Selected Issues

Abstract

Selected Issues

Banking Sector Responses to the Negative Interest Rate Policy1

1. The Swiss National Bank (SNB) set the policy interest rate at minus 0.75 percent in January 2015. The SNB first announced a negative rate of minus 0.25 percent in December 2014, which was lowered to minus 0.75 percent with the removal of the exchange rate floor.2 The negative rate applies to banks’ sight deposits at the SNB, subject to bank-specific exemption thresholds of 20 times the minimum required level of reserves (fixed as of November 2014).3 Balances below this amount carry a zero interest rate. To discourage banks from accumulating cash as way to avoid the negative rate, the exemption threshold is reduced (raised) by the cumulative increase (decrease) in a bank’s cash withdrawals.4 The interest rate and method for computing minimum exemption thresholds are unchanged since January 2015.

2. The exemption threshold makes the effective rate considerably less negative than the marginal rate, helping to insulate banks’ profits. Initially, sight deposits at the SNB jumped in response to the removal of the exchange rate floor, and thereafter, increased steadily until mid-2017. Staff estimates that sight deposits above banks’ exemption thresholds, and hence subject to the negative deposit rate, currently amount to about CHF180 billion (39 percent of total sight deposits and 27 percent of GDP).5 As a result, the effective rate on banks’ deposits at the SNB is about −0.3 percent. Anecdotal evidence suggests that most banks have exhausted their exemption thresholds, including through interbank transactions.6,7

uA03fig01

Switzerland: Money Market Rates

(Percent)

Citation: IMF Staff Country Reports 2018, 174; 10.5089/9781484362174.002.A003

Source: Swiss National Bank.1/ SARON (Swiss Average Rate Overnight) is an overnight average rate referencing the Swiss Franc interbank repo market.
uA03fig02

Switzerland: Banks’ Sight Deposits at SNB

(Billions of CHF)

Citation: IMF Staff Country Reports 2018, 174; 10.5089/9781484362174.002.A003

Sources: Swiss National Bank; and IMF staff calculations.

3. The negative interest rate policy (NIRP) was transmitted quickly to the short-end of the interbank market, although global financial conditions had already pushed down longer rates. The 3-month LIBOR converged quickly to the policy interest rate. Arbitrage opportunities encouraged banks to reallocate liquidity via the interbank market from those whose sight deposits were above their exemption threshold to those whose sight deposits were below the threshold.8 Short-term government bond yields were modestly negative even before the adoption of NIRP, but decreased further thereafter. Interest rates on longer-term instruments also decreased, although they had been on a decreasing trend prior to the adoption of NIRP, with the entire yield curve for sovereign bonds (through to 50 years) falling into negative territory on at least one occasion.

4. Banks kept most retail deposit interest rates floored at zero, while those on institutional depositors turned negative. The zero-interest-rate floor for retail customers likely reflects banks’ reliance on these deposits to fund their loan books, as well as retail depositors’ potentially higher elasticity of substitution between holding deposits and holding cash (relative to institutional investors). While no official data is available, deposit rates for institutional investors and, more recently, for large retail depositors (reportedly above CHF1 million) are thought to have declined close to the policy rate. Interest rates on new 1-year term deposits exceeding CHF100,000 turned modestly negative immediately after NIRP adoption, and have edged lower since then (but only to minus 0.25 percent). Rates on current accounts remain anchored at zero, while savings deposits maintain marginally positive rates. Overall, only 5 percent of franc-denominated deposits of domestic clients were subject to negative rates at end-2016.9

uA03fig03

Switzerland: Customer Deposits Denominated in CHF 1/

(Billions of CHF)

Citation: IMF Staff Country Reports 2018, 174; 10.5089/9781484362174.002.A003

Sources: Swiss National Bank; and IMF staff estimations.1/ Assumes households, corporates, and the public sector hold only CHF-denominated deposits and that foreigners hold only foreign-currency-denominated deposits.
uA03fig04

Switzerland: Swiss Average Rate Overnight (SARON) and Mean Interest Rates for New Transactions

(Percent)

Citation: IMF Staff Country Reports 2018, 174; 10.5089/9781484362174.002.A003

Source: Swiss National Bank.
uA03fig05

Switzerland: Bank Deposits in CHF by Domestic Customers

(Billions of CHF)

Citation: IMF Staff Country Reports 2018, 174; 10.5089/9781484362174.002.A003

Sources: Swiss National Bank; and IMF staff estimation.

