DETERMINANTS OF FIRMS INVESTMENT IN SPAIN: THE ROLE OF POLICY UNCERTAINTY

Size: px

Start display at page:

Download "DETERMINANTS OF FIRMS INVESTMENT IN SPAIN: THE ROLE OF POLICY UNCERTAINTY"

Emil Hancock
5 years ago
Views:

1 DETERMINANTS OF FIRMS INVESTMENT IN SPAIN: THE ROLE OF POLICY UNCERTAINTY Daniel Dejuan and Corinna Ghirelli Bank of Spain European Network for Research on Investment EIB - Luxemburg 9 April 018 DG ECONOMICS, STATISTICS AND RESEARCH

2 Motivation Corporate investment: key for long-run growth and main driver of business cycle Renewed interest on investment s determinants due to low recovery after Great Recession Role of uncertainty on corporate investment debate Wait and see: postpone investment until additional information comes Financial frictions: for financially constrained firms, probability of default, cost of investment

3 Long run obstacles to investment (proportion of firms %)

4 This paper Study the effect of policy uncertainty on firm investment Firm panel Static investment model with firm FE Aggregated time-varying factors explicitly accounted for Contribution: e.g. Julio and Yook (01), Gulen and Ion (016), Baker et al. (016) SME coverage, as opposed to US literature Heterogeneous effects by financial position financial frictions story

5 The uncertainty measure 1. Current political situation (CIS survey). Expectations about future political situation (CIS survey) 3. Political risk, e.g. government stability, quality of institution (PRS group) 4. EPU index (Baker et al., 015) 5. Disagreement about public deficit among experts (FUNCAS) Correlation matrix policy U pol.sit. exp.pol.sit. pol.risk EPU dis.pub.def. policy U 1.00 pol.sit exp.pol.sit pol.risk EPU dis.pub.def

6 The uncertainty measure (monthly variation) EMU constitution 09/11 terrorist attack Spanish gen. elections 1997m1 1998m1 1999m1 000m1 001m1 Irak Invasion 11/03 terrorist attack & Spanish gen. elections Prestige sunk 00m1 003m1 004m1 Bear Sterns bailout & Spanish gen. elections Lehman Brothers bailout 005m1 006m1 007m1 008m1 009m1 Greece bailout Spain bailout Spanish general elections 010m1 011m1 01m1 013m1 014m1 015m1 016m1 017m1 Brexit Spanish gen. elections more

7 The empirical model (I/K ) it = α i + β 1 U t 1 + β X it 1 + γ 3 GVA it 1 + ɛ it , 600, 000 firms more (Central Balance Sheet Data) I/K : gross fixed capital formation over total capital stock X: financial position, ROA, sales growth, size, export, quoted, corporate group Firm FE Business cycle: sector-specific GVA growth rates Tax changes indicator (Gil et al., 017) Narrative Index Standard errors clustered at year-sector level

8 Expected effects of uncertainty Leverage: if financially constrained, U, access to credit probability to default, cost of credit Profitability: No role for financial friction channel Optimal to delay when U and invest when U (?) SME: affected due to information & financial frictions Exporters: less sensitive to domestic uncertainty Quoted: less affected by credit supply shocks Corporate groups: More More affected if contagion effects Less affected if better access to credit or lower cost of debt

9 Heterogeneous effects (1) () (3) (4) (5) (6) (7) policy U (0.005) (0.006) (0.006) (0.005) (0.005) (0.005) (0.006) U x SMEs (0.003) U x debt rate p[50-100] (0.001) U x ROA p[50-100] (0.001) U x export U x quoted U x corp group tax changes post (0.00) 0.03 (0.008) 0.06 (0.004) (0.00) (0.00) (0.003) (0.00) (0.00) (0.00) (0.00) (0.007) (0.007) (0.007) (0.007) (0.007) (0.007) (0.007) Observations R X, sector-specific gross value added (GVA) growth rates included. Firm FE model. Standard errors clustered at the sector year level.

