Consumer Credit and Financial Inclusion

Size: px

Start display at page:

Download "Consumer Credit and Financial Inclusion"

Alicia Bates
5 years ago
Views:

1 Consumer Credit and Financial Inclusion Sara G. Castellanos 1 Diego Jiménez 2 Aprajit Mahajan 3 Enrique Seira 4 1 Banco de México 2 Stanford University 3 University of California, Berkeley 4 Instituto Tecnológico Autónomo de México June 28, 216

2 Motivation Results Outline: Large RCT in Mexican credit card market. Study population new to formal credit. Experimental variation in contract terms: interest rates and minimum payments. Estimate causal effects on purchases, repayment, debt and default. Policy and Academic Interest: Concern over default and overindebtedness driven by high interest rates and low minimum payments (Mexican and US Congress). Sheds light on challenges for credit expansion to (relatively) new-to-credit borrowers.... Preliminary, comments welcome! 1/16

3 Motivation Results Study Design and Data Summary Statistics Results Policy Priority: Expanding Credit Increasing access to formal credit important priority for Mexican policymakers. Important part of financial Inclusion (FI) strategies in the National Development Plan. [FI] is not only a matter of finance, it is a matter of social equality as well. (Minister of the Treasury, 214). Credit card is most often first formal loan product. Only 22% of Mexican households had a card in 212 (ENIGH) (compare to US: 77%). Government actively seeks private sector participation. Expanding credit to new borrowers: Smooth consumption, allows investment. Welfare-improving by revealed preference. However, high risk population: asymmetric information problems. But, unsophisticated or time-inconsistent borrowers could borrow too much relative to unbiased benchmark. In Mexico, proposals to regulate interest rates and minimum payments to curb default and excessive debt accumulation. 2/16

4 Motivation Results Study Design and Data Summary Statistics Results Experiment and Questions for Today Large private Mexican bank conducted experiment to understand the effects of contract terms on debt and default. Present results and discuss two issues: 1. How much risk do new borrowers represent? Document high default and card exit. 2. Can contract terms help mitigate debt and default? Effect of variation in interest rates and minimum payments on purchases, payments, debt and default. 3/16

5 Motivation Results Study Design and Data Summary Statistics Results Population and Study Sample Product: Store credit card for clients with limited credit history by one of Mexico s largest banks. Consumer credit for purchases at a large (country-wide) store. Started in 22, 3.3 million clients in 27. Card promotion done in-store. Study Population: Current card holders as of January 27. for 57% this was first credit card of any kind. for 47% this was the first banking product. Relatively new to formal credit of any sort (lowest decile of credit scores). Sampling Design: Stratified population of clients into 9 cells based on Length of card ownership (as of 1/7): 6-11 months; months; 24+ months. Repayment behavior (before 1/7): minimum-payers (<1.5MP); medium payers (>1.5MP but < full); full-payers. Sampled 18, consumers in each stratum to create study sample of 162,. Independent experiment in each stratum. Unequal size strata. Use sampling weights to construct estimators representative of population. 4/16

6 Motivation Results Study Design and Data Summary Statistics Results Study Design and Data Treatment arms: Within stratum, 8 equal-sized treatment arms of 2 clients using four interest rate and two minimum payment combinations (+ control). Interest rate: 15%; 25%; 35%; 45%. Minimum payment: 5%; 1%. Control arm: Interest rate 47% and minimum payments 4% Clients informed of new interest rate and minimum payments in usual monthly statement. Duration of new terms was not specified. Not informed that they were part of a study. Arm assignment ran from March 27 to May 29 at which point experiment ended and clients returned to standard terms. Data: Monthly bank statement data (3/7 5/9, 6/13 6/14). Annual Credit Bureau data (27 213). Match 99% of sample. ENIGH, MxFLS (unmatched) 5/16

7 Motivation Results Study Design and Data Summary Statistics Results Timeline Strata variables recorded January 27 Credit-bureau June 27 End of the experiment May 29 Bank informa on August 21 to September 214 Start of the experiment March 27 Credit-bureau June 28 Credit-bureau June 29 6/16

8 Motivation Results Study Design and Data Summary Statistics Results Summary statistics (1) Start of experiment Credit bureau-supplied information Demographic information Age 39 % Male 53 % Married 63 Monthly Income (Pesos) 13,842 (1) Beginning of the experiment Credit card information (Pesos) Payments 711 (1,473) Purchases 338 (1,23) Debt 1,198 (3,521) Credit limit 7,879 (6,117) Credit score 645 (52) 7/16

