Outline MOTIVATION A HISTORICAL TIME-SERIES FOR THE SPANISH BANK-CAPITAL RATIO - PowerPoint Presentation

1.

2. OutlineMOTIVATIONA HISTORICAL TIME-SERIES FOR THE SPANISH BANK-CAPITAL RATIOEMPIRICAL APPROACHTHE DETERMINANTS OF CREDIT GROWTH IN SPAINCONCLUSIONS AND FURTHER WORK 2

3. lessons from the global financial crisisThe Global Financial Crisis (GFC) show us very clearly how damaging a financial crisis can be:Both in terms of foregone output and……costs for the tax payers 3GDP behavior in different countries

4. new macropolicy instrumentsBanks entered into the GFC with a too low level of capitalBased on the notion that credit risk was better measured and managedIt is necessary to move from a pure microprudential view of capital requirements to a macroprudential approachThus, capital requirements need to have a countercyclical componentThis is the aim of the new policy tools introduced by Basel 3 agreementThe countercyclical capital buffer (CCyB) and……other changes in capital, liquidity and systemic banks’ regulation 4

5. What the empirical evidence tells us and what we add to the debateThere is a solid and robust empirical evidence of both the need for countercyclical tools (Jiménez and Saurina (2006)) and its usefulness to reduce credit crunches during recessions (Jiménez et al. (2017))Based on the effects of dynamic provisions in the Spanish banking lending (Saurina and Trucharte 2017)In this paper we expand quite significantly the perspective, using eight episodes of lending boom and bust in Spain (systemic banking crisis);…We use an aggregate perspective but we start our analysis almost 150 years agoIn this long period very different banking environments were present in Spain …this implies to construct a time series of bank capital (and other bank balance sheets items) back to 1880As our approach is more statistical than accountant, this long time series will be, for sure, substantially refined in the futureWe focus on Spain, where countercyclical provisions were first introduced5

6. OUR results in a nutshellWe find for Spain a much higher level of bank capital in the late XIX century, steadily declining until the Spanish Civil War (SCW)This is in line with US and UK experiences, although the level of the capital ratio was much higher in the USIn the last half a century, the capital ratio was quite stable (around 6%)It increased significantly during the GFC, reaching levels not seen after the SCW Our main conclusion is that the level of bank capital has an asymmetric effect in the credit cycle. A high capital ratio in advance of… …a credit boom reduces credit growth;… …while cater for a lower contraction of loans during a credit crunchTherefore, we find support for the usage of macroprudential tools (e.g. CCyB) in quite different economic and banking environments along nearly a century and a half6

7. OutlineMOTIVATIONA HISTORICAL TIME-SERIES FOR THE SPANISH BANK-CAPITAL RATIOEMPIRICAL APPROACHTHE DETERMINANTS OF CREDIT GROWTH IN SPAINCONCLUSIONS AND FURTHER WORK 7

8. Constructing long time series for Bank capital and total assetsWe use a statistical approach to backtrack the aggregate balance sheets of Spanish banksA very simplified version of the balance sheet is considered (we aggregate the different items to have only three in both the assets and liabilities sides)Bank capital includes both capital and reservesTotal assets (or liabilities) are enlarged using the growth rates in the different sources of informationThe three asset and liability items we aggregate are adjusted to sum to total assets and liabilities The departure point is the information available in Chapter 4 of the Banco de España Statistical Bulletin (1962-now)Old Statistical Bulletin (1952-1961)Estadísticas históricas de España volumen II, siglos XIX-XX, by Albert Carreras and Xavier Tafunell (2005) La banca española en la Restauración II, datos para una historia económica, by Anes et al. (1974)Unluckily, there is no information available for the years of the SCW8

9. A historical series of the Spanish bank capital ratioThe Spanish capital ratio, after reaching values close to 20% at the end of the XIX century, it showed a noticeable decline, registering a floor of 3.5% by 1956 The trend changed and became positive until 1990; then, it recorded a significant decline until the GFCThis behavior has been very similar to that of US and UKHowever, the US capital ratio used to be higher at least until the eighties9The bank-capital ratio in US, UK and Spain

10. OutlineMOTIVATIONA HISTORICAL TIME-SERIES FOR THE SPANISH BANK-CAPITAL RATIOEMPIRICAL APPROACHTHE DETERMINANTS OF CREDIT GROWTH IN SPAINCONCLUSIONS AND FURTHER WORK 10

