The End of Banking?
24 July 2018
Posted by: David Blair
The recent vote on fractional reserve banking in Switzerland made me reconsider last year’s article on the role of banks. Although the Swiss voted not to change their banking system, one of the effects would have been that the Swiss National Bank directly banks all citizens. This provides an interesting opportunity to consider the role of banks in the transfer and storage of value across the economy.
Voting on the unthinkable
The recent vote on fractional reserve banking in Switzerland questioned one of the fundamental building blocks of modern economies – the role of banks and the creation of money through fractional reserve banking.
Central banks can create money – so called “turning on the printing presses” – and often do so in the context of deficit spending by governments. But the most common money creation in modern economies is when banks lend to customers (as explained here) since banks need only keep a small percentage of deposits as regulatory reserves with the central bank. The central bank in turn influences the total money supply by adjusting reserve requirements and through interest rate policy.
Historically, banks preceded central banks. Central banks were created to bring more stability to existing banking systems (1694 in England and 1913 in USA) – so fractional reserve banking has evolved over the past couple of centuries. In the days before digitisation, it would have been impractical for the central bank to hold accounts for all citizens, so using commercial banks to store and transfer value was operationally natural.
But times and technology change. If we were to design arrangements for the storage and transfer of value today, it would seem strange to entrust it to a bunch of middle men who destroy value by introducing friction and risk into our arrangements. Since our store of value is fiat currency, and since our acceptance of money is an act of faith in government, it makes sense to store and transfer money directly across accounts with a government institution ie the central bank (as explained here).
Switzerland is famous for its banks, so it may surprise readers that the Swiss held a referendum to remove store and transfer of value from banks. In fact, Switzerland has a long history of financial experimentation. The most famous and long lasting is a complementary currency set up in 1934 to help SMEs through the credit limitations of the traditional banking system during the depression.
The WIR (a play on the German for “we” and “economic circle”) – ISO code CHW – has thrived over time and is still expanding today. The WIR Bank is a coop owned by members and acts as central bank for CHW so costs are low and CHW is basically interest free.
Interestingly, WIR Bank was set up not as some kind of rebellion against the elite but rather as a practical measure to increase sales for its (SME) members. In other words, WIR Bank was created to mitigate the failings of the conventional banking system which makes value transfer complicated and expensive and fails to fund the real economy in times of stress. The main source of that complexity and expense is the vast network of banks separating citizens from their central bank.
It is also interesting to note that WIR Bank was set up in 1934 and runs just fine without blockchain or other whiz bang technology. Keeping banks out of the process makes CHW simple and cheap to operate. On the other hand, conventional banks are wildly excited about applying blockchain and other technology to reduce the costs of complexity that their networks engender. They are rushing to invest in technology and consultants are gushing about 50-80% savings from blockchain, but this feels like applying band-aids to a gaping wound.
The CHW example shows that fixing the underlying problem – by removing the complexity of current banking arrangements – would be far more productive than any number of technological band-aid fixes. (It reminds me of the joy with which high tech cheque processing like photographing cheques with mobile phones are promoted, when any sane observer can see that the real solution is to eliminate cheques.)
To illustrate the problem with band-aids, consider the recent announcements that AliPay plans to cover 20 countries in Europe by end 2018. AliPay has been working on their European expansion since 2016, but it has not proceeded at internet speed because final settlement still depends on legacy banking – “Alipay has signed deals with over 100 banks and more than 40 digital wallet companies in Europe” according to one source. When we take bank intermediation out of the process and people hold accounts directly with central banks, this exercise will require signed deals with only 20 central banks.
Money has two primary functions – first the transfer of value (enabling us to buy goods and services) and second the storage of value (enabling us to time shift transactions).
When we think about value transfer, we want the simplest possible way to compensate the seller for the goods or services the buyer acquires. Cash was quite good for this purpose but no longer works online and over distance – plus, although cash seems free to consumers, it is in reality very expensive to process all those notes and coins. This is why more and more businesses are refusing to take cash. This is true not only for businesses but also for individuals – think about the hassle of keeping track of different currencies when travelling (outside of Europe). Cash is also risky because it is relatively easy to steal and counterfeit – think about pickpockets, hold ups, and massive amounts of fake notes.
The current primary alternative is cards and more recently their contactless variants. Now that security has been improved with chips and multi factor authentication, cards are reasonably fit for purpose from the consumer perspective. But for merchants, they carry significant costs and complexity, most of which is a function of the banking system (and its need to get paid) rather than anything intrinsic to the process of settlement (which is after all simply a few bytes debiting one account and crediting another).
Despite these weaknesses, current card arrangements can be used in a manner that gives us a glimpse of how settlement might look in a more rational world.
Uber is a good example. Although Uber is most famous for disrupting taxis and eventually the automobile industry, it has created notable service innovations as well. Settlement is a good example.
Although most of us take the convenience of an Uber ride for granted (when we are not complaining that the Uber took more than three minutes to arrive!), the settlement process is a wonderful user experience – primarily because there is no settlement process from the rider perspective. Uber has created a seamless settlement experience. The rider calls an Uber indicating from and to locations (the purchase order) and then takes the ride (the service delivery) and then just gets out of the car and wanders off. No wallet, no cards, no invoice, no approval (because the approval is embedded in the purchase order – the way it should be).
