[PeterisP]

$2 million in reserves seem like peanuts, any serious attempt to enter this business would imply investing enough money that mere $2 million is just a reasonable cost of doing business. Paypal needed $200m of venture capital to get started, a $2m requirement isn't prohibitive.

As for the delay, a common practice to enter a new market in finance is to "just" buy an existing bank that has all the required permits and all the legal and technical connections to the settlement infrastructure. There are thousands of banks, many of them small and relatively affordable for any investor trying to make a serious entry into the market. E.g. if some foreign financial company would want to start business in a new market, they usually would not start from scratch and bother with the licensing delays but buy some existing institution to absorb, rebrand and grow. If Amazon or Google or a new VC-backed company would want to enter the market, they're likely to do the same. If Amazon can spend a dozen billion on WholeFoods, then spending ten or fifty million on a small bank to get a banking operation running much faster is just natural, and not even a waste of money, as they'd likely need tens of millions of capital reserves anyway to justify holding immense amounts of customer money.

[dzink]

Buying a bank exposes you to being regulated and restricted as one, just as doing business on Government contracts adds a bunch of friction and overhead to every transaction and to your business (ask Google/Alphabet). The key to mass payments is lower friction per payment. Tech can fix a lot of things that regulation had to be used to prevent, but you have to pay the regulatory cost before you can remove the requirement for it. It's a bit of a Catch 22 in the US.