When I link a Capital One account, I get an OAuth-style flow and the ability to select which accounts I share directly on the C1 website. The sites I use this on show up as "Linked Apps" in my security settings in C1.
Others use the credentials-based flow.
Generally innovating through regulation is bad and I don't support it. But I think the aversion to it in the US is what creates these monopolies in the first place and in this specific use case, I think it's a faster way to solve the problem.
This is an example of what becomes possible (no card details and low fees):
https://techcrunch.com/2019/01/29/truelayers-payments-api-le...
These private apis will soon no longer be private due to Open Banking/PSD2 (europe) regulations. Plaid's biggest competitive advantage was removing the complexity of various data formats and APIs of banks. This variance will soon be gone in the coming years.
We need something like UPI (india) or Wechat (china) in the US. Zelle is a poor substitute.
Paypal uses ACH under the hood (it's one of only two options, the other being wire transfers which are expensive and often even need to have humans in the loop). So they guarantee the payment to the seller, and then as the buyer your bank account gets charged up to a day later when the ACH batch processing goes through. If you don't have enough money in your bank when that happens, it gets rejected and Paypal takes on that loss unless they can recover it from you via collections system.
To limit their risk, Paypal even uses Plaid in some cases to check your bank account balance before deciding whether to allow your payment to go through! But even that isn't a guarantee, because you could always have made multiple ACH payments that day of which some might go through and some could fail due to lack of funds. It's like trying to prevent a race condition in a multithreaded program by just checking the value again closer to where the race occurs, it might make the issue happen a little bit less often but to actually solve the problem you need certain atomic primitives which the US bank transfer system completely lacks.
> We need something like UPI (india) or Wechat (china) in the US. Zelle is a poor substitute.
Why is zelle a poor substitute? I can pay people via phone number.
From the perspective that instant payments should be an inexpensive, accessible utility for the benefit of all citizens, the Fed's FedNow efforts are preferable to Zelle or legacy ACH services.
[1] https://en.wikipedia.org/wiki/Zelle_(payment_service)
[2] https://corpgov.law.harvard.edu/2020/08/31/fednow-the-federa...
Apple and Google wallet are integrated into the credit card processing infrastructure. So they don't compete with Visa, their payments still go over Visa's rails.
Plaid's killer feature has always been their private access to APIs that the banks won't give to anyone else. It's essentially like they're using modern web APIs while everyone else is using snail mail (i.e. ACH, bank transfers that settle in batches once per day). I'm not exactly sure how they struck those deals and why there is no reasonable, open integration path or payments protocols between banks in the US like there is in every other developed country. Is it just that the banks like making their $15 per wire transfer? It feels like there's got to be more to it but I'm not sure.
But regardless, where we've ended up is the banks have fought so well to keep everybody out of their collective walled garden, but somehow Plaid has snuck in under the radar and has become so large and entrenched that they are themselves being called the monopolist, rather than the banks that own and control the system but of which there are multiple. I don't know enough about the details of Plaid's business though to weigh in on whether there's any truth to these claims.