Banks of the 21st Century: a Manifesto (I)
Banks have been sitting at the centre of the economy during the last five hundred years, transforming the intangible human innate trust about the future into available financial resources, and hence turbocharging economic growth across the world. I have spent my entire 15-year career working in and studying financial institutions, first as a strategic consultant for bank executives, then as an investment banker serving their needs, and ultimately as an investor. Today, I have the fervent belief that after successfully dodging change for a couple of decades with the help of regulatory stiffness, what we call banking will undergo the most radical revolution of our lifetimes.
This post is intended to be the first of a series about what banks might become by the second part of the 21st century, drawing a path that starts from the traditional bank made of marble and white men in suits to the fully decentralised set of protocols that we will call “banks” in the next (my guess) twenty to thirty years.
Banks of the 21st Century: a Manifesto (Updated TOC):
- Part I: from marble to protocols, an introduction (this post)
- Part II: how to build your own decentralised bank (next post)
What Is a Bank
So what exactly is a bank? In its purest existential qualities a bank is simply a trustable and (hopefully) knowledgable group of people supported by a bit (a lot) of infrastructure. Nothing more.
Let me explain. A bank does two jobs:
- Borrows funds from several sources, compensating those sources with an interest rate (the cost of borrowing) — that’s where you need trust
- Lends those funds to others, asking as compensation another interest rate (the loan yield) — that’s where you need knowledge
The name of the game is to minimise the cost of borrowing while maximising the loan yield avoiding too many bad surprises. That’s where trust and knowledge (and infrastructure) come along.
The function of providing clients with a secure and convenient place to park cash has become a highly commoditised and requested service and also for these reasons a bank, in the current macroeconomic environment, doesn’t have to pay for the money it borrows. However, in order to achieve this badge of convenience banks need to invest a lot of money in infrastructure (to offer a quick and reliable way for its clients to deposit, withdraw, and manage their cash) and in the creation of trust. The last part is the most important; in order to get the license of trust from the relevant authorities, the ultimate owners of the bank need to put skin in the game in the form of appearances and equity capital, a cushion that will absorb those unexpected losses arising from lending money and running operations. Without equity capital a bank might find itself in the situation of not having enough money to return to its depositors. Remember that depositors do not get paid anything to place their money in a bank account, and expect to see their money back in full when they need it.
Forget the maturities matching, the proprietary investments, the ancillary products, the fancy augmenting services, the different types of capital, for now. A bank is just a group of people smart enough to lend money for a good risk-adjusted return, and with enough financial and reputational cushion to gain all the badges of approval required in order to access common people’s money almost for free. And all this supported by an expensive, and often archaic, infrastructure.
The Bank v1: Internalise, Internalise, Internalise
You might be thinking that my definition of bank is quite restrictive, but this could be influenced by the fact that in our times banks have traditionally extended their activities in fields that do not pertain to their core competence. This was due to two main reasons: their difficulty to effectively control those activities (remember, trust is a fundamental aspect of what a bank is, hence a bank should focus its resources to guarantee absolute trust at the expense of anything else), and an innate desire to extract rents by leveraging their market position. All those non-core activities could be listed along a spectrum that goes from Guarantee Trust to Extract Rent.
As a consequence, guessing the valuation of Bank v1 has become a true nightmare for honest specialists: they do too many things, offer little transparency even internally on what they are doing, their regulatory landscape is fluid and continuously more stringent, and surprises have tended to be heavily asymmetric towards the negative.
Let’s look at the most classic example: lending. Banks have traditionally fully internalised the process of lending funds to customers, from product design (the introduction of a mortgage, a credit card, a personal loan), pricing (how high should the interest rate be), underwriting the quality of the borrower (what is the likelihood that she is going to repay me in part or in full), recovering the outstanding loan (how can I get the money back). There is no particular reason a bank should internalise this full process; the internalisation happened because banks wanted to have full control over whom they were giving depositors’ money, but also because they wanted to extract as much profit as possible from this activity by doing it themselves and for their own clients. Although this has worked for a long time, the framework starts showing its age: banks struggle to design products that can solve client needs growing in complexity and diversity, and they also struggle to keep the same clients from accessing other providers due to higher transparency and easier access (read Open Banking). The Return on Equity of European banks was 5.4% in 2019 (last year of the pre-COVID era) for the EU 28 countries, down from 6.1% in 2018, vs. 10%+ ahead of the financial crisis. You put your money as guarantee to run banking operations, have to swallow a lot of uncertainty and volatility without really understanding what goes on under the hood, and that investment returns you 5.4% a year. Not great, especially when you compare it to 26.2% of the Eurostoxx Index for the same period.
The same reasoning applies to all the other activities we have listed above, with the additional spin that those at the bottom (i.e. closer to Extract Rent) are more exposed to outside competition since there is no real fundamental reason a bank should offer them internally.
In an extreme scenario, you might think that all those activities could be externalised. It is my conviction that they ultimately will, and that a bank will go back being what a bank is at the core: a trustable and knowledgable group of people taking other people’s money (securely) and lending those funds to a selected group of businesses and individuals, while themselves bearing the risk of making mistakes. This combination of wisdom, experience, and commingled incentives, will be the last bit to be automatised.