5. To preserve interest margins when retail deposit rates are floored at zero, banks initially raised their mortgage lending rates. The interest rate on a 10-year fixed rate mortgage increased by 50 bps by mid-2015, although rates on shorter-duration loans increased by considerably less.10 After that, mortgage interest rates declined, partly in response to competition from nonbank mortgage lenders with funding costs close to the policy rate. Rates have seen an uptick since late-2017, broadly in line with LIBOR swap rates typically used to price fixed-term mortgages in Switzerland (Cecchin, 2011). Nonetheless, interest rates on new fixed-rate mortgages have risen by much less than market interest rates since mid-2016, highlighting the pressure faced by banks in the mortgage market.

uA03fig06

Switzerland: Interest Rates for New Mortgage Transactions

(Percent)

Citation: IMF Staff Country Reports 2018, 174; 10.5089/9781484362174.002.A003

Source: Swiss National Bank.

6. As the foregoing discussion suggests, banks are—in effect—segmenting their balance sheets into above-zero and below-zero books, corresponding respectively, to their retail and institutional customers. While money is fungible, banks’ practices suggest that pricing of mortgage lending is linked to interest rates on retail deposits, which remain around zero on average. For institutional customers and interbank transactions, deposit pricing is anchored to the cost of placing these funds at the SNB (minus 0.75 percent). Moreover, banks are reported to have become somewhat less receptive to accepting new deposits, especially from institutional clients.

7. Demand for cash increased only temporarily, and by a relatively modest amount, around the time that NIRP was introduced. The increase was concentrated in the CHF1000 note, but was more modest than for earlier demand spikes that occurred at the time of the global financial crisis (in 2008 and in 2012). More recently, the growth of currency in circulation has slowed, consistent with widening use of cashless-payment systems. Reflecting this longer-term trend, the velocity of money (nominal GDP relative to cash in circulation) has been decreasing since the turn of the decade, with a small, temporary increase in early 2015. The very limited cash hoarding likely reflects banks’ decisions not to lower retail deposit rates below zero and that cash withdrawals by banks would be factored into the calculation of their exemption threshold. However, sustained negative interest rates for an extended period could make cash hoarding more attractive.11

uA03fig07

Switzerland: Cash in Circulation

(Year-on-year growth, percent)

Citation: IMF Staff Country Reports 2018, 174; 10.5089/9781484362174.002.A003

Source: Swiss National Bank.
uA03fig08

Switzerland: Velocity of Money 1/

Citation: IMF Staff Country Reports 2018, 174; 10.5089/9781484362174.002.A003

Source: Swiss National Bank; IMF staff estimation.1/ Estimated as annual GDP divided by seasonally-adjusted cash circulation.

8. Introduction of NIRP did not raise the pace of mortgage lending in the aggregate, or accelerate the narrowing of banks’ interest margins. Mortgage lending by continued to expand—but at a slowing pace—for several years following the introduction of a series of macroprudential measures in 2012–14 and the temporary increase in mortgage interest rates. However, mortgage volume at domestically-focused banks has been growing significantly faster than at big banks since the onset of the global financial crisis in 2007.12 Aggregate mortgage lending has quickened since late-2017. The spread between rates on mortgage loans and customer deposits had been on a decreasing trend that pre-dates the introduction of NIRP, decreasing from 2.0 percentage points in 2010 to 1.5 percentage points in 2016.

uA03fig09

Switzerland: Average Interest Rates on Selected Balance Sheet Items

(Percent)

Citation: IMF Staff Country Reports 2018, 174; 10.5089/9781484362174.002.A003

Source: Swiss National Bank.
uA03fig10

Switzerland: Balance Sheet Items of Domestically Focused Banks 1/

(Billions of CHF)

Citation: IMF Staff Country Reports 2018, 174; 10.5089/9781484362174.002.A003

Source: Swiss National Bank.1/ Data break in November 2015 due to the new/accounting fules for banks.
uA03fig11

Switzerland: Mortgage Growth

(Year-on-year percentage change)

Citation: IMF Staff Country Reports 2018, 174; 10.5089/9781484362174.002.A003

Source: Swiss National Bank.

9. Sustaining the pace of mortgage loan growth has caused domestically-focused banks to assume higher credit risk and increased interest-rate risk. New lending by domestically-focused banks has been provided to borrowers with higher affordability risk. Based on survey data, the SNB finds that loan-to-income ratios on newly-issued mortgages reached a new peak in 2016 (SNB (2017a)). Also, about one third of new mortgage loans are reportedly for residential investment properties, even though vacancy rates are rising and rents are decreasing. In addition, the SNB reports that interest rate risk from maturity transformation by domestically-focused banks increased in 2015 from an already-high level and has remained elevated since then.

