I have tried to come up with a sexy title – at least I think it’s sexy: “Monetary Policy below the Zero Lower Bound,” and I intend to talk about the Swiss experience of conducting monetary policy with negative interest rates.
In the case of Switzerland, the international financial crisis of 2008 manifested itself mainly through a significant exchange rate appreciation. The Swiss franc has traditionally been considered a safe haven currency. Whenever there is a crisis, or when risk is perceived as high, the Swiss franc tends to appreciate. In Figure 1, you can see the real effective exchange rate of the Swiss franc and how the Swiss franc appreciated during the crisis. This appreciation was significant, and it clearly accelerated in the summer of 2011, which led the Swiss National Bank to introduce a temporary exchange rate floor against the euro. The exchange rate floor was in place until January 2015, when we decided to lift it. The Swiss franc experienced then another sizeable appreciation. Since then, the real effective exchange rate depreciated somewhat, partly due to a nominal depreciation, partly due to lower inflation in Switzerland, but overall the Swiss franc remains considerably stronger than before the crisis.
Figure 2 shows the bilateral nominal exchange rates of the Swiss franc against the dollar and the euro. Here too, you can see how the Swiss franc appreciated against both currencies during the crisis, with the dramatic acceleration of the appreciation during the summer of 2011.
On September 6, 2011, when we announced that the Swiss National Bank would no longer tolerate a EUR/CHF exchange rate below CHF 1.20, the Swiss franc immediately depreciated above CHF 1.20 per euro, and it stayed there for the next three years. Why did we lift the floor in early 2015? One of the questions we had to deal with when we introduced the exchange rate floor was whether we should define such a floor for an effective exchange rate, that is, against a currency basket, or for a bilateral exchange rate. While the first concept makes more sense from a purely economic point of view, the second concept is much easier to communicate and implement. And as long as the Swiss franc was appreciating against all currencies, there was no difference, and, therefore we opted for a floor against the euro. The situation changed in the second half of 2014, when the US dollar and the euro started to move into different directions. As it became clear that the Fed would start to exit from its expansionary monetary policy, and the ECB, on the contrary, prepared a further loosening of its monetary policy stance, a bilateral exchange rate floor against the euro was no longer sensible. Toward the end of 2014, we announced the introduction of negative interest rates, and we introduced them as we lifted the exchange rate floor.
Apart from such a temporary exchange rate floor, which the Swiss National Bank had already once successfully introduced in 1978, we also looked at other options, including the possibility of capital controls. However, we came to the conclusion that such controls would be more difficult to design and implement, if you want to minimize the negative spillovers to the domestic economy. We did make preparations for capital controls, but we considered them more of a last resort, if other measures proved insufficient or if the situation called for more drastic measures.
Regarding negative interest rates, as noted before, we introduced them in early 2015. However, we looked at the possibility of negative interest rates already in 2011, at a time when the concept of negative interest rates was still considered awkward, even within the Swiss National Bank. In fact, I headed the working group preparing the concept of negative interest rates, and one of the first arguments I had was whether there was such a thing as “negative interest rate” and whether we could thus use that term. I think some people remained unconvinced until the ECB finally introduced negative interest rates and used that term. But beyond terminology, there was the more serious assertion that negative interest rates were not merely a continuation of values on a scale that goes below zero. Rather, it was argued that lowering interest rates below zero was like cooling water below zero degrees Celsius: the physical properties would change, water would freeze, and you would suddenly be in a very different physical environment. But as Ulrich already noted: we learned that this is not the case. Negative interest rates are a continuation of values on a scale that goes below zero, at least within the range of negative interest rates that we have experienced so far. Nothing unexpected happened; markets and the transmission mechanism continue to work.
The first central bank to actually introduce binding negative interest rates was the Danish National Bank. They did it in 2012, and it provided a welcome case study for us. We were in close contact with the Danish National Bank during that time, and their experience confirmed that negative interest rates could be used to counter appreciation pressure on the exchange rate.
