Capital flows, monetary policy, and macroprudential policy: reflections from the Turkish experience

Capital flows, monetary policy, and macroprudential policy: reflections from the Turkish experience Hakan Kara* Central Bank of Turkey South African Reserve Bank Biennial Conference October 27-28, 2016 Pretoria, South Africa * The views expressed are those of the presenter and not necessarily reflect the official view of the Central Bank of Turkey

Outline 1. Turkey and South Africa: Fraternal twins? 2. Global financial flows and the policy dilemma 3. Turkish experience in dealing with trade-offs 4. The outcome and reflections 2

Comparing South African and Turkish Economy South Africa Commodity exporter NIIP low negative Sizeable budget deficit Low currency mismatch Differences Turkey Commodity importer NIIP high negative Low budget deficit Significant currency mismatch in NFC balance sheets Similarities Inflation targeting, low saving rate, current account deficit 3

Mar-99 Sep-99 Mar-00 Sep-00 Mar-01 Sep-01 Mar-02 Sep-02 Mar-03 Sep-03 Mar-04 Sep-04 Mar-05 Sep-05 Mar-06 Sep-06 Mar-07 Sep-07 Mar-08 Sep-08 Mar-09 Sep-09 Mar-10 Sep-10 Mar-11 Sep-11 Mar-12 Sep-12 Mar-13 Sep-13 Mar-14 Sep-14 Mar-15 Sep-15 Mar-16 GDP Growth Rates (percent, YoY) 15 Turkey S. Africa EM Average* 10 5 0-5 -10-15 -20 Source: Bloomberg, CBRT Last Observation: 2016 Q2 * GDP weighted arithmetic mean of annual GDP growth rates of China, Brazil, India, Mexico, Russia, Poland, Indonesia, Argentina, Thailand, Malaysia, Colombia, Hungary, Romania, Philippines, Chile, Peru. 4

Mar-12 Sep-12 Mar-13 Sep-13 Mar-14 Sep-14 Mar-15 Sep-15 Mar-16 GDP Growth Rates (percent, YoY) 6 Turkey S. Africa 5 4 3 2 1 0 Source: Bloomberg, CBRT Last Observation: 2016 Q2 * GDP weighted arithmetic mean of annual GDP growth rates of China, Brazil, India, Mexico, Russia, Poland, Indonesia, Argentina, Thailand, Malaysia, Colombia, Hungary, Romania, Philippines, Chile, Peru. 5

Romaina Turkey Poland South Africa Colombia India Brasil Peru Hungary Mexico Indonesia Chile Tailand Phillipines Malesia Current Account Balance (2006-2014 Average) 12 10 8 6 4 2 0-2 -4-6 -8 Türkiye benzer ülkeler içerisinde, cari işlemler açığında da enflasyonda olduğu gibi üst sıralarda yer almaktadır. Bu durum, dolarizasyonun da yüksek olmasına neden olmaktadır. 6

Poland Thailand Malaysia Peru Hungary Chile Philippines Romania Mexico Colombia Brazil South Africa Indonesia India Turkey CPI Inflation (2006-2015 Average) 9 8 7 6 5 4 3 2 1 0

Nominal Exchange Rate Volatility (2006-2016, Coefficient of Variation) Turkey South Africa Brazil Colombia India Indonesia Romania Mexico Hungary Poland Chile Malaysia China Peru Thailand Philippines 0 0.05 0.1 0.15 0.2 0.25 0.3 8

Peru Turkey Chile South Africa Indonesia Thailand South Korea Mexico Malaysia Pass-through From Exchange Rate to Inflation (at the end of 12-month period, Percent) 16 14 12 10 8 6 4 2 0 Source: Various papers 9

Jan-09 Jul-09 Jan-10 Jul-10 Jan-11 Jul-11 Jan-12 Jul-12 Jan-13 Jul-13 Jan-14 Jul-14 Jan-15 Jul-15 Jan-16 Jul-16 Jan-09 Jul-09 Jan-10 Jul-10 Jan-11 Jul-11 Jan-12 Jul-12 Jan-13 Jul-13 Jan-14 Jul-14 Jan-15 Jul-15 Jan-16 Jul-16 Exchange Rates Lira and Rand vis-a-vis US Dollar Moving Correlation Between Lira and Rand (104 weeks)* 3.5 TRY ZAR (rhs) 20 1 0.9 3 17 0.8 2.5 14 0.7 0.6 2 11 0.5 1.5 8 0.4 0.3 1 5 0.2 Source: Bloomberg * Moving coefficients of correlation between weekly changes in USD/TRY and USD/ZAR exchange rates. Last Observation: October 21, 2016 10

