Data & Analytics

Model explainability has become a hot topic as lenders look for ways to use artificial intelligence (AI) to improve their decision-making. Within credit decisioning, machine learning (ML) models can often outperform traditional models at predicting credit risk.  ML models can also be helpful throughout the customer lifecycle, from marketing and fraud detection to collections optimization. However, without explainability, using ML models may result in unethical and illegal business practices.  What is model explainability?  Broadly defined, model explainability is the ability to understand and explain a model's outputs at either a high level (global explainability) or for a specific output (local explainability).1  Local vs global explanation: Global explanations attempt to explain the main factors that determine a model's outputs, such as what causes a credit score to rise or fall. Local explanations attempt to explain specific outputs, such as what leads to a consumer's credit score being 688. But it's not an either-or decision — you may need to explain both.  Model explainability can also have varying definitions depending on who asks you to explain a model and how detailed of a definition they require. For example, a model developer may require a different explanation than a regulator.  Model explainability vs interpretability  Some people use model explainability and interpretability interchangeably. But when the two terms are distinguished, model interpretability may refer to how easily a person can understand and explain a model's decisions.2 We might call a model interpretable if a person can clearly understand:  The features or inputs that the model uses to make a decision.  The relative importance of the features in determining the outputs.  What conditions can lead to specific outputs.  Both explainability and interpretability are important, especially for credit risk models used in credit underwriting. However, we will use model explainability as an overarching term that encompasses an explanation of a model's outputs and interpretability of its internal workings below.  ML models highlight the need for explainability in finance  Lenders have used credit risk models for decades. Many of these models have a clear set of rules and limited inputs, and they might be described as self-explanatory. These include traditional linear and logistic regression models, scorecards and small decision trees.3  AI analytics solutions, such as ML-powered credit models, have been shown to better predict credit risk. And most financial institutions are increasing their budgets for advanced analytics solutions and see their implementation as a top priority.4  However, ML models can be more complex than traditional models and they introduce the potential of a “black box." In short, even if someone knows what goes into and comes out of the model, it's difficult to explain what's happening without an in-depth analysis.  Lenders now have to navigate a necessary trade-off. ML-powered models may be more predictive, but regulatory requirements and fair lending goals require lenders to use explainable models.  READ MORE: Explainability: ML and AI in credit decisioning  Why is model explainability required?  Model explainability is necessary for several reasons:  To comply with regulatory requirements: Decisions made using ML models need to comply with lending and credit-related, including the Fair Credit Reporting Act (FCRA) and Equal Credit Opportunity Act (ECOA). Lenders may also need to ensure their ML-driven models comply with newer AI-focused regulations, such as the AI Bill of Rights in the U.S. and the E.U. AI Act.  To improve long-term credit risk management: Model developers and risk managers may want to understand why decisions are being made to audit, manage and recalibrate models.  To avoid bias: Model explainability is important for ensuring that lenders aren't discriminating against groups of consumers.  To build trust: Lenders also want to be able to explain to consumers why a decision was made, which is only possible if they understand how the model comes to its conclusions.  There's a real potential for growth if you can create and deploy explainable ML models. In addition to offering a more predictive output, ML models can incorporate alternative credit data* (also known as expanded FCRA-regulated data) and score more consumers than traditional risk models. As a result, the explainable ML models could increase financial inclusion and allow you to expand your lending universe.  READ MORE: Raising the AI Bar  How can you implement ML model explainability?  Navigating the trade-off and worries about explainability can keep financial institutions from deploying ML models. As of early 2023, only 14 percent of banks and 19 percent of credit unions have deployed ML models. Over a third (35 percent) list explainability of machine learning models as one of the main barriers to adopting ML.5  Although a cautious approach is understandable and advisable, there are various ways to tackle the explainability problem. One major differentiator is whether you build explainability into the model or try to explain it post hoc—after it's trained.  Using post hoc explainability  Complex ML models are, by their nature, not self-explanatory. However, several post hoc explainability techniques are model agnostic (they don't depend on the model being analyzed) and they don't require model developers to add specific constraints during training.  Shapley Additive Explanations (SHAP) is one used approach. It can help you understand the average marginal contribution features to an output. For instance, how much each feature (input) affected the resulting credit score.  The analysis can be time-consuming and expensive, but it works with black box models even if you only know the inputs and outputs. You can also use the Shapley values for local explanations, and then extrapolate the results for a global explanation.  Other post hoc approaches also might help shine a light into a black box model, including partial dependence plots and local interpretable model-agnostic explanations (LIME).  READ MORE: Getting AI-driven decisioning right in financial services  Build explainability into model development  Post hoc explainability techniques have limitations and might not be sufficient to address some regulators' explainability and transparency concerns.6 Alternatively, you can try to build explainability into your models. Although you might give up some predictive power, the approach can be a safer option.  For instance, you can identify features that could potentially lead to biased outcomes and limit their influence on the model. You can also compare the explainability of various ML-based models to see which may be more or less inherently explainable. For example, gradient boosting machines (GBMs) may be preferable to neural networks for this reason.7  You can also use ML to blend traditional and alternative credit data, which may provide a significant lift — around 60 to 70 percent compared to traditional scorecards — while maintaining explainability.8  READ MORE: Journey of an ML Model  How Experian can help  As a leader in machine learning and analytics, Experian partners with financial institutions to create, test, validate, deploy and monitor ML-driven models. Learn how you can build explainable ML-powered models using credit bureau, alternative credit, third-party and proprietary data. And monitor all your ML models with a web-based platform that helps you track performance, improve drift and prepare for compliance and audit requests. *When we refer to “Alternative Credit Data," this refers to the use of alternative data and its appropriate use in consumer credit lending decisions, as regulated by the Fair Credit Reporting Act. Hence, the term “Expanded FCRA Data" may also apply and can be used interchangeably.  1-3. FinRegLab (2021). The Use of Machine Learning for Credit Underwriting  4. Experian (2022). Explainability: ML and AI in credit decisioning  5. Experian (2023). Finding the Lending Diamonds in the Rough  6. FinRegLab (2021). The Use of Machine Learning for Credit Underwriting  7. Experian (2022). Explainability: ML and AI in credit decisioning  8. Experian (2023). Raising the AI Bar 