11. Markov-switching REGIME models (MSRM)In order to disentangle the role played by a given macrofinancial variable in systemic banking crisis, a frequently used approach is that of Logit/Probit models. However:It requires a previous identification of the stress episodesUsually the number of stress episodes is reduced (luckily)It does not allow to analyze the depth and the duration of the crisisAssumes as given the identified systemic crisisOther (non-linear) methodology is that of MSRMThe left-hand side variable is continuousIt allows the existence of different states of the nature……where the relation among the variables analyzed can be differentThe states of the nature are endogenously determined and the probabilities of transition from one state to another are assumed to be constant 11

12. The left-hand side variableAccording to European regulation:Systemic risks to financial stability are risks of disruption to the financial system with the potential to have serious consequences for the real economyAs the flow of credit is one of the most important variable through which the financial sector influences the real sector: We consider both total and bank credit to the non-financial private sector as the relevant variable12Credit to the non-financial private sector. Annual growth

13. Is the proposed Methodology adequate?A very simple MSRM for credit growth: No explanatory variableTwo regimes with different meansIt is remarkable how the model captures the financial stress periods identified with narrative techniques13Narrative and MSRM identification of financial stress periods in Spain

14. OutlineMOTIVATIONA HISTORICAL TIME-SERIES FOR THE SPANISH BANK-CAPITAL RATIOEMPIRICAL APPROACHTHE DETERMINANTS OF CREDIT GROWTH IN SPAINCONCLUSIONS AND FURTHER WORK 14

15. Stochastic properties of the relevant variablesGiven the time span we cover, we have serious limitations to construct the most suitable explanatory variables. In particular, we consider the following:Real GDP (CPI deflated), to capture both income and wealth effectsRelative cost of lending is measured with the real interest rate Relative housing price, which is our proxy for the collateral behaviorCredit supply side effects are captured by including our key variable, the capital ratio All these variables, apart from the real interest rate, are integrated of first orderReal credit, real GDP (both in logs) and the capital ratio conform a cointegrating relationThe GDP presents a positive and higher than one elasticity, as it corresponds to an economy where progressively all the population has access to the financial sectorThe capital ratio shows a negative semi-elasticity, capturing supply/costs effects15

16. The traditional linear equation for credit growth16

17. A MSRM for credit growth in Spain17

18. Robustness analysis18

19. OutlineMOTIVATIONA HISTORICAL TIME-SERIES FOR THE SPANISH BANK-CAPITAL RATIOEMPIRICAL APPROACHTHE DETERMINANTS OF CREDIT GROWTH IN SPAINCONCLUSIONS AND FURTHER WORK 19

20. Conclusions and further workIn this paper we try to assess whether bank capital may play also a macroprudential roleThe first conclusion from this exercise is that the equity ratio has declined substantially, from figures around 20% at the end of the XIX century to below 5% by 1950 and around 8% in 2016Besides, the changes in the equity ratio have an asymmetric effect on credit growth In our view, a serious effort should be devoted to improve historical banking sector databases, so that we can ask crucial questions such as: How the implementation of a deposit guarantee scheme may have changed the leverage and the behavior of banks and depositors? What is the role that banking competition plays in financial stability?20

21.

22. Cointegration relationships22

23. Constructing long time series for Bank capital and total assets (cnt’d)Just to show a pair of examples, the growth rates of both total assets and credit in the different sources of information are quite similar in the common period23Annual growth rates in the common periodTotal assetsTotal credit

24. Robustness analysis24

25. A historical series of the Spanish bank capital ratioTotal assets over GDP remained relatively stable since 1880 until 1915 Afterwards, an upward trend was observed, only interrupted by the SCW and the GFCThe equity over GDP ratio has remained more stableAfter the SCW a sustained increase has been observed Thus the capital ratio after reaching values close to 20% at the end of the XIX century, it showed a noticeable decline, registering a floor of 3.5% by 1956 The trend changed and became positive until 1990; then, it recorded a significant decline until the GFC25Total assets and capital over GDPBank-capital ratio

26. An international comparisonThe behavior of the capital ratio in Spain, US and UK has been very similarThey share a common diminishing trend until the 1950sAfterwards, a period of stabilization or slight increase is observed……being followed by a reduction since mid-nineties, and an increase after GFCThe US capital ratio used to be higher at least until the eighties From a quantitative perspective, the Spanish equity ratio is more similar to that of UK26The bank-capital ratio in US, UK and Spain