Pali Lehohla, the former Statistician-General of South Africa, on Friday slammed the lack of methodology and transparency around clients bank account closures.
All institutions have standards that must be communicated up front to their clients, particularly the banks, which are highly regulated,' he said in an interview with Business Report.
He said banks had to ensure that clients were aware of their rights, and banks could not apply standards based on their own free will and decisions.
Therefore, when it came to closing accounts, that had to be communicated up front as part of the regulatory environment and under what conditions bank account closures happened, and what the clients were entitled to by way of appeal.
It is a highly-regulated issue and these standards are global and if anyone says that they cant provide the methodology or standards as there is a protection of clients rights, that cannot be correct. That information must be available as a general rule, as a matter of governance.
So when one asks under what conditions you are closing a bank account, the answer cannot be that we are protecting the clients. That does not serve as an answer. The answer should be these are the conditions under which this happens, Lehohla said.
Lehohla said banks needed to be more specific about reputation risk. He agreed clients names cant be disclosed. But said the statistics of the reasons for bank account closures must be in the public domain.
It cant just be public perception, it has to be backed by public standards. Allegations of corruption must be first ruled on in court. Following that the reason for closing the bank account by a perpetrator must be published due to public interest.The The Banking Association South Africa must audits if the reasons for the bank account closures are valid or not to protect the public, Lehohla said.
Another way of ensuring that clients were protected was by providing regular reports on client accounts in aggregate form and running acid tests on closed accounts, he added.
This was in response to Business Report talking to Lehohla after sharing Nedbanks refusal on Friday in response to questions on the methodology of reputational risk to say how this is being calculated, if its 7.3 million active clients are being monitored and if artificial intelligence algorithms are being used, among other questions.
The questions around reputational risk come in the wake of Nedbanks 2023 annual report, which was released this month, that revealed that it has closed nearly 200 bank accounts on reputational risk.
Its Pillar 3 Risk and Capital Management Report for the year ended December 31, 2023, gave a brief but opaque description on reputational risk, but said that in 2023: Within risk appetite at 31 December 2023, there was no material breach of risk appetite.
Nedbank said in the report that among the top 10 risks identified for 2024 was reputational risk.
Nedbank's reputational risk is informed by public perception. We have a zero tolerance for corruption, and we expect all our stakeholders, including our clients, service providers and employees, to conduct themselves ethically and with integrity, it said.
The report said a model is a quantitative method, system or approach that applies statistical, economic, financial or mathematical theories, techniques and assumptions to process input data into quantitative estimates. A model also covers quantitative approaches, where inputs are partially or wholly qualitative or based on expert judgement, provided that the output is quantitative in nature.
Nedbank was asked by Business Report to explain this via these questions:
When was reputational risk introduced as a risk metric and what prompted it?
Nedbank has 7.3 million active clients. How is Nedbank identifying perceived reputational risk of 7.3m clients? Is it accessing all their information, or what raises a red flag to certain clients? Furthermore, according to the banking code in relation to Popia, how does Nedbank get around Popia to apply this methodology?
How does Nedbank quantify the accuracy or verify public perception? Is public perception a reliable barometer of risk? Is it truthful, fair and transparent?
Are other entities asking Nedbank to look at certain individuals banked by Nedbank? If so, what is the process and procedure of external players asking Nedbank to look at certain clients?
Are you using artificial intelligence/tech algorithms on your client base to red flag potential reputational risk?
Nedbank says it closed 200 clients accounts based on their reputational risk. What is Nedbank's score card for reputational risk?
Nedbanks answer failed to disclose the modelling methods.
It said: Nedbank is bound by client confidentiality and is unable to discuss clients with third parties. It bears noting, however, that decis