Jul-10 Dec-10 May-11 Oct-11 Mar-12 Aug-12 Jan-13 Jun-13 Nov-13 Apr-14 Sep-14 Feb-15 Jul-15 Dec-15 May-16 Oct-16 Jul-10 Jan-11 Jul-11 Jan-12 Jul-12 Jan-13 Jul-13 Jan-14 Jul-14 Jan-15 Jul-15 Jan-16 Jul-16 CDS Premia and Bond Yields 5-Year CDS Premia (basis points) 10-Year Government Bond Yields (percentage points) 450 400 350 Turkey 5Y CDS S. Africa 5Y CDS EM Average* 12 11 Turkey 10Y Bond S. Africa 10Y Bond 11 10 300 10 9 250 9 200 8 8 150 100 50 7 6 7 6 0 5 5 Source: Bloomberg Last Observation: October 21, 2016 * Arithmetic mean of the 5-year CDS premia of Brazil, Chile, Colombia, Hungary, Indonesia, Mexico, Malaysia, Peru, Philippines, Poland, Romania, Russia and Thailand. 11

Dealing with capital flow and exchange rate volatility 12

The Dilemma Capital flows to emerging economies have been extremely volatile in recent years Fragile recovery during the post-crisis period Unconventional monetary policy in advanced economies Exchange rates are driven by global risk perceptions and thus do not necessarily reflect domestic fundamentals. Exchange rate as a shock amplifier? Such an environment poses complicated trade-offs and significant challenges for small open emerging economies. 13

Interaction between exchange rate, credit, and capital inflows: Amplifying channels Global Liquidity Capital Inflows Improved risk perceptions Currency Appreciation Rapid Credit Growth Balance sheet and collateral effects, further credit growth 14

Jan-07 Jul-07 Jan-08 Jul-08 Jan-09 Jul-09 Jan-10 Jul-10 Jan-11 Jul-11 Jan-12 Jul-12 Jan-13 Jul-13 Jan-14 Jul-14 Jan-15 Jul-15 Jan-16 Jul-16 US Dollar and the Commodity Prices Dollar Index and Commodity Prices (annual percentage change) 0.8 0.6 US Dollar Index Goldman Sachs Commodity Price Index 0.4 0.2 0-0.2-0.4-0.6 Source: Bloomberg Last Observation: October 21, 2016 ** Rolling coefficients of correlation between monthly changes in Dollar Index and Goldman Sachs Commodity Price Index. 15

Balance sheet effect of exchange rates in Turkey: Sensitivity Analysis The impact of a 10% depreciation on the profitability of non-financial corporates (percentage of profit excluding financial expenses) 0-5 -10-15 -20-25 -30 2008 2009 2010 2011 2012 2013 2014 2015 16

Implications for the policy design Significant policy trade-offs between price stability and financial stability Expanding the toolkit by devising multiple instruments may improve the outcome. 17

The Turkish Experience of dealing with policy trade-offs 18

1108 0209 0509 0809 1109 0210 0510 0309 0609 0909 1209 0310 0610 0910 The need for a macro approach have intensified after the QE, with rapid credit growth and marked currency appreciation. 60 Total Loan Growth Rates (13 Weeks Moving Average, Annualized, FX Adjusted, Percent) 125 Real Exchange Rate (2003=100) 45 120 30 115 15 0 110-15 -30 Beginning of QE by Fed 105 100 QE Source: CBRT 19

