Community

Aforementioned Jason Mikula just posted on X fka Twitter that end users are losing access to funds in this fracas.Which is not surprising.

When Neobank / BaaS / Fintech ("Fintech") does not have a banking license and works on top of an underlying Sponsor Bank with banking license, end user accounts in fintech do not necessarily map 1-to-1 with account(s) in Sponsor Bank. While FDIC will cover customer balances if Sponsor Bank goes bust, it won't come into the picture if Fintech goes bust. Whether end users get their money back in the Fintech's bankruptcy procedings depends on the T&Cs of the FBO account held by the Fintech at the Sponsor Bank.

Adding FCA boss Nikhil Rathi to the band of rogue regulators whom I call "Choprandos".

Banking is a confidence trick. The business model of the industry is fragile and works only because depositors think it works. There's no room for casting aspersions on a bank until its culpability is established beyond reasonable doubt (or whatever the legal standard for this industry in UK is.) Early enforcement action in banking is just as bad - or even worse - than late enforcement action.

I totally stand with UK Parliament.

We keep hearing that merchants lose their merchant accounts if chargebacks exceed a tiny percentage like 0.5% per quarter. In the case of this fraudulent merchant, Visa reported it was 29-40%, albeit over a longer period. Given that Visa was involved in each transaction from this merchant, why couldn't it block them? Even assuming that Visa "outsources" the blocking decision to acquirers / PSPs, why didn't Visa block the PSPs for continuing to acquire transactions from a merchant with such a high chargeback rate? Why did this have to go to regulator FTC?

For global payments that involve India at one end, the combination of India's A2A RTP (UPI) with the domestic payments scheme of the non-India country on the other end, has gained good traction. While it may involve SWIFT in the settlement cycle, AFAIK it disintermediates SWIFT in the basic payment workflow.

IME, banks have consistently ignored the retail cross border payments business during the last 35 years. The only sensible reason I've heard for their reluctance is too politically incorrect to be stated publicly.

On a side note, I've been hearing about SCTInst EU cross border payments for over 5 years. If that works fine, what is the need for a separate mandate for EU cross border payments w.e.f. 2024? If that does not work fine, will a separate mandate lead to a new scheme that works any better?

Woz an FI that spec'ced 2400TPS for its onramp system to an A2A RTP rail, expecting to hit that throughput in Year 3. After 15 years, the entire scheme itself hasn't reached half of that throughput! But I know exactly what would've happened if we'd moderated the customers' expectations upfront. 

While I agree with almost all your points, it's virtually impossible to win deals when the CEO / CIO has just come back fresh from attending a tradeshow where they were told by the Gartners and McKinseys and SamAs that e.g. ChatGPT can do everything faster, better and cheaper than humans. Against that backdrop, any vendor who tells them that they can get only any 2 out of 3 because some Eric Brewer / CAP Theorem said so would be branded as pessimistic and thrown out of vendor shortlist for being incompetent, unenthusiastic, or both. C'est la Vie.

Nice post!

Keen to know if there are any code analyzers that also surface (1) Bugs (2) UX deficiencies / areas of improvement.

Also, while your post already mentions that your code analyzer can identify bottlenecks impacting performance, keen to know if it can also judge whether the code will deliver a certain TPS in production - without actually running a benchmark on the required hardware infrastructure?

In an old life, just to run a benchmark to prove a certain throughput of our software, we had to send three engineers from Singapore to Germany because the hardware vendor didn't have the kind of gear required to run that kind of high TPS in any of its facilities in Asia Pacific.

Props to NatWest. First I'm hearing a bank giving loans against intangible assets. May its tribe grow. 

... which are exactly the kind of things India's programs like Atmanirbhar Bharat seek to prevent:(. 

All this is theory, what Reg ZeroMDR actually driven in India is duopoly of two American UPI apps, Walmart PhonePe and Google Pay.

While the architecture of UPI is open and permits any number of PSPs to plug into it, the lack of revenue model in UPI - due to Reg ZeroMDR - led most PSPs to soft pedal UPI, thereby leaving the market to two Big American Firms, who control over 85% of UPI volumes and TPV.

In one of the greatest ironies, the regulator, who is supposed to prevent / break up monopoly / duopoly, has actually created a duopoly in UPI. Long after the horse has bolted the barn, the regulator is trying to bolt the door by trying to impose a cap of 30% market share per UPI app. This is like trying to ban network effects, never gonna happen, no wonder the deadline for compliance with this rule has been extended thrice in three years. 

When Visa abandoned its plan to acquire PLAID, the reason given by pundits was potential antitrust objections. Won't the same antitrust objections be raised about its TINK-powered offering?

Or was antitrust just a smokescreen to avoid paying $5.3B for PLAID when it could get Tink for a substantially lower price ($2B)?

On a side note, by making an explicit pitch for Open Banking technology in its video, Visa has completely undermined the popular misconception that customers don't care about technology.