[FISV, GPN, FIS, SQ, Stripe, Adyen] On payment processors, distribution, and technology: Part 1 of 4

This is Part 1 of what I expect will be a 3-part series on merchant acquirers and issuer-side payment processors [edit: 4-part series!]. It describes the basic mechanics of the merchant acquisition model and delves into how incumbent players like First Data, Global Payments, and Worldpay got to where they were prior the flurry of megamerger announcements earlier this year. Part 2 will mostly talk about how these companies are competitively positioned relative to each other and newcomers like Adyen, Square, and Stripe. Part 3 will address issuer-side processing. There will be significant content spillover from one part into the next. 

————————————————————

In its 10Ks, Mastercard includes a diagram that succinctly illustrates the flow of money in an electronic retail transaction: 

Issuers issue credit and debit cards

to shoppers. A shopper inserts her card into

a point-of-sale terminal that collects card and transaction details. Those details are transmitted through from

the terminal to the merchant acquirer processor, who applies a series of

security checks to validate that the cardholder is who she claims to be before

then routing the payment details to the appropriate payment network (Visa,

Mastercard, Amex, STAR, etc.). 

The payment network sends that information to the card issuing bank, who

validates the cardholder’s identity, runs a compliance check on the cardholder

(i.e., is the account in good standing? does

the shopper have sufficient funds in her account? will this transaction push the cardholder

above her credit limit?) and, assuming all’s good, sends a message back through

the payment network authorizing the merchant to accept the payment. Within 2-3 days of the purchase transaction,

the issuer sends payment to the card network, which passes the funds to the

acquiring bank, who credits the merchant’s account (“settlement”). Data flows from the merchant to the issuer for

authorization; money flows from the issuer to the merchant for clearing and

settlement. 

[Merchants who accept online payments

(aka, “card-not-present” payments) require a software application called a

payment gateway that functions as the interface between the merchant’s website

and its payment processing backend. You

can think of the gateway as the online version of a point-of-sale terminal. It encrypts sensitive credit card info and

authorizes the transaction. The

processor, by comparison, transmits information to the merchant acquirer and

card networks.] 

A $100 transaction might net to $97.70 for the merchant, with the $2.30 merchant discount rate (MDR) divided among the issuer, card network, and merchant acquirer/processor like so:

Source: Heartland Payment Systems, 2015 10K

Around 70% of the MDR goes to the card issuer as interchange fees. Card networks, though sometimes vilified as extortionary middlemen, claim the smallest portion, just 10%. The merchant acquirer processor (Heartland Payment Systems in the above exhibit) accounts for the remaining 20%. The acquirer’s fee(1) flexes lower as volumes grow, blending to anywhere between $0.20 for Adyen, who acquires on behalf of large enterprises (and Adyen actually prices above others), to $0.50 (0.5% of the transaction value) for Heartland Payments, who serves small/mid-sized businesses. The disparity in take rates between the largest and smallest merchants reflects the fact that large merchants are less costly to serve as the fixed costs of service are spread over vastly more transactions. 

You will often see the terms merchant acquirer

and merchant acquirer processor used interchangeably. That’s because so many merchant services – transaction

authorization, fraud detection, data transfer, customer services, etc. – that used

to be handled by banks have either been outsourced to independents or segregated

into separate entities jointly operated by independents (Wells Fargo Merchant

Services is a JV between Wells Fargo and First Data(2)). Sometimes the merchant acquirer and processor

are housed within the same entity (Chase Paymentech is a subsidiary of JP

Morgan), but most US banks distanced themselves from merchant acquiring

activities during the ‘90s and early 2000s, and now function as a major

customer acquisition channel for independent processors (though they are losing

relevance in the face of “tech-enabled” distribution, which I’ll get into later).

[Going forward, merchant acquirer = acquirer = merchant acquirer processer = merchant processor. If what I really mean is the merchant’s bank (aka “acquiring bank”) I will use the word “bank”.]