9 Motivation Results Study Design and Data Summary Statistics Results Environment: High Rates of Card Exit 33% of control group exits bank during the 26 month experiment (15% annual exit rate). Similar rates for similar populations also in other data Card Exits Exit roughly equal parts client cancellations and bank revocations. Percentage of accounts Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 Cancelled by client Revoked by bank Other reasons Note: Other includes inactive accounts, stolen and lost credit cards 8/16

10 Motivation Results Purchases and Repayment Repayment and Delinquencies Other Findings Estimation Outline Present point estimates and Lee (29) bounds to deal with card exit. Estimate treatment effects (and Lee Bounds) month by month 8 9 Y it = β gtt ig + S is + ɛ it g=1 Graph for each treatment arm g { ˆβ gt} 26 t=1 and Lee Bounds. Set outcomes to zero (e.g. purchases, payments) or last value (debt) post exit. 8 treatment arms 9 strata 26 months = Too many ATEs! Focus today on only two contrasts : Effect of an interest decrease holding minimum payment fixed at 5%: (r = 45%, MP = 5%) vs. (r=15%, MP = 5%). Effect of a minimum payment increase holding interest rate fixed at 45%: (r = 45%, MP = 5%) vs. (r = 45%, MP=1%). Other contrasts corroborate results. s=1 Use sampling weights to be representative of the population from which study sample was chosen. 9/16

11 Motivation Results Purchases and Repayment Repayment and Delinquencies Other Findings Effects on Card Exit Effect of experimental interest-rate and minimum payment interventions small fraction of overall turnover rates. Interest rate change from 45% to 15% decreased exit by 4.5 percentage points over 26 months. Minimum payment change from 5% to 1% increased exit by 4.2 percentage points over 26 months. Implied elasticity of card exit with respect to interest rate and minimum payments are.18. Some variation across strata full payers and longer-term clients had lower turnover than minimum payers and newer clients. (2) (3) Revoked Exit I:15% P:5% -.15* -.45*** (.7) (.9) I:45% P:1%.21**.42*** (.7) (.1) I:45% P:5%.131***.344*** (.5) (.7) R-squared P-value MP P-value R.627 1/16

12 Motivation Results Castellanos cc et al. 11/16 Purchases and Repayment Repayment and Delinquencies Other Findings Effect on Purchases and Repayment Purchases MP = Purchases (23%, MXN $97 ). Why? R = Purchases (15%, MXN $63, LB include ). Repayment Behavior MP = Repayments (9%, MXN $58 ). R = Repayments (1%, MXN $65 ). Purchases Minimum payment Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 Payments Minimum payment Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 Purchases Interest rate Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 Payments Interest rate Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 Purchases by Payment Behavior Payment by Payment Behavior Treatment Estimations

13 Motivation Results Purchases and Repayment Repayment and Delinquencies Other Findings Effect on Debt and Delinquencies Debt: Both R and MP reduce debt MP = Debt (38%, MXN $84, LB includes ). Elasticity.35 R = Debt (28%, MXN $62 ). Elasticity.4 Limited effect on Delinquencies: MP = Delinquency (1pp, LB includes ). Elasticity.3 R = Delinquency (3.3pp). Elasticity.14 Debt Minimum payment Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 C. delinquencies Minimum payment Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 Debt Interest rate Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 C. delinquencies Interest rate Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 Debt by Payment Behavior Delinquencies by Payment Behavior Treatment Estimations Payment bunching by MP Other Delinquency Measure 12/16

14 Motivation Results Purchases and Repayment Repayment and Delinquencies Other Findings Effects on Bank Revenues Use different approaches to computing bank revenues (payments less purchases adjusting for balances held and defaults). Revenues Definition Across different definitions, treatment effects are negative. Implied elasticity of bank net revenues about with respect to interest rates.7. Implied elasticity with respect to minimum payments.15 Arm Revenues I:15% P:5% *** (81.63) I:45% P:1% *** (29.91) I:45% P:5% 186.9*** (163.3) R /16

15 Motivation Results Purchases and Repayment Repayment and Delinquencies Other Findings Summary Extremely high rates of card-exit. 33% of the sample exit the experiment. Exit rates comparable in Credit Bureaus for similar populations. Decreasing interest and increasing minimum payments both rates reduced debt. Elasticity of debt with respect to interest rate.4 Elasticity of debt with respect to minimum payments.35 Elasticity of card exit with respect to interest rate and minimum payments are.18. Elasticity of bank net revenues about with respect to interest rates.7 and with respect to minimum payments.15 Exit caused by contractual variation only small part of overall exit rates. 14/16

