Identifying Constraints to Financial Inclusion and their Impact on GDP and Inequality:

Transcription

1 dentifying Constraints to Financial nclusion and their mpact on GDP and nequality: A Structural Framework for Policy Workshop on Macroeconomic Policy and ncome nequality 8 September 24 dentifying Constraints to Financial nclusion

2 Roadmap Motivation Model Links to the literature ntuition Model description Data and calibration Evaluation of policy options Comparative statistics Welfare analysis Summary and next steps dentifying Constraints to Financial nclusion

3 Motivation () There is a considerable scope for nancial deepening in developing countries: deepening not equal to inclusion. Low rms access to nance. High collateral requirements and interest rate spreads. dentifying Constraints to Financial nclusion

4 Motivation (2) Smaller rms tend to be most credit-constrained, especially in developing countries. dentifying Constraints to Financial nclusion

5 Motivation (3) Constraints to nancial deepening could be country-speci c. dentifying Constraints to Financial nclusion

6 Motivation (4) Empirical evidence on the link between nancial development, growth, and inequality often inconclusive. Regression analysis: may not be suitable for developing countries. channels of transmission and causal mechanisms are hard to pin down. policy evaluation is challenging. Di erent dimensions of nancial inclusion (access, depth, e ciency) have di erential impacts. Policy impact could vary across countries. This paper: sheds light on links between nancial inclusion, GDP, inequality, and welfare through the lens of a GE model. focuses on business start ups and rm access rather than household inclusion. dentifying Constraints to Financial nclusion

7 Model () Links with the literature The model features Heterogeneous agents with respect to their wealth and talent, Occupational choice, Overlapping generations. A growing theoretical literature on the aggregate and distributional impacts of nancial intermediation Occupational choice and nancial frictions Banerjee and Newman (993), Lloyd-Ellis and Bernhardt (2), and Cagetti and Nardi (26). Relation among nancial intermediation, aggregate productivity and income Gine and Townsend (24), Jeong and Townsend (27, 28), Amaral and Quintin (2), Buera et al. (2), Moll (24). dentifying Constraints to Financial nclusion

8 Model (2) Links with the literature We focus on several dimensions of nancial inclusion within a uni ed framework and analyze how they interact. Limited commitment Evans and Jovanovic (989), Holtz-Eakin et al. (994); Banerjee and Duo (25), Jeong and Townsend (27), Buera et al. (2), Buera and Shin (23), Caselli and Gennaioli (23), Midrigan and Xu (24), Moll (24). A xed entry cost Greenwood and Jovanovic (99), Townsend and Ueda (26), D Erasmo and Moscoso Boedo (22). Asymmetric information Townsend (979), Castro et al.(29), Greenwood et al. (2, 23), Cole et al. (22). dentifying Constraints to Financial nclusion

9 Model (3) Links with the literature Unlike studies in which multiple nancial frictions co-exist, we provide normative policy assessments. Moral hazard and limited commitment Clementi and Hopenhayn (26), Albuquerque and Hopenhayn (24). Moral hazard and imperfect information Abraham and Pavoni (25), Doepke and Townsend (26). Adverse selection and limited commitment Martin and Taddei (23), Karaivanov and Townsend (24). Moral hazard, limited commitment and hidden income Kinnan (24). dentifying Constraints to Financial nclusion

10 Model (4) ntuition Greater nancial inclusiveness impacts GDP, inequality, and welfare through 3 channels: Limits waste of resources due to nancial frictions, pushing up GDP. More e cient allocation of funds increases TFP as talented agents increase the scale of production. but, untalented agents could become entrepreneurs, decreasing TFP. in some cases, there could be undesirable impact on inequality and welfare there are policy trade-o s between GDP and inequality. dentifying Constraints to Financial nclusion

11 Model (5) Overview Agents have di erent wealth (b) and talent (z), and choose their occupations between workers and entrepreneurs. Workers supply labor to entrepreneur. Entrepreneurs use labor and capital for production. n equilibrium: Untalented or talented but wealth constrained!worker. Talented with a certain level of wealth!entrepreneur. An economy with two regimes. "Savings only" regime agents cannot borrow but can make a deposit. "Credit" regime agents can borrow and make a deposit but are subject to Fixed entry cost (Greenwood and Jovanovic, 99), Limited commitment (Evans and Jovanovic, 989), Costly state veri cation (Townsend, 979), : dentifying Constraints to Financial nclusion

