Defining the issue – what is it bankers are trying to solve for

by Guest Contributor 4 min read June 10, 2013

By: Joel Pruis

What is it we as bankers are trying to accomplish?  If you have been in the industry for 20+ years, this question may sound ridiculous! We do what we do!  We are bankers!  What do you mean define what are we trying to do?  But that is the question, what is it we are trying to do?  I am going to propose we boil it down to the basic/fundamental element – Banks aggregate money from various sources and redeploy these funds to earn a return for the shareholders.  Ultimately, our objective is to generate an appropriate return for the shareholders

Getting back to the movie Moneyball, Billy Beane and Peter Brand define the objective of the Oakland A’s for the season in terms of projecting the number of wins that are needed to assure, with all probability, that the team makes the playoffs (this would be similar to the objective of banking to generate an appropriate return for the shareholders).  But Peter Brand quickly moves into very specific targets that are required for the A’s to make it to the playoffs, namely win 99 regular season games.  In order to win 99 regular season games, the A’s offense will need to score 814 runs in the season and defensively only allow 645 runs.  Plain and simple.  Very objective, very measurable and it is all based upon data, data, data.

Let’s break this down.  Based upon their conference, the teams in their conference along with the overall schedule, Peter Brand projects that 99 wins are necessary to land a spot in the playoffs.  No gut check, no darts or crystal ball but rather historical data that when analyzed provides the benchmark of 99 wins to statistically assure the Oakland A’s that they will make the playoffs.

So let’s apply this to banking.  Our objective is to generate the appropriate return for our shareholders or the old Return on Equity.  So, for example, if our targeted return on equity is 20% (making the playoffs) we need to make sure we generate enough net income (99 wins) through producing the necessary gross yield on assets (814 runs generated by the Oakland A’s offense) less the expected charge offs (645 runs allowed by the Oakland A’s defense).

For a quick dive into details, our data would provide for a margin of error on the variable to provide for statistical assurance of achieving the objective (Return on Equity).  In the movie there is no guaranty that the 814 runs will win the conference but at the same time there is no guaranty that the Oakland A’s opponents will score 645 runs.  Never in the movie does the coach, Billy Beane or Peter Brand tell the team, “You only have to score X number of runs this game, don’t score anymore.”  Or even crazier, “You are not letting the other team score enough runs, they need to score 645!”  No, the strategy is still to generate as many runs as possible while minimizing the number of runs scored by the opposition.

Rather it is the review of the total amount of earning assets of the financial institution and the overall credit quality that we must understand and control to determine our ability to generate the net yield on assets required to generate the return on equity that is required.  If we assume too much risk in the portfolio in order to generate the required yield it would be similar to having a poor pitching staff projected to allow 10 runs a game requiring the team to produce 11 runs a game in order to win.  It just is not realistic.  So basically we need to assess at the high level, are we appropriately structured to allow for the generation of enough profit to provide the appropriate return on equity.  At this point, we do not need to complicate it any further than that.

Now let’s take a look at the constraints.  We know we have them in banking, let’s take a look at probably the single biggest constraint imposed on Billy Beane and the Oakland A’s.  In the movie, before Billy Beane is even aware of the Moneyball concept, his is given his constraint by the owner.  Beane asks for more money to ‘buy players’ and is flat out rejected by the owner.  The owner, in fact, cuts Beane off by asking, “is there anything else I can do for you?”.  Net result is that the Oakland A’s have $38 million dollars for payroll vs. the New York Yankees at $120 million.  Seriously it does not seem fair.  How can you attract the needed talent when you cannot pay the type of salary needed to get the necessary players to win a championship?  Let’s rephrases this for banking…  How can a bank be expected to deploy its assets when such a high rate of return is required?  Boiling it down to a specific example, “How can I originate a commercial loan at this rate of interest when the competition is ½ to 1% lower than our rates?”

Up next – Why will 99 games get us to the playoffs?  How do we assess the environment?

Related Posts

What Is AI Decisioning?