16 Motivation Results Purchases and Repayment Repayment and Delinquencies Other Findings Other Findings and To Dos Heterogeneity: Significant heterogeneity by stratum. Negligible exit and zero ATEs for older clients who paid their balances in full pre-experiment. Strongest effects for newer clients who made low monthly payments pre-experiment. Cost of default: Bank revocation associated with a 3 times lower probability of getting a new card ±5 months from date of revocation. Credit score decreases sharply for those with revoked cards (from 62 ten months before to 55 five months after). External effects: No treatment effect on other loans or bills (e.g. phone bills). Payment habit formation: After the experiment, all cardholders were returned to the same interest rate (around 47%) and minimum payment levels (around 4%). Using 211 data to estimate the effects of previous treatment on current debt and purchase behavior (controlling for current debt?). How to reconcile large underlying default rates with insensitivity to relatively large changes in contractual terms. THANKS! 15/16

17 Appendix

18 Sampling weights by months of credit card usage and repayment behavior Cardholder s payment behavior Total Minimum payer Part-balance payer Full-balance payer (1) (2) (3) (4) Months of credit card use 6 to 11 months to 23 months months Total Return to slide

19 Payment as a proportion of debt in the beginning of the experiment Percentage of accounts (%) % 24.5% 55.1% More than 2% Payment / Amount due

20 Level and growth in credit cards by deciles Growth in credit cards by income decile (22-212) Proportion of households with a credit card Growth in CC holdings

21 Appendix Experiment description Bancarization of these clients was done through commercial stores.

22 Long Term Effects: Getting a New Card? Percentage of cardholders (new card) t 5 Months before the account attrits -5 Credit score Credit score Cancelled cards 5 Months before the account attrits Revoked cards -5 Cancelled Revoked 1 5 Months before the account attrits Med

23 Treatment Regressions Return to results (1) (2) (3) (4) (5) (6) (7) Cancellations Revoked Delinquency Delinquency Purchases Debt Payments by client by bank cummulative pct. of months I:15 P:5-4*** -1.5* 63* -63*** -65** -3.3*** -.99* (.7) (.66) (26) (123) (25) (.89) (.43) I:15 P:1-1.6* *** -98*** 18*** -2.5**.15 (.74) (.68) (3) (118) (29) (.89) (.44) I:25 P:5-2.3** -1.8** * -63* -3.2*** -.77 (.72) (.66) (23) (139) (25) (.89) (.43) I:25 P: *** -86*** 92*** * (.74) (.69) (26) (117) (26) (.9) (.45) I:35 P:5-1.7* ** (.73) (.67) (26) (128) (29) (.91) (.45) I:35 P: * 151*** -68*** 99** * (.74) (.7) (28) (125) (32) (.91) (.45) I:45 P:1.91 2** 97*** -84*** 58* ** (.76) (.7) (26) (119) (26) (.91) (.46) Cons (I:45 P:5) 16*** 14*** 415*** 2,117*** 627*** 29*** 11*** (.53) (.49) (17) (98) (19) (.64) (.31) R-squared Observations 143, ,916 87,93 87,93 87,93 144, 144, P-value MP = P-value R = Lee Bounds MP - - [ 65, 287] [ -971, 298] [ 9, 252] - - Lee Bounds R - - [ -161, 14] [ -1,512, -474] [ -243, -17] - - Note: These are cross-sectional regressions where the dependent variable is below the column number. Inverse probability weights are used according to the population. Robust standard errors are shown in parenthesis. Monetary variables are measured in 27 MXN pesos. Lee (29) bounds tightened by strata variables are provided for censored variables because of attrition.

24 Cancellations by client Return to results Minimum-payers Full-payers Cancelled by client Minimum payment Cancelled by client Minimum payment Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 Cancelled by client Interest rate Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 Cancelled by client Interest rate Mar/7 Sep/7 Mar/8 Sep/8 Mar/9

25 Revoked by bank Return to results Minimum-payers Full-payers Revoked by bank Minimum payment Revoked by bank Minimum payment Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 Revoked by bank Interest rate Revoked by bank Interest rate Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 Mar/7 Sep/7 Mar/8 Sep/8 Mar/9

26 Purchases Return to results Minimum-payers Full-payers Purchases Minimum payment Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 Purchases Minimum payment Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 Purchases Interest rate Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 Purchases Interest rate Mar/7 Sep/7 Mar/8 Sep/8 Mar/9

27 Debt Return to results Minimum-payers Full-payers Debt Minimum payment Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 Debt Minimum payment Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 Debt Interest rate Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 Debt Interest rate Mar/7 Sep/7 Mar/8 Sep/8 Mar/9

28 Payment Return to results Minimum-payers Full-payers Payments Minimum payment Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 Payments Minimum payment Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 Payments Interest rate Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 Payments Interest rate Mar/7 Sep/7 Mar/8 Sep/8 Mar/9

