Road to (Perdition) Unbundling

Road to Perdition Unbundling

We have all been proven wrong one way or another since the advent of the fintech craze. What differentiates us from one another is the degree in how erroneous we were. In no particular order, we believed Silicon Valley could disrupt financial services, bitcoin would rule the world, bitcoin would fail miserably, blockchain would solve the common cold, robo-advisory would take wealth management by storm, p2p lending would revolutionize credit, banks would be doomed to fail, banks would not be impacted by technology changes, financial chatbots would take over customer service, AI would be the secret weapon banking or insurance was waiting for, software does not need to be regulated, regtech is a regulator’s best friend, open banking and API banking is the new black, PSD2 is our Lord Savior. Although I am most assuredly missing a few nuggets, and as you may notice, some of the above are still vividly being debated and experimented with, you do get my drift. We collectively suck at predicting what is going to happen as novel ways to apply technology are put to bear on financial services.

Be that as it may, we know of two evident truths since the internet graced us with its presence. First, intermediators in any given industry will be disrupted as new business models emerge and effectively unbundle old paradigms. Second, this unbundling has not happened yet in the financial services. The question we are all working toward answering and in the process elevate to a third truth or disprove altogether is: Will a great unbundling/rebundling occur in financial services?

If you are interested in learning more about unbundling as it has and still is occurring in various industries, I urge you to follow Ben Thompson and his web site Stratechery. Suffice it to say, that, at a high level, new business models will disrupt old ones by unbundling their offering, rebundling new features/functionality and leverage at scale by aggregate attention as a result of said unbundling/rebundling.

As noted, we have seen this process at play with Google, Amazon, Netflix, Spotify and countless others. We have not seen this at play in financial services in any meaningful way. First, because financial services processes are much more complex. (There is a logical path towards greater complexity in unbundling/rebundling text, audio, pictures, simple video, music, movies, live sports events – whether the complexity is technical or commercial in nature. I argue that any process that is money centric is eminently more complex to unbundle/rebundle per se.) Second because financial services are heavily dependent on externalities such as regulation and politics. Third, subject matter expertise matters and it does take time for the right entrepreneur to build such expertise that aptly balances fin and tech – the Flynn effect is in full force in fintech and every new wave of startups will be more sophisticated than its predecessor. Fourth, because, in the majority of cases, digital transformations in financial services, whether executed by startups or incumbents, eventually turn into what Ron Shevlin describes as “newish ways of doing things that are already done”. Fourth, because, unlike other commercial activities, many financial services processes pose serious systemic risks in ways unbundling/rebundling movies or music do not. As such unbundling/rebundling in the age of the internet, where in most instances a winner takes all or most stands, may be far from desirable in lending for example. Mariano Belinky, Managing Partner of Santander InnoVentures, often stresses the point that an Uber of lending, however unlikely for many reasons, scares the living daylights out of him. For those that would like to explore further the reasons why unbundling has not occurred yet, I refer you to this excellent post by Ron which both augments some of my thoughts and serves as a healthy counterpoint.

By no means am I belittling the achievements of a few fintech firms that have reached escape velocity. PayPal (and Venmo) comes to mind as evidence that disruption at scale can happen. So does Klarna, Square or Stripe. Other firms such as Transferwise or SoFi show great promise. It is interesting to note that most of the fintech firms that have reached or are on the cusp of reaching escape velocity are in the payments sector. Arguably payments may be the easiest sector to unbundle/rebundle – as opposed to corporate lending for example - or the sector via which unbundling/rebundling will occur at scale throughout the financial services industry. Nevertheless, “Netflix” unbundling/rebundling of a retail bank, to take a specific example, has not happened yet. Why is that so?

Astute readers will have noticed bankers seem to be attracted to certain bank startups like bears are attracted to honey. It all started with BBVA’s acquisition of Simple during the Pleistocene. We were reminded of this truism in more recent times with the acquisition of Compte Nickel by BNPP. Contrary to popular opinion, bankers seldom acquire businesses for philanthropic reasons. The promise of a solid IRR needs to be included in an investment rationale, however fleetingly it is woven in an investment committee’s decision making process. Astute readers will also not have failed to notice that, after the relative lackluster performance of proto neo banks or pfm apps disguised as neo banks, the model is being refined and is delivering promising signs of success. For the doubting Thomas among you, please go check the traction achieved to date by Nubank or Compte Nickel, and even if it is too early to tell, by the likes of Monzo to name but one of the fresh-faced bank startups. (I sincerely hope UK challenger banks will be successful in the aggregate and show the way for future enterprising startups, but I digress.)