10. Consistent with the aggregate-level picture, individual bank data confirm that banks with larger exposure to NIRP adjusted their balance sheets more actively. Basten and Mariathasan (2018) find that domestically-owned retail banks13 with larger amounts of SNB sight deposits in excess of their exemption thresholds worked actively to reduce these balances by shifting into other assets (loans and financial instruments) and also by scaling-back their liabilities. Preference was given to reducing longer-term bond financing, rather than deposits, even though the former were cheaper, as banks sought to preserve their long-term relationships with retail clients, but were nonetheless less-receptive to accepting new deposits. As a result, balance sheets of these banks tended to contract.

11. During the early years of NIRP, domestically-focused banks managed to preserve, and even marginally increase, their return on assets.14,15 This result reflects a combination of: (i) continued expansion of mortgage lending, which more than compensated for the narrowing interest rate margin;16 (ii) the limited burden of NIRP due to large exemption thresholds; and (iii) higher rates on newly-issued mortgage loans in early 2015, but which can be expected to support interest income throughout the duration of these loans. Further support was provided by lower credit losses, value adjustments and provisions, and greater cost efficiency (see SNB (2017a)). However, ongoing pressure on lending rates from nonbank credit institutions providing mortgage loans, gradual dissipation of the effect of temporarily-higher mortgage lending rates, and the already very-high credit penetration of the population suggest that it may be difficult over the longer term to sustain profits at their 2015–16 levels.17

uA03fig12

Switzerland: Return on Assets, 2001–16

(Percentage)

Citation: IMF Staff Country Reports 2018, 174; 10.5089/9781484362174.002.A003

Source: Swiss National Bank.
uA03fig13

Switzerland: Income of Domestically-Oriented Banks

(Billions of CHF)

Citation: IMF Staff Country Reports 2018, 174; 10.5089/9781484362174.002.A003

Source: Swiss National Bank.
uA03fig14

Switzerland: Net Interest Income as a Ratio to Total Assets, Domestically-Focused Banks

(Percent)

Citation: IMF Staff Country Reports 2018, 174; 10.5089/9781484362174.002.A003

Source: Swiss National Bank.

References

  • Basten, Christoph, and Mike Mariathasan, 2018, “How Banks Respond to Negative Interest Rates: Evidence from the Swiss Exemption Threshold,” CESifo Working Paper 6901 (CESifo Group: Munich).

    • Search Google Scholar
    • Export Citation
  • Bech, Morten Linnemann, and Aytek Malkhozov, 2016, “How Have Central Banks Implemented Negative Policy Rates?BIS Quarterly Review, March, pp. 3144 (Bank for International Settlement: Basel).

    • Search Google Scholar
    • Export Citation
  • Cecchin, Iva, 2011, “Mortgage Rate Pass-Through in Switzerland”, SNB Working Paper 2011–8.

  • Eggertsson, Gauti B., Ragnar E. Juelsrud, and Ella Getz Wold (2017), “Are Negative Nominal Interest Rates Expansionary?NBER Working Paper 24039 (National Bureau of Economic Research: Cambridge).

    • Search Google Scholar
    • Export Citation
  • Nucera, Federico, Andre Lucas, Julia Schaumburg, and Bernd Schwaab, 2017, “Do negative interest rates make banks less safe?ECB Working Paper Series, No. 2098 (European Central Bank: Frankfurt).

    • Search Google Scholar
    • Export Citation
  • Swiss Bankers Association, 2016, Banking Barometer, September.

  • Swiss National Bank, 2017a, Financial Stability Report.

  • Swiss National Bank, 2017b, Banks in Switzerland 2016.

  • Turk, Rima A.,Negative Interest Rates: How Big a Challenge for Large Danish and Swedish Banks?IMF Working Paper, No. 16/198 (International Monetary Fund: Washington).

    • Search Google Scholar
    • Export Citation
1

Prepared by Seung Mo Choi (EUR). The paper has benefited from comments and suggestions from Toni Beutler and Thomas Nitschka at the Swiss National Bank.

2

The target range for 3-month Swiss-franc LIBOR was set between −0.25 and −1.25 percent in January 2015.

3

Foreign banks and some non-bank holders of SNB accounts have different, but constant, exemption thresholds.

4

The exemption threshold is computed as “the minimum reserve requirement of the reporting period of October-November 2014 times 20” minus/plus “an increase/decrease in cash holding from the reporting period of October-November 2014 to the current reporting period.”

