Platform Banking Taxonomy


Like drunken sailors swinging fists at one another, we have been hurling around various terms to describe new ways of banking, new ways to deliver banking services. This post attempts to sort out a taxonomy and clarify the meaning behind the most salient terms.

I am using these terms within the context of the banking world in this post. Do note they apply equally to the insurance, asset management or payments worlds, indeed to the entire financial services industry.


API Banking: Also called “open banking,” API banking is the ability, for third parties, to access a bank’s software system thereby enabling a programmatic integration between an external third party application and a bank’s internal application via bank-grade security, authentication and access management.

Within the context of PSD2 in the European Union, banks are mandated to provide access to checking accounts, which will most probably be managed via APIs. In the US, several banks are working on developing various APIs to interact with a variety of fintech startups to provide an enhanced service to their customers or end users.

For example, Capital One has launched its DevExchange for 3rd party developers to leverage APIs it has built for two-factor authentication, rewards, and offers.

In and of itself, API banking is a tactic, not a strategy, although there can be strategic components to an API tool such as key policies, access management, volume, pricing. API Banking can be either push or pull driven:

  • Push: a bank can integrate to a service it needs (for example an API integration with a compliance service provider, or
  • Pull: a bank allows integration for a service its clients want or need.

Certain banks have started to develop APIs and early indications are these APIs are part of a bigger strategic intent. In other words, a bank’s API initiative could be part of a platform strategy.


Platform Strategy: The deployment of a set of business capabilities to maximize value creation across a value chain and articulated around defining what capabilities are core and remain within the responsibility of the bank and what capabilities are given to platform partners when delivering services or products to customers or users.

Technology companies such as Intel, Microsoft, Facebook, Amazon have been very successful at prosecuting platform strategies where value is delivered to customers while the platform owner/sponsor and the platform partners share in the value creation.

Historically, banks have crafted what many believed to be platform strategies where they owned the entire value stack and did not share with partners, In effect, banks created single-brand financial supermarkets. In our view, these efforts did not (and do not) qualify as platform strategies, as the platforms did not truly enable value creation along a value chain.

Platform strategies come in multiple flavors. For example, the platform strategy of Intel was/is very different than the one followed by Amazon. It should be noted that based on size, technical sophistication, market dominance, certain banks will own a platform – in platform parlance, they will be the platform sponsor – and its strategy, while other banks may, having strategically decided so, be partners of another bank’s platform strategy.

Certain large banks have developed platform strategies not immediately apparent to the fintech community. One example is the proprietary software platforms owned by global banks in the trade finance and supply chain finance sector.


Marketplace Banking: A type of platform strategy where a bank creates a digital place where third parties can showcase and sell their products and services to the bank’s customers. In a sense, a marketplace banking strategy is akin to the eBay or Amazon’s marketplaces where buyers and sellers of products meet and transact. Certain banks have or are in the process of developing app marketplaces.

The platform strategy, for the sponsor, will consist in defining the rules of engagement, the selection of vendors allowed to the marketplace, the governance, the monetization, data privacy issues, the level of technology integration, amongst other things.

Successfully executing a marketplace banking strategy will require the sponsor to deliver “match-making” capabilities to help consumers find the right producers—and vice versa. This will become a hurdle for many existing banks as they may be inclined to push their own proprietary products and services. A startup bank may be better positioned to deliver this capability.

Presumably, marketplace banking requires APIs. Retail Banks as well as Wholesale Banks can implement marketplace banking platforms. In as much as lending is predominantly a banking activity, notwithstanding non-bank lending, marketplace lending should be viewed as either a subset or first degree cousin of marketplace banking.

One can argue (as Philippe Gelis from Kantox has) that marketplace banking could be delivered by new entrants, such as a non-bank or a fintech startup or by an incumbent bank. Some fully digital startup banks in the UK have signaled their intent to build marketplaces.

It is my view, and that of Ron Shevlin, that this will be quite challenging for a startup to effectively deliver. To be successful with a marketplace banking strategy, the platform sponsor must be a “magnet” – drawing a critical mass of both consumers and producers to the marketplace. As a new entrant into the industry, this will be quite challenging for a startup. An existing bank has a head start as it has already has a critical mass of consumers to feed the marketplace. In other words, many have tried to become eBay or Amazon starting from scratch and only eBay and Amazon have succeeded.

Smaller banks could participate as vendors within the marketplace platform of a larger bank. In addition, it may be feasible for smaller banks to pursue a marketplace banking strategy if it is focused on a specific consumer segment with unique needs. We should expect marketplace banking to develop and segment itself by size, geography, type of service, type of customers.


Bank as a Service (BaaS): The delivery of certain banking capabilities in a programmatic fashion to enable third parties to deliver their own financial products or services.

For example, a bank could deliver AML/KYC services, checking account capabilities, financial data storage, payment services via an API. These services would then be used to build and deploy “last mile” financial services by a third party, be it a fintech startup, another bank, a non-bank. An analogy would be the technology services Amazon Web Services provides to its clients.

The strategic intent behind a BaaS strategy is the creation of new non-interest income revenue opportunities, created by driving down the marginal cost of delivering a given service to near zero.

BaaS can also deliver the necessary drivers to enable a marketplace banking strategy. A bank, a startup or a non-bank can implement BaaS, although an entity that is not licensed as a bank will presumably only deliver a subset of services, compared to a licensed bank. It should be noted that we are now seeing new entrants intent on providing BaaS, notably in Europe.

As with marketplace banking we should expect segmentation and specialization in this space. The various banks that have lent their license and/or balance sheet to provide certain services to alternative lenders (p2p, marketplace) should be viewed as proto-BaaS. Finally, certain fintech startups have developed a BaaS for specific services targeted at equity crowdfunding companies.


Bank as a Platform (BaaP): Fancy term for a bank’s platform strategy, does include API banking by definition and may include BaaS or marketplace banking.


A few more important thoughts. The “platformification” of the banking industry, in one way or another – as per the above definitions – will necessarily mean different approaches to strategic thinking and technology. As far as technology is concerned, and we have seen this occur with different industries and technology giants, such as the ones referenced above, open source and open standards or standardization of either technology building blocks or data/meta data and its associated methodologies and ontologies, are necessary and required.

We should therefore expect an acceleration towards standardization. We would not be surprised if certain financial technology building blocks would end up being released as open source libraries, very similarly to what has happened to the AI world (machine learning, deep learning) thereby helping the platformification process. Whether incumbents, new entrants or technology minded third parties with an interest in market optimization and social mandates do so is anyone’s guess.

I will also note that regulatory trends in the US may force banks to pursue platformification if banks are required to provide some kind of fiduciary responsibility for providing financial services (beyond just investment services).

If you want to learn more about the subject I recommend you revisit the following posts:

Articles written by Ron Shevlin:
The Platformification of Banking

Full Stack Banking: How Fintech Will Fuel API-Based Competition

Article written by Philip Gelis:
Why “Marketplace banking” is better for newcomers while “Platform banking” fits incumbents

Articles written by David Brear & myself:
Exploring Banking as a Platform (BaaP) model

Making Bank as a Platform a reality

Finally, I owe a debt a gratitude and special thanks to Ron Shevlin for pushing me to think through my arguments as well as having provided his thoughts and comments to this article.

As usual, thoughts and comments are welcomed and highly encouraged.


The Banker Men (Lyrics)

“The Banker Men” Lyrics
heavily borrowed and distorted from Sammy Davis Jr.’s “Candy Man” song


(chorus in parentheses)

(Banker men)
(Hey, Banker men)

Alright everybody, gather ‘round
The Banker men are here
What kind of banking do you want?