While today’s consumers expect a smooth, frictionless digital experience, many financial institutions still rely on outdated technology and manual reviews to acquire new customers. These old processes can prevent lenders from making accurate and timely credit decisions, leading to lost opportunities, revenue, and goodwill. By optimizing their customer acquisition strategies, financial institutions can allocate their resources effectively and say yes to consumers faster. This guide will walk you through the current challenges facing customer acquisition and how robust optimization strategies can help. Current challenges in customer acquisition To stay competitive and engage high-value customers, you’ll need an efficient customer acquisition process that weeds out both fraudulent actors and risky consumers. However, achieving this balancing act comes with a unique set of challenges. Because today’s consumers can access goods and services almost anywhere online at any time, more than 54 percent of customers expect a heightened digital and frictionless experience. Failing to meet this expectation can lead to huge losses for lenders. Some of the most common challenges in customer acquisition include: Although 52 percent of consumers prefer digital banking options over visiting branches in person, many lenders still rely on paper documents, which can add weeks to the onboarding process. Requiring consumers to provide substantial information about themselves during an application process can lead to abandoned applications. 67 percent of consumers will leave an application if they experience complications. Verifying consumer identities is growing increasingly important. In fact, about 35 percent of customers drop out of digital onboarding because their identity can't be confirmed. Poorly defined campaign planning can cause businesses to market to the wrong population segments, resulting in wasted time and resources. What is optimization for customer acquisition?  Customer acquisition optimization is the process of implementing new methods and solutions to make acquiring new customers more efficient and cost-effective. For lenders, this means streamlining steps in the credit decisioning process to focus on the right prospects and reduce friction. What types of processes can be optimized for customer acquisition?  You might be surprised just how many processes can be optimized for customer acquisition. Here are just a few examples: Having a holistic view of consumers allows you to take the guesswork out of targeting so you can better identify and engage high-potential customers. Utilizing predictive and lifestyle data enables you to pinpoint a more precisely segmented audience for marketing. Digital application solutions that reach across multiple channels, allowing applicants to leave one channel and pick up right where they left off in another. Real-time identity verification and fraud detection during onboarding and after, helping expedite approvals and mitigate risks. Utilizing API integration to leverage multiple metrics beyond credit scores when screening applicants' financial situation. Building custom risk models that pair to your existing data so you can say yes to more customers and better manage portfolio risk. Benefits of customer acquisition optimization Optimization can bring numerous benefits to your business, providing a faster return on investment. Here are some examples. By better pinpointing your marketing through predictive and lifestyle data, you can achieve increased conversions. Faster onboarding with less friction helps retain more customers. Real-time fraud detection and identity verification reduce customer roadblocks, allowing you to realize significant growth. Custom risk models and decisioning platforms can pair your data with additional data elements, providing more than just a credit score rating for your applicants. This can help you say yes to more customers. Using AI and machine learning tools will reduce the need for manual reviews and thus increase booking rates and applications. A real-life example of these benefits can be found with the Michigan State University Federal Credit Union (MSUFCU.) With over $7.2 billion in assets and 330,000 members, the client was manually reviewing all its applications. Experian reviewed the client's risk levels and approvals, comparing their risk and bankruptcy scores to determine which were most predictive. This analysis led Experian to recommend a new decisioning platform (PowerCurve Originations®) for instant credit decisions, an alternative data score tool, and Experian Advisory Services for risk-based pricing. After implementing these optimization solutions, MSUFCU saw a 55 percent increase in average monthly automations, four times improved online application response time and began competing more effectively in the marketplace. How Experian can help Experian offers a number of customer acquisition tools, allowing companies to be more responsive in an increasingly competitive market, while still reducing fraud risk. These tools include: Acquisition optimization marketing Experian offers a web-based platform that lets clients manage their marketing efforts all in the same place. You can upload and enhance client files, identify lookalike prospects, and use firmographic and credit data to get a holistic view of your clients and your prospects. Data-driven acquisition and decisioning engine PowerCurve Originations® is a data-driven decisioning engine that accepts applications from multiple channels, automates data collection and verification and proactively monitors decision results. It's flexible enough to reach across multiple channels, letting customers set aside their application in one digital channel and resume where they left off in another. It also provides businesses with access to comprehensive data assets, proactive monitoring and streamlined development with minimal coding. Enhanced fraud detection and identity verification Experian's Precise ID® is a risk-based fraud detection and prevention platform that provides analytics to accurately verify customers and mitigate fraud loss behind the scenes, ensuring a smoother onboarding process. Robust consumer attributes for better customized models Experian gives clients access to a wider berth of consumer attributes, helping you better screen applicants beyond just looking at credit scores. Trended 3DTM attributes let you uncover unique patterns in consumers' behavior over time, allowing you to manage portfolio risk, build better models and determine the next best actions. Premier AttributesSM aggregates credit data at the most granular and meaningful levels to provide clear insights into consumer credit behavior. It encompasses more than 2,100 attributes across 51 industries to help you develop highly predictive custom models. Enterprise-wide credit decisioning engine Experian's enterprise-wide credit decision platform lets you combine machine learning with proprietary data to return optimized decisions and quickly respond to requests. Robust credit decisioning software lets you convert data into meaningful actions and strategies. With Experian's machine learning decisioning options, companies are realizing a 25 percent reduction in manual reviews, a 25 percent increase in loan and credit applications and a 26 percent increase in booking rates. Highly predictive custom models Experian's Ascend Intelligence ServicesTM can help you create highly predictive custom models that create sophisticated decisioning strategies, allowing you to accurately predict risk and make the best decisions fast. This end-to-end suite of solutions lets you achieve a more granular view of every application and grow portfolios while still minimizing risk. Experian can help optimize your customer acquisition Experian provides a suite of decisioning engines, consumer attributes and customized modeling to help you optimize your customer acquisition process. These tools allow businesses to better target their marketing efforts, streamline their onboarding with less friction and improve their fraud detection and mitigation efforts. The combination can deliver a powerful ROI. Learn more about Experian's customer acquisition solutions. Learn more

In today's fast-paced digital world, the risk of fraud across all industries is a constant threat. The traditional methods of fraud detection are no longer sufficient, as fraudsters become increasingly sophisticated in their attacks. However, with artificial intelligence (AI) and machine learning (ML) solutions, financial institutions can stay one step ahead of fraudsters. AI and machine learning-equipped fraud detection tools have the ability to identify suspicious activity and patterns of fraud that are imperceptible to the human brain. In this blog post, we’ll dive into the significance of AI and machine learning in fraud detection and how these solutions are uniquely equipped to handle the demands of modern-day risk management. Understanding artificial intelligence and machine learning AI and machine learning solutions are transformative technologies that are reshaping the landscape of many industries. AI, at its core, is a field of computer science that simulates human intelligence in machines, enabling them to learn from experience and perform tasks that normally require human intellect. Machine learning, a subset of AI, is the science of getting computers to learn and act like humans do, but with minimal human intervention. They can analyze vast amounts of data within seconds, identifying patterns and trends that would be impossible for a human to recognize. When it comes to fraud detection, this ability is invaluable.  Advantages of fraud detection using machine learning AI and machine learning have several benefits that make them valuable in fraud detection. One significant advantage is that these technologies can recognize patterns that are too complex for humans to identify. By running through a vast set of data points, these solutions can pinpoint anomalous behavior, and thereby prevent financial losses. AI analytics tools are adept at monitoring complex networks, detecting the dispersion of attacks that may involve multiple individuals and entities, and correlating activity patterns that would otherwise be hidden. Machine learning algorithms can take these patterns and turn them into mathematical models that help identify instances of fraud before the damage takes place. Secondly, they continuously learn from new data, which allows them to become more efficient in identifying fraud as they process more data. Thirdly, they automate fraud mitigation processes, which significantly reduces the need for manual interventions that may consume valuable time and resources. Another significant benefit of machine learning is its analytics capabilities, which allow organizations to gain valuable insights into customer behavior and fraud patterns. With AI analytics, they can detect and investigate fraudulent activities in real-time, and combine it with other tools to help detect and mitigate fraud risk. For example, in financial services, AI fraud detection can help banks and financial service providers detect and prevent fraud in their systems, add value to their services and improve customer satisfaction. The future of fraud detection and machine learning The rate at which technology is evolving means that machine learning and AI fraud detection will become increasingly important in the future. In the next few years, we can expect a more sophisticated level of fraud detection using unmanned machine systems, robotics process automation, and more. Ultimately, this will improve the efficiency and effectiveness of fraud detection. AI-based fraud management solutions are taking center stage. Organizations must leverage advanced machine learning and AI analytics solutions to prevent and mitigate cyber risks and comply with regulatory mandates. The benefits extend far beyond the financial bottom line to improving the safety and security of customers. AI and machine learning solutions offer accurate, efficient and proactive routes to managing the risk of fraud in an ever-changing environment. How can Experian® help Integrating machine learning for fraud detection represents a significant advancement in cybersecurity. Fraud management solutions detect, prevent and manage fraud across all industries, including financial services, healthcare and telecommunications. With the advancement of technology, fraud management solutions now integrate machine learning to improve their processes. Experian® provides fraud prevention solutions, including machine learning models and AI analytics, which can help more effectively mitigate fraud risk, streamline fraud investigations and create a more secure digital environment for all. With Experian’s AI analytics, risk mitigation tools and fraud management solutions, organizations can stay one step ahead of fraudsters and protect their brand reputation, customer trustworthiness and corporate data. Embracing these solutions can save organizations from significant losses, reputational damage and regulatory scrutiny. To learn more about how to future-proof your business and safeguard your customers from fraud, check out Experian’s robust suite of fraud prevention solutions. Want to hear what our industry experts think? Check out this on-demand webinar on artificial intelligence and machine learning strategies. Learn more Watch webinar *This article includes content created by an AI language model and is intended to provide general information.

