Europe follows the Swiss
Today, the European Central Bank (ECB) will have its regular monetary policy meeting and after Mario Draghi’s hints that the central bank will not only launch another set of Targeted Longer-Term Refinancing Operations (TLTRO) but potentially cut interest rates. The speculation is whether and by how much this will happen already today. To be fair, the economic growth in Germany and other Eurozone countries has been anaemic recently and inflation is barely above 1%, so the ECB is justified in thinking about additional monetary stimulus.
Yet, the problem is that the ECB’s discount rate is already at -0.4% and one has to wonder how much further the ECB can venture into negative territory without causing unintended consequences. If deposit rates become too negative and commercial banks pass the costs of negative rates for reserves and deposits at the ECB on to their customers, there comes a point when their customers prefer to withdraw their money from the commercial banks and rather hold it in cash in a vault instead of a bank account. After all, cash has an interest rate of 0% while the bank account may charge 1% or more per year.
Of course, if one could get rid of cash altogether, there would be no limit to negative interest rates and the ECB and other central banks could lower interest rates without any limits. In a country like Sweden or the UK where the currency in circulation is just 1.2% of GDP and 3.4% of GDP, respectively, the economy has almost gone cashless by now. The spread of all forms of electronic cash and payment systems leads to a continuous decline in currency in circulation that will probably continue unabated in coming years. But what are the Fed, the ECB or the Swiss National Bank supposed to do where currency in circulation is much more substantial?
One possibility would be to limit the amount of currency in circulation or abolish cash altogether. This seems politically impossible, since it would create a public uproar and undermine the legitimate use of cash for smaller transactions (there is also a privacy argument to be made in favour of cash but let’s ignore that for a moment).
Furthermore, the abolition of the largest denominations of paper money in India has created chaos all over the country since especially in the rural parts of the country where banks are rare, cash still plays an important role as a store of value. And even if one would somehow be able to abolish all cash or severely limit the amount of cash in circulation without side effects, it would be a boon for populist politicians who rail against financial repression by the government. Central banks who would opt for such methods would likely not remain politically independent for long.
Thus, one has to look for ways to extend negative interest rates on cash. And here the SNB has shown one possible way forward. In Switzerland, the monetary policy rate has been -0.75% for quite some time and the country has not experienced a run for cash or any other undesired side effects. This may, on the one hand be due to the relatively low rate of negative interest charged so far but it is also supported by auxiliary measures taken by the SNB in 2014.
With the introduction of negative interest rates, the SNB started to charge commercial banks a fee whenever they withdrew cash from the central bank. This fee was charged indirectly. Commercial banks who hold reserves with the SNB are charged negative interest rates only on excess reserves, while regular reserves are interest rate free. If a commercial bank withdraws cash from the central bank, the amount of regular reserves held at the central bank is reduced and converted into excess reserves. Thus, the more cash a commercial bank withdraws from the SNB, the more it has to pay in negative interest rates. This simple accounting trick has the advantage that commercial banks have to pay negative interest rates on cash holdings. Furthermore, because a certain amount of reserves is interest free, commercial banks do not have to pass on the negative interest rates to all customers. In general, only large customers like businesses and institutional investors pay a negative interest rate on their bank deposits while retail clients pay no interest. This approach by the SNB was so successful that the Bank of Japan followed suit with a similar technique in 2016.
The problem with this “fee-based” approach to negative interest rates on cash is that it has its limits because if interest rates become too low, institutional investors and business clients will start to complain and there might be approaches to circumvent the fee on cash. For example, one could create a physically backed cash ETF where banknotes are stored in a vault and the ETF trades at the local stock exchange. This way, large amounts of cash could be withdrawn from circulation. So far, however, no physically backed cash ETF has ever gained regulatory approval (one has to wonder why…).
Also, some people could just withdraw all their savings from a bank in cash and then terminate their relationship with a bank. This way, the bank would still have to pay the fees for cash to the central bank but cannot pass these costs on to end customers. In fact, the customer would have effectively created another currency, namely one that is exchangeable with digital money and normal cash at par but that is not subject to negative interest rates. In effect, this approach would then create a black market of “fee-based” cash vs. “clean cash”.
Finally, the profitability of banks may be at risk because a large source of income for most commercial banks is the net interest margin created by lending money to customers at long durations and borrowing at shorter duration. If retail customers don’t have to pay negative interest rates on their deposit accounts, the cost of capital for the bank is 0%. But if the loans handed to the customers have very low or even negative interest rates, the income generated by these loans is insufficient to cover the costs of the banks. This problem is particularly acute in Switzerland where government bond yields are negative for maturities of up to 30 years and mortgage rates for a 10-year fixed mortgage are currently around 1.1%. Thus, if the net interest margin is too low for too long, the profitability of banks may decline and their financial stability may be at risk.
It’s clear from these considerations that the approach of the SNB has its limits in real life, though we are most likely nowhere near the limit.
But what if one uses another accounting trick? Instead of charging commercial banks a fee for withdrawing cash from the central bank, the central bank could create an exchange rate between cash and digital currency. Assume, the central bank has a policy rate of -1% and wants to expand this to cash. At the beginning of the year the exchange rate between cash and digital currency is 1.00 so that commercial banks can withdraw cash at par. At the end of the year, the exchange rate has drifted lower and $1 in cash can only be exchanged for $0.99 in digital currency. After two years, the exchange rate would drop to 0.98 and so on. This accounting trick allows the central bank to charge negative interest rates at all cash holdings and it is impossible to “escape” the system by hording cash in a vault outside the system. After all, whenever the cash holdings are reintroduced into the system and converted back into digital currency, the exchange rate would make sure that the required negative interest is paid on that cash.
Ironically, it is a return to the system of variable discounts that bankers were using for many centuries before the introduction of central banks and central bank money as legal tender. Back in those days different municipalities and states issued their own currency. As usual, politicians tried to manipulate the currency and devalue it by reducing the amount of silver or gold in coins. Furthermore, transport costs from one place to another were high, so merchants who paid abroad with their home currencies had to accept a discount for their home currency that increased the farther away from home they were and the less trustworthy the currency was. Hence, every merchant in every major city had conversion tables for the most common foreign currencies, just like we have exchange rate conversions for foreign currencies today. The introduction of an exchange rate on cash vs. digital currency would simply expand this system.
Obviously, this would still cause a significant political backlash and the legal requirements for such an introduction of an exchange rate between cash and digital currency are unclear to me. But it seems that this could be a feasible way to open the path to unlimited negative interest rates in the future – for better or for worse.
Currency in circulation as share of GDP
Source: OECD, Federal Reserve, ECB, Bank of England, SNB, Riksbank.