Every business makes decisions about people and transactions all day long. Should we approve this loan? Is this purchase fraud? Which customer should get this offer, and what should it be? For a long time, those decisions were made in one of two ways: a person reviewed each case by hand, or the company wrote fixed rules, like "approve anyone with a credit score above 700." Both work. Both also leave value on the table. The manual review is slow and hard to scale. The fixed rule can turn away good applicants and is slow to adapt when the market shifts. AI decisioning is a third way. What makes AI decisioning work Instead of relying on a single reviewer or a rigid rule, automated decisioning uses models that learn from data — studying how thousands of past cases turned out, finding the patterns that predict an outcome, and applying them to each new decision, often in real time. The result is faster, more consistent decisions. But a model on its own isn't the whole story. Getting real value from AI decisioning takes good data to learn from, AI analytics to generate insights, the tools to act on it and the governance to keep it compliant. What we've found is that the pieces only pay off when they work together, and that is where we're built differently. A model is only as good as what it learns from, and we pair your data with one of the deepest views of consumer and commercial credit: decades of full-file history and vetted attributes. Then we give you the tools to act on it. Use cases across your business Whether you're trying to grow your customer base, reduce fraud, manage lending risk, or improve collections, automated decisioning brings all the pieces together to make more accurate, consistent and explainable decisions at scale. Fraud and Identity A fraudulent transaction that slips through costs money and erodes trust. Rules are static, and fraudsters move fast. They'll probe boundaries, find the blind spots and move to the next scheme. By the time the rules are updated, they're already three steps ahead. How AI decisioning changes this: AI fraud detection with real-time risk scoring and decisioning across transactions and customer interactions Intelligence that continuously learns from results to help adapt fraud strategies as threats evolve Reduced false positives and less friction for customers at account opening and checkout Identity verification tools that confirm someone is who they say they are without slowing down the experience Credit and Lending Loan approval is where the relationship begins. Credit risk decisioning helps lenders find that delicate balance between approving enough people to grow, but carefully enough to manage risk. Missing that balance means turning away good customers or taking on losses that are difficult to absorb. How AI decisioning changes this: Increased approval opportunities for creditworthy applicants without increasing overall risk Models you can update and deploy quickly as market conditions change, rather than waiting months Ability to run "what-if" scenarios to test how a new strategy would have performed on your historical data before putting it live Collections Which customer should your team reach out to today? Through which channel? What kind of message? If you reach out too aggressively, you push someone who might have recovered into default. If you wait too long, you lose them. If you call someone at work, they resent you; if you text, they might ignore it. If you offer a payment plan, they might accept it, but only if the terms make sense to their financial situation. How AI decisioning changes this: Optimized next-best-action and contact-channel strategies for each individual customer Improved recovery potential through better targeting Less time spent on accounts with a lower propensity to pay, freeing your team for higher-impact cases Ability to segment and test new strategies before rollout Customer Acqusition Finding the right customers is about reaching the right people with the right offer at the right time. To stay competitive, it’s now a requirement to balance growth with risk while creating a seamless experience converting prospects into customers. How AI decisioning changes this: More precise prospect targeting using credit, behavioral, and alternative data, where permitted, to identify consumers most likely to respond Personalized offers delivered in real time Dynamic decision strategies that can be updated quickly as market conditions and customer behavior change Ongoing testing and optimization of acquisition strategies to improve campaign performance and support customer lifetime value Driving results with AI decisioning Every customer interaction is a decision. Businesses that can adapt quickly will be better positioned to grow, manage risk, and deliver the experiences customers expect. The technology will continue to evolve, but the goal remains the same: making informed decisions that balance business objectives, risk, and customer experience. Learn more about our decisioning software

July 27, 2026 by Zohreen Ismail
Why Innovation Matters for Members First Credit Union

Learn how Members First Credit Union uses innovation and data-driven insights to better serve members and expand financial opportunity.

July 24, 2026 by Scarlet Nickel
Ask the Expert: Unlocking the ROI of alternative data with Natasha Madan and Julius Heim