Sweet banking?
Sour banking?
Long or short banking?

Anything you want…

You’ve come to the right men
‘cuz we are the Banker men


Who can take the fed rate, push it oh-so-low
Then wonder why the rate of GDP growth is so slow?
The Central Bank can, any Central Bank can
Who makes monetary easing, oh-so-very pleasing
The Central Bank plan, the Central Bank plan

Who can take a yield curve, sprinkle it askew
Shower it with rate cuts and a miracle or two
Kuroda-san can, Kuroda-san can
‘cuz he floods us all in yen, then he floods us all again
The Kuroda-san con, the Kuroda-san con


(The Central Banker makes everything he monetizes satisfying and delicious
Now you talk about your retirement wishes, better hope for loaves and fishes…)

(Yeah, Yeah, Yeah)


Who takes the interbank rate, smashes it to rubble
To fertilize and monetize a bloated asset bubble?
The Bernanke man can, the Bernanke man can
He waves his magic wand and he floods the world with bonds
It’s the Bernanke man scam, the Bernanke man scam

Who can take the Euro, prop it up just so
Relax a bit, and then the bloody Brits decide to go
The Draghi man can, the Draghi man can
His yield curve is inverted, and policies perverted
It’s the Draghi man flam, the Draghi man flam


(The Central Banker makes everything he monetizes satisfying and delicious
Now you talk about your retirement wishes, better hope for loaves and fishes…)

(Yeah, Yeah, Yeah)


Who can take deposits, give interest close to naught
Keep the party going, and just hope that she’s not caught?
The Yellen-melon can, oh the Yellen-melon can
Doesn’t matter what we save, our pensions still get shaved
It’s the Yellen-melon slam, the Yellen-melon slam

If revenue’s a dollar, who can borrow three?
Then pass the bill to you, your kids and all your family.
Your Congressman can, your Congressman can
No point in getting angered, his district’s gerrymandered
It’s the Congressman sham, the Congressman sham


(The Central Banker makes everything he monetizes satisfying and delicious
Now you talk about your retirement wishes, better hope for loaves and fishes…)

(Yeah, Yeah, Yeah)


Give us a safe haven, please who can it be?
Forget our indiscretions and protect us while we flee
The Swiss Franc man can; the Swiss Franc man can
Zurich has its faults, but there’s gold in them thar’ vaults
It’s the Alpine-man band, the Alpine-man band

By punning on an old song, I‘ve bored you all to tears
But still, this took more effort than the Fed has made for years
‘cuz Alan Greenspan, Alan Greenspan
From boom to bust to boom, he’s the smartest man in the room?
It’s the Greenspan-man scam, the Greenspan-scam


(The Central Banker makes everything he monetizes satisfying and delicious
Now you talk about your retirement wishes, better hope for loaves and fishes…)

(Yeah, Yeah, Yeah)


(a-Central Banker can, a-Central Banker can, a-Central Banker can)
(a-Central Banker can, a-Central Banker can, a-Central Banker can)


Entrepreneurship and the Id Machine



Slavoj Zizek coined the term Id Machine to describe an engine that allows for the materialization of one’s desires. Id is the unorganized part of one’s personality which contains one’s most instinctual drives. The Id contains the libido which is the source of energy that is unresponsive to reality.

Zizek applied the Id Machine to two movies by Andrei Tarkovsky: Solaris and Stalker.

In both movies, protagonists are faced with an “area” – in Stalker the area is called the room, located within the zone – where their desires are materialized. Zizek names this “area” the Id Machine. The Id Machines are different in each movie. In Solaris, protagonists do not have any control over which desire materializes itself once they are in the Id machine, thus leading to terrifying realizations. In Stalker, protagonists need to figure out what they desires. The realization they sometimes do not know what they desire also leads to terrifying realizations.

Tarkovsky’s philosophical musings, as interpreted by Zizek give us one Id Machine that reveals the perils of our passive nature and another Id Machine that reveals the perils of our active nature. we get tripped by the unknown parts of our desires, by what we think are our desires and by our inability to formulate our desires.

I think parallels can be drawn with entrepreneurs, startups and venture investing. VCs and entrepreneurs all enter an Id Machine at some point, where our desires materialize. Outcomes are never certain. Some outcomes are unexpected, others should have been expected, very few come out as expected.

A Startup’s main protagonists – entrepreneurs and venture investors – need to go through much introspection to sift through their desires. By desires I mean goals, visions, strategy, tactics. Clarity and transparency are paramount. Paradoxically, as the libido is the source of energy unresponsive to reality; the entrepreneur – and to a lesser extent the venture investor – also needs to be “unresponsive to reality”. In other words, the entrepreneur needs to be unshackled from the constraints of reality in order to achieve his dreams. However, she should not make complete abstraction of reality. Complete abstraction from reality leads to either Solaris or Stalker’s Id Machine, with suboptimal results.

I was recently asked what I actively sought in blockchain startups when investing. My answer was interoperability with the real world, pointing to the necessity for a blockchain startup to take into account the realities of the law, especially in the context of securities law in the capital markets space.

I thought further about my interoperability answer in light of Tarkovsky’s movies and Zizek’s interpretation and believe I apply it to all startups. Further elaborating on interoperability, I define it as the quality to will a new reality while understanding the constraints of a current reality, incorporating these constraints within one’s thought processes, and using them to the best of one’s advantages. This is the quality I seek in an entrepreneur and in a startup. In Freudian terms, I seek an entrepreneur who can apply the right ego touches to her id. Too much ego touches and the id’s desires never materialize, too little ego touches and the traps of the Id machine come in play. The right ego touch is also essential in regulating the id’s tendency for instant gratification. Organic growth with the right tempo is often not recognized as one factor of success with startups. There are many pitfalls with fast growth and/or high valuations within short periods of time (too little growth also being a killer). This I view as being part of a certain interoperability with the real world, or with certain natural laws of organic growth.

I find the above amusing on a personal note as I have always been more Jungian than Freudian in my interpretations. That may provide me material for another post.


Financial Services Productivity


It’s a funny thing, productivity. Very easy to define – producing more with less – but more difficult to measure. Productivity is easier to define for a given company (revenue per employee for example), somewhat easier to compare amongst like minded companies in the same industry, more difficult to use as a metric across industries, complex when applied to services as opposed to manufacturing industries and utterly bewildering when taking into account qualitative factors.

Productivity occurs when firms “innovate”. I use quotation marks because there are so many different ways to segment and qualify innovation. At meta level, innovation is the application of better technologies to an economic process – a technology being a technique or collection of techniques invented by man.

The narrative unfolds as such: a) we invent new technologies, then b) we innovate by applying these new technologies, and as a result c) we become more productive.

Over long periods of time this cycle benefits societies as goods and services in a given industry become better and cheaper. As an example and as a result of productivity gains, there are now less individuals engaged in food production and the cost of food in our daily budgets has plummeted. In other words agriculture has become vastly more productive.

Contrary to what many may think, the financial services industry has always been a heavy user of technology. To name but a few, advanced telecommunications applied to financial services have facilitated cross border transactions, advanced computing has helped the securitization industry, advanced data science has helped intricate trade and investment strategies. I would therefore state that financial services firms, on the aggregate, have always been innovators.

Have they become more productive though? In absolute, or relatively speaking?