5

This estimate is for the banks which are subject to the reserve requirements (i.e., blue bars in the text chart) and is obtained from two separate approaches: (i) information on minimum required reserves as of November 2014 and subsequent changes in banks’ cash holdings; and (ii) the SNB’s 2017 income statement which indicates a “net result from Swiss franc positions” which is assumed to reflect mainly interest collected on sight deposits. For all banks, according to the Swiss Banking Association (SBA), CHF234 billion of sight deposits at the SNB were subject to negative interest rates at the end of 2016.

6

According to the Swiss Banking Association (SBA), a market for liquidity developed between banks whereby those with liquidity exceeding their exemption threshold (e.g., cantonal banks), transferred their excess liquidity for a fee to banks that were below their exemption threshold. Private banks are reported to have been most impacted by the NIRP primarily because they are subject to lower minimum statutory reserve requirements.

7

The SNB (2017a) reported that as of 2016, “… Most of these [domestically-focused] banks’ sight deposits remain below or at the exemption threshold, in spite of the steady increase observed since the introduction of negative interest. …” (footnote 26).

9

Nonresidents’ deposits are only about 6 percent of total CHF-denominated deposits.

10

The limited pass-through to lending rates was also observed in other countries that adopted a NIRP, such as Sweden and Denmark. See Eggertsson, Juelsrud and Wold (2017) and Bech and Malkhozov (2016).

11

The cost of hoarding cash includes an upfront fixed cost (e.g., building secure storage facilities), in addition to per-period variable costs (e.g., security and insurance). Therefore, the breakeven interest rate at which it becomes cost effective to convert to cash will be lower the longer that negative rates are expected to persist.

12

The Swiss banking sector consists of G-SIBs (owning about a half of banking sector assets), domestically-oriented banks, defined here to include cantonal, regional, savings, and Raiffeisen banks (about one fourth of sector assets), and others, including private banks, foreign-controlled banks, and branches of foreign banks. This definition is not fully consistent with the SNB’s (2017a) definition of “domestically-focused commercial banks” (banks where domestic loans exceed 50 percent of total assets or with a prominent role in the domestic deposit market).

13

The analysis is based on 50 banks that fulfill the supervisory authority’s definition of a retail bank. The two big banks and cooperative banks—which together account for 40 percent of the Swiss loan market—are excluded from the analysis.

14

Turk (2016) documents that bank profits were broadly stable in Denmark and Sweden after the introduction of the NIRP. In these countries, lower interest income was offset by lower wholesale funding costs and higher fee income.

15

G-SIBs have a different business model and showed more volatile returns.

16

The average interest margin on outstanding claims of domestically-focused banks decreased by 0.3 percentage points (equivalent to 2.5 percent) in 2016, while mortgage lending grew by 4.1 percent, resulting in an increase in net interest income.

17

At the time of writing, profit data for 2017 was not yet available.

Switzerland: Selected Issues
Author: International Monetary Fund. European Dept.
  • View in gallery

    Switzerland: Money Market Rates

    (Percent)

  • View in gallery

    Switzerland: Banks’ Sight Deposits at SNB

    (Billions of CHF)

  • View in gallery

    Switzerland: Customer Deposits Denominated in CHF 1/

    (Billions of CHF)

  • View in gallery

    Switzerland: Swiss Average Rate Overnight (SARON) and Mean Interest Rates for New Transactions

    (Percent)

  • View in gallery

    Switzerland: Bank Deposits in CHF by Domestic Customers

    (Billions of CHF)

  • View in gallery

    Switzerland: Interest Rates for New Mortgage Transactions

    (Percent)

  • View in gallery

    Switzerland: Cash in Circulation

    (Year-on-year growth, percent)

  • View in gallery

    Switzerland: Velocity of Money 1/

  • View in gallery

    Switzerland: Average Interest Rates on Selected Balance Sheet Items

    (Percent)

  • View in gallery

    Switzerland: Balance Sheet Items of Domestically Focused Banks 1/

    (Billions of CHF)

  • View in gallery

    Switzerland: Mortgage Growth

    (Year-on-year percentage change)

  • View in gallery

    Switzerland: Return on Assets, 2001–16

    (Percentage)

  • View in gallery

    Switzerland: Income of Domestically-Oriented Banks

    (Billions of CHF)

  • View in gallery

    Switzerland: Net Interest Income as a Ratio to Total Assets, Domestically-Focused Banks

    (Percent)