While the reaction to the financial crisis has lead to many proposals to restrict, regulate, or otherwise reorganise, the way banks operate, hoping to prevent future financial crisis, the idea of getting rid of the banking sector altogether remains fringe. There seems to be a broad consensus, across the political spectrum, that banks have some useful functions that we can’t do without, and that the problem is to get them to accomplish these useful functions without causing unwanted damage.
But, are banks actually net useful? Do they produce more good than damage across a full banking/business cycle? I remain to be convinced. I think it’s relevant to ask the question: can a modern economy operate without a banking sector? I suspect the answer is yes, and to try to understand why, let’s see what banks do, and for each function of a modern bank, which currently existing, or imaginable, institution could take over that function.
- Payment system: this consists of operating cash deposit accounts, electronic fund transfers, payment cards (debit card, or the payment function of credit card less the borrowing), and handling cash (operating ATMs and cash deposits). This could be operated by pure payment processors, of which there are already examples, online (e.g. PayPal) or offline (prepaid debit card operators). Narrowly defined payment processors would be required to hold 100% of customer balances with the regulating sovereign, so that the government’s guarantee on deposits is explicit and enforced in the structure, and that no loss is possible outside of pure fraud cases. Customers are not allowed debit (negative) balances, and positive balance pay no, or less than the short term rate, or even cost some fee to maintain, because the processors get the short term rates, and deduce their operating costs as fees. People are not expected to keep their savings in these institutions, and if they do, they pay for it through the low returns.
- Brokerage: this allow people to buy and sell credit instruments (shares, bonds, etc) emitted by third parties in one convenient place. At its core, it’s a financial plumbing service business that requires little capital. Banks have broker department but brokers can operate perfectly well independently, and some of the largest brokers are not part of a bank (e.g. Interactive Brokers). There is no essential problem in requiring that brokers are not part of banks. Brokers still have a shadow banking function via (a) holding cash balances at banks (b) offering margin trading where they take credit risk. The former can be delegated to a payment process as above, or internally via sweep to money market or government bonds based instrument which nowadays can be realised with minimal trading costs. The margin trading function can also be made illegal, and replaced by equivalent products with third parties (see below).
- Savings accounts: deposit accounts at traditional banks with a government guarantee are in effect very similar to a government bond, they are a sort of derivative product: a synthetic treasury bill (to use US terminology). Like many derivatives, the systemic social utility of the product is probably negative. This could be dealt with by abolishing the government guarantee. Savers wanting deposit account-like products could buy government paper, or approximations thereof, without explicit guarantees but usually higher rates, directly through brokers.
- Credit intermediation: this is perhaps the function that people have the hardest time imagining how to do without. One of the core function banks operate is risk pooling and control: you lend to the bank and they asses the credit worthiness of borrowers and pool the risk among many borrowers for you. But there is already plenty of alternatives structures that also fulfil that role: large companies most often raise money directly from investors via share and corporate bond issuance. The technology is now there to expand those markets to smaller businesses. Peer-to-peer lending platform already do this for credit to individual or single person businesses. Funds of corporate credit or private equity fill a similar role for mid-size companies. There are also so called non-bank lenders who raise money from the market (shares, bonds) and then lend to individuals or small or medium size businesses, without being involved in guaranteed deposits. So the role of credit intermediation that banks currently do could be entirely taken over by a combination of direct markets, aggregation platforms, and specialist intermediary businesses. All these businesses are funded via equity and corporate credit, whose failure is usually non-systemic and where the creditors know they are taking risk.
- Issuance placement: this is old school investment banking, where banks play the role of an intermediary to have large borrowers meet with investors and solve the somewhat tricky issue of pricing bulk issues of credit instruments. I’m not sure whether non-bank companies are involved in this business, but I don’t see either that it needs to be part of banks at all. This is not capital intensive (the intermediary is paid via a fee) and there’s little systemic advantage to have that role entangled with banking. Indeed, conflicts of interests are rife.
- Market making and derivative issuance: banks trading department often have inventory of financial instruments (aka trading positions) via which they provide liquidity to non-bank parties who may otherwise not find someone to match their trade. Several firms operate as independent market makers successfully. The barriers to entry are low on electronic markets with central clearing. There’s no reason to think the market would not fill that niche should banks be banned from proprietary trading of all forms, including market making (the current attempts at trying to limit proprietary trading while allowing market making are inadequate, because systematic rules to distinguish the two activities are virtually impossible to devise).
- Monetary policy transmission: in the payment processor model above, the processors are not allowed to borrow from the central bank, and thus only pre-existing money circulates in the system. There may be a shortage, that is today implemented by central banks via the operation of bank reserves and open market monetary operations, where the central bank controls the quantity of core money to match the financial activity without triggering too much inflation or deflation. This is the one function that is today almost exclusively done by banks, but could they be circumvented? The Brazilian central bank, cleverly, operates some of its policy implementation via open trading of interest rates derivatives. The Swiss central bank is today implementing policy via the cap on the EUR/CHF exchange rate, which they put into effect by trading directly on the foreign exchange market. I don’t see any reason why the entire monetary policy couldn’t be thus implemented. There’s a wide variety of instruments that are broad enough to prevent becoming an active investor, which is indeed not a desirable function of central bankers. Intervention on the forex market; interest, equity indices, maybe property index derivatives would achieve the same function as reserve management, and probably with greater precision to boot.
So, if we ban banks for operating the functions above where narrow institutions can replace them, are we left with anything? I’d like to be corrected but I can’t think of a useful function that banks do better than specialists institutions.
In conclusion, how should we ensure that banks do not cause the next financial crisis? by making banking an illegal business to be in, after allowing a long period for winding down existing banks in an orderly manner.
The invisible hand will swiftly replace bankers with less toxic operators. It will not prevent all types of future systemic crisis — you can still get unwieldy financial entanglements without banks — but it should make it easier to understand and regulate the system and avoid the worst of the unwanted socialisation of losses that banks can inflict of society via their privileged role as de-facto issuers of synthetic sovereign debt.