29 Delinquencies Return to results Minimum-payers Full-payers C. delinquencies Minimum payment Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 C. delinquencies Minimum payment Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 C. delinquencies Interest rate Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 C. delinquencies Interest rate Mar/7 Sep/7 Mar/8 Sep/8 Mar/9

30 Percentage of delinquent months Return to results This measure has a similar trend to cumulative delinquencies, but different magnitudes. After 26 months of treatment: Increasing MP leads to a 1.36*** pp increase (13%) in the percentage of months delinquent per account. Decreasing R leads to a 1** pp decrease (9%) in the percentage of months delinquent per account. Months delinquent (%) Minimum payment Mar/7 Sep/7 Mar/8 Sep/8 Mar/9 Months delinquent (%) Interest rate Mar/7 Sep/7 Mar/8 Sep/8 Mar/9

31 Payment as a proportion of amount due Fraction of cardholders Beginning of the experiment Payment as a percentage of amount due Fraction of cardholders months after 6 months after I:45% P:5% I:45% P:1% Payment as a percentage of amount due Note: (1) Bins have a width of 2.5 pp each. (2) The rightmost bin of each graph includes those who pay more than 5 pp.

32 Implications for Borrower Welfare (1) Arrears in telephone I:15, P:5 1.1 (6.4) I:15, P:1 6.1 (6.6) I:25, P:5-6.1 (6) I:25, P:1 1.2 (6.5) I:35, P:5-3.1 (6.3) I:35, P:1 2.4 (6.4) I:45, P:1-4.2 (6.4) Cons (I:45, P:5) 71*** (4.4) R-squared.1 Observations 143,916 P-value of IR.656 P-value of MP.5446 Amount in arrears Jul/7 Jan/9 Jul/1 Jan/12 Jul/13 I:15% P:5% I:15% P:1% I:45% P:5% I:45% P:1%

33 Implications for Borrower Welfare (1) (2) (3) (4) All type of loans Only credit cards At least one Total amount At least one Total amount Dependent variable: delinquent loan in arrears delinquent loan in arrears I:15, P: (.66) (536) (.57) (413) I:45, P: (.66) (492) (.57) (384) Cons (I:45, P:5).28*** 9,45***.19*** 6,741*** (.47) (351) (.4) (272) R-squared Observations 143, , , ,916 P-value of IR P-value of MP

34 Credit score (1) (2) (3) (4) (5) (6) (7) (8) (9) (1) (11) Jun/7 Jun/8 Jun/9 Jun/1 Jun/11 Dec/11 Jun/12 Dec/12 Jun/13 Dec/13 Apr/14 I: 15, P: (.68) (1.2) (1.18) (1.26) (1.3) (1.33) (1.33) (1.31) (1.3) (1.31) (1.32) I: 45, P: (.68) (1.) (1.17) (1.25) (1.3) (1.33) (1.33) (1.31) (1.3) (1.32) (1.32) Constant (I:45, P:5) *** 66.61*** *** *** *** *** 635.9*** *** *** *** 639.2*** (.48) (.72) (.84) (.9) (.93) (.95) (.94) (.93) (.91) (.93) (.93) Observations 142, , , , , ,84 132, ,28 13, , ,684 R-squared P-value of IR P-value of MP Return to slide

35 Prediction of NPV of Revenue by Credit Score NPV of Revenue Credit score in June 27 95% CI lpoly smooth kernel = epanechnikov, degree =, bandwidth = 3.39, pwidth = 5.9 Return to slide

36 Credit limit and duration of the card in the market Credit cards from all banks opened between 24 and (card closes before 27 months) Mean initial credit limit for the experiment 3, 6, 9, 12, Credit limit in pesos 95% CI lpoly smooth kernel = epanechnikov, degree = 3, bandwidth = , pwidth = Return to Revenues Return to slide

37 Other Findings: Bank Revenues Consider the identity for month t Due t = Due t 1 + Buy t Pay t + (i/12) Debt t + Fees t where Debt t is the average (over the month) of the daily amount due. Rewrite Pay t Buy t = Due t 1 Due t + (i/12) Debt t + Fees t Consider an agent observed from t = 1 and is in the experiment until T when the card exits or the experiment ends. Then, given a discount rate β T β t (Pay t Buy t) = t=1 T β t (Due t 1 Due t) + t=1 T β t ((i/12) Debt t + Fees t) t=1 LHS is a measure of discounted net revenue accruing to the bank. We begin by analyzing this measure of revenue to the bank.

Bancarizing with Credit Cards: Experimental Evidence on Interest Rates and Minimum Payments Elasticities for New Clients

Banco de México Documentos de Investigación Banco de México Working Papers N 2015-11 Bancarizing with Credit Cards: Experimental Evidence on Interest Rates and Minimum Payments Elasticities for New Clients