Looking at revenue per employee and operating income (OI) per employee as crude productivity metrics for 2015, Bank of America delivers $392k in revenue and $104k in OI per employee vs Facebook which clocks $1.24m in revenue and $435k in OI per employee. On the face of it, Facebook is vastly more productive than Bank of America. The comparison gets more interesting when adding Goldman Sachs with $1m in revenue and $240k in OI per employee and Visa with $1.23m in revenue and $796k in OI per employees. Obviously some financial services firms are more productive than others and rival tech giants such as Facebook. These comparisons are unfair though as the business models are vastly different. Still, firms like Visa – other likes MasterCard or the CME Group come to mind –  are more technology-intensive companies while any bank – with the exception of Goldman – are less technology-intensive.

From an empirical point of view, we know a majority of consumers and entreprises are dissatisfied with their financial institutions. Quality of service as well as user experience are poor, services are slow and inefficient, products and services are costly. Even if it is difficult to gauge the qualitative and quantitative impact Wikipedia has had on our productivity, it is undeniable there has been a positive impact. Even though it is difficult to gauge the qualitative impact of the financial services industry on financial wealth and health in the aggregate, it is undeniable the impact has been in certain instances negative.

From a macro-industry point of view, Thomas Philippon, a professor of finance with NYU Stern School, recently wrote that, as per his analysis and research, the unit cost of financial intermediation had remained constant at 2% from 1886 til 2015, see here. This means that any intermediated asset has cost users 2 cents for every dollar AND has remained constant for well over a century! This is actually the greatest indictment of the financial services industry one could every come up with. Arguably, during this period the costs of many products and services in other industries have dropped while quality increased. Not so for financial services.

To be clear, financial services firms have been innovators and they have become more productive as evidenced by the massive profits the industry has experienced. Shareholders and employees have benefited. (Even after the 2008 financial crisis, financial services profits have reached record highs. The banking industry alone hit a record of $1 trillion in profits worldwide in 2014.) The costs of financial products and services has not decreased for the end users. Further, as profits were more than adequate, the costs of delivering products and services did not decrease either.

Innovation and productivity did occur, only for the industry itself though.

When taking into account the massive scale of the financial services industry – between 15% and 17% of total GDP depending the economic cycle and the exuberance of financial markets – this has to result in major challenges for any economy. The primary function of financial services is to optimally allocate capital. In other words the industry needs to help us spend money, send money, receive money, invest money, save money, insure, in the best possible way. If the process whereby all these activities is essentially “rigged”, economic activity suffers. There are obviously high level considerations – fiscal, monetary and political – when analyzing the efficiency of financial services, especially from a macro point of view. I only focus on technology, innovation and the resulting business models that can emerge once “real” productivity takes hold, as opposed to “rent-seeking” productivity.

One can argue that Venture Capital is one of the enablers of sea-changing innovation with the systemic application of new technologies. Indeed, the first wave of fintech, emanating from the Silicon Valley and focused on backing direct to consumer models bent on competing against financial services incumbents was based on the oft successful VC/Entrepreneur strategy applied to other industries. That this first wave was not as successful as it was originally thought does not mean “end-user centric” productivity will not finally permeate financial services. On the contrary, it was a necessary first wave that shook the industry into action.

Whether incumbent will successfully reinvent themselves, startups will win meaningful market share or partnerships between incumbents and startups is the way of the future is opened for debate. What is not open for debate, is the unavoidable imperative towards finally lowering the marginal cost of delivering financial products or services and eventually lowering the cost of products or services (within reason as one cannot lower the cost of borrowing for example).

All the narratives unfolding under our very eyes – digitization, platform as a service, chatbots, roboadvisory, alternative lending, APIs, cognitive banking or insurance, blockchain, faster payments… – are emanations of this unavoidable imperative.

I recently checked US financial services payroll on the Bureau of Labor and Statistics’ website. Interestingly enough the US financial services industry employed approximately 8.5m people prior to the 2008 crisis. Employment stands now at around 7.9m and is expected to grow to 8.4m by 2020. I am puzzled by this forecast as I expect financial services industry payrolls to continue to decrease in developed countries (US and Europe included) as more inefficiencies are weeded out of the system. (Facebook employs 14,500, Visa employs 11,300 while BoA employs 210,000; there is still much to do.)

No discussion about financial services productivity would be complete without mentioning regulation. Indeed, regulators can be viewed as having been complicit in the building of a rent-seeking industry. The rate of change of technology has accelerated to such a degree and consumer behaviors and expectations have changed to such an extent that financial services regulators cannot afford business as usual. Thusly the novel approach to innovation the Financial Conduct Authority has taken in the UK or the Monetary Authority of Singapore. Every regulator is now actively thinking or devising new ways of engaging the eco-system they regulate and this includes how innovation impacts these eco-systems.

The lesson here is everyone is breathing fintech, from service providers to incumbents to regulators and startups, as a vector to deliver productivity gains.

I want Thomas Philippon to run the numbers in 5 and 10 years from now, and I will be crushed if the cost of intermediating an asset will not have dropped to below 1%. How low can we go?


Differential Diagnostics, Venture Capital & Zebras


Yesterday evening I had dinner with a good friend of mine who is a world renowned cardiothoracic surgeon. I asked him if he followed a framework when dealing with each patient and he brought up the subject of differential diagnosis. At its core, differential diagnosis is a method used to identify a disease when alternatives are possible while utilizing a process of elimination. A doctor will assess a patient in context (symptoms, patient’s history) and taking into account medical knowledge, go through a decision tree, starting from most likely diagnosis, eliminating each alternative until the right diagnosis is reached.

There are two approaches to differential diagnosis. The specialist and the generalist approach. The specialist approach – used by a surgeon for example – utilizes a sharp shooter technique, selecting from the most likely to the least alternative, one alternative at a time. The specialist approach is narrow and deep. The generalist approach – used by a family doctor for example – utilizes a broad brush technique, also selecting from the most likely to least likely alternative yet considering a group of alternatives together. The generalist approach is broad and shallow (and I do not mean this in a negative way).

Medical doctors have to learn an incredible amount of historical knowledge and then have to practice extensively in live conditions, in hospitals, before becoming experts in their fields. The body of knowledge at their disposal does not change markedly – it is not like we are inventing new diseases, ailments, different ways of breaking a bone on a regular basis. The medical tools, medical drugs at their disposal, and the medical techniques do change. So there is a constant “on the job” training occurring.

The framework I use in venture capital strikes me as eerily similar to differential diagnostics. First, I  am a specialist venture investor as I only invest in fintech. It goes without saying that I need to develop a very deep understanding of the financial services world in order to be effective at my job. Without explicitly knowing – it until now – I have developed a sharp shooter approach, akin to the one used by my surgeon friend, that allows me to very quickly assess the merits of a payments startup for example. For each of the five sectors that comprise fintech – lending, capital markets, insurance, asset management and payments – I have a top 10 of “things” I look for for which the presence or the absence are a deal killer. I rarely need to go past thing 3 or 4.

I use the sharp shooter differential diagnostic approach when I first encounter a startup. it is a way for me to eliminate the noise and get to the signs fastl. If I am still interested and impressed past this first stage, I will switch to a generalist differential diagnostic approach where I bunch groups of “things” and attempt to figure out, holistically adds systemically, patterns I like/do not like or that make sense/do not make sense, repeating the process until I eliminate the startup as a potential investment or I confirm my initial positive signal.

Much like my surgeon friend who has to go through thousands of cases per year to hone his skills, I go through approximately 1,000 business models per year. This is the material I need, along with historical knowledge base I built over the years – a mix of theoretical knowledge and many years of practice as both an operator and investor – to keep current. The number of business models does not change at the margin that much, the number of ways a team should be built, how a startup should be scaled, a board should be architected – all the business aspects of building a business –  do not vary that much. What changes are the the technologies and how they are applied to specific business models. So I need to constantly learn that aspect to stay ahead.How AI, quantum computing, AR will be applied to fintech are my learning curves.