Merchant processors, like the card

networks they interact with, is a scale-based infrastructure business, grabbing

10s of basis points on transactions flowing across a largely fixed cost base. But the scale economies that Visa and

Mastercard enjoy are a natural outgrowth of network effects: merchants are

compelled to accept Visa cards because consumers reliably carry them; banks

issue MA/V cards and consumers carry them because most merchants accept them. Growth in transaction volumes directly

reinforces the moat, giving rise to winner take most outcomes. 

This feedback dynamic is not really present among merchant acquirers. The fact that I process payments through Braintree (who, in turn, indirectly outsources the processing to First Data via Wells Fargo Merchant Services) does not spur other merchants do likewise. There is no direct connection between merchants and issuers, and shoppers don’t know or care what processor the merchant has running under the hood. For card networks, scale economies emerged naturally. Visa and Mastercard came to their current size and ubiquity predominantly through organic growth. By contrast, merchant transaction processors, engaged in a low-risk activity with few ways to differentiate, were forced into consolidation by the competitive demands of a commodity service. The downward trajectory of processor take rates tells this tale.

Source: McKinsey on Payments: Innovation

and disruption in U.S. merchant payment

An acquirer’s ability to offer low

prices ties back to relative cost efficiency, which is achieved by spreading fixed

costs over more transactions than competitors. 

The idea is that by combining, two merchant acquirers can remove

overhead, improve sales productivity through cross-selling, and aggregate

transactions across a rationalized set of platforms. Others in the space are then compelled to

merge to remain cost competitive. 

After decades of M&A, the merchant acquirer space has consolidated into a handful of players in the US. In 2018, the 10 largest merchant acquirers accounted for 92% of US card acceptance volumes, up from 84% in 2008 and 50% in 1989(3). The 3 largest independents at the end of the last year were First Data (with nearly twice as many transactions as the next largest acquirer when its JV alliances are counted), Global Payment Networks, and Worldpay/Vantiv. 

[Based on number of purchase

transactions using credit, debit, and prepaid cards that companies

processed. First Data includes

partnerships with Bank of America, Wells Fargo, and other banks. Elavon is owned by US Bancorp. Paysafe is owned by Blackstone. Source: The Nilson Report (via The Wall Street Journal)]

2 of those 3 were once owned by

financial institutions. First Data was acquired

by American Express in 1980 and then spun-off 12 years later. Vantiv (now WorldPay) was the acquirer

processing unit of Fifth Third Bank before Advent International purchased

majority ownership in June 2009. A year

later, Advent, along with Bain Capital, LBO’ed Worldpay, which was divested by

a financially devastated RBS so that RBS could meet the conditions for joining

the UK government’s asset protection scheme. 

First Data too was taken private. KKR shelled out $29bn for the company in September 2007, jamming First Data with 10 turns of leverage as the Great Recession beckoned. Nearly $2bn that could have been spent on R&D and sales were absorbed by interest payments just as consumer spending turned south. But the ensuing economic challenges and precarious balance sheet only exacerbated major structural rot that was festering well before KKR’s involvement. First Data had just gotten too big, too unwieldy. Its technology decayed, costs bloated, and platforms multiplied. From 2004 to 2013, the company cycled through 6 CEOs before settling on Frank Bisignano, who de-levered the balance sheet and revived a company(4) that had up to then missed e-commerce, under-indexed to small business, botched its overseas strategy(5), and failed to capitalize on high growth ISV-driven distribution.  

Banks were and still are First Data’s most important distribution channel, accounting for ¾ of its North American merchant processing revenue. In addition to referral agreements with some banks, First Data has joint ventures with Bank of America, Wells Fargo(6), and PNC in North America and with other major financial institutions overseas. In these JVs, both the bank and First Data contribute merchant contracts valued in rough proportion to their respective ownership. The bank, through its extensive network of branches, refers merchants to First Data, who exclusively processes all the JV’s merchant transactions. First Data brings a breadth of POS services to the bank’s merchants and relies on the bank to fund settlements.

