Practical Monetary Policy I: Unconventional Policies PhD Course in Monetary Economics Deputy Governor Martin Flodén 15 May 2017
Overview Background: recent monetary policy Unconventional policies
Background: Recent monetary policy
Negative repo rate and government bond purchases Purchases of government bonds Repo rate -0.50% 300 300 250 250 200 200 150 150 100 100 50 50 0 February March April July October April December April 2015 2016 2017 0 Note: SEK billion and per cent. Source: The Riksbank
Difficult situation 2014/15 Inflation had been far too low for many years Weaker confidence in monetary policy and the inflation target Falling inflation expectations Good growth but still weak resource utilisation Monetary policy made even more expansionary abroad ECB Switzerland and Denmark
Inflation had been too low for many years 3.0 3.0 2.5 2.5 2.0 2.0 1.5 1.5 1.0 1.0 0.5 CPIF excluding energy 0.5 0.0 10 11 12 13 14 15 16 17 0.0 Note. Per cent. The CPIF is the CPI with a fixed mortgage rate. Source: The Riksbank
Falling inflation expectations 2.6 2.4 5 years ahead 2 years ahead 2.6 2.4 2.2 2.2 2.0 2.0 1.8 1.8 1.6 1.6 1.4 1.4 1.2 10 11 12 13 14 15 16 17 1.2 Note. Per cent. Expectations refer to money market participants Source: TNS Sifo Prospera
Still low resource utilisation 2014/2015 1.0 1.0 RU-indicator (Feb 2015) 0.5 0.5 0.0 0.0-0.5-0.5-1.0-1.0-1.5-1.5-2.0 10 11 12 13 14 15 16 17-2.0 Note. The Riksbank s indicator for resource utilisation according to the assessment in February 2015. The RU indicator is normalised so that the mean value is 0 and the standard deviation is 1. Source: The Riksbank
Low policy rates abroad Note. Per cent. Source: The Riksbank, Denmark s Nationalbank, Swiss National Bank, ECB
Also extensive asset purchases Central banks assets as a share of GDP 120 100 80 Swiss National Bank Bank of Japan Bank of England Federal Reserve ECB Riksbank 120 100 80 60 60 40 40 20 20 Note. Per cent 0 08 09 10 11 12 13 14 15 16 17 0 Source: respective country's central bank
Exchange rate important Note. The KIX is an aggregate of the krona exchange rate against currencies in countries that are important for Sweden's international transactions. Source: The Riksbank
Other considerations Underlying international interest rates had fallen for several decades Demographics, self-insurance (emerging economies), technological change, low public investment,... Limited problems with negative rate and bond purchases No increased demand for cash The banks are profitable Households and companies are optimistic, have good access to credit Household debts? The Riksbank s balance sheet? Distributional effects? Bond market... strained but functioning satisfactorily according to SNDO Monetary policy has had an impact
Downward trend in global real interest rates 6 4 United Kingdom USA Sweden 6 4 2 2 0 0-2 -2-4 99 03 07 11 15 Note. 10-year yield on real government bonds in Sweden, the United Kingdom and the United States, per cent. Swedish real interest rates are zero coupon yields interpolated from bond prices using the Nelson-Siegel method. -4 Sources: Bank of England, Federal Reserve, Thomson Reuters and the Riksbank
Swedish banks are profitable Return on equity Cost structure in the banks Cost/assets 3.0 2.5 2.0 1.5 1.0 0.5 Spain Finland Belgium Sweden Netherlands Austria Denmark Ireland Portugal United States Italy France United Kingdom Switzerland Germany 0.0 30 40 50 60 70 80 90 Cost/income Return on equity measured in rolling four quarters, per cent. Unweighted mean value. The red line refers to a selection of major European banks. Cost in percent of assets and percent of income. Sources: The banks interim reports, the IMF and SNL Financial
Good sentiment in the economy 130 120 Micro index Macro index Economic Tendency Indicator 130 120 110 110 100 100 90 90 80 10 11 12 13 14 15 16 17 Note. Index, average = 100, standard deviation = 10, seasonally adjusted figures. 80 Source: National Institute of Economic Research
Worrying trends in house prices and household debt House prices rise faster than income Household debt is increasing In the left-hand figure, house prices and disposable income have been indexed according to 1986 Q1 = 100. House prices refer to single family houses. In the right-hand figure, debt-to-income refers to debt as a percentage of disposable income. Sources: Statistics Sweden and the Riksbank
Much slower price increases in Germany despite similar monetary policy Short-term interest rates Housing prices 6 5 Tyskland Sverige 6 5 200 180 Tyskland Sverige 200 180 4 4 160 160 3 2 3 2 140 140 1 1 120 120 0 0 100 100-1 05 07 09 11 13 15-1 80 05 07 09 11 13 15 80 Note. Per cent and index for real housing prices, 2005 Q1 = 100. Sources: The BIS and the OECD