Well-designed underwriting strategies are critical to creating more value out of your member relationships and driving growth for your business. But what makes an advanced underwriting strategy? It’s all about the data, analytics, and the people behind it. How a credit union achieved record loan growth Educational Federal Credit Union (EdFed) is a member-owned cooperative dedicated to serving the financial needs of school employees, students, and parents within the education community. After migrating to a new loan origination system, the credit union wanted to design a more profitable underwriting strategy to increase efficiency and grow their business. EdFed partnered with Experian to design an advanced underwriting strategy using our vast data sources, advanced analytics, and recommendations for greater automation. After 30 months of implementing the new loan origination system and underwriting strategies, the credit union increased their loans by 32% and automated approvals by 21%. “The partnership provided by Experian, backed by analytics, makes them the dream resource for our growth as a credit union. It isn’t just the data… it’s the people.” – Michael Aubrey, SVP Lending at Educational Federal Credit Union Learn more about how Experian can help you enhance your underwriting strategy. Learn more

If you’re a manager at a business that lends to consumers or otherwise extends credit, you certainly are aware that 10-15% of your current customers and prospective future customers are among the approximately 27 million consumers who are now – or will soon be -- fitting another bill into their monthly budgets. Early in the COVID-19 pandemic, the government issued a pause on federal student loan payments and interest. Now that the payment pause has expired, millions of Americans face a new bill averaging more than $200. Will they pay you first? If this is your concern, you aren’t alone: Experian recently held a webinar that discussed how the end of the student loan pause might affect businesses. When we surveyed the webinar attendees,  nearly 3 out of 4 responses included Risk Management as a main concerns now. Another top concern is about credit scores. Lenders and investors use credit scores – bureau scores such FICO® or VantageScore® credit score or custom credit scores proprietary to their institution – to predict credit default risk. The risk managers at those companies want to know to what extent they can continue to rely on those scores as Federal student loan payments come due and consumers experience payment shock. I’ve analyzed a large and statistically meaningful sample (10% of the US consumer population in Experian’s Ascend Sandbox) to shed some light on that question. As background information, the average consumer with student loans had lower scores before the pandemic than the average of the general population. One of my Experian colleagues has explored some of the reasons at https://www.experian.com/blogs/ask-experian/research/average-student-loan-payments). Here are some of the things we can learn from comparing the credit data of the two groups of people. I looked at a period from 2019 and from 2023 to see how things have changed: Average credit scores increased during the pandemic, continuing a long-term trend during which more Americans have been willing and able to meet all their obligations. During the COVID Public Health Emergency, consumers with student loans brought up their scores by an average of 25 points; that was 7 points more than consumers without student loans. Another way to look at it: in 2019, consumers with student loans had credit scores 23 points lower than consumers without. By 2023, that difference had shrunk to 16 points. Experian research shows that there will be little immediate impact on credit scores when the new bills come due. Time will tell whether these increased credit scores accurately reflect a reduction in the risk that consumers will default on other bills such as auto loans or bankcards soon, even as some people fit student loan bills into their budgets. It is well-known that many people saved money during the public health emergency. Since then, the personal savings rate has fallen from a pandemic high of 32% to levels between 3% and 5% this year – lower than at any point since the 2009 recession. In an October 2023 Experian survey, only 36% of borrowers said they either set aside funds or they planned using other financial strategies specifically for the resumption of their student loan payments. Additional findings from that study can be found here. Furthermore, there are changes in the way your customers have used their credit cards over the last four years:   Consumers’ credit card balances have increased over the last four years. Consumers with student loans have balances that are on average $282 (4%) more now than in 2019. That is a significantly smaller increase than for consumers without student loans, whose total credit card debt increased by an average of $1,932 (26%). Although their balances increased, the ratio of consumers’ total revolving debt balances to their credit limits (utilization) changed by less than 1% for both consumers with student loans and consumers without. In 2019, the utilization ratio was 9.8 percentage points lower for consumers with student loans than consumers without. Four years later, the difference is nearly the same (9.6 points). We can conclude that many student loan borrowers have been very responsible with credit during the Public Health Emergency. They may have been more mindful of their credit situation, and some may have planned for the day when their student loan payments will be due. As the student loan pause come to an end, there are a few things that lenders and other businesses should be doing to be ready: Even if you are not a student loan lender, it is important to stay on top of the rapidly evolving student loan environment. It affects many of your customers, and your business with them needs to adapt. Anticipate that fraudsters and abusers of credit will be creative now: periods of change create opportunities for them and you should be one step ahead. Build optimized strategies in marketing, account opening, and servicing. Consider using machine learning to make more accurate predictions. Those strategies should reflect trends in payments, balances, and utilization; older credit scores look at a single point in time. Continually refresh data about your customers—including their credit scores and important attributes related to payments, balances, and utilization patterns. Look for alternative data that will give you a leg up on the competition. In the coming weeks and months, Experian’s data scientists will monitor measures of performance of the scores and attributes that you depend on in your data-driven strategies — particularly focusing on the Kolmogorov-Smirnov (KS) statistics that will show changes in the predictive power of each score and attribute. (If you are a data-driven business, your data science team or a trusted partner should be doing the same thing with a more specific look at your customer base and business strategies.) In future reports and blog posts, we’ll shed light on the impact student loans are having on your customers and on your business. In the meantime, for more information about how to use data and advanced analytics to grow while controlling costs and risks, all while staying in compliance and providing a good customer experience, visit our website.