In Figure 3, you can see the policy rates of different central banks. As you can see, the Swiss National Bank lowered its policy rate quickly in the autumn of 2008, like the Fed and the ECB, and we reached zero in 2011. Given that in our case the policy rate is the three-month Swiss franc Libor, and not an overnight rate, the Swiss monetary policy stance is even somewhat more expansionary at a given policy rate relative to the Fed and the ECB. The ECB did not go all the way to zero at first, but it started to lower interest rates again in 2012, and finally it went negative in 2014. We waited longer, but lowered interest rates further, all the way to −0.75 percent.
By introducing negative interest rates, the Danish National Bank, the ECB, and the Swiss National Bank confirmed that the effective lower bound for monetary policy is below zero. However, this does not mean that there is no lower bound. The reason for such an effective lower bound is twofold, as pointed out by Ulrich: the existence of paper currency and bank profitability and, therefore, financial stability more broadly.
The existence of paper currency entails a risk of a run for paper currency because banknotes guarantee a zero nominal return, giving people an alternative to bank accounts or other financial assets with a negative nominal interest rate. At the same time, however, holding large amounts of paper currency is costly and inconvenient. At the time, when we lowered interest rates to −0.75 percent, the conventional wisdom was that it would not be possible to lower interest rates below 0.50 percent without risking a run on paper currency. So, it took a little courage to do it – and analysis. In particular, we figured out that insurance costs for paper currency are not easily scalable, because insurance companies put limits on how much cash they are ready to insure per unit, be it a safe or a vehicle. You cannot take a truck or a vault and fill it to the top with banknotes and get full insurance. The insurance company will only insure a certain amount per truck or vault. As a consequence, the costs for storage and insurance are higher than usually assumed. In addition, it is logistically and administratively cumbersome to store and handle large amounts of cash, as you have to deal with anti-money laundering regulation and so on.
As to the concern about bank profitability, we took that into account when we designed our concept of negative interest rates with relatively large interest exemptions. In our case, as in the case of Denmark, negative interest rates were implemented to counter the appreciation pressure on the exchange rate. We figured out that what mattered for the transmission of the negative interest rates to money market was the interest rate, that is, the marginal rate that investors or banks have to pay for holding an additional Swiss franc, and not the total amount of interest that the banks have to pay on their sight deposits with the Swiss National Bank. Accordingly, we set a relatively high exemption threshold, which limited the financial burden for the banks while, at the same time, it made Swiss franc hoarding less attractive relative to other currencies. In order to determine the threshold for each bank, we took the minimum reserve requirement, and multiplied it by 20. This was the most straight-forward way legally. Needless to say that not all banks were equally happy with this approach, especially banks with low minimum reserve requirements that had accumulated significant sight deposits with the Swiss National Bank during crisis.
Another feature of our negative interest rate concept was that we introduced a monthly adjustment for cash holdings. Banks have to report their cash holdings on a monthly basis as part of their minimum reserve requirement. We announced that if we registered an increase in cash holdings, adjusted for seasonal factors, we would reduce the amount of their sight deposits that was exempt from negative interest rates by the same amount. As a result, banks were indifferent between holding paper currency and holding sight deposits with a negative interest rate. This provided another disincentive for banks to exchange their bank deposits for paper currency, apart from the other costs of holding paper currency.
Now, let us turn to the transmission mechanism. Like the ECB, we saw an immediate transmission of our negative interest rate to the money market. Particularly, the three-month Swiss franc Libor moved in line with our deposit rate to −0.75 percent. As you can see in Figure 4, there was also a clear transmission to capital markets. Both government bond yields and corporate bond yields came down. The only interest rates that did not follow, and even increased somewhat after the introduction of negative interest rates, were mortgage rates. Our interpretation of this is that the banks – being reluctant to move deposit rates of their retail customers into negative territory – increased mortgage rates to protect their interest rate margins.