10 Conclusions and policy implications Effect of policy uncertainty on firm investment, On average: policy U by standard deviation, I/K by 4 pp More negative effect for financially constrained firms, small firms (access to credit) Belonging to corporate group seems to improve access to credit for SMEs Financial friction story Policy implications: decision, coherence, clarity, transparency in public policy

11 THANK YOU FOR YOUR ATTENTION! DG ECONOMICS, STATISTICS AND RESEARCH

12 Short run obstacles to investment (net effect %)

13 Counterfactual exercise Counterfactual - policy uncertainty in avg. observed investment rate avg. predicted investment rate with policy U fixed in 006 as from

14 Firm FE model. Standard errors clustered at the sector year level. (1) () (3) (4) (5) (6) (7) ROA (0.003) (0.003) (0.003) (0.004) (0.004) (0.004) (0.004) debt burden (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) debt rate (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) sales growth rate (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) empl. growth rate (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) (0.001) policy U (0.005) (0.005) (0.00) (0.00) (0.006) (0.006) GVAsec (0.000) (0.000) GVAsec (0.000) (0.000) GVAsec (0.001) (0.001) GVAsec (0.001) (0.001) GVAsec (0.001) (0.001) GVAsec (0.001) (0.001) ESI (0.00) (0.00) GDPgrowth (0.001) (0.000) tax changes (0.00) (0.004) (0.00) Time FE yes no no no no no no Observations R

15 Robustness exercise Add time FE with U X (as is Baker et al., 016) β 1 absorbed by time FE Omitted variable bias ruled out Can still identify the interaction Shift interest to channels through which effect of U materializes

16 Heterogeneous effect including time FE (1) () (3) ROA (0.003) (0.003) (0.004) debt rate (0.00) (0.003) (0.00) corp group (0.003) (0.003) (0.003) U x ROA U x debt rate (0.00) (0.003) Observations R X included. Firm FE model. Time FE included. Standard errors clustered at the sector year level.

17 Balanced panel: Heterogeneous effects (1) () (3) (4) (5) (6) (7) policy U (0.007) (0.007) (0.007) (0.007) (0.007) (0.007) (0.007) U x SMEs (0.003) U x debt rate p[50-100] (0.00) U x ROA p[50-100] (0.004) U x export (0.003) U x quoted (0.01) U x corp group (0.005) tax changes (0.006) (0.006) (0.006) (0.006) (0.006) (0.006) (0.006) Observations R Based on sub-sample of firms observed throughout the entire period (balanced panel). X included, sector-specific gross value added (GVA) growth rates included. Firm FE model. Standard errors clustered at the sector year level. more

18 Balanced panel: Heterogeneous effects including time FE (1) () (3) ROA (0.0) (0.03) (0.0) debt rate (0.013) (0.014) (0.013) corp group (0.006) (0.006) (0.006) U x ROA (0.014) U x debt rate (0.005) Observations R Based on sub-sample of firms observed throughout the entire period (balanced panel). X included. Firm FE model. Time FE included. Standard errors clustered at the sector year level.

19 Robustness: placebos (1) () (3) policy U (t-1) *** *** *** (0.0055) (0.005) (0.0053) policy U (t+1) (0.0036) policy U (t+) (0.009) ROA *** *** *** (0.0035) (0.0035) (0.0035) debt rate *** *** *** (0.00) (0.00) (0.00) tax changes *** *** *** (0.005) (0.008) (0.005) debt burden *** *** *** (0.0004) (0.0004) (0.0004) sales growth rate *** *** *** (0.0008) (0.0008) (0.0008) empl. growth rate *** *** *** (0.0008) (0.0008) (0.0008) Observations 3,41,365 3,41,365 3,41,365 R-squared Number of id 618, , ,811 Based on entire sample. X included, sector-specific gross value added (GVA) growth rates included. Firm FE model. Standard errors clustered at the sector year level.

20 Descriptive Statistics Des. stat. firm data: full sample mean sd min max N ROA ,41,365 debt burden ,41,365 debt rate ,41,365 sales growth rate ,41,365 empl. growth rate ,41,365 SMEs ,41,365 export ,41,365 corp group ,41,365 quoted ,41,365 1(Gross Inv.> 0) ,41,365 I/K ,41,365 back

21 Average investment rate

22 The uncertainty measure (yearly) back more m6 1998m6 1999m6 weighted avg Pol.Eco.U Pol.Eco.U 000m6 001m6 00m6 003m6 004m6 005m6 006m6 007m6 008m6 009m6 010m6 011m6 01m6 013m6 014m6 015m m6

23 Heterogeneous effects of group by size All No SME SME (1) () (3) policy U (0.015) (0.05) (0.015) U x corp group (0.013) (0.01) (0.03) corp group (0.017) (0.016) (0.05) debt rate (0.00) (0.016) (0.00) ROA (0.004) (0.05) (0.004) debt burden (0.000) (0.00) (0.000) sales growth rate (0.001) (0.004) (0.001) empl. growth rate Time FE no no no Observations R Firm FE model. Standard errors clustered at the sector year level.