Over the past few decades, the financial industry has gone through significant changes. One of the most notable changes is the use of alternative credit data1 for lending. This type of data is becoming increasingly essential in consumer and small business lending. In this blog post, we’ll explore the importance of alternative credit data and the insights you can gain from our new 2023 State of Alternative Credit Data Report. Benefits and uses of alternative credit data and alternative lending Alternative credit data and alternative financial services offer substantial benefits to lenders, borrowers, and society as a whole. The primary advantage of alternative credit data is that it provides a more comprehensive and accurate credit history of the borrower. Unlike traditional credit data that focuses on a borrower’s financial past, alternative credit data includes information from non-traditional sources like rent payments, full-file public records, utility bills, and income and employment data. This additional data allows you to gain a better understanding of financial behavior and assess creditworthiness more accurately.Alternative credit data can be used throughout the loan lifecycle, from underwriting to servicing. In the underwriting phase, alternative credit data can help lenders expand their pool of potential borrowers, especially those who lack or have limited traditional credit history. Additionally, alternative credit data can help lenders identify risks and minimize fraud. In the servicing phase, alternative credit data can help lenders monitor financial health and provide relevant services and an enhanced customer experience.Alternative lending is critical for driving financial inclusion and profitability. Traditional credit models often exclude individuals who have limited or no access to credit, causing them to turn to high-cost alternatives like payday loans. Alternative credit data can provide a more accurate assessment of their ability to pay, making it easier for them to access affordable credit. This increased accessibility improves the borrower's financial health and creates new opportunities to expand your customer base. “Lenders can access credit data and real-time information about consumers’ incomes, employment statuses, and how they are managing their finances and get a more accurate view of a consumer’s financial situation than previously possible.”— Scott Brown, President of Consumer Information Services, Experian State of alternative credit data Our new 2023 State of Alternative Credit Data Report provides exclusive insight into the alternative lending market, new data sources, inclusive finance opportunities and innovations in credit attributes and scoring that are making credit scoring more accurate, transparent and inclusive. For instance, the use of machine learning algorithms and artificial intelligence is enabling lenders to develop more predictive alternative credit scoring models and enhance risk assessment.  Findings from the report include: 54% of Gen Z and 52% of millennials feel more comfortable using alternative financing options rather than traditional forms of credit.2 62% of financial institution firms are using alternative data to improve risk profiling and credit decisioning capabilities.3 Modern credit scoring methods could allow lenders to grow their pool of new customers by almost 20%.4 By understanding the power of alternative credit data and staying on top of the latest industry trends, you can widen your pool of borrowers, drive financial inclusion, and grow sustainably. Download now 1When we refer to “Alternative Credit Data,” this refers to the use of alternative data and its appropriate use in consumer credit lending decisions, as regulated by the Fair Credit Reporting Act. Hence, the term “Expanded FCRA Data” may also apply in this instance and both can be used interchangeably.2Experian commissioned Atomik Research to conduct an online survey of 2,001 adults throughout the United States. Researchers controlled for demographic variables such as gender, age, geographic region, race and ethnicity in order to achieve similar demographic characteristics reported in the U.S. census. The margin of error of the overall sample is +/-2 percentage points with a confidence level of 95 percent. Fieldwork took place between August 22 and August 28, 2023. Atomik Research is a creative market research agency. 3Experian (2022). Reaching New Heights with Financial Inclusion 4Oliver Wyman (2022). Financial Inclusion and Access to Credit

Growing deposits from existing customers and members is an ongoing priority for banks and credit unions. However, it can be challenging to identify the best candidates. Who among our customer base has significant deposit growth potential? Who among our member base has the financial capacity to take advantage of special offers? With an effective deposit growth strategy, you can find the best customers and members to engage.  What does an effective deposit growth strategy look like?  An effective bank and credit union deposit growth strategy is powered by differentiated data and digital engagement. Let’s take a closer look at each element:  Data: A comprehensive measurement of consumers’ income and insights into their banking behaviors can help you identify those with the greatest deposit growth potential. You can then use supplemental data, such as lifestyle and demographic data, to customize deposit offers based on your customers or members’ unique needs.  Digital engagement: To further personalize this experience, consider sending deposit offers through your mobile or online banking platforms when there are triggering events on their account. Not only does this optimize the digital experience, but it also helps boost the chances of your customers or members responding.  Finding the right partner  Experian’s solutions can help your business secure deposits and customer relationships in today’s crowded market, including Banking InsightsTM. Banking Insights provide greater visibility into integrated demand deposit account activity, such as checking and saving account inquiries, to help you better assess consumers’ financial stability. By using these insights to power your banking growth strategies, you can identify those with the financial capacity to bring in more deposits.  Read our e-book to learn about other solutions that can help you boost deposits, strengthen existing relationships, and provide seamless digital experiences. Read e-book

With great risk comes great reward, as the saying goes. But when it comes to business, there's huge value in reducing and managing that risk as much as possible to maximize benefits — and profits. In today's high-tech strategic landscape, financial institutions and other organizations are increasingly using risk modeling to map out potential scenarios and gain a clearer understanding of where various paths may lead. But what are risk models really, and how can you ensure you're creating and using them correctly in a way that actually helps you optimize decision-making? Here, we explore the details. What is a risk model? A risk model is a representation of a particular situation that's created specifically for the purpose of assessing risk. That risk model is then used to evaluate the potential impacts of different decisions, paths and events. From assigning interest rates and amortization terms to deciding whether to begin operating in a new market, risk models are a safe way to analyze data, test assumptions and visualize potential scenarios. Risk models are particularly valuable in the credit industry. Credit risk models and credit risk analytics allow lenders to evaluate the pluses and minuses of lending to clients in specific ways. They are able to consider the larger economic environment, as well as relevant factors on a micro level. By integrating risk models into their decision-making process, lenders can refine credit offerings to fit the assessed risk of a particular situation. It goes like this: a team of risk management experts builds a model that brings together comprehensive datasets and risk modeling tools that incorporate mathematics, statistics and machine learning. This predictive modeling tool uses advanced algorithmic techniques to analyze data, identify patterns and make forecasts about future outcomes. Think of it as a crystal ball — but with science behind it. Your team can then use this risk model for a wide range of applications: refining marketing targets, reworking product offerings or reshaping business strategies. How can risk models be implemented? Risk models consolidate and utilize a wide variety of data sets, historical benchmarks and qualitative inputs to model risk and allow business leaders to test assumptions and visualize the potential results of various decisions and events. Implementing risk modeling means creating models of systems that allow you to adjust variables to imitate real-world situations and see what the results might be. A mortgage lender, for example, needs to be able to predict the effects of external and internal policies and decisions. By creating a risk model, they can test how scenarios such as falling interest rates, rising unemployment or a shift in loan acceptance rates might affect their business — and make moves to adjust their strategies accordingly. One aspect of risk modeling that can't be underestimated is the importance of good data, both quantitative and qualitative. Efforts to implement or expand risk modeling should begin with refining your data governance strategy. Maximizing the full potential of your data also requires integrating data quality solutions into your operations in order to ensure that the building blocks of your risk model are as accurate and thorough as possible. It's also important to ensure your organization has sufficient model risk governance in place. No model is perfect, and each comes with its own risks. But these risks can be mitigated with the right set of policies and procedures, some of which are part of regulatory compliance. With a comprehensive model risk management strategy, including processes like back testing, benchmarking, sensitivity analysis and stress testing, you can ensure your risk models are working for your organization — not opening you up to more risk. How can risk modeling be used in the credit industry? Risk modeling isn't just for making credit decisions. For instance, you might model the risk of opening or expanding operations in an underserved country or the costs and benefits of existing one that is underperforming. In information technology, a critical branch of virtually every modern organization, risk modeling helps security teams evaluate the risk of malicious attacks. Banking and financial services is one industry for which understanding and planning for risk is key — not only for business reasons but to align with relevant regulations. The mortgage lender mentioned above, for example, might use credit risk models to better predict risk, enhance the customer journey and ensure transparency and compliance. It's important to highlight that risk modeling is a guide, not a prophecy. Datasets can contain flaws or gaps, and human error can happen at any stage.. It's also possible to rely too heavily on historical information — and while they do say that history repeats itself, they don't mean it repeats itself exactly. That's especially true in the presence of novel challenges, like the rise of artificial intelligence. Making the best use of risk modeling tools involves not just optimizing software and data but using expert insight to interpret predictions and recommendations so that decision-making comes from a place of breadth and depth. Why are risk models important for banks and financial institutions? In the world of credit, optimizing risk assessment has clear ramifications when meeting overall business objectives. By using risk modeling to better understand your current and potential clients, you are positioned to offer the right credit products to the right audience and take action to mitigate risk. When it comes to portfolio risk management, having adequate risk models in place is paramount to meet targets. And not only does implementing quality portfolio risk analytics help maximize sales opportunities, but it can also help you identify risk proactively to avoid costly mistakes down the road. Risk mitigation tools are a key component of any risk modeling strategy and can help you maintain compliance, expose potential fraud, maximize the value of your portfolio and create a better overall customer experience. Advanced risk modeling techniques In the realm of risk modeling, the integration of advanced techniques like machine learning (ML) and artificial intelligence (AI) is revolutionizing how financial institutions assess and manage risk. These technologies enhance the predictive power of risk models by allowing for more complex data processing and pattern recognition than traditional statistical methods. Machine learning in risk modeling: ML algorithms can process vast amounts of unstructured data — such as market trends, consumer behavior and economic indicators — to identify patterns that may not be visible to human analysts. For instance, ML can be used to model credit risk by analyzing a borrower’s transaction history, social media activities and other digital footprints to predict their likelihood of default beyond traditional credit scoring methods. Artificial intelligence in decisioning: AI can automate the decisioning process in risk management by providing real-time predictions and risk assessments. AI systems can be trained to make decisions based on historical data and can adjust those decisions as they learn from new data. This capability is particularly useful in credit underwriting where AI algorithms can make rapid decisions based on market conditions. Financial institutions looking to leverage these advanced techniques must invest in robust data infrastructure, skilled personnel who can bridge the gap between data science and financial expertise, and continuous monitoring systems to ensure the models perform as expected while adhering to regulatory standards. Challenges in risk model validation Validating risk models is crucial for ensuring they function appropriately and comply with regulatory standards. Validation involves verifying both the theoretical foundations of a model and its practical implementation. Key challenges in model validation: Model complexity: As risk models become more complex, incorporating elements like ML and AI, they become harder to validate. Complex models can behave in unpredictable ways, making it difficult to understand why they are making certain decisions (the so-called "black box" issue). Data quality and availability: Effective validation requires high-quality, relevant data. Issues with data completeness, accuracy or relevance can lead to incorrect model validations. Regulatory compliance: With regulations continually evolving, keeping risk models compliant can be challenging. Different jurisdictions may have varying requirements, adding to the complexity of validation processes. Best practices: Regular reviews: Continuous monitoring and periodic reviews help ensure that models remain accurate over time and adapt to changing market conditions. Third-party audits: Independent reviews by external experts can provide an unbiased assessment of the risk model’s performance and compliance. These practices help institutions maintain the reliability and integrity of their risk models, ensuring that they continue to function as intended and comply with regulatory requirements. Read more: Blog post: What is model governance? How Experian can help Risk is inherent to business, and there's no avoiding it entirely. But integrating credit risk modeling into your operations can ensure stability and profitability in a rapidly evolving business landscape. Start with Experian's credit modeling services, which use expansive data, analytical expertise and the latest credit risk modeling methodologies to better predict risk and accelerate growth. Learn more *This article includes content created by an AI language model and is intended to provide general information.

