crazed, booze-fueled and logic challenged personalities have made some classic quotes lately. In fact, it is now a cottage industry predicted how long each will last in the spotlight. Unfortunately for all of us, both have unlimited access to airtime.
In similar fashion, a whole cottage industry has cropped up in banking that is getting good at predicting how long deposits will last when purchasing an FDIC closed bank. The exercise is instructive even if your bank is not looking at an acquisition, as it gives insight into building franchise value.
In an FDIC assisted transaction, bids are placed for both assets and deposits. When it comes to deposits, the liabilities are either bid at par or a premium. While the average premium for 2011 is 0.2%, most transactions go at par. When bid at par, the acquiring bank gets a period of time to reset the stated rate down on all or a portion of the liabilities. In addition, the acquiring bank may also choose to impose fees. Of course, dropping the rate and increasing fees, means the acquiring bank will face run-off of volume so one of the key assumptions in an acquisition is what the predicted deposit defection or retention will be.
To make this calculation, banks need to know the origination of the deposit, current balance, current rate, market rate (or alternatives) and the amount of other services utilized by the account. While our suggestion is to do this on the customer level, most acquiring banks don't want to get that granular and do it by deposit category (checking, savings, etc.). Once each factor is known, a defection probability can be applied to each and then weighted. While the exact probability and weighting is dependent on the bank, market and demographic they do all follow normal distributions and tend to similar. For example, for most banks, the length of time with a bank has a larger impact than the size of the balances which are both dwarfed by the number of other products utilized by the account or account type. A set of accounts with online bill pay and/or remote deposit capture is more than 2x more likely to stay than an account that has been there for 4 or more years.
Finally, some acquirers break it down between the "initial defection period" and the "secondary defection period." Here, the initial defection period is normally thought to occur over the first 3 months. This is when about 70% of defections occur with the remaining 30% occurring over the time frame between 3 and 18 months after the acquisition/reset period.
By way of an high level example, a bank that acquires $100mm of deposits that has industry standard account balances, 20% in DDAs, 5Y+ of account seasoning and has an average of 1.7 products each has about an average structure of a failed bank. Here, if you impose an average decrease of 45bp in rate and a $2 per month average increase in account fees per month, the acquiring bank can expect to lose about $17mm immediately and another $6mm over the next 18 months. Of course, these numbers are just ballpark estimates. More of a rate adjustment would mean more run-off as would newer accounts or more balances in CDs without other utilized products. In general, the range of run-off is normally between 10% and 55% of the acquired bank.
If you have had other experiences with FDIC acquisitions or have other factors in your model, we would love to hear about them. Until we get more data, we admit that predicting run-off is more art than science. However, we will say models are getting more accurate by the month. While estimating deposit defection is a whole lot more accurate than forecasting how long Gaddafi or Sheen can hold the crazy spotlight, our prediction is that Sheen can outlast Gaddafi.
Subscribe to the BID Daily Newsletter to have it delivered by email daily.