Int Inter erdep epend endenc ence b e between een Fi Fiscal - PowerPoint PPT Presentation

Int Inter erdep epend endenc ence b e between een Fi Fiscal cal an and d Mo Monetar ary Poli licy cy: th the case se for or Costa Costa Rica Valerie Lankester Catalina Sandoval CEMLA XXV Meeting of the Central Bank Researchers Network October 30 th , 2020

The ideas expressed in this paper are those of the authors and not necessarily represent the view of the Central Bank of Costa Rica.

Motivation • Relationship between the effectiveness of monetary policy and fiscal policy coordination (Sargent and Wallace, 1981). • Tradeoffs between the degree of independence of the policies and their effectiveness (Aiyagari and Gertler, 1985). • This is especially relevant for Costa Rica where the central government’s debt level has reached levels over 50% of its GDP and the Central Bank has made remarkable efforts to strengthen its independence. 3

Objectives Analyzing the interdependence between monetary policy and fiscal policy in Costa Rica in period 1991-2019: 1. Fiscal dominance test: analyze the relationship between primary fiscal balance and public sector liabilities. 2. To estimate the effect of fiscal variables on the Central Bank's monetary policy rate. 3. To estimate the effect of the fiscal deficit on the inflation rate. 4

Inflation and fiscal deficit, 1991-2019 14 12 10 Percentages (%) 8 6 4 2 0 -2 -4 1990q1 2000q1 2010q1 2020q1 Quarters Series Inflation (QoQ variation) Fiscal deficit (% of GDP) 5 Source: Central Bank of Costa Rica.

Costa Rica: monetary policy and fiscal events MONETARY POLICY Central Bank Law Managed Inflation Monetary Openness Crawling Capitalization of float target policy rate of capital band BCCR deficit (80s) account 1992 1994 1995 2006 2009 2011 2015 2018 Banco Anglo’s Fiscal reform Expansionary bankruptcy fiscal policy FISCAL EVENTS 6

Inflation and exchange rate regime Central Bank has modified its monetary policy regime toward inflation targeting 110 90 Crawling peg Crawling band Managed float Mean inflation: 13.4% Mean inflation: Mean inflation: 70 5.5% 1.3% 50 30 10 Apr. 83 Jun. 85 Apr. 96 Jun. 98 Apr. 09 Jun. 11 Jan. 80 Feb. 81 Mar. 82 May. 84 Jul. 86 Aug. 87 Sep. 88 Oct. 89 Nov. 90 Dec. 91 Jan. 93 Feb. 94 Mar. 95 May. 97 Jul. 99 Aug. 00 Sep. 01 Oct. 02 Nov. 03 Dec. 04 Jan. 06 Feb. 07 Mar. 08 May. 10 Jul. 12 Aug. 13 Sep. 14 Oct. 15 Nov. 16 Dec. 17 Jan. 19 Feb. 20 -10 Inflation Period average Source: Central Bank of Costa Rica 7

De Deficit of of the Central Bank (% of of GDP) GDP), 1983 983 - 20 2019 BCCR has a deficit since the crisis of the 80’s, but it has decreased over time 5.0 Deficit (% GDP) Average deficit 4.0 3,0% 3.0 Percentages 1,5% 2.0 0.6% 1.0 0.0 1983 1984 1985 1986 1987 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 2017 2018 2019 8 Source: Central Bank of Costa Rica

Ce Central Government Debt, 2000-2020* Public finance’s behavior changed significantly after the financial crisis in 2008 70% 10% 8% 60% 6% Central Goverment Debt (% of GDP) 50% 4% 40% YoY change 2% 30% 0% 20% -2% 10% -4% 0% -6% 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020* Debt(% of GDP) GDP growth Note: *IMF projection 9 Source: Central Bank of Costa Rica

Fisc Fi scal and d pr primary balance e of the e Cen entral Gover ernmen ent (% of GDP), 2000-2020* The country has reached a critical fiscal situation 4.0 2.0 0.0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020* Percentajes of GDP -2.0 -4.0 -6.0 -8.0 -10.0 Primary balance(% of GDP) Interest payment Fiscal deficit Note: *IMF Projection 10 Source: Central Bank of Costa Rica

Data suggests a different relationships through time Policy interest rate vs primary deficit (% GDP), 1991-2019 35 30 Banco Anglo’s bankruptcy Policy interest rate 25 Policy rate (%) 20 15 10 5 0 -5 0 5 10 Primary deficit (% of GDP) Period 1991-1999 2000-2008 2009-2019 11 Source: Central Bank of Costa Rica.