Measures are needed for a better-functioning housing market Review of property taxation Continued reform of the rent-setting system Review of municipal regulations for construction Increased competition in the construction and civil engineering sector
Macroprudential measures are also needed Speed limits Incentives to accumulate debt Debt-to-income limit Loan-to-value limit Tax deductability of interest expenditure Risk weights
The Riksbank s balance sheet SEK billion 900 900 800 Assets Liabilities Equity 800 700 600 500 400 Foreign exchange reserve Other Foreign exchange loans 700 600 500 400 300 200 100 Gold 0 Assets Foreign exchange reserve Liabilities Equity Banknotes and coins Government bonds Gold Certificates and fine tuning Banknotes and coins August 2008 September 2016 300 200 100 0 Source: The Riksbank
Swedish QE modest relative to GDP 25% 20% ECB Sweden 25% 20% 15% 15% 10% 10% 5% 5% 0% okt-14 apr-15 okt-15 apr-16 okt-16 apr-17 okt-17 0% Bond purchases as a percentage of GDP in 2015. With regard to the ECB, all programs in their extended asset purchase program (APP) are included. The projections are linear from today s level and do not necessarily reflect the actual rate of purchase. Sources: ECB and the Riksbank
but large relative to stock of government bonds 50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 0% ECB - government bonds Sweden -Real government bonds Sweden - Nominal government bonds 14 15 15 16 16 17 17 50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 0% Percent of stock. For the ECB it is assumed that 70% of the total purchases will comprise nominal and real government bonds. The projections are linear from today s level and do not necessarily reflect the actual rate of purchase. Sources: ECB, Reuters and the Riksbank
Daily turnover in relation to stock 7% 7% 6% Government bond 6% 5% 4% Covered bond 5% 4% 3% 3% 2% 2% 1% 1% 0% 0% Note. Per cent, 3 months moving average. Sources: Macrobond, the Swedish National Debt Office and the Riksbank.
Monetary policy has had an impact Impact on market rates Confidence in the Riksbank s focus on the inflation target High growth, rising resource utilisation Rapid appreciation has been avoided Higher inflation expectations Higher inflation
Monetary policy has had an impact Stable exchange rate Rising resource utilisation Higher inflation expectations Higher inflation Note. Index, 18 Nov 1992 = 100 The KIX is an aggregate of countries that are important for Sweden's international transactions. The CPIF is the CPI with a fixed mortgage rate. Source: Statistics Sweden and TNS Sifo Prospera
Continued expansionary monetary policy is required to stabilize inflation around 2 per cent Inflation has risen and inflation expectations are close to 2% But Excluding energy prices, inflation is still low It will take some time before inflation stabilises around 2 per cent Uncertainty about policy abroad
Some questions
Suppose that the policy rate has been cut to zero but that further monetary stimulus is motivated Unconventional policies? Negative interest rate Asset purchases, QE, maturities Forward guidance, communication Helicopter money Fiscal policy, fiscal theory of the price level What are the transmission mechanisms? Understand money and operational frameworks Problems and unintended side effects? Where is the lower bound?
Forward guidance and liquidity traps
Eggertsson & Woodford (2003) PC: ππ tt = κκxx tt + ββee tt ππ tt+1 IS: xx tt = EE tt xx tt+1 1 σσ (ii tt EE tt ππ tt+1 rr tt nn ) Normal regime: rr tt nn = ρρ > 0 Deflationary regime: rr tt nn = rr dd < 0, exit with Pr = αα, never return. In normal times, MPR: ii tt = ρρ + φφ ππ ππ tt, and φφ ππ > 1. See also Eggertsson (2006) and Eggertsson (2008)
Eggertsson & Woodford (2003) Under discretion, solution to normal regime is ππ tt = xx tt = 0 and ii tt = ρρ. Solution to deflationary regime is then ii tt = 0 and ππ tt = κκ χχ rrdd <0 xx tt = 1 ββ(1 αα) χχ rr dd <0 where χχ = σσσσ 1 ββ 1 αα κκ(1 αα) and where the solution assumes that χχ > 0.
Eggertsson & Woodford (2003) 0 Inflation β=0.99, σ=1, η=1, θ=2/3 0 Output gap -2-5 -4-6 -10-8 -κ/χ -10 -(1-β*(1-α))/χ -15-12 -20-14 -16-25 -18-20 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Persistence: 1-α -30 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Persistence: 1-α
Eggertsson & Woodford (2003) Simplify: αα = σσ = 1 (so that χχ = 1), and assume deflationary regime in tt = 0. Discretionary solution is then ππ 0 = κκrr dd, xx 0 = rr dd, ππ tt = 0, and xx tt = 0, for tt 1. Suppose cb can commit to ii 1 = ρρ εε. Then ππ 0 cc = κκ 1 + κκ + ββ εε + κκrr dd, xx 0 cc = 1+κκ εε+rr dd, ππ 1 cc = κκκκ, xx 1 cc = εε, ππ tt = 0, and xx tt = 0, for tt 2.