24 The uncertainty measure (monthly) Correlation matrix policy U ESI financial U Vol.index 10 years gov.bond Vol.index IBEX35 policy U 1.00 ESI financial U Vol.index 10 years gov.bond Vol.index IBEX Vol.index 10 years gov.bond: calculated by IESI, based on daily growth rate 10 years gov.bond returns Vol.index IBEX35: monthly average of daily volatility of IBEX35 index reported by Bloomberg back

25 Index of permanent and exogenous tax changes Constructed by Gil et al. (017) for Spain Narrative approach (Romer and Romer, 010) Based on record of legislated tax changes Each change is attributed to year of adoption t Quantified as change in tax revenues collected in t wrt counterfactual Classified as permanent/temporary, exogenous/endogenous All changes in t are expressed as % of GDP of t 1 Back

26 Indicator of tax changes Tax Revenues[t] as % of GDP[t-1] Back

27 Financial vs political uncertainty Correlation matrix (monthly) financial U vol.ibex35 ind.vol.ibex35 ind.vol.exchange ind.vol.brent ind.vol.10y.bond financial U 1.00 vol.ibex ind.vol.ibex ind.vol.exchange ind.vol.brent ind.vol.10y.bond Financial U: High frequency shocks, not persistent

28 Financial vs political U (yearly) Here: yearly data, RHS lagged yearly avg. U in t 1 explains investment in t Policy uncertainty better suited since more persistent m6 1998m6 1999m6 000m6 001m6 00m6 003m6 004m6 005m6 weighted avg Pol.Eco.U Pol.Eco.U 006m6 007m m6 009m m6 011m m6 013m m6 015m6 016m m6 1998m6 1999m6 000m6 001m6 00m6 003m6 004m6 005m6 weighted avg Fin.U Fin.U m6 007m m6 009m m6 011m m6 013m m6 015m6 016m back

29 Cost of debt and % of firms belonging to corporate group median cost of debt year Group=1 Proportion Group=1 Group= prortion of firms (%) cost of debt (1) corp-group (0.001) Large company (0.001) debt burden (0.000) sales growth (0.000) ROA (0.001) Observations R OLS with time FE and sector FE. Standard errors clustered at the sector year level. Back

30 Debate EC & OECD: high uncertainty is all the more damaging for growth as it magnifies the effect of credit constraints and weak balance sheets, forcing banks to rein in credit further and companies to hold back investment (Buti and Padoan, 013) ILO: indecision of policy makers in several countries has led to uncertainty about future conditions and reinforced corporate tendencies to increase cash holdings or pay dividends rather than expand capacity and hire new workers (ILO, 013) back

31 Q&A, Malta, October ECB Question: Could you comment on the possible political uncertainty in Spain and Portugal and if that could derail the process of reform that you ve alluded to so many times today? And is it a worry for the ECB, the fact that some of the parties that may be called to form a government, especially in Portugal, reject the euro? Draghi: Well, the answer is no, I can t comment. But a more gentle way of saying no would be the following: uncertainty, for economics, is bad. It s bad for consumption; it s bad for investment. Political uncertainty, however, is part of democracy. back

32 Table: Correlation matrix pol U fin U ESI GVAsec1 GVAsec GVAsec3 GVAsec4 GVAsec5 GVAsec6 pol U 1.00 fin U ESI GVAsec GVAsec GVAsec GVAsec GVAsec GVAsec The table reports correlations of the time-varying factors used in the estimations. All variables are defined yearly. pol U fin U are the policy and financial uncertainty measures. They are expressed as the log of the yearly weighted average of monthly original indexes. ESI is the yearly weighted average of monthly ESI, as provided by the European Commission. GVA refers to gross value added growth rate at the sectoral level. Sectors from 1 to 6are as follows: Energy, Industry, Construction and real estate, trade and hotel, informatics and communications, Others. more

33 Table: Descriptive Statistics: sub-sample of firms observed between mean sd min max N ROA , debt burden , debt rate , sales growth rate , empl. growth rate , SMEs , export , corp group , quoted , (Net Inv.> 0) , (Gross Inv.> 0) , Gross Inv.Rate , Gross Inv.Rate-Intang.Assets , Gross Inv.Rate-Tang.Assets , back

34 THANK YOU FOR YOUR ATTENTION! DG ECONOMICS, STATISTICS AND RESEARCH

Unconventional Monetary Policy and Bank Lending Relationships

Unconventional Monetary Policy and Bank Lending Relationships Christophe Cahn 1 Anne Duquerroy 1 William Mullins 2 1 Banque de France 2 University of Maryland BdF-BdI Workshop - June 9, 2017 1 / 43 Motivation