But JVs seem to be an increasingly tenuous revenue source as banks assume greater control of their merchant accounts. Citigroup, which sold its Citi Merchant Services unit to First Data in 2005 as part of an alliance agreement, is reportedly starting a new merchant acquirer unit for large merchants; this past July, Bank of America announced that it would be dissolving its joint venture, Bank of America Merchant Services, with First Data. BAMS was formed in 2009, when First Data took advantage of Bank of America’s crisis to secure generous terms for itself. According to The Wall Street Journal:

“The First Data partnership has been fraught with issues for years, people familiar with the matter said. Clients have experienced delays getting money from customers, the people said, and outages have kept merchants from accepting cards for short periods. Clients complaints have caused concern within Bank of America because many of these merchants have other relationships with the bank, one of the people said”.

One Bernstein analyst cited in the article estimates that BAMS accounts for “10% to 12% of First Data’s revenue”. When JPMorgan and First Data terminated their joint venture in 2008, First Data’s domestic transactions, which otherwise would have grown by 5%, fell by 6% y/y in the quarter the Chase Paymentech JV was terminated and 14% the following quarter. (JP Morgan is now bundling its Treasury Services and merchant acquiring units to offer merchants a wider breadth of services). JP Morgan is particularly interesting because I’m thinking that ChaseNet, its closed loop processing network, could serve as a model for Bank of America. In 2013, JPM struck a 10-year deal with Visa, in which Visa charges Chase fixed fee for use of its network and Chase, Visa’s largest issuer, directs more of its volumes over VisaNet. With a white-labeled card network in place, a closed loop is created when JP Morgan serves as the issuer for cardholder and as the acquiring bank for the merchant that the cardholder transacts at. By “owning” its own card network, JP Morgan can bypass the operating rules set by Visa and Mastercard and negotiate deals directly with merchants, deals that pass through the savings from the Visa deal and/or collect data that merchants can use for marketing. ChaseNet was a major area of concern for the merchant acquirer space at the time it was announced, but despite some traction with big box retailers, it doesn’t appear to have had a material effect on the independents’ merchant retention.    

Anyways, at the same time that major banks

are getting more serious about merchant services, they are a less relevant distribution

channel for independent acquirers than they were 10 or 20 years ago, when a merchant

might default to the POS device and transaction processor offered by their bank,

who would bundle those things with lending and other corporate banking services. Banks (and by extension the legacy acquirers

that they refer customers to or JV with) are being disintermediated by payment

facilitators like Square, Stripe, and PayPal, who can onboard merchants in

minutes compared to days for a traditional acquirer(7), and by ISVs, who are

bundling payments into mission critical business applications. Software, hosted on the cloud, available to

small merchants at low cost and little risk, has become a critical customer

acquisition channel for payments processing over the last 5 years. Developers are tailoring modern business

management suites to different verticals and bringing payments along for the

ride. “Integrated payments” – what

merchant acquirers call it when payments functionality is integrated into

critical business management software – renders software more useful by automatically

reflecting transactions in the general ledger and updating customer data in a

CRM system. These partnerships have

become an essential growth vector for merchant processors, with volumes from

the ISV channel growing 3x-4x times faster than the acquirer volumes taken as a

whole(8). 

The ISV channel also retains merchants better than other channels – Global Payments reports merchant retention rates of ~90%-95% for the ISV channel, compared to ~70%-75% for merchants sourced through ISOs and ~85% for merchants sourced through banks(9). That’s because 1/ a merchant is unlikely to frivolously swap out her practice management system, especially one that bills patients and reconciles payments, and 2/ the payments capabilities from the merchant acquirer are tightly fused with the software and cannot be easily ripped out. This isn’t like dealing with ISOs who can and did regularly hunt for the cheapest processing base. But even as ISVs bring about higher volume growth and better retention, they also provide an opening for modern players like Stripe, Adyen, and PayPal, who can use ISV partnerships as a way to wrest payments from incumbent acquirers long dependent on bank referrals and traditional ISOs for distribution.