This article was updated on November 9, 2023. Fraud – it’s a word that comes up in conversations across every industry. While there’s a general awareness that fraud is on the rise and is constantly evolving, for many the full impact of fraud is misunderstood and underestimated. At the heart of this challenge is the tendency to lump different types of fraud together into one big problem, and then look for a single solution that addresses it. It’s as if we’re trying to figure out how to un-bake a terrible cake instead of thinking about the ingredients and the process needed to put them together in the first place. This is the first of a series of articles in which we’ll look at some of the key ingredients that create different types of fraud, including first party, third party, synthetic identity, and account takeover. We’ll talk about why they’re unique and why we need to approach each one differently. At the end of the series, we’ll get a result that’s easier to digest. I had second thoughts about the cake metaphor, but in truth it really works. Creating a good fraud risk management process is a lot like baking. We need to know the ingredients and some tried-and-true methods to get the best result. With that foundation in place, we can look for ways to improve the outcome every time. Let’s start with a look at the best known type of fraud, third party. What is third-party fraud? Third-party fraud – generally known as identity theft – occurs when a malicious actor uses another person’s identifying information to open new accounts without the knowledge of the individual whose information is being used. When you consider first-party vs third-party fraud, or synthetic identity fraud, third-party stands out because it involves an identifiable victim that’s willing to collaborate in the investigation and resolution, for the simple reason that they don’t want to be responsible for the obligation made under their name. Third-party fraud is often the only type of activity that’s classified as fraud by financial institutions. The presence of an identifiable victim creates a high level of certainty that fraud has indeed occurred. That certainty enables financial institutions to properly categorize the losses. Since there is a victim associated with it, third party fraud tends to have a shorter lifespan than other types. When victims become aware of what’s happening, they generally take steps to protect themselves and intervene where they know their identity has been potentially misused. As a result, the timeline for third-party fraud is shorter, with fraudsters acting quickly to maximize the funds they’re able to amass before busting out. How does third-party fraud impact me? As the digital transformation continues, more and more personally identifiable information (PII) is available on the dark web due to data breaches and phishing scams. Given that consumer spending is expected to increase1, we anticipate that the amount of PII readily available to criminals will only continue to grow. All of this will lead to identity theft and increase the risk of third-party fraud. More than $43 billion in total losses was reported due to identity theft and fraud in the U.S. in 2022.2 Solving the third-party fraud problem We’ve examined one part of the fraud problem, and it is a complex one. With Experian as your partner, solving for it isn’t. Continuing my cake metaphor, by following the right steps and including the right ingredients, businesses can detect and prevent fraud. Third-party fraud detection and prevention involves two distinct steps. Analytics: Driven by extensive data that captures the ways in which people present their identity—plus artificial intelligence and machine learning—good analytics can detect inconsistencies, and patterns of usage that are out of character for the person, or similar to past instances of known fraud. Verification: The advantage of dealing with third-party fraud is the availability of a victim that will confirm when fraud is happening. The verification step refers to the process of making contact with the identity owner to obtain that confirmation and may involve identity resolution. It does require some thought and discipline to make sure that the contact information used leads to the identity owner—and not to the fraudster. In a series of articles, we’ll be exploring first-party fraud, synthetic identity fraud, and account takeover fraud and how a layered fraud management solution can help keep your business and customers safe and manage third-party fraud detection, first-party fraud, synthetic identity fraud, and account takeover fraud prevention. Let us know if you’d like to learn more about how Experian is using our identity expertise, data, and analytics to create robust fraud prevention solutions. Contact us 1 Experian Ascend Sandbox 2 2023 U.S. Identity and Fraud Report, Experian.