I continue to apply both differential diagnostics frameworks during the lifetime of an investment, constantly toggling from one to another.

I believe the best VCs are good at differential diagnostics. Not only because they master the framework and have built their own heuristics in their particular domains, but because they also know when to switch from sharp shooter to generalist differential diagnostics. That is a crucial skill. I also believe top VCs are more adept at applying differential diagnostics in context. By that I mean that – taking a fintech example – a US payments company may need a different sharp shooting approach than a EU payments company, while one may need the same generalist approach for both. It all depends on nuances relating to culture, jurisdiction, consumer/user behaviors, market structure. I tend to call these nuances “terroir”. Yes, I like wine. Knowledge of terroir will help you choose the right differential diagnostics approach at the right time, and load the right decision trees.

I also believe specialist VCs have an edge over generalist VCs. To be clear, both need to master the two differential diagnostic techniques. The specialist VC will always have an edge with the sharp shooter technique given the required deep knowledge she needs to operate in only one field. This is especially important considering the changing VC landscape is currently experiencing: the rise of crowdfunding and angel investing on one end of the spectrum and that of corporate VCs, sovereign wealth funds, mutual funds and large PE funds on the other end of the spectrum may force traditional VC funds to specialize in order to retain an edge. Specialized VCs may be the way of the future.

I am also well aware that medical doctors have an edge over venture capital investors when it comes to track records. On the evidence, declining mortality rates and improved longevity beat hands down VC-backed startup survival rates. This means that even with the best differential diagnostics tools and the most astute and timely ways to apply said tools and make a decision, venture investing is an extraordinarily difficult business to succeed in. There is much literature attesting to this fact. VC investing and startups building are ruled by power laws.

I do not pretend to disprove nor fight this fact. What I do is try to refine the odds ever so slightly. For me this means to always have Zebras in mind.

Theodore Woodward, a 1940s professor of medicine coined the aphorism “When you hear hoofbeats, think of horses not zebras.” He meant that if you diagnose something “normal” applying your diagnostic tools, there is a great chance it is indeed a “normal” thing and not something else, something “exotic”.

This works well in the medical field. Not so well in venture capital.

Hence, if there is one thing that keeps me up at night, it is Zebras. Due to the unfathomable emerging properties of large systems, venture investing breeds many more Zebras than horses, even though you may have correctly diagnosed a horse from the beginning. By that I mean that you may start with a horse, but due to unforeseen circumstances, you end up with something else, a Zebra. Very few Zebras end up with positive outcomes. The great majority of Zebras experience neutral to negative outcomes.

Thusly it is imperative to be paranoid about Zebras. I endeavor to excel at differential diagnostics which is a necessary requirement but not a sufficient one. Additionally I try to take risks I can measure in ways that attempt to mitigate negative Zebra effects. I shy away from entrepreneurs and startups that open themselves to fragility. I favor entrepreneurs and startups that strive to capture optionality and build antifragility. This means favoring entrepreneurs and startups that exhibit the right mix of technology, business and talent (the necessary requirements) AND that will thrive during volatile business conditions OR that do not include business variables whose rate of change increases negatively as business conditions fluctuate. Examples of fragility would be a cost of acquisition that increases as the startup increases traction, churn that increases the more clients are acquired, a loan default rate that increases as interest rates increase, a technology build that increases in complexity even as the startup matures. I picked up fragility and antifragility concepts from Nassem Taleb, and encourage anyone involved in investing and startups to read his work. Much more could be written about how one can apply antifragility thinking to startup investing; for another post maybe.

In as much as I apply differential diagnostics techniques to scrutinize the form and substance of a startup, my Zebra heuristics helps me understand the likelihood such form and substance will behave positively in dynamic situations. Not a perfect approach for sure.

The best VCs excel at diagnosing the right horses then shunning the patently negative Zebras. This still leaves the field wide open for a variety of surprises.


Fintech Food for Thought


Statement: It is cheaper to create a fintech startup today than 15 years ago, yet very few fintech startups reach escape velocity and have been able to build a sustainable business yet. There are plenty of fintech unicorns but there is only one PayPal to date.

Question(s): Does the fintech startup scene obey an even more severe power law of success or is it too early to tell?


Statement: Financial Services incumbents continue to be hurt by a low interest rate environment that hurts their profitability and severely constrains them in the marketplace.

Question(s): Would a high interest rate environment limit financial services innovation to systematic progress, to the detriment of systemic progress? Would interest rates increases limit the ability disruptive fintech startups have at competing against financial services incumbents?


Statement: Incumbents notoriously do poorly with innovation. They are beset by agency issues, inflexibility, bureaucracy. They are also the first to retrench when failures arise.

Question(s): Will incumbents exhibit the same tendencies at such a pivotal point of transition to the new digital age? Or will they exhibit more resiliency as a matter of survival.


Statement: New technologies, new behaviors, new business models are giving rise to the omnipresence and the power of networks and platforms in an industry where very complex processes are the norm and where mastering these processes require depth and breadth of knowledge.

Question(s): Which is more likely, a) disruption coming from fintech startups alone, b) fintech startups failing to dislodge financial services incumbents, or c) collaboration between startups and incumbents?


Statement: Many fintech startups are building businesses in either fragile activities (lending) or “race to the bottom” activities (remittances, p2p payments)

Question(s): How difficult will it be for these startups to build resilient businesses long term? Will financial services incumbents be negatively impacted?


Statement: Most if not all financial services operations are eminently complex, standards and regulatory rules add to the cost of doing business, even more so when cross border processes are taken into account.

Question(s): Does this mean the capital requirements to build a sustainable fintech startup at scale – and the current size of financing rounds – is too high or too low? How will financing rounds size trend going forward?


Statement: Financial services incumbents boards are light on technology gravitas and knowledge. Fintech startup boards are light on deep financial services knowledge and understanding.

Question(s): Which will close the knowledge and experience gap first? Can the gap be closed?


Statement: Innovation is about taking risk. Running a financial services business is about managing risk.

Question(s): Can these two activities be reconciled? Under what circumstances?


Statement: The financial services industry is undergoing profound change and is also under tremendous stress. Banks, Insurance, Asset Managers are faced with existential threats – real or perceived. Every participant in the industry is responding to change, even forward thinking regulators in certain jurisdictions – UK, Singapore.

Question(s): Can financial services regulators avoid further change to their own business models? Can they get away with systematic change or will they have to contemplate systemic change? Are the equipped to innovate within their midst? What will be the consequences if they do not change and adapt?


Statement: Financial services participants such as PayPal in the US, Starbucks – and others – “hold” sometimes more money on behalf of their customers than certain banks do. These actors do not hold bank licenses nor are they subjected to the same level of scrutiny as banks.

Question(s): Will this trend increase, both in terms of quantity of money held and number of participants? If so, will regulators pay a closer look at these participants and will regulation take into account the weight these participants hold within the overall market structure?


Statement: Bank or Insurer owned Venture Capital firms invest with a strategic mandate. Independent Venture Capital firms are not encumbered but such constraints.

Question(s): Which yields the best outcomes? For investors, for the incumbent parent? Is it sufficient for a bank or insurer to own its own venture fund? Should it be better for a bank or insurer to invest in an independent venture fund? Would both owning a venture fund and investing in an independent fund be optimal?


