The once immutable banking sector has experienced significant disruption and transformation over the past few years. One such transformation has been the widespread use of sophisticated risk management tools that pair advanced computing technologies, such as artificial intelligence (AI) and machine learning (ML), with statistical analysis. Though the goal of this technology was to make risk management processes smarter and more efficient, its implementation has instead created a global echo chamber in the financial services industry. The resulting bubble is practically leaving many current and emerging market segments out in the cold.
How We Got Here
Straight after the GFC (Global Financial Crisis), banks and other traditional financial institutions immediately sought ways to shed risk. Their exceptionally low-risk tolerance meant discontinuing many banking staples, such as access to personal or business lines of credit. It led to the subsequent debilitating credit crunch.
Part of the problem was that financial institutions didn’t know how to step outside of their legacy systems to measure risk in the new world order. Instead, they maintained the same rigid protocols that had been in place virtually unchanged for decades. This lack of adaptability meant they could not quickly respond to the financial crisis by shifting priorities and reconsidering the risk variables that would protect their institution yet still allow them to service some of the markets. It was not just a big bank problem, either; it was a common issue among financial institutions of all shapes and sizes.
As the GFC was unfolding, a growing wave of alternative finance companies– the precursors to today’s fintech sector– stepped in to fill the credit gap. Many of these companies used data-based algorithms to assess risk and match customers to products. They also ate up the lion’s share of the market. The big banks and other traditional financial institutions were initially content to watch from the sidelines.
It took several years for established organisations to incorporate the same algorithms into their risk assessment, KYC, and AML processes. In many cases, the growing availability of stand-alone automated risk-management systems and services made the transition possible. It was and continues to be coupled with the extensive acquisition of small fintech companies.
As traditional organisations were busy upping their game, the world continued to evolve. Data pools have expanded exponentially. Customers and the financial products they need increased in complexity. Smart technologies, such as AI and ML, have also seen rapid advancement.
However, as smart, automated risk assessment systems may have become more sophisticated and ubiquitous, the outcomes they produce have become more standard… and predictable.
These risk assessment processes have seen exponential growth in the last five years by integrating smart statistical methodologies into basic computing. The capacity of computers to process or sift through terabytes of data and run AI-powered statistical analysis, as a matter of course, has also put these same cutting-edge tools in the hands of millions of analysts. Institutions and entities big and small can now arrive at regression and correlation conclusions at lightning speed.
The downside of this extreme standardisation of risk assessment processes and methodologies is that it has led to confirmation bias in making investment decisions. The conclusions of risk assessments are frighteningly similar across the globe. It is because everybody is running the same model on the same data.
The current global, synchronised approach to risk assessment practically eliminates the competitive edge of one financial institution over another. It also generates conclusions that often are out of touch with the needs and realities of specific markets or industries.
Here is a rundown of how the risk/reward ratio has been impacted in today’s market:
- The information premium is mostly non-existent (except if we are talking about illegal insider information).
- Everybody in the market has the same tool and same data.
- The risk/return balance is valid only for a very small range.
- The key risks are now even more shifted in the tail. So, the tails have gotten fatter.
Breaking Free from the Herd
So, how can investors outperform a very well synchronised herd– especially if:
- They stay away from the herd and risk losing out to it as the herd has a very synchronised and robust momentum that cannot be matched.
- They are part of the herd, and they have to be ready for a significant fall when it comes.
Before I end, let me make one thing clear: I am not against automatic risk management tools, nor do I believe that these tools should not be widely available to analysts all over the world. What I am against is the homogenisation of algorithms that lead to global, synchronised results. These results can alter the way business clients, in particular, operate, leading to inefficiencies and lost opportunities for growth.
TuningBill: a New Approach to Risk Allocation
TuningBill is dedicated to global SME traders left behind by the current mainstream financial system. Our neobanking services revolve around the specific needs of the trader network, and we have no intention of mainstreaming them.
Our goal is to add financial wings to SME traders. These wings will allow them to fly above their fellow traders, see a risk first and act first. We give them financial tools that help them remain flexible in an unpredictable world while building a foundation for improved operations and future growth.