As the sophistication of fraudulent schemes increases, so must the sophistication of your fraud detection analytics. This is especially important in an uncertain economic environment that breeds opportunities for fraud. It's no longer enough to rely on old techniques that worked in the past. Instead, you need to be plugged into machine learning, artificial intelligence (AI) and real-time monitoring to stay ahead of criminal attempts. Your customers have come to expect cutting-edge security, and fraud analytics is the best way to meet — and surpass — those expectations. Leveraging these analytics can help your business better understand fraud techniques, uncover hidden insights and make more strategic decisions. What is fraud analytics? Fraud analytics refers to the idea of preventing fraud through sophisticated data analysis that utilizes tools like machine learning, data mining and predictive AI.1 These services can analyze patterns and monitor for anomalies that signal fraud attempts.2 While at first glance this may sound like a lot of work, it's necessary in today's technologically savvy culture. Fraud attempts are becoming more sophisticated, and your fraud detection services must do the same to keep up. Why is fraud analytics so important? According to the Experian® 2023 US Identity and Fraud Report, fraud is a growing issue that businesses cannot ignore, especially in an environment where economic uncertainty provides a breeding ground for fraudsters. Last year alone, consumers lost $8.8 billion — an increase of 30 percent over the previous year. Understandably, nearly two-thirds of consumers are at least somewhat concerned about online security. Their worries range from authorized push payment scams (such as phishing emails) to online privacy, identity theft and stolen credit cards. Unfortunately, while 75 percent of surveyed businesses feel confident in protecting against fraud, only 45 percent understand how fraud impacts their business. There's a lot of unearned confidence out there that can leave businesses vulnerable to attack, especially with nearly 70 percent of businesses admitting an increase in fraud loss in recent years. The types of fraud that businesses most frequently encounter include: Authorized push payment fraud: Phishing emails and other schemes that persuade consumers to deposit funds into fraudulent accounts. Transactional payment fraud: When fraudulent actors steal credit card or bank account information, for example, to make unauthorized payments. Account takeover: When a fraudster gains access to an account that doesn't belong to them and changes login details to make unauthorized transactions. First-party fraud: When an account holder uses their own account to commit fraud, like misrepresenting their income to get a lower loan rate. Identity theft: Any time a person's private information is used to steal their identity. Synthetic identity theft: When someone combines real and fake personal data to create an identity that's used to commit fraud. How can fraud analytics be used to help your business? More than 85% of consumers expect businesses to respond to their security and fraud concerns. A good portion of them (67 percent) are even ready to share their personal data with trusted sources to help make that happen. This means that investing in risk and fraud analytics is not only vital for keeping your business and customer data secure, but it will score points with your consumers as well. So how can your business utilize fraud analytics? Machine learning is a great place to start. Rather than relying on outdated rules-based analytic models, machine learning can vastly increase your speed in identifying fraud attempts. This means that when a new fraudulent trend emerges, your machine learning software can pinpoint it fast and flag your security team. Machine learning also lets you automatically analyze large data sets across your entire customer portfolio, improving customer experiences and your response time. In general, the best way for your business to use fraud analytics is by utilizing a multi-layered approach, such as the robust fraud management solutions offered by Experian. Instead of a one-size-fits-all solution, Experian lets you customize a framework of physical and digital data security that matches your business needs. This framework includes a cloud-based platform, machine learning for streamlined data analytics, biometrics and other robust identity-authentication tools, real-time alerts and end-to-end integration. How Experian can help Experian's platform of fraud prevention solutions and advanced data analytics allows you to be at the forefront of fraud detection. The platform includes options such as: Account takeover prevention. Account takeovers can go unnoticed without strong fraud detection. Experian's account takeover prevention tools automatically flag and monitor unusual activities, increase efficiency and can be quickly modified to adapt to the latest technologies. Bust-out fraud prevention. Experian utilizes proactive monitoring and early detection via machine learning to prevent bust-out fraud. Access to premium credit data helps enhance detection.  Commercial entity fraud prevention. Experian's Sentinel fraud solutions blend consumer and business datasets to create predictive insights on business legitimacy and credit abuse likelihood. First-party fraud prevention. Experian's first-party fraud prevention tools review millions of transactions to detect patterns, using machine learning to monitor credit data and observations. Global data breach protection. Experian also offers data breach protection services, helping you use turnkey solutions to build a program of customer notifications and identity protection. Identity protection. Experian offers identity protection tools that deliver a consistent brand experience across touchpoints and devices. Risk-based authentication. Minimize risk with Experian's adaptive risk-based authentication tools. These tools use front- and back-end authentication to optimize cost, risk management and customer experience. Synthetic identity fraud protection. Synthetic identity fraud protection guards against the fastest-growing financial crimes. Automated detection rules evaluate behavior and isolate traits to reduce false positives. Third-party fraud prevention. Experian utilizes third-party prevention analytics to identify potential identity theft and keep your customers secure. Your business's fraud analytics system needs to increase in sophistication faster than fraudsters are fine-tuning their own approaches. Experian's robust analytics solutions utilize extensive consumer and commercial data that can be customized to your business's unique security needs. Experian can help secure your business from fraud Experian is committed to helping you optimize your fraud analytics. Find out today how our fraud management solutions can help you. Learn more 1 Pressley, J.P. "Why Banks Are Using Advanced Analytics for Faster Fraud Detection," BizTech, July 25, 2023. https://biztechmagazine.com/article/2023/07/why-banks-are-using-advanced-analytics-faster-fraud-detection 2 Coe, Martin and Melton, Olivia. "Fraud Basics," Fraud Magazine, March/April 2022. https://www.fraud-magazine.com/article.aspx?id=4295017143

In a series of articles, we talk about different types of fraud and how to best solve for them. This article will explore first-party fraud and how it's similar to biting into a cookie you think is chocolate chip, only to find that it’s filled with raisins. The raisins in the cookie were hiding in plain sight, indistinguishable from chocolate chips without a closer look, much like first-party fraudsters. What is first-party fraud? First-party fraud refers to instances when an individual purposely misrepresents their identity in exchange for goods or services. In the financial services industry, it's often miscategorized as credit loss and written off as bad debt, which causes problems when organizations later try to determine how much they’ve lost to fraud versus credit risk. Common types of first-party fraud include: Chargeback fraud: Also known as "friendly fraud," chargeback fraud occurs when an individual knowingly makes a purchase with their credit card and then requests a chargeback from the issuer, claiming they didn't authorize the purchase. Application fraud: This takes place when an individual uses stolen or manipulated information to apply for a loan, credit card or job. In 2023, the employment sector accounted for 45% of all false document submissions — 70% of those who falsified their resumes still got hired. Fronting: Done to get cheaper rates, this form of insurance fraud happens when a young or inexperienced individual is deliberately listed as a named driver, when they're actually the main driver of the vehicle. Goods lost in transit fraud (GLIT): This occurs when an individual claims the goods they purchased online did not arrive. To put it simply, the individual is getting a refund for something they actually already received. A first-party fraudster can also recruit “money mules” — individuals who are persuaded to use their own information to obtain credit or merchandise on behalf of a larger fraud ring. This type of fraud has become especially prevalent as more consumers are active online. Money mules constitute up to 0.3% of accounts at U.S. financial institutions, or an estimated $3 billion in fraudulent transfers. How does it impact my organization? Firstly, there are often substantial losses associated with first-party fraud. An imperfect first-party fraud solution can also strain relationships with good customers and hinder growth. When lenders have to interpret actions and behavior to assess customers, there’s a lot of room for error and losses. Those same losses hinder growth when, as mentioned before, businesses anticipate credit losses that aren’t actually credit losses. This type of fraud isn’t a single-time event, and it doesn’t occur at just one point in the customer lifecycle. It occurs when good customers develop fraudulent intent, when new applicants who have positive history with other lenders have recently changed circumstances or when seemingly good applicants have manipulated their identities to mask previous defaults. Finally, first-party fraud impacts how your organization categorizes and manages risk – and that’s something that touches every department. Solving the first-party fraud problem First-party fraud detection requires a change in how we think about the fraud problem. It starts with the ability to separate first- and third-party fraud to treat them differently. Because first-party fraud doesn’t have a victim, you can’t work with the person whose information was stolen to confirm the fraud. Instead, you’ll have to implement a consistent monitoring system and make a determination internally when fraud is suspected. As we’ve already discussed, the fraud problem is complex. However with a partner like Experian, you can leverage the fraud risk management strategies required to perform a closer examination and the ability to differentiate between the types of fraud so you can determine the best course of action moving forward. Additionally, our robust fraud management solutions can be used for synthetic identity fraud and account takeover fraud prevention, which can help you minimize customer friction to improve and deepen your relationships while preventing fraud. Contact us if you’d like to learn more about how Experian is using our identity expertise, data and analytics to improve identity resolution and detect and prevent all types of fraud. Contact us