Approving payment is nonsense – we need to control the order stage because that is when we are legally bound to pay (assuming proper delivery). Refusing payment for a legally ordered and correctly delivered good or service will just land us in court with low to zero odds of winning.
Approval is embedded in the Uber process by first making a contractual agreement for services when the rider provides valid credit card details and second making a valid purchase order when specifying the ride start and end points. Delivery is evidenced by GPS tracking rider’s and driver’s phones.
The concept of seamless (or frictionless or invisible) payments is receiving wider attention now. BBVA’s cooperation with Sodexo for invisible canteen payments is a good example.
Because we have not had better ways of doing it, we tend to believe that payment is an essential aspect of control over transactions. As stated above, this is a dangerous delusion, and much better processes can be imagined and have already been implemented.
Low tech examples include using direct debit mandates to pay utility bills. Higher tech examples will include various kinds of smart contracts. Smart contracts themselves are merely digitised versions of old human powered arrangements like letters of credit.
When we drop our obsession with approving payments – either by opening our wallet or with multi factor authenticated digital banking – we can embrace seamless settlement.
To facilitate seamless settlement, we need higher simplicity and lower costs. The easiest path to simple and cheap is to reduce the transaction participants to a minimum. This will look like buyer and seller holding accounts with the central bank.
Settlement becomes a consequence of validated intentions between parties. The first is a valid order and the second is an appropriate delivery. Settlement is an automatic consequence of these two. This can easily be implemented with smart contracts, for example.
Seamless settlement is equally or more applicable for corporates. When the order is valid and the delivery is correct, payment should be automatic. Most payment factories already work along these lines with far less frauds and errors than the traditional model of finance directors signing cheques (and infinitely higher productivity).
Order validity and delivery correctness are not traditional banking strengths. When central banks can easily avail compute power and network capacity to handle citizens’ transactions directly, and other service providers can better provide transaction validation, there is no need for banks in this process. On the contrary, banks add unnecessary cost and risk to the process – imagine the millions of pages of bank regulation that could be consigned to the dustbin of history.
Store of value
Fiat currency as a store of value is basically the same wherever and however it is held. Some of the risks of fiat currency are intrinsic to its faith based nature – devaluation through inflation and markets, expropriation, etc. Other risks do depend on externalities – cash under the mattress has a high risk of theft, cash at banks is subject to credit risk, cash invested may be subject to securities risks, etc.
At the height of the global financial crisis, some large European corporates were grateful to be able to use their banking subsidiaries to place deposits with the central bank. Why should such safety be denied other citizens?
Having accounts directly with the central bank need not limit people’s freedom to invest where they please. Presumably the central bank would pay little or no interest (as many currently do on regulatory reserves) and people would be free to invest elsewhere and in securities – just as is the case today with bank accounts.
Because settlement will be frictionless and free, sweeping can be democratised allowing people to manage their money efficiently as they see fit. People so inclined could even transfer their salaries by auto sweep to bank accounts.
Banks can still provide investment products and funding to their customers, though one would expect the markets to become more competitive.
Banks currently have a role in the central bank management of the economy, but a recent BIS report makes clear that alternatives exist albeit requiring some changes in central bank practices.
KYC and AML
Banks are currently central to KYC and AML. This allows regulators to pass responsibility to banks when something goes wrong, but it can hardly be described as efficient. Customers are driven to distraction by bank compliance officers continuously re-inventing the KYC wheel to meet their shifting interpretations of regulators’ gnostic rulings and the latest billion dollar fine. This process is hugely expensive for the economy as a whole, and the results are very poor.
Although privacy advocates may object, it is basically necessary for all economic actors to be known to their governments. We have identity cards, company registrations, tax numbers, and so forth. Such authorities are in a much better position to validate identity than banks. Indeed, banks usually purvey validation from such authorities, again creating an extra layer of cost and complexity that adds no value whatsoever to the resilience of the system as a whole.
The grey and black markets will presumably always find ways to thrive, but using the banking system to police them is expensive and futile.
The basic job to be done is to debit the buyer’s account and credit the seller’s account when the buyer so intends. Banks add little value to this process and as explained above generally add risk and cost. That is why banks struggle to differentiate themselves in transaction banking.
Banks claim that they provide value add in for example their e-banking, but most users find fintech offerings much more impressive. Further, banks’ isolation from the real economy makes it hard to implement seamless or invisible settlement – so we are stuck with e-banking approval processes as the only way for buyers to signal their intention to pay.
Most bank value add comes in the form of funding, which post GFC regulation is making harder for them. Whatever value add bank software provides can work equally well – and with greater transparency – when the actual settlement is removed to people’s accounts at central banks.
Banks will still have a role in investment and funding products, derivatives, and value added services and software. The advantage of people holding accounts directly with central banks is that such products and services can be unbundled so that users can chose the solution sets that best meet their needs.
Some users may prefer to have a bundled bank solution looking the same as what we see now. Others may prefer to shop around for individual services that best meet their needs. At a macro level we see this kind of distinction between Chinese internet bundled platforms like WeChat compared to western internet specific solutions like Venmo (although with Facebook moving into payments we may be witnessing a convergence).
Using banks to intermediate the transfer and store of value for fiat currencies no longer makes sense. Technology now allows central banks to hold accounts for all citizens both legal and physical. Removing the basic transfer and store of value from banks will free them to build value added services without worrying about the basic and highly regulated mechanics, which are better handled at scale by central banks.