A Game of Take Rates

When I first read Jevgenijs Kazanins’ very excellent fintech take on why Cash App Pay could be the next major driver of Block’s growth in the face of some challenging headwinds for Block, I was floored by the initially compelling case. He argues that if Cash App can incentivize more payments volume to go through Cash App Pay, there is an opportunity for massive gross profit uplift, because Cash App Pay’s take rate is substantially higher than Cash Card’s take rate on purchase volume.

Let’s just pause on that take rate point for a minute to let it sink in. When the average Cash App user purchases something with their Cash App card, Cash App earns about 0.90% in interchange. But when they make that same purchase and simply scan a QR code or click the Cash App Pay button, it seems as though Cash App suddenly earns 2.6% - 2.9% (+$0.10 - $0.30) on the same transaction, up to a 3x monetization uptake. It seems too good to be true, magical even.

But if it were true, why hasn’t Cash App, which is normally almost too good at understanding their unit economics, taken advantage? Whereas Jevgenijs is more optimistic about the ~$2.5bn in GPV that Cash App Pay has generated in its first ~2.5 years of existence, I find it downright embarrassing. That’s less than 1% of Square’s GPV after two years in the wild? That pales in comparison to the take-up of most of their other products. And that’s despite it being essentially pre-installed on every terminal with no change in hardware required.

I started to have a sinking feeling that maybe something was missing in the calculus. And as I dove deeper, it became clear that the “magic” take rate uptick we thought we’d found was not quite as it seems. Our first hint that we might be missing comes from Jevgenijs’s seemingly reasonable suggestion to “leave the topic of Square’s growth potential for later, and today will focus on Cash App.” The problem is that when we’re talking about money movement between Cash App user and Block merchant, we can’t simply leave one out. But to understand that, we need to talk more about cake.

As easy as a piece of cake?

First, let’s try to understand where the perceived uptake in take rates comes from. To do that, we need to dig into the Payments Layer Cake, which Finix describes so well that you should just read their post and come back if this is unfamiliar, but here’s a diagram that sums up the layers.

Let’s just say that normally, when a Cash App user makes a purchase, Cash App acts as the Issuer / Issuer Bank which means there are several intermediaries between what the merchant is charged and what Block earns on the transaction (acquiring banks, acquiring processors, card networks, issuing processors), meaning its ultimate take rate is relatively low.

But in the case of Cash App Pay, there’s potentially a few magic tricks at work. The first is that in this case, Block is both the acquiring processor + bank AND the issuing bank, because it’s facilitating the Square merchant’s acquiring POS and it’s issuing the Cash App user’s debit card, thereby cutting the intermediaries down in the typical payments flow and seemingly increasing Cash App take rates by 35-50% (instead of just interchange, you’re now getting interchange and the markup).

But then it seems to get even better with the second trick, because if you’re both the acquirer and the issuer, you don’t even need the payment network at all. You can simply move money by book transfer between your merchant’s bank account and your user’s bank account in a perfectly closed loop payment network. Suddenly, Cash App Pay has a magical 2.9% take rate with no money movement costs at all - pure magic right (I’m not actually sure if Block is doing this, but why wouldn’t they?).

The problem with these magic tricks, like the ones in life, is that they are actually using some deception to hide what’s truly happening. In this case, it’s technically true that Cash App Pay now earns a 2.90% take rate on this transaction, but by that logic, the Square side of the business would have to have a take rate of 0% as the entirety of the take rate would be transferred to the issuing side of the business. That doesn’t make a whole lot of sense so we need to add up both sides of the business to understand the economics at play.

Let’s play it out. In the status quo, without Cash App Pay, Square would earn ~1.85% in acquirer markup and Cash App would earn 0.90% in interchange for a total of 2.75% in take rate. In the Cash App Pay case, Square and Cash App would split the 2.90% Merchant Discount rate they artfully charged without having any money movement. So instead of the 2-3x markup, the uptick is more like +15bps (+5%). (Why is this math different from the diagram? Because Square / Stripe and other cheekily charge a flat rate for credit and debit cards even though debit cards have lower network fees and interchange, thereby pocketing the difference).

Let It Flow, Let It Flow

So maybe there isn’t a massive take rate arbitrage to take advantage of, but there’s still a huge opportunity to disintermediate and own this payment flow that warrants a further push. Just look at American Express and potentially CapOne/Discover as examples of companies that already benefit or stand to benefit massively from owning the acquirer and issuer relationship. Owning the payment network delivers all sorts of value from pricing power to instant settlement, risk mitigation, full lifecycle transaction analytics and of course, the bragging rights of being one of the few fintechs not beholden to the whims of the card networks.

But how could Block get more aggressive with its burgeoning closed loop payment network? I can tell you one way it won’t - with its current non-compelling value proposition. This incredibly long help article details a value proposition that essentially boils down to “you can scan a QR code or tap a button to pay instead of swiping or tapping your card.” I’m not sure why anyone would choose to do that instead of using Apple Pay / Google Pay / PayPal which has multiple cards to choose from at any given time (including the Cash Card), and already or is soon to be adding the QR code feature.

Instead, with the pricing power and network ownership, Block could offer cashback on every Cash App user’s purchase at a Square seller. Or it could offer a processing discount every time a Square merchant sells to a Cash App user. Or both - 1% back to merchant and user would still yield a 0.90% net take rate (not to mention the second order value of the cashback being reused in the ecosystem).

Sure it’s aggressive, but given the reduced operating expense, risk mitigation and zero money movement costs, it’s not going to be underwater on the unit economics side even if its initially gross margin dilutive. It’s not even as aggressive as Square’s own version of this initiative, where they gave Square Card business users the full 2.75% discount on purchases made at Square sellers (because there wasn’t in fact a charge when money was book transferred from business bank account to business bank account).

And sometimes, it takes an aggressive play to win market share and build out cross-sell as we saw first with Chase Sapphire Reserve’s too-good-to-be-true 100K points bonus that ultimately racked up at least a 20% increase in overall credit card market share, and now with the Robinhood Credit Card which will rely on the same cross-sell opportunity. This could be the lever Block needs to be able to radically change the status quo for the ailing Square business while expanding Cash App’s available TAM. Let’s see if Jack agrees.

(Side note: one of the things, I just couldn’t understand was how Cash App Pay worked when it was initiated by merchants who used Stripe or Adyen as Jevgenijs mentioned in his post. If the payment network remains open instead of closed, without Block as acquirer and issuer, what is the advantage? At that point, the benefit would have truly been only the reduced friction of a QR code / button click for some marginal and it didn’t seem worthwhile to get in bed with your competitors for it. But sure enough, when you dig through the integration details, it becomes clear that the only way to actually integrate Cash App Pay into your Stripe Seller checkout flow is… to become a Square seller. In fact, the integration reads like an ad for Cash App which is Stripe’s competitor in this case. Certainly somebody had fun with this strategic partnership (how did Stripe okay this??). I’m curious if the juice was worth the squeeze so if you know more, please let me know!).