Model governance is growing increasingly important as more companies implement machine learning model deployment and AI analytics solutions into their decision-making processes. Models are used by institutions to influence business decisions and identify risks based on data analysis and forecasting. While models do increase business efficiency, they also bring their own set of unique risks. Robust model governance can help mitigate these concerns, while still maintaining efficiency and a competitive edge. What is model governance? Model governance refers to the framework your organization has in place for overseeing how you manage your development, model deployment, validation and usage.1 This can involve policies like who has access to your models, how they are tested, how new versions are rolled out or how they are monitored for accuracy and bias.2 Because models analyze data and hypotheses to make predictions, there's inherent uncertainty in their forecasts.3 This uncertainty can sometimes make them vulnerable to errors, which makes robust governance so important. Machine learning model governance in banks, for example, might include internal controls, audits, a thorough inventory of models, proper documentation, oversight and ensuring transparent policies and procedures. One significant part of model governance is ensuring your business complies with federal regulations. The Federal Reserve Board and the Office of the Comptroller of the Currency (OCC) have published guidance protocols for how models are developed, implemented and used. Financial institutions that utilize models must ensure their internal policies are consistent with these regulations. The OCC requirements for financial institutions include: Model validations at least once a year Critical review by an independent party Proper model documentation Risk assessment of models' conceptual soundness, intended performance and comparisons to actual outcomes Vigorous validation procedures that mitigate risk Why is model governance important — especially now? More and more organizations are implementing AI, machine learning and analytics into their models. This means that in order to keep up with the competition's efficiency and accuracy, your business may need complex models as well. But as these models become more sophisticated, so does the need for robust governance.3 Undetected model errors can lead to financial loss, reputation damage and a host of other serious issues. These errors can be introduced at any point from design to implementation or even after deployment via inappropriate usage of the model, drift or other issues. With model governance, your organization can understand the intricacies of all the variables that can affect your models' results, controlling production closely with even greater efficiency and accuracy. Some common issues that model governance monitors for include:2 Testing for drift to ensure that accuracy is maintained over time. Ensuring models maintain accuracy if deployed in new locations or new demographics. Providing systems to continuously audit models for speed and accuracy. Identifying biases that may unintentionally creep into the model as it analyzes and learns from data. Ensuring transparency that meets federal regulations, rather than operating within a black box. Good model governance includes documentation that explains data sources and how decisions are reached. Model governance use cases Below are just three examples of use cases for model governance that can aid in advanced analytics solutions. Credit scoring A credit risk score can be used to help banks determine the risks of loans (and whether certain loans are approved at all). Governance can catch biases early, such as unintentionally only accepting lower credit scores from certain demographics. Audits can also catch biases for the bank that might result in a qualified applicant not getting a loan they should. Interest rate risk Governance can catch if a model is making interest rate errors, such as determining that a high-risk account is actually low-risk or vice versa. Sometimes changing market conditions, like a pandemic or recession, can unintentionally introduce errors into interest rate data analysis that governance will catch. Security challenges One department in a company might be utilizing a model specifically for their demographic to increase revenue, but if another department used the same model, they might be violating regulatory compliance.4 Governance can monitor model security and usage, ensuring compliance is maintained. Why Experian? Experian® provides risk mitigation tools and objective and comprehensive model risk management expertise that can help your company implement custom models, achieve robust governance and comply with any relevant federal regulations. In addition, Experian can provide customized modeling services that provide unique analytical insights to ensure your models are tailored to your specific needs. Experian's model risk governance services utilize business consultants with tenured experience who can provide expert independent, third-party reviews of your model risk management practices. Key services include: Back-testing and benchmarking: Experian validates performance and accuracy, including utilizing statistical metrics that compare your model's performance to previous years and industry benchmarks. Sensitivity analysis: While all models have some degree of uncertainty, Experian helps ensure your models still fall within the expected ranges of stability. Stress testing: Experian's experts will perform a series of characteristic-level stress tests to determine sensitivity to small changes and extreme changes. Gap analysis and action plan: Experts will provide a comprehensive gap analysis report with best-practice recommendations, including identifying discrepancies with regulatory requirements. Traditionally, model governance can be time-consuming and challenging, with numerous internal hurdles to overcome. Utilizing Experian's business intelligence and analytics solutions, alongside its model risk management expertise, allows clients to seamlessly pass requirements and experience accelerated implementation and deployment. Experian can optimize your model governance Experian is committed to helping you optimize your model governance and risk management. Learn more here. References 1Model Governance," Open Risk Manual, accessed September 29, 2023. https://www.openriskmanual.org/wiki/Model_Governance2Lorica, Ben, Doddi, Harish, and Talby, David. "What Are Model Governance and Model Operations?" O'Reilly, June 19, 2019. https://www.oreilly.com/radar/what-are-model-governance-and-model-operations/3"Comptroller's Handbook: Model Risk Management," Office of the Comptroller of the Currency. August 2021. https://www.occ.treas.gov/publications-and-resources/publications/comptrollers-handbook/files/model-risk-management/pub-ch-model-risk.pdf4Doddi, Harish. "What is AI Model Governance?" Forbes. August 2, 2021. https://www.forbes.com/sites/forbestechcouncil/2021/08/02/what-is-ai-model-governance/?sh=5f85335f15cd

Data-driven machine learning model development is a critical strategy for financial institutions to stay ahead of their competition, and according to IDC, remains a strategic priority for technology buyers.  Improved operational efficiency, increased innovation, enhanced customer experiences and employee productivity are among the primary business objectives for organizations that choose to invest in artificial intelligence (AI) and machine learning (ML), according to IDC’s 2022 CEO survey.   While models have been around for some time, the volume of models and scale at which they are utilized has proliferated in recent years. Models are also now appearing in more regulated aspects of the business, which demand increased scrutiny and transparency.   Implementing an effective model development process is key to achieving business goals and complying with regulatory requirements. While ModelOps, the governance and life cycle management of a wide range of operationalized AI models, is becoming more popular, most organizations are still at relatively low levels of maturity. It's important for key stakeholders to implement best practices and accelerate the model development and deployment lifecycle.   Read the IDC Spotlight Challenges impeding machine learning model development  Model development involves many processes, from wrangling data, analysis, to building a model that is ready for deployment, that all need to be executed in a timely manner to ensure proper outcomes. However, it is challenging to manage all these processes in today’s complex environment.   Modeling challenges include:  Infrastructure: Necessary factors like storage and compute resources incur significant costs, which can keep organizations from evolving their machine learning capabilities.   Organizational: Implementing machine learning applications requires talent, like data scientists and data and machine learning engineers.  Operational: Piece meal approaches to ML tools and technologies can be cumbersome, especially on top of data being housed in different places across an organization, which can make pulling everything together challenging.  Opportunities for improvement are many While there are many places where individuals can focus on improving model development and deployment, there are a few key places where we see individuals experiencing some of the most time-consuming hang-ups.   Data wrangling and preparation   Respondents to IDC's 2022 AI StrategiesView Survey indicated that they spend nearly 22% of their time collecting and preparing data. Pinpointing the right data for the right purpose can be a big challenge. It is important for organizations to understand the entire data universe and effectively link external data sources with their own primary first party data. This way, stakeholders can have enough data that they trust to effectively train and build models.   Model building  While many tools have been developed in recent years to accelerate the actual building of models, the volume of models that often need to be built can be difficult given the many conflicting priorities for data teams within given institutions. Where possible, it is important for organizations to use templates or sophisticated platforms to ease the time to build a model and be able to repurpose elements that may already be working for other models within the business.   Improving Model Velocity Experian’s Ascend ML BuilderTM is an on-demand advanced model development environment optimized to support a specific project. Features include a dedicated environment, innovative compute optimization, pre-built code called ‘Accelerators’ that simply, guide, and speed data wrangling, common analyses and advanced modeling methods with the ability to add integrated deployment.  To learn more about Experian’s Ascend ML Builder, click here.   To read the full Technology Spotlight, download “Accelerating Model Velocity with a Flexible Machine Learning Model Development Environment for Financial Institutions” here.  Download spotlight *This article includes content created by an AI language model and is intended to provide general information. 