So then starting around 2015, First Data began deepening its ties with software vendors. After abstaining from major M&A during over the previous decade, it spent $1.5bn buying CardConnect and BluePay, further extending its reach into the ISV channel. The company revitalized Clover, a sleek point-of-sale terminal hardware/software vendor, analogous to Square. Clover heralded First Data’s major move from a commodity transaction processor to a more vertically integrated payments provider that could help merchants grow their revenue and manage their back-office operations. It was positioned as a platform on which developers could build front-end applications and leave the nuts and bolts of payments to First Data (for instance, Bypass, which sells cloud-based point-of-sale and back office software to stadium food vendors, integrates with Clover devices and runs Clover’s payment software in the background). More recently, taking its cues from Square, Clover has been developing its own in-house services (instant deposits, third party financing); migrating upmarket to larger, more stable merchants; and customizing for verticals, like full-service restaurants. 

There were many other Square copy-cats when First Data acquired Clover in late 2012, but they didn’t have First Data’s installed base of 6mn merchants or its longstanding distribution arrangements with major banks. The growing merchant base drew in third party developers to create apps, which in turn made the product more valuable for more merchants. From early 2016 to 2018, the number of Clover devices sold exploded from 150k to over 1mn, while volumes processed on those devices ballooned from less than $20bn to $70bn(10). Benefits flowed the other way too. Clover, along with solutions obtained through other acquisitions, expanded the set of offerings that First Data could sell through its moribund bank channel and helped the company win bank partnerships too. Notably, in late 2018 Royal Bank of Scotland replaced its existing merchant services supplier with First Data, largely due to Clover.  

Global Payments

Global Payments is another processor that

emerged in the early days of credit card acceptance, incubated as part of

National Data Corp. until being spun-off in 2001 (unlike incumbent peers First

Data, Vantiv, and Worldpay, Global Payments was never owned by a bank). Almost as soon as it achieved independence, Global

embarked on an aggressive international expansion policy, moving first into

first Canada(11), then Central and Eastern Europe(12), Asia(13), the UK(14),

and Brazil, mostly by purchasing the merchant portfolios of major banks and

then rounding out inaugural positions in those regions with smaller

acquisitions and organic efforts. Today

the company operates out of more than 30 countries. While the US makes up most its revenue,

Global commands more share in foreign markets – where it is usually the #1, 2,

or 3 acquirer – than it does in the US. Management

has long promulgated its global footprint one of the key things that

differentiates it from US-centric peers like First Data and Vantiv who can’t

carry domestic merchant relationships overseas or leverage technology and solutions

across different markets.

Whereas in international markets outside of maybe the UK and Canada, Global’s merchant acquiring is still very much either owned by or sourced through banks, in the US, up until ~6 years ago, Global Payments sourced most of its revenue through Independent Sales Organizations. ISOs, as the name implies, can re-sell payment processors and combine services from different vendors, as well as develop and bundle their own value-added services. GPN pays the ISO a contractually agreed upon fee per transaction and the ISO sells GPN’s services to merchants for whatever price it can get. For about a dozen years from the time it was spawned in 2002, GPN’s direct ISO channel was a major growth driver for Global, accounting for over half the company’s US revenue and 20%-25% of its operating income by 2013. Then, suddenly, ISO revenue plateaued. This channel never regained its former glory and likely never will. 