Statement: Financial services incumbent IT/IS staff are usually convinced they are better at building new products, services, platforms. Fintech Startups are usually convinced they are better at going to market first.

Question(s): Which is the most value destructive behavior? Which behavior is the easiest to correct?


Statement: Fundamental and economically productive product or service or business model innovation in the financial services industry has been scarce- e.g. mortgages, ATM, securitization. Most innovation has benefited the speculating activities prevalent in asset management, trading, capital markets.

Question(s): Will new technologies and their application via fintech further this trend or invert it?


Statement: Many seasoned and reputable venture capital investors have gone on record stating corporate venture firms do not know how to invest and incumbents have a poor record with innovation. Most corporate venture capital investors are convinced fintech startups know little about the financial services industry.

Question(s): Which belief is the most erroneous? If true, which is easiest to upgrade?


Statement: In part due to local legislative and regulatory DNA, in part due to entrepreneurial genius, in part due to the size of their market, Chinese fintech firms (pure plays or children of Chinese tech giants) are ahead compared to their Western brethren. Further, based on recent evidence, cracking the Chinese market is a non trivial endeavor for a US or a European startup. US and European fintech actors do not enjoy the same advantages Chinese fintech actors do.

Question(s): Will Chinese fintech actors expand to Europe and the US? If so, how will Western regulators and legislators react? Will Chinese financial services markets mature to the point of being opened and interoperable with the outside world?


Statement: To date, the vectors of financial services industry disruption and innovation have been technology, a change in consumer and enterprise habits, the Great Recession, strengthened regulatory oversight, entrepreneurial spirit and a low interest rates environment. These have, to a large extent been forced upon the industry and its incumbents. Notably absent has been the political sphere – executive or legislative.

Question(s): Will the political sphere engage with fintech and the financial services industry transformation? What will be the likely effects?


Statement: Fintech innovation needs both talent and capital.

Question(s): Which of talent or capital is more constrained? Are we faced with a demand or a supply issue? How will this change in the future?


Statement: Transitioning from the industrial age to the digital age induces profound implications. The way we organize ourselves, transact with one another, interact with one another are and will be drastically different. So will the skills, business architectures, mustering of resources and capital to sustain new models. Particularly so in the financial services industry. Incumbents have the advantage of political clout, access to high level spheres of power and decision making. Startups and entrepreneurs master the art of creation – sometimes successfully. Be that as it may both need to see the future differently than they experienced the past.

Question(s):  Is that transformation purely technology and business dependent? If not can either startups or incumbents transform the industry for the digital age without political leaders that understand what the digital age needs? Have political leaders emerged in any country or continent that understands the new age we are entering and its implications to the financial services industry and fintech as its enabler?


Statement: We are witnessing many changes within the financial services industry. Yet, Money, the concept of money has not changed for may generations.

Question(s): Should the concept of Money change? If not, why? If so, which is the most likely vector to effect a change; technology, politics?


You are welcome to come up with your own statements and associated question(s). Please comment and share.


Individual Identity Rights


I wrote about new business opportunities for financial services incumbents, specifically banks, in my previous post. More notably, I posited that 1) because banks were in the Trust Business 2) they have an opportunity to expand their offerings by 3) protecting their customers’ IDENTITY and DATA much like they protect customer’s money today.

Soon after I published that post, I came across a short video by Tyler Cowen (see here) in which he discusses the importance of trust in the banking relationship. He points out that trust is made possible by a shared understanding that individual property rights exist and will be enforced by the state.  A bank that takes customer money can’t just keep it, and has legal obligations to protect it. Tyler’s video reminded me the multifaceted aspects of trust and that I had only touched on the trust between a bank and its customers.

Given that I believe in how technology is and will enable individuals to utilize their identities contextually and enable them to monetize their own data, that video spurred me to think about data and identity ownership. To be clear, this post is not about exploring new business models, rather it is about understanding what data means legally to us and the implications of ownership and rights associated with data.

Our lives are increasingly defined by the electronic data stored in third party databases that is generated by day to day activities, for which limited records existed even a decade ago.  Drive your car, by groceries, visit a web-site, pay a toll electronically – data is harvested, data is stored. When aggregated, these prosaic electronic breadcrumbs have massive economic value. Indeed, considering that our economies are undergoing a massive realignment and restructuring, moving away from the industrial age towards the digital age, it is easy to realize that the data and metadata we generate (about ourselves, our behaviors, our habits, our consumptions) or that of our own physical assets generate will be increasingly valuable.

And the amount of data we generate is increasing, not decreasing. If our data and our identities are already valuable today, they will be more so tomorrow. At this particular historical moment, the commercial value of consumer data is a one-way street. Once a business has your data, they may have legal obligations to you, depending on the state or country where you live (HIPPA and Graham Leach Bliley are two U.S. examples). But you don’t have a financial stake in the data.  Say, (for example) that an advertiser makes money by sending an ad targeted to you because of knowledge about your purchasing preference. You’re not going to receive a commission for the use of or reliance on your data. Where does personal data fit into the framework of traditional property law?  This is an admittedly broad question, but we can make some general observations. Why does this matter?  Most economists – or so I hope – agree that a strong protection of property is one of the most important vectors driving economic activity and wealth creation. Western industrial capitalism is premised on the understanding that individuals have the right to enjoy their private property without fearing it being stolen or misappropriated by a third party, let alone a government.

Generally speaking, there are two types of property. Tangible property, refers to physical things, (a house, a plot of land, a car, physical cash, gold…). Intangible Property, refers to incorporeal assets (intellectual property (“IP”, copyrights, patents, trademarks), corporate good will, securities, security interests,  and dematerialized investments, money, …).

So – what is “personal data”, the stuff that makes up our identities. It’s definitely intangible, but it is certainly not a dematerialized investment or money. Could it and should it be considered and individual’s IP? The answer is most probably not. Could it be a corporation’s IP? Maybe so. The lack of clarity on data and its ownership is indeed tricky.

The Merriam-Webster dictionary defines IP as “something (such as an idea, invention or process) that comes from a person’s mind”. Modern IP laws arose out of the need to protect personal creation. The printing press, mass media, the internet are technology vectors that increased the value of one’s creation. Commercial interests required strong protection and licensing laws. As such, traditional IP comes out of active creation.

Can we say the same of all the data and metadata associated with our health our payments history, our interactions with our social media/networks, our apps, our smartphones and IoT? Or are we faced with passive creation. Would these types of data and metadata be treated as IP or are they in a class of their own? My non-legal-expert view is that we are dealing with a new class of property borne out of new ways to create it, enabled by new technologies and ultimately supporting new economic activity which demands new legal constructs.

The same questions and comments apply to our Identities – physical, digital, private, health, financial.

Clarity on what personal data is leads to clarity on what types of rights can be associated with it, and to the extent there are gaps, what types of rights should be developed.

Ownership is equally important. Who owns our data? In some instances we do, in others we cede control as part of a Term of Service we barely read, and in yet others we probably wade in a grey area where those that use and monetize our data are more than content to keep the status quo and not explicitly spell out ownership.

I strongly believe data ownership frameworks need to be brought up to the level of sophistication of data privacy laws. How our data can be used, how it should be protected data is a national and international discourse our governments, the corporations we interact with and ourselves are engaged in continuously and for many good reasons. No one can use or misuse private information without prior consent, no one can handle our private information carelessly. We already have the right to digital seclusion (i can restrict access to my Facebook or Twitter identity to a handful of trusted friends, or altogether shut it completely) and are slowly gaining the right to be forgotten digitally.

Rights associated with having, owning and securing a personal identity are intertwined with self-determination, basic human rights and freedom of speech.

