Risk management 2.0

Balancing human touch and automation

‘The things that make me different are the things that make me, me.’?

Immortal words often attributed to that most eminent philosopher, Winnie the Pooh. And though a bear (not a human) and one of very little brain, his words feel more pertinent now than ever. Because the essence of the human experience is that no one else can think or see things from the same perspective. And that's precisely what makes humans indispensable in the world of risk management.

Yet, to streamline processes and boost efficiency, financial institutions are increasingly turning to AI and predictive analytics to assess and mitigate risk. These automated systems, while efficient, lack the distinct qualities that make humans so powerful – empathy, ethical judgment and an understanding of the evolving ways of the world.

In this blog, we'll delve deep into the evolving landscape of risk management – exploring how automation and human expertise can coexist harmoniously to create more robust, responsible, and ultimately successful risk management strategies.

The evolution of risk management ?

To understand the current state of risk management, we need to look at where it all began. Traditionally, it relied heavily on manual processes and historical data analysis, which, although valuable, were time-consuming and prone to human bias.

The introduction of automation marked a turning point. Advanced algorithms and machine learning promised faster, data-driven decision-making while reducing the potential for human error. Risk managers were excited about streamlined processes and more accurate predictions.

Since then, we’ve seen a huge upswing in automated risk management processes. In fact, a PwC report has estimated that by 2030, 30% of roles in the finance sector could be automated.

The rise of automation ?

In recent years, automation has transformed risk management practices, including the use of predictive modelling to identify patterns and anomalies that humans might otherwise miss. For example, AI can quickly scan social media activity and online behaviour, providing a holistic view of applicants for more accurate risk predictions.

Automated systems can also process swathes of data in seconds, providing risk managers with real-time insights into market trends or potential threats allowing portfolio managers to rebalance investments more effectively than ever.

However, automated systems lack human thought and empathy which can pose its own challenges.

Where automation falls short?

1. Lack of context: Automated systems may struggle to grasp nuanced situations, potentially leading to misjudgements. For instance, during the GameStop price surge in 2021, many systems continued selling GameStop shares, failing to understand that it was driven by retail investors on social media. This resulted in substantial losses for hedge funds that had short positions in the games retailer.

2. Data quality: Automated decisions rely on data quality; inaccurate inputs result in inaccurate outputs.

3. Out of the blue events: Automated models might not account for unprecedented events, leading to unexpected vulnerabilities. When, in 2020, every sector – from travel to healthcare and hospitality experienced unique challenges because of the pandemic, AI models, which tend to focus on individual sectors, could never have predicted the cascading effects on the entire economy.?

4. Ethical concerns: AI and automation raise ethical questions, especially when it comes to decision-making with significant consequences for people.

AI models may use proxy variables that indirectly capture sensitive attributes like race or gender, leading to bias. For example, a model might consider postcode or educational background as proxies for race or socioeconomic status, leading to discriminatory loan denials.

The human touch??

All of which means there’s still a host of reasons why people power still matters in the world of risk management. Human risk managers bring a wealth of skills and attributes to the table that simply can’t be replicated by machines:

1. Contextual understanding: Humans assess situations within a broader context. For instance, they understand how the Russia-Ukraine war affects global energy markets, a factor an AI system might overlook.

2. Ethical decision-making: Humans navigate complex ethical dilemmas. For example, by empathising with and understanding the perspective of the client in front of them, they can avoid making cultural generalisations, ensuring legal and moral decisions.

3. Adaptability: Humans are naturally adaptable and can respond effectively to unforeseen events or emerging risks that automated models may miss.

4. Communication: Effective communication with clients often requires human judgment and empathy. If you want a customer to come back to you time and time again, you need to build a rapport with them on a human level.

5. Learning and improvement: Human risk managers can seek feedback and recognise their shortcomings, helping them to continuously learn and improve their decision-making processes.

Striking the right balance

So, how do financial institutions strike the right balance between automation and human expertise in risk management? Here are some strategies:

1. Augmented intelligence: Embrace a model of "augmented intelligence," where humans and machines work alongside one another. This way, automation can aid human decision-making without replacing it entirely.

2. Continuous training: Invest in training and upskilling your risk management teams to harness the full potential of technology, while retaining their essential human skills.

3. Ethical frameworks: Develop clear ethical guidelines for the use of automation in risk management. Ensure that decisions align with both legal requirements and morality.

4. Regular audits: Implement regular audits of automated systems to identify and rectify biases, errors, or shortcomings.

5. Collaboration: Foster a culture of collaboration between data scientists, technologists, and risk managers to ensure seamless integration of technology into risk management processes.

The value of weaving the human touch into use of automated systems

Imagine a bank using an automated algorithm to assess a business loan application. The algorithm would evaluate credit scores, financial histories, and debt ratios but may overlook other circumstances and life events that impact a company’s ability to repay the loan. Perhaps a particular news story drove a sudden spike in demand, or a senior manager retired leading to a period of transition.

Whatever the situation, a human loan officer could step in, consider the broader context, and gain a better understanding of the situation. This crucial human touch would result in a more informed decision, potentially saving the bank from issuing a loan that could turn sour.

Automation has brought huge advantages, but it must be matched with human expertise to deliver accurate and fair analysis. The right tools combine these two facets with cutting edge technology and established risk modelling. The resulting solutions streamline and simplify your processes while offering you the kind of rich, detailed insight to help you make confident decisions, fast.

To see how Wiserfunding can help you make smarter credit decisions click here .