ISOs, who were responsible for a great

deal of industry growth in the ‘90s and early 2000s, are losing share to

technology-driven distribution, guys like Square and other payment facilitators

who can onboard merchants within minutes and offer increasingly sophisticated

self-serve POS systems at a fraction of the price. Merchants are no longer forced to default to payment

solutions supported by ISOs. Why spend thousands

for a heavy POS install that needs to be regularly maintained by an ISO when

you can just purchase Square Register for $799 or buy some iPads and $50 Square

card readers? Moreover, generic payments

solutions from ISOs are being disintermediated by business management software,

natively integrated with payments, that meets the idiosyncratic needs of

specific industries. Wix, a website

builder that is more accurately described as business management applications

for small merchants, embeds its in-house payment system into software tailored

to meet the unique demands of e-commerce sellers, musicians, hotels, event

managers, etc. Mindbody, a SaaS platform

that wellness providers use to run their businesses, also natively integrates

payments into its business management software.

The ISO growth issues came at the tail end of a 7-year stretch – one that included significant pricing compression in Canada, Global’s second largest country, significant margin contraction in North America(15), and a data breach that compromised up to 1.5mn credit cards – where Global’s stock price went nowhere. Damn near the whole executive suite was replaced over the next few years. Global’s former CEO, Paul Garcia, who ran the company since its IPO, retired in October 2013. By 2014, 7 of the 8 executive officers were appointed within the previous 2 years. 

With the ISOs fading, Global would spend $1bn+ acquiring growthier channels over the next few years. In 2012, it spent $413mn for an ISO called Accelerated Payment Technologies that embedded payment solutions into enterprise software geared towards small vets, retailers, restaurants, and auto repair shops(16). APT was Global’s first meaningful entry into integrated payments. It would be combined 2 years later with the $427mn purchase of an integrated payments provider called PayPros, the duo rebranded as OpenEdge. APT and PayPros would be followed by the acquisitions of Realex (European payments gateway; $119mn) and Ezidebit (APAC version of OpenEdge; $266mn). Financially, these acquisitions boosted Global’s growth and margin profile(17). Strategically, they gave the company greater control over distribution.

Global’s shift to direct distribution

was jolted forward again when the company acquired Heartland Payments for $4.4bn

(~20x pre-cost synergies EBITDA; ~13x post) in April 2016. Heartland was the 9th largest US processor by

volume at the time, over-indexed to SMEs in the restaurant (35% of volumes),

retail (17%), and convenience/liquor store (11%) verticals. It was set-up differently from its others in

that it acquired customers through the industry’s largest commissioned salesforce,

with 1,400 salespeople on staff compared to only ~50 for Global. Heartland’s founder, Bob Carr, appalled by

the aggressive and exploitative sales tactics of ISOs – who, as he saw it, hid

behind the complexity of interchange rates to surreptitiously jack up fees and

take rates – organized the company around a merchant-friendly direct sale model

early in its life. When the Durbin

Amendment took hold, Heartland famously passed all the debit interchange

reductions to their merchants while peers exploited non-transparent pricing and

merchant ignorance to keep the savings for themselves. Testament to Heartland’s service culture, the

company experienced far lower attrition rates than the industry despite its

dilapidated technology (which was upgraded after Global acquired them). 

Given Heartland’s strident direct sales approach, there was some concern that Global’s ISO model was just too culturally incongruent for the acquisition to make sense. But Global never went into the deal embracing both distribution modes. The new management team stopped looking at ISOs as a growth driver and was intent on maintaining the double-digit growth being delivered by Heartland’s direct sales apparatus. Despite their SME orientation, the two companies had few overlapping verticals and Heartland’s revenue was entirely US-based, so the cross-selling opportunities proved ripe. 

By the time 2016 rolled around, Global

had dramatically diluted its exposure to the low-margin wholesale channel,

which fell from 25% of US sales in 2012 to 11% by 2015 (pro-forma for

Heartland), and from ISOs in general(18). 

Organic revenue growth accelerated from an estimated mid-single digits

pre-Heartland(19) to double-digits in 2017. 

Direct distribution from Heartland, indirect distribution through ISV

partnerships, and e-commerce gateways from Realex, Ezidebit, and eWAY reinvigorated

the company’s growth trajectory. Right

around the time every company began insisting they were a tech company that

just happened to sell shoes or pizzas or whatever, Global’s management began to

say stuff like: “We are not a processor, we are not a payments company, we are

not an acquirer. We are a technology provider to these customers who need to

have these types of integrated interactions with their consumers” [2015

Investor Conference].

