Welcome to the Pollinate’s ‘Fintech Deepdives’, a series of informative articles and blogs about some key issues in merchant payments. There’s a vast quantity of change happening right now. The global pandemic accelerated a switch to physically remote payments (both online and in-store); regulatory pressures to improve how merchants verify and authenticate the payment identity of their customers (and what does that even mean, payment identity?); new technology in stores that have traditional terminal vendors looking nervously at their inventory lists; and shifts in the market place as providers consolidate through a series of huge acquisitions, and changes in schemes and technology strengthen their hold globally.
Will payments shift back to being in-store when the pandemic is over? Will changes to ecommerce push more people to using their phones for payments? Do you need to be offering a “buy now, pay later” service for your higher-priced goods and services? These are just a few of the questions that need answering.
It’s a tough environment to stay on top of all of this change. At Pollinate, we know that it’s often hard to get your head up and look ahead even six months when you’re facing all the challenges that today’s world brings. But over the next weeks we’re hoping to delve into some of the changes that we see happening both now and in the coming years, to challenge some orthodoxies and to ask some hard questions.
We start this week by looking at a potentially huge change in the dynamic of card acquiring, with the card schemes increasingly promoting something called “Switch to Issuer” which cuts acquirers entirely out of real time payments.
How do two models work?
Traditionally, in the “four party” model, the merchant connects to their acquirer, who connects via a scheme to the card issuer of the consumer.
As the industry evolved, an ecosystem of gateways came into being, sitting between the merchant and the acquirer, particularly when processing ecommerce transactions.
But even though the gateways sat between the merchant and their acquirer, the card authorisations still went via the acquirer’s systems out onto the card networks, where they would be routed to the issuer for an authorisation decision. Over time, the data associated with these authorisations has become the fuel for many of the value-added services that now make up big chunks of acquirer’s revenue streams – fraud detection, dashboards, data insights and analytics and so on – with the critical point being that the data was real-time.
The authorisation requests went to the scheme via the acquirer’s servers (old timers like me still call these “hosts” or “switches” because we remember when they were big old mainframes, but they’re just as likely now to be some skinny hardware in a cloud somewhere).
The authorisation responses went back to the merchant via the acquirer’s servers.
And so the acquirer was able to act on the data in near real-time.
(I should note that this isn’t talking about clearing – which happens much more slowly, with the actual *money* not making its way back into the bank accounts of merchants for _at best_ 24 hours and sometimes very much longer than that, but that could be the subject of an entirely different article).
That’s the way things always were.
What are the key differences?
But all business models evolve, and something that I think has been around for a little while, but now seems to be really gaining some traction, is that the schemes (certainly Mastercard, Visa a little more quietly) have been looking at the model, and wondering why those authorisation requests need to travel all the way to the acquirer before going to the scheme. Why not have the gateway (Mastercard and Visa both own several gateways) pass the authorisation request directly to the scheme, who can then route the transaction directly to the issuer?
In many ways it makes sense. For the merchant and the cardholder, there’s one fewer processing system involved, which means fewer hops for the traffic, which should mean faster authorisation times. For an acquirer, if all the traffic could flow this way, then they wouldn’t need to maintain their authorisation switch any more. Keeping up to date with mandated change (typically on a twice yearly basis) is an expensive business, especially on older, legacy systems. They keep their role in clearing (and crucially therefore, the money flows back through them), without the high availability, always on, must respond in milliseconds costs of running authorisations. In other words, Switch to Issuer lets Acquirers save some cost – fewer developers, fewer expensive old systems, and less to go wrong.
However, in gaining this potential efficiency, there is much also to be lost. For the merchant, it means complexity if things go wrong. Is the problem with the gateway, the scheme, or the acquirer? For the merchant, it feels like a very big step. Traditionally, acquirers like secretly to complain that they are rather taken for granted in the cards ecosystem, doing important but unlovely work like KYC and AML on merchants, managing and maintaining compliance, making sure that merchant gets paid, and keeping the authorisation services fast and available, while the schemes (so they say) lavish their attention on the flashier issuers, who get to deal with consumers, and cards, and cool apps. But you can’t help wondering if their concerns about being undervalued might not be true if they find themselves comprehensively disintermediated by a shift in the model so that they can no longer see the real time data that lets them provide the Value Added services that represent a growing proportion of their revenues. If left with just signing up merchants and paying out the settlement funds, the future for acquirers looks pretty grim – surely it’s worth maintaining their own infrastructure to be able to stay a critical part of the value chain?
How widespread is this “Switch to Issuer” model?
Well, it’s difficult to tell. I’ve heard about it being deployed in North America, in Africa and in Oceania, but it can also be used to describe switching to domestic, rather than “international brand” schemes too. I am not aware of any examples yet in the European model. I think it’s safe to assume, though, that as processing margins continue to be driven down by competition in the market, the idea of cutting out some of the participants that share in the revenue pool will become increasingly attractive. Acquirers need to make sure that they are developing their platforms to keep the cost of change low, and providing irreplaceable value to their customer, while casting one eye in the direction of newly emerging payment methods like Request to Pay.
The really big question is what competition regulators might think about this, especially in the heavily regulated European market. It’s difficult not to see the essential problem in the schemes selling “Switch to Issuer” as a solution to the problem “It costs too much to meet all the changes that the schemes require us to make”. In the end, I suspect this is why we’ve not yet seen this model emerge strongly in the EU.
There’s no doubt that being an acquirer is tough. New entrants can move faster. Many alternative payment methods have fewer regulatory requirements. But giving up processing of the authorisation and so losing the real-time data about who is spending right now feels to me like an admission of defeat – a retreat from fintech into just finance. Or maybe just financial plumbing.
By our payments advisor, Nick Telford-Reed