Up to now the sum total of rights associated with data, which I label Individual Identity Rights have not coalesced into a systemic societal issue. Too many interested parties want their hands on our data with as little friction as possible. Enterprises because of monetization potential, Governments because of their thirst for transparency and control. The early stage of the digital age have mirrored the industrial age from a centralization point of view. Large intermediators such as Ebay, Facebook, Amazon or Google have dominated – and will continue to do so for many years to come. Be that as it may, the potential of blockchain technology is enabling decentralized business models to emerge. Soon we may have the choice to conduct our private business (sharing with friends, buying, selling, creating) with a decentralized marketplace, a decentralized social network, a decentralized search engine – the list goes on. The data we generate on these platforms will be our own, and we better have ownership rights that reflect such an unequivocal fact.

Up to now the ways and frequency we have needed to produce a form of identity to gain access to a service, a product or a place has been limited. Both will increase and with them the complexity of provisioning and managing our identities. The multiple identities we will create and inhabit better have the same ownership rights that reflect how central identity will be in our post-industrial world.

Up to now we have not paid much attention to our data and have been more than content to cede its monetization to third parties in exchange for convenience or entertainment. As data will rise as one of the central vectors of our economic and social engines we will want to control and share in the wealth creation, we will demand more transparency with regards to who will use our data, for how long and in what capacity, and we better have ownership rights that reflect these value chains.

Individual property rights have been essential to wealth creation in the industrial age. Individual identity rights will be essential to wealth creation in the digital age.


I would like to thank Stephen Palley for helping me think through my arguments, providing invaluable feedback and editorial support.


That Banking Moment


Today there are more bankers convinced of the need to transform their businesses than those that are not. This is no small matter as realizing the need to change is half the battle. The other half of the battle is to find the right solutions and implement accordingly.

In order to find the right solution one has to ask the right questions. I have struggled to find the right framework for these questions until I came upon this article by Scott Anthony.

Scott outlines three main questions:

- What business are we in today?

- What new opportunities does the disruption open up?

- What capabilities do we need to realize these opportunities?

Here is my attempt at answering these questions for a Bank.

- What business is a Bank in today?
Taking my cue from Scott, I will avoid the obfuscating and basic answers such as “offer accounts”, “lends”, “makes payments” which are either technology based or category based. More abstractly, a bank acts as an intermediary by linking depositors and borrowers. In comes deposits, safely tucked in accounts, out comes loans safely underwritten to borrowers – or so we hope. This intermediation role creates various benefits: a) spurs economic activity and supports the community in which the bank operates, b) safeguards and protects money entrusted by customers, c) provides access and convenience to money and how it is transacted, d) builds wealth directly (lending activity needs to be profitable) and indirectly (savings, investments). Abstracting further, a bank is in the business of providing trusted services around a customer’s money. Abstracting even further, a bank is in the “TRUST” business. Do note there is a major difference between being in the money business and being in the trust business. Thinking of being in the money business forces you to think in terms of products and services around how money is stored (checking account), transferred (payments), invested (assets) or lent (loans). The outcome of such a paradigm is to sell products. Such outcome may not have been explicit when banks operated in small environments, serving defined geographies where the relationship a banker had with his community was the vector that enabled all. This outcome is explicit in modern banking however. Therein lies the conundrum and the creative/destructive tension. Banks have ended up engaging in the business of selling products that serve a function around money whereas their existential function is to extend and project TRUST. Many pundits have recently declared banks need to be less product centric and more customer centric as a result of this tension. I agree and will unequivocally and irrevocably state that a Bank needs to reclaim and redeploy TRUST. Without trust, there are no bankers. Without trust there is no bank.

- What new opportunities does the disruption open up for a Bank?
In an era where new ways to invest, underwrite risk, lend, transfer money are being rolled out, all of which necessitating less knowledge centralized in an individual’s brain (a banker) or an organization (a bank),  where the way we spend our time and our money occurs less within the constraints of the physical world and more via digital means; a Bank is rapidly finding itself threatened and ultimately disintermediated as an agent handling our money. We also live a paradox where we do not “like” our Bank – we spend less and less time in contact with its employees or its branches and we profoundly dislike the excesses of some bankers and the opacity, applicability or utility of many banking products – while we “like” our new sacred cows – we spend more and more time on our beloved social media apps, marketplaces, social messaging apps, social gaming apps, business apps – yet we TRUST our Bank more than we trust our new sacred cows. Lonely is the pundit advocating we store our money with Facebook or the customer ready to do so. Banks have so far treated this phenomenon as an existential threat. I posit this phenomenon is actually an opportunity. A major opportunity.

As a result of our digital engagement we have experienced an explosion in the amount of data we generate. We are drowning in data and metadata. Our identities have multiplied to the point where our confusion about their management is only surpassed by the threats we face every day from hackers. Whereas software and hardware are the vessels, arteries and vital organs of any functioning business, data has become its lifeblood. The second coming of artificial intelligence will only further the point I want to make: Data has become an asset class and will become more and more valuable, unlocking a multitude of values we cannot begin to imagine today, for us and those we engage with.

Tying TRUST and DATA together, I come to the inevitable conclusion that today’s opportunity for a Bank is to provide TRUST services around its customers data. Data is what you do, who you are and how you evolve today. It will be what you monetize tomorrow. So far, we, the real owners of data, have been cut off from its monetization, with consent – engaging in a quid pro quo with a social network – without consent – with little control over how one’s data is used to price a loan for example.

Let’s imagine a Bank offering its clients a master account, part checking account where a client will entrust money, part data account where a client will entrust data. Let’s further imagine this Bank will monetize the data residing in the data account and – much like with different flavors of traditional bank accounts – will offer a cut off the revenue generated. Little to nothing if the customer consents to narrow use cases, narrow data sets or anonymized data. Much more if the customer consents to wide use cases, wider data sets or personalized data. Let’s further imagine this Bank will also provide services around a customer’s identity: verifying one’s identity based on the requirements of third party services, individuals or entities. Imagine that and you have imagined a Bank reinventing its core tenet, TRUST in the age of DATA and IDENTITY. In a subtle way, this reinvention is akin to a Bank finding back its original roots. Indeed, an old school banker was entrusted with his customers data when interacting with them in the community. The data resided in the banker’s head, shared only because of the trust factor. Tomorrow, the data will reside in the cloud, protected by one’s Bank, with a trust factor.

To convince you further of the validity of such a thesis, consider what the likes of Google, Amazon or Facebook are interested in? Are they rushing to obtain a bank license to handle money or are they focusing on harnessing the power of data? I will leave you to answer this question on your own and ponder the competitive pressures banks are and will face whether they choose to own and manage trusted data or not.

The other major opportunity I see for a Bank resides with the ability to orchestrate a value chain – instead of the old paradigm of owning the entire value chain. I analyzed this opportunity in previous post. The concept of Bank as a Service, Bank as a Platform, the Platformification of Banking is slowing taking hold in the ecosystem. A few startups have capitalized on this trend already, a few Tier 1 Banks have made preliminary moves. I do not pretend there will be only one new Banking reality of course and some banks will not chose the “value orchestration” path. What I am convinced of is that “value orchestration” is a major opportunity. The shear amount of data and transactions we are and will continue to generate within the context of heterogenous and diverse technology ecosystems we elect to engage with requires a new breed of Banks adept at organizing, servicing, facilitating and sharing work flows and processes across a financial services value chain.

