BID® Daily Newsletter
Mar 1, 2012

BID® Daily Newsletter

Mar 1, 2012

LEAPING INTO SERIAL CONTINGENT RISK


We get confused every Leap Year, as we think it should be a normal year and the other 3Ys should be considered "Leap" since you go leap over Feb. 29th to get there. Worse yet, like a time change, Feb. 29th always throws our finely tuned bodies off, so today feels like a Friday. Tossing in an extra day has created a domino effect, also called serial contingent risk, which brings us to today's topic. Having a better appreciation for this concept will explain why we often miss our budgetary goals, why most technology projects run behind schedule and why most new product launch schedules are flawed. To understand serial contingent risk let's use an example - say you and your spouse have an important awards banquet to go to where you are going to get Banker of the Year. Both you and your spouse drive from different parts of the city and have an average commute time from work of 30 minutes. You both agree that you need to leave the house by 6:00pm in order to make it to the ceremony on time at 6:30pm. Working backwards, you deduce that you and your spouse should leave work at 5:30pm to meet and then depart. Sounds like a good plan, except for the fact that you are probably going to be late. If you wonder why, consider this breakdown: Because you and your spouse's commute times are an average of 30 minutes, either one of you is going to go over that about half the time. Thus, to be on time, both you AND your spouse need to make it home at or below your average commute time. Note the contingent serial risk here. If just one of you is late, or you are both late, then you are both going to be late to the awards party. All things being equal, you are going to be late 75% of the time. Whenever you have a series of events that are combined together, you need to take into account the probability or the variation of each occurring. Thus, if your bank is going to move locations, convert to a new core system or launch a new product; chances are you can't simply add all the time and money from all the minor tasks together to arrive at an answer. You have to wait for one part of the project to finish before you complete the next part and that adds serial risk. This is why airlines and shippers build in a 10% to 20% variance to schedules to compensate. Of course, one of the largest implications relates to budgeting. Take loan production for example. You start with the question of how much risk you want to take for the year and then figure out how to maximize that risk in the form of a return. If loan production is the best way to achieve that return, then you back into the resources (business development, underwriting, systems, etc.) required to achieve that production. Your CEO turns to you and says what can you produce? Being an astute banker, you deal in probabilities, so you look at the variation of the past and size it to expected production and credit availability. So you say, I can grow loans between 5% and 15%. The CEO can't process a range, so they force you to pick a number and you then take the average of 10%. You just assured yourself that you will likely be under budget. Similar to the awards banquet commute, half the time you will be under loan growth from a statistical point of view and half the time over it. However, should credit availability provide you with 13% or 15% growth like you predicted it might, you are not going to have the business development officers or underwriters to handle the overage. This creates an asymmetrical risk profile where you have more downside than upside to your process. Of course, experienced bankers know this intuitively, which is why they go around sandbagging their budgets each year - but that is a different resource allocation issue for another time. One way around serial contingent risk is to recognize it, work with probabilities, build extra resources into plans and prepare for alternative scenarios. In our budget example, a better approach would be to look at a variety of different loan growth scenarios, assign probabilities, take the most likely (which was between 5% and 15%) and then set triggers so if loan growth is running above 5% by March, extra capacity is brought on to support. By better understanding serial contingent risk, bankers can leap over problems and stay on track.
Subscribe to the BID Daily Newsletter to have it delivered by email daily.