2007:09 2007:11 2008:01 2008:03 2008:05 2008:07 2008:09 2008:11 2009:01 2009:03 2009:05 2009:07 2009:09 2009:11 2010:01 2010:03 2010:05 2010:07 2010:09 2010:11 2011:01 Sharp widening in the current account deficit, financed with short-term inflows called for an immediate policy action. Current Account Balance (Seasonally Adjusted, Quarterly Average, Billion USD ) 0-1 -2-3 -4-5 -6-7 80 70 60 50 40 30 20 10 0-10 Main Sources of External Financing* (12-months Cumulative, Billion USD) Portfolio and Short-Term* FDI and Long-Term** Current Account Deficit -8 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2008 2009 2010 2011-20 Source: TURKSTAT, CBRT. *Short-term capital movements are sum of banking and real sectors' short term net credit and deposits in banks. Long-term capital movements are sum of banking and real sectors long term net credit and bonds issued by banks and the Treasury. Source: CBRT. 20

Credit booms, exchange rate misalignments, and financial stability Credit booms and an overvalued exchange rate are the most robust and significant predictors of financial crises. Borio and Lowe (2002), Reinhart and Rogoff (2009), Jorda et al (2011), Gourinchas and Obstfeld (2012), Schularick and Taylor (2012) A rise in the household debt to GDP ratio predicts lower output growth over the medium-run. Mian, Sufi and Verner (2015) 21

1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016* Historically, sharp capital flow reversals (sudden stops) in Turkey are associated with large output losses. 10 8 6 4 2 0-2 -4 Net Capital Flows / GDP GDP Growth Rate -6-8 -10 Crisis Year Cumulative Output Loss 1994 14,3% 2001 12,3% 2008-2009 15,1%. * As of Q2, annualized. Source: CBRT. 22

Monetary Policy Framework: Financial stability augmented IT Conventional Policy Unconventional Policy Objectives Price Stability Price Stability Financial Stability Policy Tool(s) Policy Rate Policy Rate Interest Rate Corridor Reserve Req. Policy 23

Using monetary policy tools as macroprudential instruments Rapid credit growth and widening current account deficit have required an urgent response by the end of 2010. No institution had the mandate and/or explicit objective at the time to deliver the necessary MaP response. The CBRT stepped in by Incorporating financial stability into the IT framework Redesigning the toolkit to respond to macro financial risks Reserve requirements and interest rates were used jointly for both monetary policy and MaP purposes. 24

Jun-10 Aug-10 Oct-10 Dec-10 Feb-11 Apr-11 Jun-11 Jun- Jun- Jun- Jul-10 Jul-10 Aug Aug Sep- Sep- Oct- Oct- Nov Nov Nov Dec- Dec- Jan- Jan- Feb- Feb- Mar Mar Apr- Apr- May May May Jun- Jun- 12 Initial reponse: a flexible interest rate policy joint with tigher Reserve Requirements 14 10 8 6 4 2 Overnight Interest Rate Lending Corridor Rate Overnight Rate (%,5-day MA) One-Week Repo Rate (%) 12 10 8 6 4 2 140 14 0 12 10 8 6 4 2 Reserve Requirement ratio (RRR) Effective RRR 12 10 8 6 4 2 0 0 25

Yet, the new approach posed challenges for policy communication. Complex transmission mechanism of new tools Inherently vague nature of financial stability: difficulty of linking each tool to objectives (unlike IT) Even so, this episode has increased the awareness for the need to adopt a more explicit institutional MaP framework. Paved the way for the foundation of a financial stability committee. 26

Second Phase of MaP Implementation: Financial Stability Committee 27

Financial Stability Committee (FSC) was a significant step for the establishment of a formal institutional framework for MaP. Organization Structure CENTRAL BANK OF THE REPUBLIC OF TURKEY CAPITAL MARKETS BOARD UNDERSECRETARIAT OF TREASURY FINANCIAL STABILITY COMMITTEE CHAIR: DEPUTY PRIME MINISTER BANKING REGULATION AND SUPERVISION AGENCY SAVINGS DEPOSIT INSURANCE FUND Founded in June 2011. Enhances information sharing, coordination, and cooperation. Main duties are to assess the systemic risks, identify necessary measures, and make relevant policy recommendations. No decision power or tools; the power rests with the authorities represented in the Committee. Each institution has its own mandate and responsibility. 28

Broad Objectives of Macroprudential Policy Dampen the financial amplification channels Interaction between capital flows, credit, and exchange rate Lower the probability of a sudden stop Contain credit growth and household overborrowing Improve the quality of financing 29

