Convergence, capital accumulation and the nominal exchange rate

Transcription

1 Convergence, capital accumulation and the nominal exchange rate Péter Benczúr and István Kónya Magyar Nemzeti Bank and Central European University September 2

2 Disclaimer The views expressed are those of the author and do not necessarily reflect the official view of the Bank. Péter Benczúr and István Kónya Nominal growth 2

3 Our Goal Study the effect of the exchange rate regime on economic growth and convergence, through nominal asset holdings and nominal asset accumulation Why is it relevant? Péter Benczúr and István Kónya Nominal growth 3

4 Stylized Facts Households accumulate financial assets as their country develops (also holds in time series) 2 A large fraction of these assets are denominated in domestic currency The convergence process of open economies may be influenced by the exchange rate regime Péter Benczúr and István Kónya Nominal growth 4

5 Money (M2) and GDP Péter Benczúr and István Kónya Nominal growth 5

6 What We Do We take these facts as given and study their implications Does the exchange rate regime affect real convergence in the medium and long run, where price rigidities are not relevant? If yes, what is the quantitative significance of such effects? We build a two-sector, small open economy growth model with a motive to hold money (own-currency financial assets) to answer this question Counterfactual: compare flexible and fixed exchange rate regimes and their impact on the accumulation paths of capital and financial assets Péter Benczúr and István Kónya Nominal growth 6

7 Related literature Two-sector calibrated real models: Fernandez de Cordoba-Kehoe (2), Bems-Hartelius (26) We add nominal assets and look at the effects of the nominal exchange rate Exchange-rate based disinflations (Rebelo and Végh 995) Different objective, the role of money, risk premium Literature on valuation effects (Lane and Milesi-Ferretti 25, Tille 25) Péter Benczúr and István Kónya Nominal growth 7

8 Model Assumptions Two sectors, traded and non traded Sources of growth are exogenous productivity growth and induced capital accumulation Sector-specific capital with investment adjustment costs Money-in-the-utility in local currency, foreign borrowing in foreign currency No helicopter drop Debt-dependent interest rate Péter Benczúr and István Kónya Nominal growth 8

9 Households Households consume, accumulate capital and real money, borrow in foreign currency max t= ( β t log c t + γ log H ) t P t s.t. k j,t = ( δ) k j,t + i j,t bt h bt h + H t H t = w t + R t s t rj,tk k j,t P t j=t,n PI t s t j=t,n ( + φ 2 i j,t k j,t s t c t ) i j,t + T t Péter Benczúr and István Kónya Nominal growth 9

10 Firms Final composite goods ( + φ 2 Intermediate goods i j,t k j,t ( c t = κ ct T ) i j,t = κ I (i T j,t ) λ ( c N t ) λi ( i N j,t ) λ ) λi Y T t Y N t = A T,t k α T T,t l α T t = A N,t k α N N,t ( l t) α N Péter Benczúr and István Kónya Nominal growth

11 The Central Bank The central bank balance sheet: Assets s t bt c s t ht c s t dt c Liabilities H t Focus on case when only b c t is held Explore general case in the future Péter Benczúr and István Kónya Nominal growth

12 The Central Bank (2) Assume CB profit rebated to households: T t = s t bt c + H t H t s tbt c Rt ( c = H t ) Rt c ( = H t s ) t/s t+ Rt c Currency union: seignorage exogenous (to be explored) Péter Benczúr and István Kónya Nominal growth 2

13 Interest Premium We assume foreign borrowing by households is subject to a debt-dependent interest premium Preferred interpretation: real costs of monitoring to avoid default Interest rate d = ] [e β ψ (b/y b/ȳ ) Important: central bank only earns risk-free rate (/β) on foreign bonds! Péter Benczúr and István Kónya Nominal growth 3