Changes in your portfolio are a constant. To accelerate growth while proactively identifying risk, you’ll need a well-informed portfolio risk management strategy. What is portfolio risk management? Portfolio risk management is the process of identifying, assessing, and mitigating risks within a portfolio. It involves implementing strategies that allow lenders to make more informed decisions, such as whether to offer additional credit products to customers or identify credit problems before they impact their bottom line. Leveraging the right portfolio risk management solution Traditional approaches to portfolio risk management may lack a comprehensive view of customers. To effectively mitigate risk and maximize revenue within your portfolio, you’ll need a portfolio risk management tool that uses expanded customer data, advanced analytics, and modeling. Expanded data. Differentiated data sources include marketing data, traditional credit and trended data, alternative financial services data, and more. With robust consumer data fueling your portfolio risk management solution, you can gain valuable insights into your customers and make smarter decisions. Advanced analytics. Advanced analytics can analyze large volumes of data to unlock greater insights, resulting in increased predictiveness and operational efficiency. Model development. Portfolio risk modeling methodologies forecast future customer behavior, enabling you to better predict risk and gain greater precision in your decisions. Benefits of portfolio risk management Managing portfolio risk is crucial for any organization. With an advanced portfolio risk management solution, you can: Minimize losses. By monitoring accounts for negative performance, you can identify risks before they occur, resulting in minimized losses. Identify growth opportunities. With comprehensive consumer data, you can connect with customers who have untapped potential to drive cross-sell and upsell opportunities. Enhance collection efforts. For debt portfolios, having the right portfolio risk management tool can help you quickly and accurately evaluate collections recovery. Maximize your portfolio potential Experian offers portfolio risk analytics and portfolio risk management tools that can help you mitigate risk and maximize revenue with your portfolio. Get started today. Learn more

From science fiction-worthy image generators to automated underwriting, artificial intelligence (AI), big data sets and advances in computing power are transforming how we play and work. While the focus in the lending space has often been on improving the AI models that analyze data, the data that feeds into the models is just as important. Enter: data-centric AI. What is a data-centric AI? Dr. Andrew Ng, a leader in the AI field, advocates for data-centric AI and is often credited with coining the term. According to Dr. Ng, data-centric AI is, ‘the discipline of systematically engineering the data used to build an AI system.’1 To break down the definition, think of AI systems as a combination of code and data. The code is the model or algorithm that analyzes data to produce a result. The data is the information you use to train the model or later feed into the model to request a result. Traditional approaches to AI focus on the code — the models. Multiple organizations download and use the same data sets to create and improve models. But today, continued focus on model development may offer a limited return in certain industries and use cases. A data-centric AI approach focuses on developing tools and practices that improve the data. You may still need to pay attention to model development but no longer treat the data as constant. Instead, you try to improve a model's performance by increasing data quality. This can be achieved in different ways, such as using more consistent labeling, removing noisy data and collecting additional data.2 Data-centric AI isn't just about improving data quality when you build a model — it's also part of the ongoing iterative process. The data-focused approach should continue during post-deployment model monitoring and maintenance. Data-centric AI in lending Organizations in multiple industries are exploring how a data-centric approach can help them improve model performance, fairness and business outcomes. For example, lenders that take a data-centric approach to underwriting may be able to expand their lending universe, drive growth and fulfill financial inclusion goals without taking on additional risk. Conventional credit scoring models have been trained on consumer credit bureau data for decades. New versions of these models might offer increased performance because they incorporate changes in the economic landscape, consumer behavior and advances in analytics. And some new models are built with a more data-centric approach that considers additional data points from the existing data sets — such as trended data — to score consumers more accurately. However, they still solely rely on credit bureau data. Explainability and transparency are essential components of responsible AI and machine learning (a type of AI) in underwriting. Organizations need to be able to explain how their models come to decisions and ensure they are behaving as expected. Model developers and lenders that use AI to build credit risk models can incorporate new high-quality data to supplement existing data sets. Alternative credit data can include information from alternative financial services, public records, consumer-permissioned data, and buy now, pay later (BNPL) data that lenders can use in compliance with the Fair Credit Reporting Act (FCRA).* The resulting AI-driven models may more accurately predict credit risk — decreasing lenders' losses. The models can also use alternative credit data to score consumers that conventional models can't score. Infographic: From initial strategy to results — with stops at verification, decisioning and approval — see how customers travel across an Automated Loan Underwriting Journey. Business benefit of using data-centric AI models Financial services organizations can benefit from using a data-centric AI approach to create models across the customer lifecycle. That may be why about 70 percent of businesses frequently discuss using advanced analytics and AI within underwriting and collections.3 Many have gone a step further and implemented AI. Underwriting is one of the main applications for machine learning models today, and lenders are using machine learning to:4 More accurately assess credit risk models. Decrease model development, deployment and recalibration timelines. Incorporate more alternative credit data into credit decisioning. AI analytics solutions may also increase customer lifetime value by helping lenders manage credit lines, increase retention, cross-sell products and improve collection efforts. Additionally, data-centric AI can assist with fraud detection and prevention. Case study: Learn how Atlas Credit, a small-dollar lender, used a machine learning model and loan automation to nearly doubled its loan approval rates while decreasing its credit risk losses. How Experian helps clients leverage data-centric AI for better business outcomes During a presentation in 2021, Dr. Ng used the 80-20 rule and cooking as an analogy to explain why the shift to data-centric AI makes sense.5 You might be able to make an okay meal with old or low-quality ingredients. However, if you source and prepare high-quality ingredients, you're already 80% of the way toward making a great meal. Your data is the primary ingredient for your model — do you want to use old and low-quality data? Experian has provided organizations with high-quality consumer and business credit solutions for decades, and our industry-leading data sources, models and analytics allow you to build models and make confident decisions. If you need a sous-chef, Experian offers services and has data professionals who can help you create AI-powered predictive analytics models using bureau data, alternative data and your in-house data. Learn more about our AI analytics solutions and how you can get started today. 1DataCentricAI. (2023). Data-Centric AI.2Exchange.scale (2021). The Data-Centric AI Approach With Andrew Ng.3Experian (2021). Global Insights Report September/October 2021.4FinRegLab (2021). The Use of Machine Learning for Credit Underwriting: Market & Data Science Context. 5YouTube (2021). A Chat with Andrew on MLOps: From Model-Centric to Data-Centric AI *Disclaimer: When we refer to “Alternative Credit Data," this refers to the use of alternative data and its appropriate use in consumer credit lending decisions, as regulated by the Fair Credit Reporting Act. Hence, the term “Expanded FCRA Data" may also apply in this instance and both can be used interchangeably.