So far we see several trends unfolding: a) the buildup of an ecosystem of fintech startups, b) the strong gravitational pull of social networks + messaging apps (soon to be joined by the full force of AI powered chatbots) exercised over our daily attention, c) a secular trend towards peer to peer relations or horizontal networks (sharing/renting economy, blockchain, cryptocurrencies…) d) the resulting arms race all banks have undertaken to digitize their customer touch points.

This arms race is the result of the mistaken assumption that retaining customer attention by owning it fully is the main way to continue delivering value creation. I am not convinced and even if I were, competing for attention against nimble upstarts, savvy tech giants or the secular horizontal network trend is a strategy I do not like the odds of – few banks will survive doing so. Rather, refocusing one’s strategy on value orchestration to facilitate and enable the seamless inclusion of financial services conversations where we spend most of our time, the new nature of the transactions occurring during these conversations and their seamless operational orchestration and provisioning seems to be a much more fertile ground to mine.

We have yet to see a Bank owning the “value orchestration” mantle. I believe that will change soon. How soon? Within less than 5 years is my bet. I am convinced this will happen because the Internet has fundamentally altered the way we can do business. Achieving near zero marginal cost of delivering any product or service will occur in every industry. I am convinced this has not happened yet because the financial services industry is unique, complex and heavily regulated.

If you think that only large banks can and will capitalize on the “value orchestration” opportunity you are wrong. In my view, although there will be few “value orchestration” or platform owners, there will be many smaller banks that federate and participate as platform partners. Further, if you think this platformification may lead to what I refer to the “dumb pipes” syndrome, you are wrong again. The age of dumb pipes is long gone, smart pipes is what you need to think through and digest – the variety of services at both end of the pipes and within the pipes themselves is underestimated by many.

A more appropriate concern is how will disruption and the resulting opportunity of “value orchestration” impact the direct relationship a Bank has with its customers? Will that relationship be maintained, shared or broken and to what extent? Could we see “Intel Inside” models emerging, capitalizing on implicit trust and technology prowess augmented by value orchestration without the necessary immediacy of a direct to consumer experience?

- What capabilities does a Bank need to realize these opportunities?
I will limit myself to a high level analysis.

First, let’s rifle through some important existing capabilities.

a) Regulated and Licensed: Although viewed as a constraint by some, I view these as assets. The trick will be to educate regulators as to the need for innovation. Different licenses will be needed, changes to existing licenses too. Different regulatory frameworks will need to be adopted.

b) Security, Cybersecurity, Authentication, Authorization, Identification: Banks invest heavily in these area. Again they shall need to add new technologies to the mix, which they are already in the process of doing. I would not be surprised to see a Bank acquiring a cybersecurity firm for example. Core competencies need to be brought in.

As for some of the new capabilities.

a) UX/UI: We are now used to sleek experiences and interfaces in our digital & data worlds. Nothing short of closing the gap and excelling is acceptable for a Bank going forward. I view this capability as core actually. I would advocate acquiring best of breed UX/UI practices, hiring leading designers. That capability, that talent needs to be acquired and treated well as it will be too time consuming to grow it internally.

b) Data Analytics: If your business is TRUST + DATA, you better be good at analyzing the latter to back up the former. Certain banks already have data science talent in house and are uniquely positioned to understand their own as well as their customers’ data. Still more needs to be done. I can see home-growing talent into specialized units, even spinning off these units to better grow them – at least one bank has done so I believe – or acquiring best of breed startups.

c) Artificial Intelligence: Arguably a wide field. There is an arms race going on. Google, Facebook, Amazon, Apple are snapping up talent in the US and I am sure European companies are doing the same in their respective countries. In a way AI and Data Analytics are intertwined, thusly AI is as important when one is dealing with data. Again, acquire!

d) Cutting edge Technology: One need not acquire all cutting edge technology capabilities (cloud, blockchain, quantum computing, AR, VR, IoT, API…), partnering will do for most, understanding, mastering and managing is a must though. To be fair, many banks have started learning and closing the gap here.

e) HR Skills: Hire, hire, hire from outside banking to acquire mindsets that live and breathe either data or complex networks… technologists, executives familiar with platform strategies, data experts, software entrepreneurs, p2p and/or network specialists, experts that understand and study the emerging properties of large systems (biologists, behavioral scientists…) . Basically, hire less bankers, more non-bankers.

If the above spurs your imagination, please share other opportunities you may find attractive, as well as capabilities I have not thought of.


The next Banking (R)evolution


The introduction of new technologies has facilitated new consumer and customer behaviors. These new behaviors have facilitated the adoption of new technologies. The resulting virtuous circle has ushered a period of rapid change which has profoundly change one industry after another. Industry incumbents have had to face a new reality where vertical integration, a fancy word for “owning the entire value chain” has turned into a liability. Indeed, the virtuous circle I mention has allowed new competitors to deliver value at one point of the value chain, without owning the entire value chain. Take the media and entertainment industries as an example. It used to be that “content was king” and “pipes were dumb”. Based on these heuristics Hollywood studios ruled over an entire value chain and were comfortable living in a world where the only thing they needed to do was to deliver their content to movie theaters. This is no longer true. Even though original content still rules, pipes are not dumb anymore. Pipes are actually smart, and that are built on top of platform strategies. Content is important, but so is how you create content, how you deliver it, with what and to whom, how you measure how it is delivered, plus the balkanization of communities of users make it eminently more difficult for a vertically integrated entertainment business to remain at the top of the food chain without profound changes. Witness the rise of Netflix, Amazon with their different value propositions around entertainment content and compare to how the main Hollywood studios are armed for the future.

The financial services industry in general, and the banking industry in particular are now faced with the same tectonic changes other industries have faced. For banks, this is an even more perilous exercise as most of them have never faced a breakdown of their value chain in the past and have enjoyed “near” monopoly in their geographies thanks to accommodating regulatory frameworks.

For simplicity’s sake, I break down a bank’s business into four layers (borrowing from a Boston Consulting Group framework):

  • Infrastructure: comprised of IT hardware (mainframes, cloud, hosted) and software (core banking system, CRM, client reporting, transaction/payment processing, analytics)
  • Products: comprised of three parts which are accounts, lending and the rest (payments, savings, investments, brokerage, advisory)
  • Interface: comprised of branches, web apps, mobile apps, customer service centers
  • Clients ecosystems: comprised of retail, SME and enterprise.

Yesterday’s bank owned each layer. Clients dutifully visited their branches or relationship managers to consume products created by their bank which were delivered by the infrastructure owned by the same bank.

To the extent that banks faced competition it was from another bank which also owned its entire vertical stack end to end, which was operating in the same geography. Oligarch banks ruled.

Today’s bank is under threat at each layer of its stack instead which makes for a much more complex competitive landscape.

