Could The Mets Teach You Underwriting?
I’m a lifelong Yankee fan. I don’t exactly hate the Mets but their success bothers me the way I’m bothered by flies on a beautiful day. They’re a minor anoyance. Still, their success has lessons for the baseball fan and credit union executive alike.
Baseball is undergoing a statistics driven revolution: its savviest executives are using Big Data to make decisions, ranging from whom to draft to where infielders should be positioned, that used to be considered non quantifiable.
The Mets, who are headed to the World Series, have an Analytics Department that uses a top secret algorithm to help pick their lineup (No joke). Joe Madden of the Cubs was one of the first managers to position his infield based entirely on a batter’s statistically validated hitting tendencies, which is why it’s not uncommon to see a second baseman in short right field. And the Astros GM is so successful at drafting players – just ask the Yankees –that the FBI is investigating whether the St. Louis Cardinals tried to break into a database he uses. In fact, the most successful teams of the last few years haven’t been the ones with the most money to spend but the ones that aren’t afraid to use the analysis provided by an explosion of data points to challenge conventional wisdom. All the teams that I mentioned were better known for loosing than winning until they took new approaches to analyzing an old game.
The same thing is rapidly happening to lending. Those institutions that are willing to quickly embrace new data driven techniques and combine them with traditional banking knowledge might well be able to punch above their weight when it comes to performance and growth
To be sure, banking has always been numbers driven but crusty old-timers who started in the lending industry before computers really took hold love to tell me that, at its best, underwriting is an art rather than a science. But now, every day brings new data and evidence that a willingness to apply it can result in a more effective underwriting framework.
For example, Fannie Mae announced that by the middle of 2016 it would start making it easier to analyze the credit worthiness of “nontraditional borrowers.” It is also going to start using “trended data” in analyzing an applicant’s credit worthiness. In a press release it explains that:
“Currently, credit reports used in mortgage lending only indicate the outstanding balance and if a borrower has been on time or delinquent on existing credit accounts such as credit cards, mortgages or student loans. With trended credit data, lenders will have access to the monthly payment amounts that a consumer has made on these accounts over time. Among other benefits, this will allow lenders to determine if the borrower tends to pay off revolving credit lines such as credit cards each month, or if the borrower tends to carry a balance from month-to-month while making minimum or other payments.” (http://www.fanniemae.com/portal/about-us/media/corporate-news/2015/6305.htm).
Baseball and banking have another thing in common: they are both grappling with how to handle the transition to this new type of analysis. For example, even as statistics make it easier to predict borrower behavior, credit unions are still valued for the fact that they “know their members.”. Are sophisticated statistical models really consistent with a “know your member” approach to underwriting?
The Mets may provide one possible answer to that question. One of the reasons I they have become so successful is that their GM Sandy Alderson has been able to bridge the gap between the world of the crusty old-timers who actually played the game and the stat geeks who have gotten no closer to a baseball diamond than staring at the back of a baseball card.
I don’t know how to strike a balance between the old and new; it’s an issue that all boards and CEOs should grapple with as they try to position credit unions for future growth.