Eggertsson & Woodford (2003) Loss function: L = 1 2 ββtt (ππ tt 2 + λλxx tt 2 ). Compare discretion (εε = 0) to commitment (εε 0): 2L(εε) = κκ 1 + κκ + ββ εε + κκrr dd 2 + λλ 1+κκ εε+rr dd 2 + ββ κκκκ 2 + ββββ εε 2 We immediately see that L εε εε=0 < 0 (because rr dd <0) So, try to commit to keep rate low long to get out of liquidity trap!
Committment is powerful in standard NK model PC: ππ tt = κκxx tt + ββee tt ππ tt+1 ππ tt = κκee tt jj=0 ββ jj xx tt+jj IS: xx tt = EE tt xx tt+1 1 (ii σσ tt EE tt ππ tt+1 rr nn tt ) xx tt = 1 EE σσ tt jj=0 rr tt+jj where nn rr tt+jj ii tt+jj EE tt+jj ππ tt+1+jj rr tt+jj Use xx tt in expression for ππ tt to get: ππ tt = κκ σσ EE tt[ rr tt + 1 + ββ rr tt+1 + 1 + ββ + ββ 2 rr tt+2 + ] Note: Committing to a change in rr tt+jj affects the output gap in all j+1 periods up until that period. And the (discounted) accumulated changes in the output gap affect today s inflation. So, committing to policy in the distant future will, in the standard model, have a large immediate impact on inflation. This is known as the forward guidance puzzle.
Some reflections on forward guidance Information about policy rule? Delphic Central banks must forecast their own policy rate. (Or use market expectations) Should this forecast be published? Commitment to deviate from normal policy rule? Odyssean How? Time inconsistency?
Forward guidance: How? Forecast of policy rate New Zealand (1997), Norway (2003), Sweden (2007) Scenarios in inflation reports At zero lower bound: for how long? BoJ, Fed, BoE, BoC,
Forward guidance Time or state contingent? Time contingent: Fed 2008-2012, ECB 2013-, State contingent: Fed 2012-, BoE 2013- Triggers or thresholds?
Forward guidance: Fed Source: IMF (2013a), Table 1
Forward guidance: BoE At its meeting on 1 August 2013, the MPC agreed its intention not to raise Bank Rate from its current level of 0.5% at least until the headline measure of the unemployment rate had fallen to a threshold of 7% subject to the conditions below. Knockouts: Medium-term inflation forecast above 2.5% Medium-tem inflation expectations not well anchored Stance of monetary policy is a significant threat to financial stability, and other policies cannot mitigate this Source: BoE (2013)
Forward guidance: BoE intentions Clarity about the MPC s view on the appropriate trade-offs How fast to reduce slack in the economy? Risks to overriding objective of price stability? Reduces uncertainty about the future path of monetary policy Reduces the risk that market interest rates rise prematurely Source: BoE (2013)
The Riksbank s repo rate path Is this forward guidance? No explicit commitment ( a forecast not a promise ) Some elements of conditional commitment The Riksbank has not seen much need for commitment Not restricted by ELB 2011-2014 Market expected lower future rates than the Riksbank 2010-2016
Market pricing lower than the repo path 3 2,5 Repo Riksbank Market, after decision 2 1,5 1 0,5 0 2012 2013 2014 2015 2016 Repo rate, the Riksbank s repo rate path, and market pricing in February 2013. Percent.
even when UMP was launched 2 1,5 Repo Riksbank Market, after decision 1 0,5 0 2014 2015 2016 2017 2018-0,5-1 Repo rate, the Riksbank s repo rate path, and market pricing in February 2015. Percent.
Would policy have been more expansionary if the repo path had not been published? Impact on market expectations? Difficult to know Should not be confused with outcome if the Riksbank s forecasts and monetary plans had been different Impact on repo rate decisions? Some indications that it has been easier to decide on changes in the repo forecast than in the repo rate. Has the path been an unhelpful substitute for repo rate decisions?
Different situation today market pricing now indicates earlier rate hike 1 0,8 0,6 Repo Riksbank Market, after decision 0,4 0,2 0-0,2 2016 2017 2018 2019 2020-0,4-0,6-0,8-1 Repo rate, the Riksbank s repo rate path, and market pricing in April 2017. Percent.
Different situation today Market pricing now indicates earlier rate hike than the Riksbank s forecast So, stronger argument for committing to the path today But I cannot see myself committing unconditionally And committing with escape clauses is basically what we already do (repo path + report + minutes + other statements) There are other ways to push market prices (and expectations) closer to our path if that becomes an issue E.g. lending at fixed rate at longer horizons as in 2009
So why publish the repo forecast? Facilitates communication Compare e.g. with complicated use of code words by some other central banks Important part of our forecast Better external, often critical, discussion of monetary policy Not a powerful high-frequency tool for monetary policy, but still helpful for illustrating reactions function