That banks did not pass on the negative interest rates to their retail depositors can be seen in Figure 5. We know that banks passed on negative interest rates to institutional investors, as well as to high net worth individuals, but always with a large exemption threshold. It seems that the banks were concerned that passing on negative interest rates to retail depositors could lead to a run on cash, given that it is less costly to store smaller amounts of cash. But a lot of smaller amounts of cash can still be a problem for banks.
Interestingly, we had one bank in Switzerland that passed negative interest rates on to their retail depositors. But that was a special case. It was the Alternative Bank Switzerland. It is a bank specialized in ethical and sustainable finance. According to them, they gained more new customers than they lost when they introduced negative interest rates for their retail depositors. Of course, it was all over the news when they announced the introduction of negative interest rates for retail depositors, and it seems that this extra publicity helped them to attract new customers who did not know about this bank before and who apparently valued certain ethical principles over profit maximization. This case can, thus, not be generalized.
So, what did banks do if they could not pass negative interest rates on to their retail depositors? As noted before, it looks like they protected their interest rate margin by not lowering or even increasing mortgage rates. Figure 6 shows the 10-year mortgage rate, which used to move parallel to the equivalent Libor swap rate, which gave the banks a constant margin. With the introduction of negative interest rates, that margin increased, as the mortgage rate was not lowered in line with the Libor swap rate. The fact that the mortgage rates did not fall further was not at all problematic in Switzerland, given that we were worried about the real estate market overheating. As I noted at the beginning, the problem in Switzerland was really the exchange rate and not a lack of domestic demand. So, the objective of negative interest rates in Switzerland was not the stimulation of the credit channel, but rather to make the exchange rate less attractive to safe haven flows.
What happened to the demand for paper currency when negative interest rates were introduced? Did we see increased demand? You may say that this is a special concern in Switzerland, given that we have a 1,000-franc note, a banknote with a very high denomination. As you can see in Figure 7, the growth rate for 1,000-franc banknotes in circulation increased with the introduction of negative interest rates at the beginning of 2015. However, this increase was not unprecedented and therefore not unusual. We already saw a similar increase earlier in the crisis, namely in 2011, at the same time we saw an acceleration of the Swiss franc’s appreciation. Moreover, the increase in demand for banknotes peaked in late 2015, at a lower level than in 2012, and it slowed down again afterward.
Switzerland’s nominal interest rates are usually lower than nominal interest rates in the euro area, even lower than nominal interest rates in Germany. So, when all interest rates came down during the crisis, Switzerland’s interest rates reached zero much earlier. Before we introduced negative interest rates, one of the problems that contributed to the appreciation pressure on the Swiss franc was that the traditional interest rate differential with the euro zone disappeared. As you can see in Figure 8, the introduction of negative interest rates at the beginning of 2015 helped to partially restore the historic interest rate differential with Germany. Since then, the interest rate differential decreased again, which is one of the reasons for the continued appreciation pressure on the Swiss franc.
In terms of inflation, the ultimate objective of our monetary policy, you can see in Figure 9 that inflation in Switzerland was negative for several years since 2011. An expansionary monetary policy was therefore clearly necessary. Inflation has noticeably increased at the beginning of 2017, back into positive territory, partly aided by the increase in oil prices.
In Figure 10, you see our latest conditional inflation forecast. It is conditional in the sense that the policy rate is fixed at −0.75 percent throughout the entire period. Even at that interest rate level, there is no inflationary pressure visible. Our models suggest that price stability will be preserved even without a monetary policy tightening over the medium term.
The question going forward, therefore, is whether we have to get used to an environment with much lower nominal interest rates? Will negative interest rates stay with us or become a recurring feature over the coming years? If this is the case, we should be thinking about how to deal with the lower effective bound on nominal interest rates. If monetary policy below the zero lower bound is to become a regular occurrence, we need to be able to lower interest rates much further into negative territory than we have been able to do so far. This is something we should be working on now.