The Bank v2: Accept You’Re Doing a Bad Job
With time it became obvious that banks do a bad job in several of those non-core activities:
- First, it appeared clear they were too creative in trying to find ways to boost their returns by investing their own capital (originally intended merely as a guarantee) or by selling aggressively their products and services to their clients — and so it came the Dodd-Frank Act
- Then, it appeared obvious they were gloriously ineffective at creating and managing their IT infrastructure when compared to tech giants (that weren’t too keen to get involved in a heavily regulated business such as banking) — today, try to type “bank in a box” on Google
- And ultimately they emerged in no way fit to innovate their product offering in order to keep up with the diverse and always evolving needs of their individual and business clients — compare HSBC’s mobile banking experience with Revolut’s, try to get an eCommerce-focused loan like the ones offered by Wayflyer at a traditional bank, or compare the fees a bank charges you on international transfers with TransferWise’s (now just Wise)
Effectively, a bank today might already externalise most of those activities.
It seems, however, that those new players haven’t learned the lessons either, and each is trying to expand its footprint vertically to incorporate upstream or downstream what banks used to do. I believe this is more a product of inadequate pricing for their services (as part of the just-grow-and-dump-to-a-buyer recent mantra sponsored by Softbank and the likes) than of rational expansion.
A bank, however, has the opportunity today to go back to its core: sitting on top of a series of effective solution providers, it can re-align its incentives to those of the depositors while increasing profitability by a factor of 3–5x.
Back to the most classic example: lending. Instead of directly engaging clients by developing internally a series of (necessarily generic) products, a bank could select a specialised lender to partner with. Specialised lenders have better knowledge of a specific segment and are more flexible to develop the customer experience, and as a consequence they can extract higher risk-adjusted returns via more targeted offerings — and not by providing riskier loans. Specialised lenders, on the other hand, lack the deep pockets (i.e. the access to the deposits) that a bank has thanks to that trust badge, and they do not necessarily have the expertise and right incentives to keep the quality of their borrowers in check (hint: when a FinTech startup pitches you that banks are old and stupid and they are smart, and use machine learning, etc, walk away).
The financial advantages are obvious and ubiquitous, especially when the potential savings of a lighter infrastructure are taken into account. We are in the middle of this transformation process. In 2019, alternative lenders originated €6.7bn in the EU, almost 2x what was originated in 2018 and 4x what the volume of 2017. Banks v1 keep losing pieces of business while retaining the least profitable bits (namely current accounts). I look at the set up of my personal finances and reflect: my daily banking activities are managed through Revolut, I transfer money internationally via Wise, invest on Hargreaves Lansdown and Interactive Brokers, experiment alternative lending via Mintos, buy crypto on Coinbase Pro — when off-chain, more on DEXs and AMMs in another post, place my state-guaranteed deposits on Raisin at the best offeror, purchase insurance online on various websites. HSBC, my traditional long-standing banking relationship, simply holds my current account on which they technically lose money. Nothing against them.
The Bank v3: Decentralising Trust
Rule of thumb: if you believe (deep inside yourself) that any inexperienced apprentice could be easily trained to do your job sufficiently well, a machine (or more generally an algorithm) will get soon pretty close in performing the same task (I am quoting myself).
The scattered landscape described above works, at least for me as a user, because there is still somebody out there guaranteeing I can trust the different components of the system. Most importantly, I know that all the banks where I have parked liquidity benefit of few deposit insurance schemes, and that the custodians of my investment assets are heavily regulated by relevant authorities. With all the caveats in place, I trust that those central authorities can do a good job for me. Often this trust is overstated, to the point that several depositors of failed retail banks didn’t even know they were running the risk of losing their money beyond what guaranteed by the states if their banks went under as they sometimes did. In the services I use that are instead not regulated or guaranteed, I consciously run a risk and demand a commensurately higher return.
The explosion of blockchain technologies with their decentralised computing and consensus mechanisms have shaken some of the fundamentals of centralised systems we thought inescapable. It is not my intention to offer an introduction to Bitcoin or Ethereum, although I believe would be a good use of everybody’s time to get an understanding of the basics of Ethereum. Banking, being the industry that more than any other is founded on the pillars of unequivocal trust, took a while to be impacted until ultimately change knocked at the door.
The DeFi (“Decentralised Finance”) ecosystem today aims at replicating the different components of traditional finance by getting rid of as many overwhelmingly powerful central authorities as possible. Albeit in its infancy, it has done it rather successfully through one of the several Layer 1 blockchains out there (Layer 1 blockchains are some sort of virtual distributed computer operating systems — of which Ethereum is so far the most successful):
- I can deposit my collateral guarantee funds and raise additional liquidity through protocols offering reasonably robust confidence I won’t run away with the money ( e.g. MakerDAO)
- Invest this liquidity in other people’s projects (tokens) via Compound or AAVE
- Swap those assets with unknown but trustable counterparts on Uniswap
- And participate to the governance and maintenance of the whole system, all this without the existence of a centralised authority that often extends beyond what would be deemed efficient in microeconomic theory
We are at the infancy of the DeFi industry, with great ideas lost under the storm of speculation. We don’t really know the projects that will succeed and those that will succumb to fate, human flaws, or lack of adoption, but it is my conviction that DeFi is here to stay. Banks, especially those built from scratch, should see this as an opportunity to go back being (and being remunerated for) something closer to what they should be: a group of people leveraging their skills in securing cheap sources of funding while looking for promising high-returning projects where to deploy those same funds.
Next -> Part II: how to build your own decentralised bank (next post)