First, clients spend more time somewhere else than with a bank. We all know the relative decline of branches. Not only are retail consumers not visiting their branches as much as they used to, but they are also increasingly spending time in completely different ecosystems than in the past; communities where a local bank relationship manager has little leverage if any. These ecosystems are called Facebook, Google, Amazon, WhatsApp, Snapchat, Instagram, Pinterest. (Even though such change is not as pronounced with SME and enterprise clients, there is also change with these segments.) Second clients are used to a different customer experience based on the service they are getting from these digital communities, thereby making bank web apps and mobile apps always play catch up. In other words, clients are moving banks, and bank customer interfaces are under threat. Third, products are under threat although we have to nuance this statement and look at lending separate from the rest. Let’s look at the rest first. Accounts are being loosened from the tight grip of Mr Banker – PSD2 in Europe, the open bank initiative in the UK will take care of that – allowing, under consent, third party access to account data and meta data. Payments is experiencing the highest level of competition given it has the lowest barrier to entry, either from fintech startups endogenous to the industry, new entrants exogenous to the industry (Amazon, Apple, Google, Facebook) or grown up startups (PayPal). Brokerage and Investments are prone to the same opening to multi-competition. This leaves us with lending which I believe should be analyzed completely differently than the rest because no one will ever be able to come up with a “zero marginal cost” lending product. Indeed, the cost of borrowing is comprised of the bank’s cost of borrowing and a margin to compensate for risk and provide adequate profit. That cost will never scale to zero or near zero. This, in my view is the main reason why lending will never experience an “Uber” moment where banks will be completely disintermediated – further, think of the unintended negative consequences of a massively large lender for example – whereas the main cost of the “rest” is that of delivery and marginal cost of delivery can and should be driven down to near zero. Fourth, infrastructure is where there has been to date the least disruption and competition, notably around core banking systems and CRM, even though blockchain technology holds the promise of much change in asset servicing.

To date the overwhelming number of competitors attacking the above layers have not been successful. Fintech startups focused on investments (robo advisory), brokerage, lending have not reached escape velocity and acquired meaningful market share to the detriment of banks. Some pundits believe it is because banks have much more defensible business models (regulation, licenses…). Although I do agree most startups have failed so far, I also know not to discount the entrepreneur/startup threat over the long run on the basis of a failed first wave. I am actually paranoid for banks as the overwhelming types of strategies banks have put in place to deal with change are in my opinion either inadequate or short term focused.

Indeed, banks have focused on revenue optimization strategies (pricing, cross selling, upselling, margins) or cost reduction strategies (layoffs, better hardware, better software) by applying concepts (digital banking, API banking, mobile banking, cognitive banking) on existing business models. To the exception of a few banks who recently started working on a platform strategy – which forces them to address the competition they are will face at each of the four layers – all other banks are still in a “vertical integration” paradigm. This will change – the market will force that change, some banks will adapt, other competitors will rise to the challenge.

I view all these bank moves as incremental evolutionary steps, good enough to compete another day, not good enough to reinvent banking drastically.  A digital bank – and there are many startup digital banks in the UK for example – is still vertically integrated, even though it holds the promise of being a “better” bank.

Incumbents will have to choose how they want to compete going forward. Below are some of the potential options available:

  • The “Better” Vertically Integrated Bank: Essentially more of the same, that is a bank that still owns the entire stack, will compete against a multitude of competitors, but will do so better armed marginally – digitally so, less siloed, better hardware, better software, less employees. Although I believe some will be successful at this strategy, I am afraid it will be a very risky one. No network effects to speak of, no ability to drive to meaningful zero marginal cost of delivery for all products such a bank would offer
  • The “Platform” & Vertically Integrated Bank: Same as above but with some type of platform strategy that will allow a bank to partner with third parties and share the value created by delivering better product and service to consumers. Probably less risky than the above and one many banks will want to deliver. Still a difficult proposition in a world where modularity will be more and more important.
  • The partially Vertically Integrated Bank: Whether traditional or platform driven, this Bank will drop a few non core activities, not enough to not be vertically integrated but enough to reach another level of rationalization. I expect tier 2 and tier 3 banks with limited resources to be the best candidates to follow this model and some shrewd tier 1 banks to make a hard turn towards this model. Very interesting as a platform.
  • The “Interface” Bank: No more vertical integration for this type of bank. To date we have only seen Interface examples (Simple is but one of the examples). The Interface specialists have suffered from a disconnect with the ecosystems where users gravitate and have not been successful to date. They key to success will lie with how an Interface bank partners with these digital ecosystems. My gut tells me AI powered virtual assistants may have a shot at being very successful Interface Banks. Strong potential for network effects and driving to zero marginal cost of delivery
  • The “Product” Bank: By far the most intriguing layer strategy. Product banks focused on innovating only on one particular product or a family of products (when was the last time the financial services industry came up with an innovative lending product tailored to someone’s cash flow patterns for example). A Product bank would partner with Interface providers and/or ecosystems of users for example. Not network effect to be expected for lending products – definitely for other products – but the benefits of innovation and differentiation can be powerful. I would even expand the horizon of what a product could be by including “data”. Data being the new hot asset class and data management as well as identity management being crucial in our digital age, why not see the emergence of data banks.
  • The ”Infrastructure” Bank: I see three separate models. First, the generalist “Bank as a Service” (BaaS) model that will deliver services to Product Banks, Interface Banks, startups, partially vertically integrated banks, fintech startups, enterprises. BaaS is the most promising bank model of the future as the focus is on the provisioning of products as a service, or of services. We are not dealing with lending here, we are dealing with delivering the building blocks to enable lending – the same applies to all other activities. As such there is a very high probability for this model to drive to near zero marginal cost of delivery. In this context, we can apply the “Uber” label. Second, the differentiated specialist BaaS. This model is particularly relevant for high value add services such as advanced data analytics, underwriting analytics, risk analytics. Remember one of the points I made at the beginning of this post: there are no dumb pipes anymore, only smart pipes. To date banks are arming themselves with the services startups specialized in data analytics can offer (CRM, fraud…) but it is conceivable the specialization will be so important going forward and the pipes so strategic that a “Bank” will provide this as a service going forward instead of a non-licensed startup. Third, the commoditized specialist BaaS. I expect some infrastructure services to become commoditized faster than others. Think hardware fine tuned for banking use cases or core banking systems. Think about an AWS offering but for banking. Much like there are core processors for specific activities (video, gaming, AI tomorrow), there may very well be core infrastructure providers for banks.

I have to make several additional comments to tie loose ends.

If the above vision comes to fruition and we do see a segmentation of banking, I fully expect the regulatory and licensing landscape to change. In other words, we will see a new regulatory approach where different types of banking licenses will be issued based on the business model and its implicit and explicit risks to the market and to clients/consumers. Just to give one example, an Interface Bank as an AI powered Virtual Assistant may have to meet certain licensing requirements around providing financial advice to its clients but may not need to comply with lending requirements. To be clear, some fintech startups competing or providing services at each layer level may not require the same type of banking licensing as the Banks that will operate at each layer level.

Further, competition at each layer level forces one to think platform strategy which results in either developing and implementing one’s own platform strategy or becoming one of the building blocks of someone else’s platform strategy. There is no escaping platform strategies.

Additionally, layer specialization, other than with Lending, and I repeat myself here, can deliver very strong network effects enabled buy near zero marginal cost of delivery. This I believe will be in and of itself a revolutionary paradigm for banking.

Finally, the bank that will successfully partner and integrate with ecosystems of users, regardless of the approach taken, will stand a higher chance of success than trying to create their own new communities or continue with existing ones. Like it or not, social networks are here to stay and will take on a greater importance in our lives going forward.

Trying to craft a roadmap for the above vision is tricky. We are in the early innings of platform strategies or API/marketplace strategies for banks and much remains to be done – no one has declared a BaaS for example. I venture that we shall see increased activity along these vectors in the next 5 years – the actions of Facebook, Google, Amazon, Apple, Alibaba (and Snapchat, Instagram, WhatsApp, WeChat….) will make that absolutely inevitable. Incumbents may also naturally gravitate towards a few of the six options I laid out above – either as a result of further divestitures, acquisitions or mergers – leaving space for new entrants (large tech companies, fintech startups). In other words, the industry is large enough to see various participants succeed and avoid a banks lose, new entrants wine scenario, or vice versa.

Last parting thought. I strongly believe the above also applies to the insurance industry – with the appropriate tweaks.