New tools were developed to dampen the amplification mechanisms. MAIN TOOLS: Wide Interest Rate Corridor Flexible interest rate policy against sudden shifts in capital flows Reserve Option Mechanism Build reserves with low sterilization costs during inflows. Reduce sensitivity of credit and exchange rates to capital inflows. 30

The policy toolkit has been expanded to weaken the amplification mechanisms triggered by global liquidity cycles. Global Liquidity Flexible Interest Rate Corridor Capital Inflows Reserve Option Mechanism Improved risk perceptions Currency Appreciation Rapid Credit Growth Balance sheet and collateral effects, further credit growth 31

MaP measures were taken to contain credit growth and change the composition of credit to reduce the share of household borrowing. MAIN TOOLS: Financial Institutions Based Risk weights Loan-loss provisions Borrower Based LTV ceilings Maturity limits Limits on # of installments 32

12/06 06/07 12/07 06/08 12/08 06/09 12/09 06/10 12/10 06/11 12/11 06/12 12/12 06/13 12/13 06/14 12/14 06/15 12/15 06/16 Household borrowing displays a strong relationship with the current account deficit. Consumer Loans and the Current Account (Change YoY, % of GDP for Consumer Loans, % of GDP for Current Account Deficit) 10 5.0 9 4.5 8 4.0 7 6 5 4 3.5 3.0 2.5 2.0 1.5 3 1.0 2 CAD/GDP ΔConsumer Loans Stock/GDP(RHS) 0.5 1 0.0 Source: TURKSTAT, CBRT. 33

The macroprudential measures to smooth credit cycle and to contain household debt were implemented in two major steps. First Round (2011) Higher risk weights and provisions for consumer loans. Limits to credit card payments LTV cap for housing loans Second Round (2013-2014) Caps, limits, and higher risk weights on credit cards Maturity restrictions (36 months) for uncollateralized consumer loans LTV cap for vehicle loans 34

0110 0111 0112 0113 0114 0115 0116 Macroprudential policies have succeeded in containing loan growth. 50 I Loan Growth Rates (Annual percentage change) 50 45 45 40 35 30 Two Rounds of MaP Policy Tightening II 40 35 30 25 Commercial 25 20 20 15 15 10 Consumer 10 5 5 0 0 Source: CBRT. Last Observation: October 14, 2016. Inclusive of loans extended by all types of banks (deposit banks, Participation banks, and development/investment banks). FX adjusted. 35

06/06 12/06 06/07 12/07 06/08 12/08 06/09 12/09 06/10 12/10 06/11 12/11 06/12 12/12 06/13 12/13 06/14 12/14 06/15 12/15 06/16 The share of consumer loans in total loans have declined considerably since 2011. 27 26 Consumer Loans/Total Loans I 25 II 24 23 MaP to curb consumer loans 22 21 20 36

12.10 12.11 12.12 03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 08.16 The MaPs have been instrumental in reversing the upward trend in the household indebtedness ratio in recent years. Household Assets and Liabilities (Billion TL, Percent) 1000 Total Assets Total Liabilities Liabilities/Assets (RHA) 60 750 55 500 50 250 45 0 40-250 35-500 30 Source: CBRT Financial Stability Report May 2015 Last Observation: August 2016 37

01/03 06/03 11/03 04/04 09/04 02/05 07/05 12/05 05/06 10/06 03/07 08/07 01/08 06/08 11/08 04/09 09/09 02/10 07/10 12/10 05/11 10/11 03/12 08/12 01/13 06/13 11/13 04/14 09/14 02/15 07/15 12/15 05/16 140 130 Real exchange movements since 2011 have boosted the competitiveness of the economy. Real Effective Exchange Rate (CPI Based. 2003=100) New Policy Mix 120 110 100 90 80 Source: CBRT. Last Data : July 2016. 38

Reserve requirement ratios were used to increase the quality of financing (liabilities of the banking system) MAIN TOOLS: reserve requirement policy Strategy: adjust RRR and remuneration rates to favor Core liabilities over non-core liabilities Long-term over short-term TL over FX 39

