It's easy to get distracted by the value of the transactions, but they're just the payload. Nobody is "buying" a product worth those billions in transactions, the same way someone bought $N worth of books for a bookseller to have $N revenue.

A shipping company is evaluated on its ability to move containers, and nobody cares (well, the insurer does) if the containers are filled with iPhones or cardboard, as long as the customer pays the shipping fee.

> So the question is if the margins are so low that they need a billion dollars to make a profit, what kind of scale do they need to be worth a billion dollars.

Well, if we're assuming P/E=20 then they need to be heading for 50 million dollars in profit. Let's guess their "gross margin" from their spread is 2%, so they're currently at top-line revenue of 2 million dollars (and the same in costs). So if their costs were completely flat they'd need to grow to be processing 25 billion dollars of transfers to justify the valuation; in practice growth won't be completely free of costs so we're talking about a bit more, maybe 50 billion dollars. Does that sound vaguely plausible?

> And it somehow it makes me question their business model as well. Sooner or later they will be tempted to raise fees and get more profitability.

I doubt it; there are a lot of competitors in the same space. At the moment TransferWise has the name recognition, but that won't let them raise costs a lot.

They don't need to sell a billion dollars, they need to move a billion dollars

And yes, they might increase their fees when they get bigger, still, if it's smaller than what the banks charge, it's a win

(However, since money movements are not balanced they still need to "move back" some of it so they can settle their accounts - but I guess they already have some bulk currency buyng/selling bank to do that)