The problem with data-driven decisions is that we dont always like the result. Credit scores are recently falling into that trap.
A Montana state legislator, who apparently disliked having a low score because of excessive inquiries, proposed a law that would restrict the use of inquiries in credit scoring. This reminds me of the Indiana bill of 1897 that would define Pi as 3.2, avoiding the annoyance of an irrational number. Math doesnt care about legislative fiat. It does, however, care about incomplete or misrepresentative data. That is why the evolving focus on alternative data really does matter for bankers.
In recent years weve seen a growing number of immigrants and young people seeking credit in an environment that has had minimal growth since the recession. As lenders become more amenable to risk, they need a means of assessing risk beyond what traditional credit data can support. They need data that will enable scoring of the large portion of the population that lack traditional data.
New models that incorporate alternative data have been emerging for years. The FICO Expansion Score was an early entrant into this field offering an alternative data-driven score nearly 10 years ago. Alternative data providers have introduced scores of their own. For example, the LexisNexis RiskView alternative data score was introduced in 2007 and today provides a significant lift when used in conjunction with traditional credit scores (allowing lenders comfortable with their current score to keep it and still get better results).
The Vantage Score 3.0, announced with great fanfare last year, incorporates alternative data into their core score suggesting that non-traditional data has finally gone mainstream. The problem is that many lenders are reluctant to adopt new scores.
All of this brings me back to data-driven decisions. While there are credit card issuers taking full advantage of increasingly available alternative data to improve their profitability, most lenders are not. There is no disputing that incorporating the right alternative data into scoring models allows more consumers to be scored and all consumers to be scored more accurately. While there is a slightly higher cost for the model, it generally pales in comparison to the value of more accurate decisions. So why havent we reached the tipping point of market adoption?
Some institutions are trapped in legacy systems that burden them with high switching costs. Ive worked with bankers who found that replacing their decisioning environment was a comparable cost to upgrading their models. For these institutions, the only feasible way to maintain competitiveness is to schedule replacement of these legacy systems.
An increasingly common objection is that the current regulatory environment has shifted priorities from profitability to compliance. Even if a solution is known to have a good ROI, there arent enough qualified employees left to implement it. Explaining any change to regulators opens the door to additional scrutiny, even if the change is an improvement. While the pain of this argument is real, at some point banks will need to adapt to the new environment or sell the bank to someone who has.
Lenders who are the early adopters are creating custom models that provide a notable lift over what can be accomplished with traditional data. Those institutions are profitably offering credit to consumers that other lenders would consider high risk due to the lack of complete information. This is not only allowing them to grow their portfolio, but skim the cream off the top of the previously unscorable market. Late comers will see far less advantage.
There is no data-driven reason why every lender isnt already using these new tools. It may be difficult or expensive to make changes to existing systems, but those costs will be incurred regardless. Delaying the decision any longer will keep otherwise healthy institutions from competing effectively in an increasingly competitive market. Managing credit risk has become a data-driven business with promising results. If youre making decisions without the best data available, youre not likely making the right decisions. Your profitability depends on using the most predictive data available. If your credit risk strategy is stuck in the 1990s, it is time to bring your bank into the 21st century.
Eric Lindeen is marketing director for Zoot Enterprises Inc., a provider of loan origination, account acquisition and credit risk management solutions for large financial institutions. You can follow him on Twitter @EricLindeen.