At this point, Global Payments was a

few years into integrated payments. Heartland

had a software-driven distribution part of its business too, but in a deeper

sense. In 5 years preceding its

acquisition, Heartland had rolled up a random assortment of POS and other

payments-centric software companies that specialized in K-12 schools, colleges,

restaurants, and liquor retailers that collectively came to comprise ~30% of

total net revenue(20). Whereas Global’s integrated

payments approach entailed selling payment solutions to an ISV, who then integrated

those capabilities into their own software and then sold that software to the

merchant, Heartland owned the actual software application that embedded its

payment solution(21). In other words, it

was both the payments provider and the ISV. 

[I suspect that Heartland skipped over integrated payments because

selling through a third party, while fine for software, would have created

channel conflicts with its payments salesforce].

Global Payments soon replicated this approach, presumably thinking that vertically integrating applications would deliver new revenue streams and protect its payments solutions from the disintermediating impact of payment facilitators. A payment facilitator assumes responsibility for the complex requirements that its sub-merchants – a health and wellness business on Mindbody’s platform is an example of a sub-merchant – would otherwise need to bear to accept electronic payments (online payments, for the most part). It onboards sub-merchants; maintains KYC and AML compliance; ensures that its sub-merchants comply with standards set by the card networks; integrates with payment processors (i.e. First Data and Global Payments); bears responsibility for financial losses related chargebacks and merchant fraud(22); aggregates payments under its own merchant account, and disburses those payments to sub-merchants(23). 

Source: Stripe

Besides horizontal facilitators like

Stripe and Braintree, wildly popular among startups and online merchants, there

are many of others you’ve never heard of (here’s a list of payment facilitators registered with

Mastercard). 

Merchant acquirer processors like First Data, Global Payments, and

Worldpay/Vantiv run behind payfacs in the payments stack. Stripe and Braintree are often referred to as

payment processors, but that’s a bit of a misnomer since they outsource the

actual transaction processing part(24). 

This doesn’t alter the payfac value proposition, which really hinges on

rapid onboarding, simple integration, transparent pricing, and everything else

that ISVs and merchants care about. 

Processing, on the other hand, is among the most commodified parts of

the payments stack. Merchants generally don’t

know or care if Vantiv or First Data or whoever ultimately processes the

transaction. 

The legacy acquirers, of course, would object to being labeled commodity processors as they’ve been layering on value-added services like secure card storage, PCI compliance validation(25), encryption, loyalty, reporting, POS hardware/software, customer support, and so on, for years. Their panoply of services includes tools that allow some of small, more industry focused payfacs to do payfac-y things. Global Payments, for instance, offers some combination of merchant underwriting, sub-merchant funds disbursement, and risk management solutions to payfacs like PayLease, Yapstone, and Property Solutions, who specialize in the property management sector. To be clear about the schematic, a vacation rental platform like VRBO uses Yapstone as its payment facilitator and Yapstone, in turn, uses tools from Vantiv to onboard and distribute funds to the property owners (“sub-merchants”) who list their properties on VRBO.

It requires considerable time and

investment to obtain the proper certifications and licenses, underwrite

sub-merchants in ~real-time, manage risk, detect fraud, maintain compliance,

and handle payouts. Some card networks

require quarterly or annual reports attesting to PCI DSS compliance. A SaaS platform or an ISV, rather than assume

control of payments arcana outside its realm of expertise, might outsource this

stuff to specialists like Stripe, Square, Adyen, and PayPal, who skim basis

points on the payment flow. 

But these basis points add up to real

money at some point. The SaaS platform

starts thinking “hey, I have enough volume running through my system to scale my

own payments infrastructure”. 