12 Model (6) ndividuals Each agent lives for 2 periods. First period credit participation, occupational choice, and investment decisions. Second period consumption (c) and bequest (b ) decisions to maximize utility, u(c; b ) = c! b!, such that c + b = W (second period wealth). The optimal bequest rate is!! u(c; b ) is a linear function of W!the agent is risk neutral! max E (u) max E (W ): Each agent has an o spring, with wealth b and talent (z) which is either inherited from parents (with prob. ) or drawn from a stochastic process. dentifying Constraints to Financial nclusion

13 Model (7) Occupational choice Occupational choice between being a worker or an entrepreneur Each worker supplies one unit of labor and earns w when production is successful. The entrepreneur invests in capital and labor, and obtains income through business pro t. The production technology is f (k; l) = z(k l ). Production fails with probability p, in which case the output is zero and only a fraction () of installed capital is recovered. dentifying Constraints to Financial nclusion

14 Model (8) Credit participation decision All agents can make a deposit, but need to pay a xed cost ( ) to borrow. f the agent doesn t pay the cost and can thereby only save savings only regime. f the agent pays the cost and can thereby borrow credit regime. Two steps: n equilibrium, is more likely to exclude poor entrepreneurs from nancial markets as this amounts to a larger fraction of their wealth. First, the agent chooses occupation conditional on the regime she is living in. Second, the agent chooses the underlying regime by comparing the expected incomes that can be obtained in each regime. dentifying Constraints to Financial nclusion

15 Model (9) Savings regime ndividuals in savings only regime cannot borrow n the rst period, the agents wants to maximize expected income given the initial wealth, max expected income max W S ( + r W S = d )b + ( p)w for workers S (b; z) for entrepreneurs S (b; z) = max k;l f( p)[z(k l ) {z wl k + k] + p( )k } {z } if production succeeds + ( + r d )(b k)g subject to k b. {z } wealth not used in production dentifying Constraints to Financial nclusion if production fails

16 Model () Credit regime Agents in the credit regime have access to external credit by paying a participation cost ( ). ( + r W C = d )b + ( p)w for workers C (b; z) for entrepreneurs. The agent chooses to pay only if W C > W S : n order to borrow, agents need to sign a nancial contract! the amount of borrowing (), the value of collateral (), and the face value of the contract (). mplicit lending rate r l =, and leverage ratio = f production fails and the entrepreneur cannot pay!default. dentifying Constraints to Financial nclusion

17 Model () Financial frictions Limited commitment Contract enforcement is imperfect entrepreneur can abscond with a fraction (=) of rented capital. Entrepreneurs do not abscond only if = <! the bank is only willing to lend. Asymmetric information Whether production fails or not is only known to entrepreneur. Banks have a monitoring technology, with a cost proportional to the scale of production () paid by the lender. The bank s optimal veri cation strategy follows Townsend (979), which occurs if the entrepreneur cannot pay the face value of the loan and default. dentifying Constraints to Financial nclusion

18 Model (2) Optimal loan contract Collateral is interest bearing (r d )! = b Entrepreneurs borrow to increase production scale! = k(b; z) Financial sector is perfectly competitive! zero pro t condition pins down (b; k) if production succeeds z } { ( p) + if production fails z } { p min(; ( )k + ( + r d )(b )) = ( + r d )k + pk [ if ( )k + ( + r d )(b ) < ] {z } otherwise loan value {z } expected cost of monitoring dentifying Constraints to Financial nclusion

19 Model (3) Optimal loan contract Entrepreneur of type (b; z) chooses k and l to max pro t C (b; z) = max k;l f ( p)[z(k l ) wl + ( {z )k + ( + r d )(b )] } if production succeeds +p max(; ( )k + ( + r d )(b {z ) )g } if production fails subject to k (b ) {z } credit constraint where is the solution to the bank s zero pro t condition. dentifying Constraints to Financial nclusion

