Community

One point that's lost in the popular "fintech will disrupt bank" posturing is heterogeneity. Even a simple A2A fund transfer goes through at least three organizations: Sending Bank, Scheme Operator, Receiving Bank. Whatever jingbanggizmo a Fintech can do inside its own company in the capacity of Receiving Bank, the sad reality is that it has virtually no control over what the other two parties do - or don't do. 

10-20 years ago, when I'd receive a credit in my bank account that I did not recognize, I'd ask my bank who the Payor was, I'd get no answer. The payments industry that "has never lacked ambition" would tell me that this was because of limited character length of memo field in ISO8583 and assure that me that the problem would get sorted out when ISO 20022 with longer memo field was implemented. 

ISO 20022 has come and gone but, even today, when I get a credit that I don't recognize in my bank or neobank account and I ask the bank / neobank who the Payor is, the answer is the same: No answer. I'm sure the payments industry will now tell me now that all my problems will be solved by genAI!

Another example is recovery of APP Scam payments. Once the victim has sent out the payment, it's as hard to recover the payment today as it was 5-10 years ago, despite the fact that Payee is fully KYCd entity and the Receiving Bank / Neobank knows her name, address, etc. 

IMO the much vaunted disruption of banks by neobanks / fintechs did not happen because neobanks / fintechs overestimated the role of technology within their own companies, underestimated the role of technology with the other parties, and totally missed privacy regulations that forbid one party from sharing full transaction information with the other parties. 

Okay thanks, I've heard about this PFM / MoMMA use case for AI, and have shared my thoughts below https://www.finextra.com/newsarticle/46803/bank-profits-face-170bn-hit-from-ai---mckinsey.

My conspiracy theory: This will be driven more by politics than technology. Big banks will block Agentic AI from automatically switching accounts or whatever; without automation, Agentic AI will lose its compelling value, become a damp squib, and fade away like the MINTs and WeSabes of the past 20 years. 

But time will tell.

On the contrary, I think AI will help banks more than nonbank competitors.  

10-15 years ago, finsurgents said Online P2P Lending, Neobanks, Mobile Payments, etc. will unbundle banking and inflict death to banks by thousand cuts. The story was compelling but banks weathered all those storms or fake storms and have emerged more profitable than 10 years ago.

Banks have been heavy users of traditional AI for NBO, Churn Prediction, Fraud Detection, etc. for decades. The only new thing in AI is genAI and genAI has so far been a game of incumbents like Microsoft, Amazon and Google in BigTech and JPMC, Goldman Sachs, CapitalOne in Financial Services. Apart from customer service chatbots, I haven't seen fintechs doing anything sensational with genAI - and the chatbots of banks are equally good / bad / ugly.  

Can you think of three use cases / lines of businesses where genAI will spawn fintech competitors and cut deeper into banking? I can't.

Let's also not forget that, in the last 5 years, many new age fintech competitors that were created by technologies like Open Banking, BaaS, etc. have lost credibility e.g. Synapse. 

tl;dr: The threat to banks has diminished from 10-15 years ago, not increased, genAI notwithstanding. Not surprisingly, for the last 5 years or so, I hear fintechs talking about partnership with banks rather than disruption of banks.