For a growing number of vertical software vendors with ostensibly small addressable markets, integrating payments processing and taxing the flows running through their systems is a key component of the revenue model. MindBody derives ~40% of its revenue from payments. Wix last year released a proprietary payments system that natively integrates into web apps tailored to specific verticals (Wix Hotels, Wix Restaurants, Wix Stores, etc.). 

Large platforms like Amazon and AirBnB

will build the whole stack in-house. Smaller

ISVs might handle frontend payfac functions like customer service and white label

underwriting and funds distribution tools from a legacy acquirer like Vantiv or

Global, or from one of these newer payfac-in-a-box providers like Finix or NMI. 

But you can also see how ISVs wanting

to own more of the payments flow could metastasize into a threat. So, perhaps wary of disintermediation risk

posed by selling payment solutions into customers increasingly looking to

capture additional value for themselves, Global Payments looked to own the

software piece too. And so began the

next leg of Global’s acquisition splurge. 

In the 13 months from September 2017 to October 2018, Global spent

$2.3bn ($600mn of that in stock) purchasing 3 SaaS companies. The first one, ACTIVE Network, is a system of

record for health and fitness programs. 

It’s what kids camps and charity 5ks and the YMCA use to manage

registrations and accept payments(26). AdvancedMD,

which sells practice management and patient engagement software to SME

physicians, was acquired a year later. A

month after that, in October 2018, Global scooped up SICOM, a kitchen

management and analytics software for restaurants.

These were good purchases. Altogether, Global paid ~5x revenue and something like low/mid-teens EBITDA for mission critical SaaS companies growing revenue by mid-teens, generating 30%-40% EBITDA margins. The synergies make sense too. Global is leveraging its worldwide distribution to find new markets for all three companies, who source ~all their revenue from the US. It has done the same for Heartland’s vertical software and its e-commerce gateway acquisitions, with apparently good results. In the case of AdvancedMD, Global was only capturing 20% of the company’s $3bn annual payment volume prior to the acquisition; it is now positioned to pick up the other 80%. SICOM’s kitchen management (middle-of-house) and analytics (back-of-house) solutions complement the Xenial, the POS system that Global acquired through Heartland. Together, the two companies cross-sell into each other’s installed bases and offer a complete solution to restaurants of all sizes.         

To recap, 7 years ago Global was a

merchant acquirer that sourced most its business from ISOs and whose point of

differentiation was a global footprint. Then

it spent $1.1bn to buy its way into integrated payments (2012-2014); $169mn to

buy its way into e-commerce (2015-2016); $4.4bn to bolster the durability of

its US distribution and pick up some POS software (2016); and $2.3bn to evade

ISV disintermediation and own more of software stack (2017-2018)(27). The $8bn of cash and stock that went into

these acquisitions compares to Global’s enterprise value of just $3.4bn in

mid-2012, prior to the start of its transformative journey. “Tech-enabled” revenue – that is, revenue

from integrated payments, e-/unified commerce, and vertical software – was

nonexistent in 2012 and grew to account for nearly half of Global’s revenue by

the end of 2018. The low-margin ISO

wholesale revenue that once comprised 25% of revenue has been whittled down to

a sliver. With a few exceptions – like

restaurant POS/management software, where Square, TouchBistro, LightSpeed, and many

others are tailoring POS systems for the restaurant vertical and grabbing share

from legacy vendors like MICROS and NCR, who cover ~30% of the market –

Global’s software verticals and integrated payments markets (dentists, vets,

K-12 elementary schools, small physician practices, health and fitness

programs) seem like defensible niches without a whole lot of economic

sensitivity.

Ok, I’m going to stop there and leave Worldpay/Vantiv for Part 2. 

Footnotes

(1) mechanically, the way this works

is the issuer will deduct its interchange from the $100 transaction and send

the acquiring bank $98.34 ($100 minus $1.66). 

The acquiring bank will credit the merchant’s account for the full $100,

with the acquirer processor advancing the interchange fee when settling daily

transactions. At the beginning of the

next month, the acquirer processor collects the $2.16 from the merchant ($1.66

for the advanced interchange + $0.50 in processing fees).