12/2011 03/2012 06/2012 09/2012 12/2012 03/2013 06/2013 09/2013 12/2013 03/2014 06/2014 09/2014 12/2014 03/2015 06/2015 12/2011 03/2012 06/2012 09/2012 12/2012 03/2013 06/2013 09/2013 12/2013 03/2014 06/2014 09/2014 12/2014 03/2015 06/2015 12/2011 03/2012 06/2012 09/2012 12/2012 03/2013 06/2013 09/2013 12/2013 03/2014 06/2014 09/2014 12/2014 03/2015 06/2015 12/2011 03/2012 06/2012 09/2012 12/2012 03/2013 06/2013 09/2013 12/2013 03/2014 06/2014 09/2014 12/2014 03/2015 06/2015 20.0 19.0 18.0 17.0 16.0 15.0 14.0 13.0 12.0 11.0 10.0 Reserve Requirements were used in several dimensions to improve the quality of external finance and bank liabilities. TL vs FX (short term) TL short FX short 20.0 19.0 18.0 17.0 16.0 15.0 14.0 13.0 12.0 11.0 10.0 Core vs Non-core (FX) Core FX Non-Core FX 12.0 11.0 10.0 9.0 8.0 Short vs Long (core) Core TL short Core TL long 20 18 16 14 Short vs Long (Non-core) Non-Core short Non-Core long 7.0 12 6.0 10 5.0 8 4.0 6 40

02/14 04/14 05/14 07/14 08/14 09/14 11/14 12/14 01/15 03/15 04/15 06/15 07/15 08/15 10/15 11/15 01/16 02/16 03/16 05/16 06/16 07/16 09/16 RR measures have been effective in extending the maturity of the noncore liabilities Maturity Breakdown of Non-Core FX Liabilities (Percentage Share) 70 RR Measures to encourage maturity extension 60 50 Up to 1-Year 40 30 Longer than 3-Years 20 10 1 to 3-Years 0 Source: CBRT. Last Observation: September 9, 2016. 41

01/10 04/10 07/10 10/10 01/11 04/11 07/11 10/11 01/12 04/12 07/12 10/12 01/13 04/13 07/13 10/13 01/14 04/14 07/14 10/14 01/15 04/15 07/15 10/15 01/16 04/16 07/16 as well as stabilizing credit to deposit ratios. 130 Credit/Deposit Ratio (Percent) 130 120 120 110 110 100 90 RR Measures to Support Core Liabilities 100 90 80 80 70 70 Source: BRSA Last Observation: October 5, 2016. 42

Macro Outcomes of MaP Implementations 43

2004Q3 2005Q1 2005Q3 2006Q1 2006Q3 2007Q1 2007Q3 2008Q1 2008Q3 2009Q1 2009Q3 2010Q1 2010Q3 2011Q1 2011Q3 2012Q1 2012Q3 2013Q1 2013Q3 2014Q1 2014Q3 2015Q1 2015Q3 2016Q1 2016Q3 2017Q2 The interaction between capital flows, exchange rate and bank loans have been dampened considerably since the implementation of MaPs. 2 1.5 1 Capital flows, exchange rate, and credit cycles (HP filtered, standardized) Loans (t) REER (t+3) Inflows (t+3) 0.5 44 0-0.5-1 -1.5 The new policy mix -2 Source: CBRT. 44

0208 0808 0209 0809 0210 0810 0211 0811 0212 0812 0213 0813 0214 0814 0215 0815 0216 0816 Current account balance has shown a persistent and sizeable improvement in recent years, despite the volatility in capital flows. Current Account Deficit and Net Capital Inflows (12 Month Cumulative, Billion USD) 100 90 80 70 60 50 40 30 20 10 0 Net Capital Inflows CAD (excluding gold, right axis) MaP CBRT starts using ROM actively 100 90 80 70 60 50 40 30 20-10 10 Source: CBRT. 45

0107 0707 0108 0708 0109 0709 0110 0710 0111 0711 0112 0712 0113 0713 0114 0714 0115 0715 0116 0716 The share of long-term and FDI inflows have increased since 2011. Net Capital Inflows (12-Months Cumulative, Billion USD) 100 80 60 Portfolio and Short Term** New Policy Mix FDI and Long Term* 40 20 0-20 Source: CBRT. *Long term inflows are sum of banking and real sectors long term net credit and bonds issued by banks and the Treasury. **Short term capital movements are sum of banking and real sectors' short term net credit and deposits in banks. Last Observation: August 2016. 46