Theoretical framework: consolidated government's budget identity • Budget identity of the government for one period (Wahls,2010): 𝑕 ! + 𝑠 !"# 𝑐 !"# = 𝜐 ! + 𝑐 ! − 𝑐 !"# + 𝑡 ! (1) Expenditures Revenues • Intertemporal budgetary balance: ( $%& + $%& , $%& ' ' ' 1 + 𝑠 𝑐 !"# + ∑ $%& #)* & = ∑ $%& #)* & + ∑ $%& (2) #)* & (𝒉"𝝊"𝒕) $%& ' 𝑆𝑐 !"# = −∑ $%& 𝑆 = 1 + 𝑠 and primary deficit = 𝑕 − 𝜐 − 𝑡 2 & 12

Theoretical framework: consolidated government's budget identity • Government budgetary constraint: ' 𝑆 "$ (𝜐 − 𝑕) !($ + 𝑆 "# ∑ $%& ' 𝑆 "$ 𝑡 !($ 𝑐 !"# = 𝑆 "# ∑ $%& (3) • If debt is positive ( 𝑐 >0) the present value of incomes ( 𝜐 , 𝑡 ) should be higher than expenditures ( 𝑕 ). • The adjustment can be done through reducing expenditures or increasing revenues from taxes or seigniorage. • Who adjusts to maintain balance define dominance : • Monetary dominance (MD) • Fiscal dominance (FD) 13

Previous literature from developing economies Evidence suggests that the scope for monetary policy has been contingent on fiscal policy: • Primary balance is found to be exogenously determined from public liabilities (Tanner and Ramos, 2005; Jevđović and Milenković, 2018) • Monetary policy rate reacts to fiscal variables: • Positively (Kuncoro and Sebayang (2013), Ahmed et al. (2019)) • Not significantly (Zoli, 2005) Afondo et al. (2019) • Fiscal deficit seems to have a significant long-run effect on inflation (Catao and Terrones, 2005; Jalil, Tariq and Bib, 2014) 14

Quarterly data from 1991-2019 o Monetary variables: monetary policy interest rate 1 , inflation rate, core inflation rate, inflation target 2 , monetary base, nominal exchange rate, international reserves o Fiscal variables: Fiscal deficit , primary deficit, central government debt (total, external, and internal debt) o Other variables: Product gap, public liabilities, Central Bank deficit, WTI oil prices 15

Objective 1: Fiscal dominance test To estimate the relationship between primary balance and liabilities • VAR Model: 𝑞𝑠𝑗𝑛𝑐𝑏𝑚 ! = 𝛽 ' + ∑ ()# 𝛽 ( 𝑞𝑠𝑗𝑛𝑐𝑏𝑚 !"( + ∑ ()# 𝜸 𝒌 𝑚𝑗𝑏𝑐 !"( +𝜁 ! (1) 𝑚𝑗𝑏𝑐 ! = 𝛿 ' + ∑ ()# 𝜺 𝒌 𝑞𝑠𝑗𝑛𝑐𝑏𝑚 !"( + ∑ ()# 𝛿 ( 𝑚𝑗𝑏𝑐 !"( +𝜕 ! (2) • Classifying the results on fiscal or monetary dominance regimes: Impulse Response Functions Granger causality test Sign Prim. Balance Liabilities Granger causality Dominance response ( 𝜸 𝒌 ) response( 𝜺 𝒌 ) Primary balance à Public liabilities Fiscal Zero FD FD Public liabilities à Primary balance Monetary Negative FD MD Positive FD or MD MD 16

Objective 1: Fiscal dominance test Granger causality test results Empirical considerations: Variables as GDP percentages and in first differences (unit root) • • Number of lags: 4 according to HQIC and SBIC information criteria • Controls for seasonality effects including dummies • Controls for Banco Anglo’s bankruptcy in 1994, international financial crisis in 2008 and 2009 fiscal events Granger causality test results by period H0 of NO causality Period Chi2 P-value Conclusion A à B Liabilities à PB 2.21 0.70 1991-2019 Ambiguous PB à Liabilities 6.47 0.17 Liabilities à PB 5.32 0.26 Fiscal 1991-2007 dominance PB à Liabilities 10.32 0.04 Liabilities à PB 3.73 0.44 2008-2019 Ambiguous PB à Liabilities 3.29 0.51 17 Notes: VAR satisfices stability condition. Source: Central Bank of Costa Rica.

Objective 1: Fiscal dominance test Impulse Response Function test Impulse-Response Functions 1991-2019 Sign Prim. Balance Liabilities response ( 𝜸 𝒌 ) response( 𝜺 𝒌 ) Response of PB to Liabilites var1, pasivo_var1, pasivo_var1 Response of Liabilities to PB Zero FD FD .4 1 2 Negative FD MD .2 .5 1 Positive FD or MD MD 0 0 0 -.2 -.4 -1 -.5 Results suggest fiscal dominance, 0 2 4 6 8 10 12 0 2 4 6 8 10 12 step but: 95% CI orthogonalized IRFs • PB response to cicles Notes: VAR satisfices stability condition. Source: Central Bank of Costa Rica. • Identification problem 18

Objective 2: reaction function of the Central Bank General approach: considering the fiscal space • Taylor Rule (1993) • Evidence for Costa Rica between 1991-2002: a positive and significant effect of domestic debt (0.23) on the basic passive interest rate was found (Muñoz and Sáenz, 2003). • Model 𝑗 " = 𝛾 # + 𝛾 $ 𝑗 "%$ + 𝛾 & (𝜌 − 𝜌 ∗ ) "%$ + 𝛾 ( (𝑧 − 𝑧 ∗ ) "%$ +𝛾 ) 𝑓 "%$ + 𝜸 𝟔 𝑔𝑗𝑡𝑑𝑏𝑚 "%$ + 𝛾 + 𝑒𝐽𝑆 "%$ + 𝑣 " • Empirical considerations: • Controls for seasonality effects including dummies • Controls for Banco Anglo´s bankruptcy in 1994, international financial crisis in 2008 and 2009 fiscal events, also for exchange regime • Trend variable included • Newey-West standard errors 19