(2) See this Merchant Services Agreement from First Data

(3) The

Nilsen Report

(4) Under Bisignano’s

oversight, First Data’s international and issuer processing operations turned a

corner, but merchant acquiring revenue through its bank JV channel contracted.

(5) First Data was

running its international division out of Denver, CO and emphasizing its issuer-side

product (that large, card issuing financial institutions to authorize credit

cards) at a time when banks more inclined to build in-house.  

(6) In 1993, Wells Fargo formed a JV

alliance with merchant processor Card Establishment Services, which was

acquired by First Data. Sometimes First

Data’s alliances are majority owned/controlled and consolidated, sometimes not.

(7) Part of how they do this is by

saving KYC and other compliance underwriting steps until after the merchant has

started processing transactions

(8) Transaction Trending EP18: An

Overview of the Evolving ISV Marketplace, with Marc Abbey, Accenture

(9) though I’m guessing that

differences in industry mix across channels also drives the pronounced variance

in retention

(10) Compared to $26bn of volume

across 67k merchants for CardConnect and $19bn of volume across 77k merchants

for BluePay at the time of those companies were acquired.

(11) In 2001/2002, GPN acquired the

merchant services businesses of the Canadian Imperial Bank of Commerce (CIBC)

and National Bank of Canada to become the largest independent Visa and

Mastercard acquirer in Canada.

(12) In February 2004, Global Payments

paid $64mn for MUZO, the largest indirect payment process in the Czech Republic

(13) In July 2006, Global Payments

entered into a JV agreement with HSBC, purchasing a 56% interest for $67mn.

(14) In 2008 and 2009, GPN paid $746mn

to HSBC Bank plc to acquire HSBC Merchant Services, which provides payment

processing for UK and online merchants

(15) on account of a mix shift towards

lower-margin ISOs

(16) standalone APT was already

processing 90% of its volumes with GPN. 

Moreover, it was growing much faster (mid-teens) and generated higher

margins than Global Payments. 

(17) Global’s high-teens margins in

North America were particularly shitty due to the concentration of ISO

business. In Europe, where Global leaned

on bank distribution, the company generated margins close to 50%.

(18) Heartland was strong restaurant,

hospitality, and education, domains where Global had little presence.

(19) GPN lamely stopped disclosing

organic growth figures when its ISO channel started to slow, so this is just a guess. Unsurprisingly, they resumed disclosure when

organic growth returned to double-digits

(20) Heartland also bought a POS

tablet company called Leaf whose development efforts were curtailed a year

later

(21) Clover and Square sort of fall

into both categories. They are

integrated solutions in the sense that they are payment systems that an ISV can

integrate into an existing POS via an API. 

But they are also themselves POS vendors. 

(22) Because the payfac, not the

acquiring bank, has the direct relationship with the sub-merchant, the payfac

takes on these risks.

(23) the sub-merchant does not require

its own merchant ID. It accepts payments

from its customers under the payfac’s master account. Hence why payfacs are also called payment

aggregators

(24) From Braintree’s bank agreement with Wells Fargo (First Data): “You authorize and instruct us

to allow Braintree to direct all amounts due to you for credit or debit card

processing through Wells Fargo. Braintree will serve as your agent for purposes

of directing your proceeds from credit and debit card funded processing

services”.

(25) Payment

Card Industry Data Security Standard:

“a set of security standards designed to ensure that ALL companies that accept,

process, store or transmit credit card information maintain a secure environment”

(26) According to management, even

while it is by far the largest player in the space, ACTIVE has a small share

(5%) of the overall TAM. 

(27) Global is now pursuing SaaS companies in other nichey/payments heavy areas like government and supplementing existing companies with analytics and customer engagement features. 

Disclaimer: At the time this report was posted, Forage Capital held shares of MA but did not hold shares of Adyen, FIS, FISV, GPN, or SQ. This may have changed at any time since.