14 Money and the Exchange Rate Consolidated current account (b t = b h t + b c t ) b t R t b t = TB t ( R c t R t ) Ht s t+ Money demand H t s t+ = γx t s t+ /s t /R t Péter Benczúr and István Kónya Nominal growth 4

15 Main results The convergence path depends on the exchange rate regime if the interest rate on reserves is different from the interest rate households pay Foreign risk-free bonds and foreign cash When domestic money is only partially backed by foreign reserves, unexpected changes in the (fixed) exchange rate have valuation effects Currency mismatch (To be proved later...) Péter Benczúr and István Kónya Nominal growth 5

16 Mechanism Convergence implies an increase in desired real money balances Under a flexible exchange rate regime, this takes place through an exchange rate appreciation since the nominal money supply is constant Holding money yields a real return due to nominal appreciation Under a fixed exchange rate, money balances accumulate through an increase in reserves without appreciation Holding money is perceived to yield no return by households Thus HH money demand is different across regimes This implies a different path for foreign bonds and the interest rate The interest rate feeds into capital accumulation Péter Benczúr and István Kónya Nominal growth 6

17 Quantitative evaluation We calibrate the model to the CEE region to evaluate the quantitative significance of our mechanism We explore the effects of unexpectedly fixing the exchange rate at the current flexible level We carry out robustness checks Péter Benczúr and István Kónya Nominal growth 7

18 Calibration (in progress) Hungary, Czech Republic, Slovakia, Poland and the Baltics in 27 Sectoral shares in consumption, investment and income shares in sectors are taken from national accounts and are averaged across countries The importance of money in the utility function: Euro area average M2/GDP ratio Interest premium: DSGE estimate (consistent with NFA developments) TFP path and initial conditions A gradually evaporating excess TFP growth both in T and NT Wealth: 27 M2/GDP and NFA data Capital: steady state given TFP Péter Benczúr and István Kónya Nominal growth 8

19 Parameters Parameters Initial Conditions and Steady State β.96 NFA per GDP -.45 δ.6 Money per GDP.7 φ 5 T capital, (% of s. s.).5 α T.42 NT capital (% of s. s.).5 α N.37 Initial TFP in T.6 η.36 Initial TFP in N.8 η I.44 Money/GDP in s.s..7 TFP half-life in T 5 NFA/GDP in s.s. -.3 TFP half-life in N Péter Benczúr and István Kónya Nominal growth 9

20 Convergence: Flexible Regime GDP in Euros.5 T production N production T consumption Consumption N consumption T labor.5 T capital N capital Wage.5 T rental rate.45 N rental rate Relative price of N.3 NFA per GDP Money in Euros Péter Benczúr and István Kónya Nominal growth 2

21 Fixed vs. Flexible Regime T production 2 N production T labor T consumption 2.5 N consumption.5 Relative price of N Capital NFA per GDP Money in Euros Péter Benczúr and István Kónya Nominal growth 2

22 Results Fixing the exchange rate: sizable and persistent differences relative to the flexible case Initial increase in consumption, sectoral reallocation to NT Big impact on portfolio, away from holding money Fixing produces traditional symptoms of overvaluation: consumption boom, switch towards NT But our mechanism does not rely on declining interest premium! Robustness: ψ = {.,.5,.} b /y = {,.45,.9} h/ȳ = {.4,.7, } Péter Benczúr and István Kónya Nominal growth 22

23 Robustness (baseline:, low value: - -, high value: -.-) 3 Consumption.5 Labor in T 2 NFA per GDP Interest premium Consumption Labor in T 3 NFA per GDP Initial NFA Consumption.5 Labor in T.5 NFA per GDP Importance of money Péter Benczúr and István Kónya Nominal growth 23

24 Conclusion Our model has interesting and empirically relevant implications to evaluate the impact of the exchange rate regime on medium and long run convergence Still to explore: the asset composition of the central bank and valuation effects Application: studying the impact of the crisis on small open economies with currency mismatch Péter Benczúr and István Kónya Nominal growth 24