1998 1999 2000 2001 2001 2002 2003 2004 2004 2005 2006 2007 2007 2008 2009 2010 2010 2011 2012 2013 2013 2014 2015 2016 GDP growth stayed relatively stable since 2011, despite the heigthened volatility in capital flows. 15 10 5 0-5 -10-15 -20 Annual Change in Net Capital Flows/GDP (percent) GDP (real, annual growth, percent, right axis) 20 15 10 5 0-5 -10-15 -20 Source: CBRT. 47

12/08 06/09 12/09 06/10 12/10 06/11 12/11 06/12 12/12 06/13 12/13 06/14 12/14 06/15 12/15 06/16 The contribution of global factors to Turkey s monetary and financial conditions have declined after the implementation of MaP policies. 100 90 80 70 % Share of Global Factors* in Explaining Turkey s Financial Conditions 60 MaP measures 50 40 30 20 10 *Global Factors are VIX index, global growth, US treasury 10-year yield and 10-2 year spread. Source: Kara, Özlü and Ünalmış (2015) 48

01/10 05/10 09/10 01/11 05/11 09/11 01/12 05/12 09/12 01/13 05/13 09/13 01/14 05/14 09/14 01/15 05/15 09/15 01/16 05/16 09/16 Inflation expectations improved during the initial stages; yet deteriorated since the taper tantrum due to persistent depreciation in TL. 10.0 Medium-term (2-year ahead) Inflation Expectations (Percent) 9.0 8.0 7.0 Adoption of the New Policy Framework Taper tantrum and political uncertainty 6.0 5.0 Target 4.0 3.0 2.0 1.0 0.0 Source: CBRT. Last Observation: October 2016. 49

06Q2 06Q4 07Q2 07Q4 08Q2 08Q4 09Q2 09Q4 10Q2 10Q4 11Q2 11Q4 12Q2 12Q4 13Q2 13Q4 14Q2 14Q4 15Q2 15Q4 16Q2 06Q2 06Q4 07Q2 07Q4 08Q2 08Q4 09Q2 09Q4 10Q2 10Q4 11Q2 11Q4 12Q2 12Q4 13Q2 13Q4 14Q2 14Q4 15Q2 15Q4 16Q2 06Q2 07Q1 07Q4 08Q3 09Q2 10Q1 10Q4 11Q3 12Q2 13Q1 13Q4 14Q3 15Q2 16Q1 06Q2 07Q1 07Q4 08Q3 09Q2 10Q1 10Q4 11Q3 12Q2 13Q1 13Q4 14Q3 15Q2 16Q1 Estimated coefficients from a Time Varying Parameter Phillips curve 0.30 FX Pass-through coefficient 0.25 Import Price Pass-through 0.25 0.20 0.15 0.10 0.05 Filtered Smooted 0.20 0.15 0.10 0.05 Filtered Smooted 0.00 0.00 0.25 Output Gap Coefficient 0.35 Unit Labor Cost Coefficient 0.20 0.15 0.30 0.25 0.20 0.10 0.05 Filtered Smooted 0.15 0.10 0.05 Filtered Smooted 0.00 0.00 50

Contributions to inflation during the last 10 years 14 12 10 8 6 4 2 0-2 -4-6 -8 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 (F) Rigidity (constant term) Unprocessed Food (Direct and Indirect) Exchange Rate ($/TL) Import Prices Output Gap Unit Labor Cost Tax Other CPI Inflation 51

Final Remarks Macroprudential policies in Turkey have engineered a soft landing and alleviated the adverse impact of capital flows. The Turkish experience suggests that targeted macro prudential policies can improve the policy tradeoffs. Building buffers in good times, using them in bad times Yet, neither monetary policy nor macroprudential policy can substitute for deeper structural reforms. 52

Main Lessons Turkey has tried various forms inflation targeting in the past 10 years. Multiple tools - multiple instruments framework eases some of the tradeoffs posed by the global financial cycle but may introduce additional challenges especially under low credibility of inflation targets. Complexity and communication challenges 53

The bottom line Simplicity is the ultimate sophistication. Leonardo Da Vinci 54