20 Model (4) Occupational choice and access to credit When an agent obtains external credit, the occupation map changes the area of constrained workers shrinks, and that of unconstrained entrepreneurs increases. dentifying Constraints to Financial nclusion

21 Model (5) Competitive equilibrium Given an initial joint probability density distribution H (b; z), a competitive equilibrium consists of allocations fc t (b; z); k t (b; z); l t (b; z)g t=, sequences of joint distributions fh t (b; z)g t= and prices fr d (t); w(t)g t, such that Agent of type (b; z) optimally chooses the underlying regime, occupation, c t (b; z); k t (b; z); l t (b; z) to maximize utility at t ; Capital market clears at all t ; Labor market clears at all t ; fh t (b; z)g t= evolves according to the equilibrium mapping: H t+ (b; z) Z = (z)db fb =bg H t(b; z)dz + ( ) fb =bg H t(b; z)dbdz: z dentifying Constraints to Financial nclusion

22 Data and Calibration () World Bank enterprise surveys (micro data) Provide rm-level cross-section data. Cover a broad range of nancial access measures. World Bank development data platform (macro data) Gross savings rate, non-performing loan, and interest rate spread. Six countries at various stages of economic development Three LCs Uganda in 25, Kenya in 26, and Mozambique in 26. Three EMs Malaysia in 27, Philippines in 28 and Egypt in 27. dentifying Constraints to Financial nclusion

23 Data and Calibration (2) Overview of data Financial deepening in LCs is more constrained across all dimensions, but there is signi cant heterogeneity within the country groups. dentifying Constraints to Financial nclusion

24 Data and Calibration (3) Data, model, and calibrated parameters dentifying Constraints to Financial nclusion

25 Comparative Statistics () Reducing the participation cost.5. Uganda Malaysia GDP..5 TFP ψ nterest rate spread ψ Gini coefficient ψ Percent of firms with credit ψ Non performing loan ratio ψ ψ.5.25 dentifying Constraints to Financial nclusion

26 Comparative Statistics (2) Relaxing collateral constraints.4.3 Uganda Malaysia GDP.4.3 TFP λ nterest rate spread λ Percent of firms with credit λ Gini coefficient λ Non performing loan ratio λ λ dentifying Constraints to Financial nclusion

27 Comparative Statistics (3) ncreasing intermediation e ciency.2. Uganda Malaysia GDP.5. TFP χ nterest rate spread χ Gini coefficient χ Percent of firms with credit χ Non performing loan ratio χ χ.4.2 dentifying Constraints to Financial nclusion

28 Comparative Statistics (4) mpact on GDP and inequality The impact of nancial deepening on GDP and inequality vary with its form and country-speci c characteristics. Relaxing generally o ers the greatest bene ts in terms of GDP, but inequality responds more to lower. dentifying Constraints to Financial nclusion

29 Change in GDP Comparative Statistics (5) nteractions among nancial constraints is relaxed by 2 percent for di erent and (Philippines) dentifying Constraints to Financial nclusion

30 talent Comparative Statistics (6) Welfare analysis The impact of nancial deepening on welfare (Philippines) partial equilibrium wealth general equilibrium dentifying Constraints to Financial nclusion

31 Summary and next steps () We develop a tractable micro-founded GE model with features speci c to developing countries to evaluate nancial deepening policies. Highlight access, depth and e ciency dimensions of nancial deepening. Analyze transmission channels and the impact of di erent forms of inclusion on GDP and inequality. Emphasize how country speci c features play a role through the process of nancial development. A tool for policy analysis Allows identifying the bottleneck factor in the nancial system. Provides quantitative policy evaluations. dentifying Constraints to Financial nclusion

32 Summary and next steps (2) There are several caveats in applying the framework Does not provide guidance on HOW to promote di erent forms of nancial deepening. Does not directly examine issues of household nancial inclusion or mobile banking. Next steps will include: Multi-sector model to study formal/informal sector, structural transformation. Monopolistic banking structure. dentifying Constraints to Financial nclusion