I note Nickel, Nubank, Monzo and their brethren have a few things in common: a) repeat entrepreneurs/founders, b) deeply experienced financial services executives, c) opened to more rather than less regulation, d) positioned their value proposition precisely where banks have been ineffective and where retail customers feel the most pain; and last but not least, e) developing new core banking capabilities or even new core banking systems altogether. These startups also have benefitted from the lessons learned from previous fintech waves as well.

The core banking point I allude to is essential. A core banking system is made up of a deposit module, a lending module, a general ledger and a CRM module (KYC module if you prefer). To these core modules, one of course has to add a myriad of other functionalities. Let’s keep things simple for the purposes of this article – even at the risk losing the core banking purists amongst us. To my knowledge, there is no legacy core banking system that can handle real time processing, although some startups are trying to address the issue lately. All are architected around batch processing in one way or another. Now, picture Transferwise successfully achieving the throughput of Netflix or Facebook for example. There is no core banking system in the world that could process such a volume in real time. The same applies to the first startup that successfully unbundles/rebundles a bank checking account at scale and finds itself with 200 million users overnight. As an aside, the CEO of one of the small banks that provides services to the likes of Transferwise admitted the limitations of existing core banking technology to me recently. Framing the issue another way, we can all be assured that Facebook’s or Netflix’s “core banking system” equivalent is architected, built and managed for the massive unbundling and rebundling success they have achieved. Core banking innovation will happen and will be essential to further banking disruption – I will attempt to address this specific point in another post.

We are now at an intriguing junction in the fintech driven evolution of the financial services. Retail customers are relatively more aware of new services than in the past. Banks are more aware of the threats they face. New fintech startups are more sophisticated both from a fin and a tech point of view. Regulators and Legislators are more “engaged”, from the FCA’s willingness to foster competition by nudging a new generation of challenger banks, to the EU’s willingness to pry open the banking industry with its PSD2 directive. Technology companies are establishing themselves as formidable competitors in the payments space – not a day goes by with an announcement from Facebook, Whatsapp, Amazon, Apple or Google. Further, we all are scrutinizing the phenomenal success of WeChat and, redolent with envy, trying to emulate it in our respective geographies while warily noting WeChat’s slow encroachment in the US and Europe.

I argue that all of the above trends are coming to maturity in a synchronistic way and raising the probability that further unbundling/rebundling will occur in financial services in ways that make the past attempts pale in comparison. I also argue that many banks are not prepared for such possibility and the ones that are the most prepared are located in the EU as they are forced to face the consequences of PSD2 – API banking, open banking here we come. Still, most of the strategic moves have been either defensive in nature or prophylactic. Acquiring Compte Nickel may be viewed as a defensive move with optionality for example. Preparing for PSD2 by ensuring one meets a mandate while limiting potential damage may be viewed as a defensive move too. Building an app marketplace and tying it to one’s existing client base may also be viewed as a defensive move.

For the purposes of our discussion, let me outline what I think would be an offensive move while taking full advantage of both attention aggregation and unbundling/rebundling of a feature/functionality set. I will not attempt to weigh in on who may be able to successfully prosecute such a blue print. A well-heeled startup, a forward-thinking incumbent or an ambitious tech giant could equally achieve such a vision with the proper execution and necessary luck.

a.             Isolate and unbundle a checking account. Define the core functionality you want to offer around as simple and functional of a checking account as possible. Simple is good! Basic functionality even better. Not too basic that you do not appeal to a wide enough group of individuals to start with. Compte Nickel is a perfect example.

b.             Build a new core banking system around this unbundled checking account, comprised of a core deposit module, a customer centric general ledger, a KYC module, a kick ass set of APIs, a marketplace interface and real time processing capabilities – no batch please. The goal here is to build the core tech platform that fits a checking account unbundle, no more no less PLUS the ability to plug in n+1 services to that checking account. Note that I do not mention a lending module (I will get back to this later). I believe this may be Monzo’s vision, also what Fidor tried to do before it got acquired.

c.             Once the unbundling occurs, then rebundle by offering, in a marketplace environment, best of breed services. There is a long tail for this, just check Amazon or Netflix’s offerings to convince yourselves of such assertion.