A visibility gap lenders can't afford to ignore Alternative data is often associated with thin-file or credit invisible consumers. But its value extends far beyond those segments. Experian's Clarity Services database includes approximately one in five credit-active consumers, including one in four consumers with prime-and-above credit profiles. That means lenders may be missing important signals, not only for emerging borrowers, but also for applicants who appear well qualified using traditional bureau data alone. Consider two consumers with the same credit score. Based on traditional credit data, they may appear equally creditworthy. But when Clarity data is added, one consumer may demonstrate stable repayment behavior while another shows recent defaults on alternative finance products. The credit score hasn't changed, but the decisioning context has. That's where alternative data creates value: helping lenders distinguish between consumers who look similar on paper but represent very different levels of risk and opportunity. In this Ask the Expert session, Experian’s Julius Heim, Vice President of Analytics Product Build, Innovation and Scores, and Natasha Madan, Senior Director, Analytics Consulting, explain how different alternative data assets solve different business challenges and why the greatest return comes from using them together throughout the credit lifecycle. What that visibility gap is really costing lenders Better visibility matters because every lending decision carries consequences. Without alternative data, lenders may approve applicants whose repayment behavior suggests elevated risk but isn't reflected in a traditional credit file. Without cash flow insights, they may decline consumers who appear thin file on bureau data despite demonstrating strong income and responsible financial management. The result is a two-sided cost: avoidable bad debt on one side and missed growth opportunities on the other. But ROI extends beyond approvals alone. It also appears through stronger marketing strategies, improved conversion, reduced friction and more precise risk segmentation throughout the lending lifecycle. "ROI can mean many things ... marketing to the right people, achieving better approval rates, reducing risk, getting less friction and overall profitability."Julius Heim, Vice President of Analytics Product Build, Innovation and Scores Where alternative data creates ROI Improve approval strategies Use additional consumer signals to recover creditworthy applicants while avoiding unnecessary declines. Reduce portfolio risk Identify elevated repayment risk earlier through enhanced visibility beyond traditional bureau data. Improve portfolio performance Increase conversion, reduce friction and strengthen profitability across the credit lifecycle. Different data. Different jobs. Not all alternative data solves the same problem. Clarity Services can help lenders strengthen decisions early in the customer journey. It provides additional visibility during prospecting and acquisition, helping identify potential risk before an application moves through the underwriting process. Cash flow insights can provide value in a different way. When traditional credit information offers part of the picture, consumer-permissioned cash flow data can provide greater insight into income, spending patterns and financial capacity. That makes it especially valuable as a second look during underwriting. Together, these complementary data assets help lenders improve decisioning throughout the credit lifecycle. They can support acquisition, underwriting, account management and collections while building on the trusted foundation of traditional bureau data. Research also continues to demonstrate measurable lift when cash flow insights are combined with traditional credit information. "I recently did a study with a client where we actually saw a 20% lift in KS [Kolmogorov-Smirnov] above and beyond credit bureau data. Again, the bureau data itself was very predictive. But even from the cash flow data, we still got a 20% lift, which is an amazing stat." Julius Heim, Vice President of Analytics Product Build, Innovation and Scores The greatest value comes from using these data sources together for a more holistic consumer view. Start with proof, then build Adopting alternative data doesn't have to begin with a large transformation. A practical first step is a data study. By comparing current decision strategies with enhanced data, lenders can identify where additional visibility creates measurable lift within their own portfolios. This approach allows institutions to validate results before making broader operational changes. Every lender has different workflows, technology environments and business priorities. A flexible implementation strategy helps organizations incorporate new data in ways that support existing processes rather than disrupting them. Three ways to get started Run a data study Benchmark current decision strategies and quantify potential lift. Start simple Begin with targeted data attributes or proven scores before expanding to more advanced use cases. Build with confidence Scale implementation based on measured business outcomes and organizational priorities. This approach allows lenders to validate results, build confidence and expand their strategy over time. Explore alternative data with a trusted partner Every lending decision benefits from better consumer insight. Experian helps lenders combine trusted credit data with alternative data, cash flow insights and advanced analytics to strengthen decisioning, improve portfolio performance and uncover new opportunities for growth. Whether you're evaluating alternative data for the first time or expanding an existing strategy, Experian can help you identify where additional consumer insight can create measurable business value. Learn more Contact us About our experts Julius Heim Vice President of Analytics Product Build, Innovation and Scores, Experian Julius Heim works at the intersection of financial services, analytics and innovation. He focuses on leveraging data to drive smarter decision-making and support more inclusive financial ecosystems. Julius brings a practical perspective on how organizations can translate insights into real-world impact, with particular interest in emerging trends across fintech, credit, and the use of alternative data, such as cash-flow data, across the credit lifecycle. Previously, he served as Head of Analytics on the lender side and held roles in insurance analytics earlier in his career. Natasha Madan Senior Director, Analytics Consulting, Experian Natasha Madan partners with lenders to drive smarter, data-driven credit and risk decisions. She specializes in leveraging alternative data and advanced analytics to help organizations improve portfolio performance, optimize customer acquisition, and expand responsible access to credit. During her 15 years at Experian, Natasha has held leadership roles spanning data analytics, product analytics and consulting, giving her a broad perspective of how data can be leverage to solve complex business challenges. She has worked with a diverse range of lenders – including banks, credit unions, fintechs and specialty finance companies to develop analytics strategies that optimize customer acquisition, underwriting and portfolio management. Natasha is passionate about helping organizations unlock the full potential of data to improve both business outcomes and consumer financial inclusion.

July 24, 2026 by Julie Lee

Subscribe to our Newsletter

Enter your name and email for the latest updates.

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.

Subscribe to our Newsletter

Don't miss out on the latest industry trends and insights!
Subscribe