Financial Development, Volatility, and GrowthPhilippe Aghion George-Marios Angeletos Abhijit Banerjee Kalina ManovaHarvard and NBER MIT and NBER MIT and NBER Harvard and NBERMOTIVATION• early RBC theory dichotomy between long-run growth and business cycle• data (e.g. Ramey and Ramey, 1995) volatility has a negative effect on growthP. Aghion, G.M. Angeletos, A. Banarjee, K. Manova: Financial Development, Volatility, and Growth p. 2Independent variable (1) (2) (3) (4) (5) (6) (7) (8)initial income-0.0019 -0.0129 -0.0158-0.0110 -0.0230 -0.0258(-0.69) (-4.02)*** (-4.76)***(-3.49)*** (-8.00)*** (-8.65)***volatility -0.2465 -0.2796 -0.2286 -0.3205 0.27120.0370 -0.2561 -0.2303(-2.60)*** (-2.63)*** (-2.48)** (-2.91)*** (1.41)(0.22) (-2.08)* (-1.42)pop growth-0.0087 -0.00940.0022 -0.0003(-3.25)*** (-3.39)***(0.92) (-0.12)sec school enrollment 0.0281 0.0201 0.0095 0.0046(2.09)** (1.49) (1.98)* (1.08)government size 0.00004 -0.00011(0.10) (-0.42)inflation 0.0001 -0.0011(1.05) (-2.39)**black market premium -0.0203 -0.0317(-2.28)** (-0.41)trade openness 0.00011 -0.00006(1.88)* (-1.88)*R-squared 0.0904 0.0969 0.3734 0.5445 0.0829 0.4194 0.8397 0.9324N 70 70 69 62 24 24 21 20***,**,*,^ significant at the 1%, 5%, 10% and 11% respectively.Table 1a. Ramey-Ramey revisitedNote:Dependent variable is average growth over the 1960-1995 period. t-statistics in parenthesis. Constant term not shown. 19601995 sample period.Dependent variable: Growth 1960-1995Whole sample OECD countriesMOTIVATION• early RBC theory dichotomy between long-run growth and business cycle• data (e.g. Ramey and Ramey, 1995) volatility has a negative effect on growth• endogenous growth theory (AK, precautionary savings, investment risk) ambiguous effect of volatility on growth, via savings/investmentP. Aghion, G.M. Angeletos, A. Banarjee, K. Manova: Financial Development, Volatility, and Growth p. 3Independent variable (1) (2) (3) (4)initial income-0.0094 -0.0161-0.0123 -0.0258(-3.89)*** (-5.63)***(-4.25)*** (-8.23)***volatility -0.1829 -0.25890.0142 -0.2295(-2.14)** (-2.70)***(0.09) (-1.35)investment/GDP0.1742 0.11590.0662 0.0036(6.47)*** (4.42)***(2.43)** (0.18)pop growth-0.0076-0.0001(-3.16)***(-0.06)sec school enrollment 0.0074 0.0047(0.62) (1.04)government size -0.00013 -0.00010(-0.37) (-0.35)inflation 0.0001 -0.0011(0.80) (-2.02)*black market premium -0.0178 -0.0333(-2.32)** (-0.41)trade openness 0.00010 -0.00006(1.86)* (-1.70)R-squared 0.4472 0.6687 0.5515 0.9326N 70 62 24 20***,**,*,^ significant at the 1%, 5%, 10% and 11% respectively.Table 1b. Ramey-Ramey revisited (controlling for average investment/GDP)Note: Dependent variable is average growth over the 1960-1995 period. All regressors are averages OECDWhole SampleMOTIVATION• early RBC theory dichotomy between long-run growth and business cycle• data (e.g. Ramey and Ramey, 1995) volatility has a negative effect on growth• endogenous growth theory (AK, precautionary savings, investment risk) ambiguous effect of volatility on growth, via savings/investment• data (e.g. Ramey and Ramey, 1995) most of the effect via a different channel, not savings/investmentP. Aghion, G.M. Angeletos, A. Banarjee, K. Manova: Financial Development, Volatility, and Growth p. 4THIS PAPERtransmission channel:cyclical composition of investmentexogenous shocksproductivity-enhancing investmentsTFP, growth, and volatilitydifferential effects depending on credit marketsa theory for the Solow residualP. Aghion, G.M. Angeletos, A. Banarjee, K. Manova: Financial Development, Volatility, and Growth p. 5RESULTS• complete markets  productivity-enhancing investment countercyclical mitigates business cycle likely positive relation between growth and volatility• tight borrowing constraints  productivity-enhancing investment procyclical amplifies business cycle likely negative relation between growth and volatilityP. Aghion, G.M. Angeletos, A. Banarjee, K. Manova: Financial Development, Volatility, and Growth p. 6EMPIRICAL FINDINGS• cross-section and panel 46 countries/OECD 1960-2000/1973-1999• tighter credit constraints  lower growth and more volatility stronger effect of volatility on growth not via total investment higher sensitivity of growth to shocks more countercyclical R&DP. Aghion, G.M. Angeletos, A. Banarjee, K. Manova: Financial Development, Volatility, and Growth p. 7LAYOUT1. Introduction2. Model3. Investment in Capital and R&D4. Growth and Volatility5. Empirical Findings: Cross-Section6. Empirical Findings: PanelP. Aghion, G.M. Angeletos, A. Banarjee, K. Manova: Financial Development, Volatility, and Growth p. 8THE MODEL• two types of investment• type 1: “working capital” short horizon: little time-to-build, low adjustment costs• type 2: “productivity-enhancing investment (R&D, technology adoption, etc.)” long horizon: more time-to-build, high adjustment costs• cyclical variation in cost/return vs  liquidity riskP. Aghion, G.M. Angeletos, A. Banarjee, K. Manova: Financial Development, Volatility, and Growth p. 9PRODUCTIVITY GROWTH• aggregate TFP At(Solow residual)ln At ln Tt ln atTt: level of technologyat: exogenous shock• the shock follows an AR1ln at ln at−1tt N−½2, 2 ∈ 0, 1   0P. Aghion, G.M. Angeletos, A. Banarjee, K. Manova: Financial Development, Volatility, and Growth p. 10INDIVIDUAL ENTREPRENEUR• a mass one of agents (entrepreneurs) is born in each period t• agents leave for two periods¾ period-t agents born ¾ investment choices ¾ capital produces ¾ liquidity shock realized ¾ R&D produces iff liquidity shock met ¾ consume and die Period t+1 Day Period t Day Night • credit markets ˆ investment choices ˆ TFP growth and volatilityP. Aghion, G.M. Angeletos, A. Banarjee, K. Manova: Financial Development, Volatility, and Growth p. 11INDIVIDUAL ENTREPRENEURday of period t• agent i born withWti wTt• trade in “day” credit market and make investment choices• budget constraintKti Zti Bti≤ Wti• equivalentlykti zti bti≤ wkti, zti, btiKtiTt,ZtiTt,BtiTtP. Aghion, G.M. Angeletos, A. Banarjee, K. Manova: Financial Development, Volatility, and Growth p. 12nightofperiodt• capital producesti Atkti• liquidity shock is realizedCti ctiTtwhere ctii.i.d. with c.d.f. Fc• trade in “overnight”