The subtlety here lies in starting with a simple offering, the core checking account, with a defined targeted audience (where traditional banks do not offer good service cost effectively) and turn the PSD2 mandate (if no PSD2, the market will mandate it) upside down. Rather than meet it defensively, make it a core strategic proposition for your customers.

Think of a traditional bank value proposition as a bundle that includes certain features provided for free, others subsidized, and yet others that are revenue/profit leaders. Retail bank customers all use their checking accounts and use to varying degrees the services associated with their account. Some are heavy users of certain functionalities, other light users of other functionalities. The current bundling obfuscates the asymmetry between customers needs and uses and what the bank charges. The entity that first unbundles the checking account and provides it at a cost level that attracts users at scale (first order of aggregation and the beginnings of network effects) while providing a choice of “a-la-carte ancillary services at scale (second order of aggregation and network effects) will blow up the pricing models banks have successfully used to sustain their business model. This I believe lies at the core of a true unbundling/rebundling with proper user aggregation. Customers pay for the bundles of services that more closely meet their needs at an appropriate cost. It should be noted that both the manufacturing capability and the distribution capability of a bank get upended in this scenario.

A few more questions are worth considering:

-      Can this scenario occur if the unbundler does not own the new core banking system? Several pundits believe there is no need to own a core system pointing that a third party could own and operate a “banking infrastructure”. Before hastily answering ask yourself this: Could Netflix or Facebook have achieved their phenomenal success by only focusing on the client interface and customer service while a third party owned and took care of their own core systems? I think not. Actually a new core bank system is one of the competitive advantages necessary to capture value.

-      Can this scenario occur without the need for a license? Several pundits believe so, arguing, in similar fashion to Silicon Valley advocates that new distribution models are only software code and thusly not worthy of licensing requirements. While I am sympathetic to this line of thinking and while I explicitly excluded lending activities from the blueprint above, relegating it to the marketplace, I think some type of bank licensing is needed to reflect the intricacies of segregating, securing, handling consumers’ money and facilitating a host of third party services. One could think of a new type of banking license, less than a full banking license, more than a narrow payment or money license.

Let us now switch our focus to how value is created with the new model I propose from a strategic point of view. The traditional banking model is predicated upon ownership of both manufacturing – production of products and services - and distribution – whether via brick and mortar branches or digital channels. Ownership of the entire stack and mingling of both distribution and production is what used to create value and what, many bankers hope, will continue to create value. The only rub is if consumers stop using traditional points of distribution or if distribution gets to be reinvented. With the new model, the point of contact is the checking account per se as opposed to a branch or an app and the value springs from two very specific actions which reinforce one another. On the one hand integration between the checking account and a variety of third party services. On the other hand aggregation of a plethora of third party services available to the checking account holder. These two reinforcing mechanisms not only create value differently but they also, relatively speaking, concentrate power with the owner of this new operating chain to the detriment of the legacy producer/distributor. In a model with a long tail of options and informed choice the value of a bank brand diminishes.

Do keep in mind the above can be tailored for many different target markets in banking: SMEs, HNWI, corporate, mass affluent, non-banked. Not only can this blueprint be applied to different segments, I also believe it can be applied to specific subject matters, one example being regtech as a service where a platform could serve as the integration/aggregation layer for many regtech solutions on the supply side and many financial institutions big and small on the demand side.

The Compte Nickel acquisition will make many take notice. BNPP signaled to the market this approach may be the most significant threat to banking – from a bank that already has its very own “digital bank startup” no less; proving that innovation is easier achieved outside of a corporation than inside. On this latter point, further proof of Conway’s law.

I expect the above blueprint to be copied, tweaked, optimized and deployed in the coming years. I expect EU banks to hold an inherent advantage compared to US banks in so doing, due to the pressure they face with PSD2. I expect both EU & US banks to weigh with all their might in favor of delaying regulatory or legal initiatives that facilitate this model. I expect the WeChat success, whether delivered by WeChat or copied by GAFA, to keep incumbents honest and help them recognize defensive plays alone are losing propositions. Basically, I expect the unbundling/rebundling of financial services for the coming 5-8 years to be drastically more vivid than it has been over the past 5-8 years. One last parting thought, it is inevitable that credit intermediation will be upended as a result of this blueprint. How money is created as well as how capital requirements are engineered in the aggregate will have to be revisited should this model take hold at scale – food for thought for another post.