Abstract
This paper examines whether weak central bank finances affect inflation by scrutinizing the key rationale for such a relationship: that the absence of Treasury support makes central bank finances relevant for price stability. Specifically, I ask whether central banks which are not likely to enjoy fiscal support when needed experience higher inflation as their financial situation deteriorates. I find this to be true among a large sample of 82 countries between 1998 and 2008. De facto potential fiscal support appears relevant, while de jure fiscal support, which I survey analyzing 82 central bank laws, does not appear to matter. The results also bring forward an explanation for the conflicting results of the previous empirical studies, which neglected this key component.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Reuters (2005)
Francisco de Paula Gutiérrez, President of the Central Bank of Costa Rica (Central Banking, Vol. XV No.4, May 2005, page 82)
Financial strength can be defined as “the extent to which an entity is constrained by its financial situation in pursuing its strategic goals or policies” (Stella 2008).
The database and the corresponding law extracts are available on the author’s webpage.
See Ize (2005) for empirical examples.
Jeanne and Svensson (2007) report examples of capital-adequacy ratio targets for the Bank of Japan, the Bank of India and the Bank of Indonesia. The latter aims for example at a ratio of capital and reserves to total monetary liabilities equal to 10%.
Regarding major central banks, the recapitalization of the European Central Bank on December 2010 is an example of a recent recapitalization episode, although the structure of the ECB makes this case peculiar compared to the ones mentioned here.
With this assertion we deliberately disregard the possibility that the central bank either changes its investment policy or resorts to financial repression to boost profits. The former possibility can provide some leeway only when the current investment policy is not efficient (or not constrained by law): when it is, it cannot be considered as a straightforward solution anymore, which is why I will ignore this possibility as in the connected literature. The latter is likely to face limits and can hardly be considered as a long-run solution (Klüh and Stella, 2008) in contrast with the possibilities we mentioned here.
The choice of focusing on Latin America is not explicitly defended but implicitly justified by the particularities of these countries: central banks in this region experienced severe losses and the issue of central bank independence is presented as an historically controversial issue for these countries.
They take a 3-years moving average for their CBFS variable but consider yearly moves.
Note that initially I also included a square term for CBFS1 (demeaned to avoid collinearity) to take into account potential non-linear effects. As the coefficient for the square term was not significant in almost every case, this specification was not judged worth being reported.
An alternative approach would be to perform subsamples estimates for countries for which \( {\mathrm{I}}_{\mathrm{i}}^{\mathrm{N}}=1 \). The method I use here has the relative benefit of increasing the precision of the estimates given the larger number of observations resulting, which is why it is preferred. Subsamples estimates however led to the same conclusions.
Note that if the fact that country i shares the characteristics N has per se an effect on inflation, this effect is in theory captured by the fixed effects present in Eq. (2).
The OIN account is the residual item after taking into account the asset items (foreign assets, claims on central bank government, claims on other levels of government, claims on financial institutions, claims on the private sector, etc.) and the liability items (reserve money, foreign liabilities, central government deposits, monetary authority securities, etc.) that are explicitly identified.
“Effective equity” refers to the difference between assets and liabilities of the balance sheet valued at market prices (Buiter 2008). Due to different accounting treatments it can differ from the equity figures found in financial reports.
These cases are double-checked, and in some cases, central banks confirmed by email that there was no such mention in the law.
Fiscal space can be defined as “the availability of budgetary room that allows a government to provide resources for a desired purpose without any prejudice to the sustainability of a government’s financial position” (Heller 2005). I follow this definition in the rest of the paper.
More specifically, I take the rating on foreign currency debt, which is available for a larger number of countries in the sample. Note that it is widely correlated with the local currency rating (98% on average).
To minimize the weight of discretionary judgments, for each country I take an average of the ratings provided. For a small number of countries there are no rating provided by any of the rating agency: I chose to generate missing ratings relying on the main studies documenting the determinants of sovereign ratings. In total I have generated ratings for 18% of the observations. See Online Appendix 1 for details.
Here countries with low fiscal space are the ones for which the value for these variables exceeds the third quartile value of my sample. In so far as comparing Debt-to-GDP ratios or Debt-to-Revenue ratios between countries with different fundamentals might be misleading, when using this proxy I put aside advanced economies. Those are indeed likely to be able to run sustainable fiscal policy at significantly higher debt thresholds than developing or emerging economies. The results are not dependent upon this choice.
Using the log(1 + π) instead yields similar results. When doing so, CBFS1 was also taken in log values, and I considered the logarithm of (1 + CBFS1) in order to take the observations for which CBFS1 was negative.
I use the variables I judge the most relevant for this study in order to keep the model parsimonious and to avoid potential multicollinearity issues. Initially I also considered openness and money growth. As the former variable barely entered with the expected sign and in a significant way in the regressions, I didn’t consider it. Money growth was highly correlated with inflation (>0.5) so that I didn’t include it. The inclusion of these variables didn’t change the main result though.
It is computed as the residuals of the regression of real GDP on a trend normalized by the corresponding fitted values.
As CBFS also has had a tendency to decline in the 2000’s (Klüh and Stella 2008), this also allows us to avoid any bias in the estimates.
The economic independence score is computed based on seven criteria. For example, the fact that the central bank credit facilities cannot be automatically used by the government, that the government pays market interest rates if using such a facility, that the use of such a facility is temporary, that it involves limited amounts: all are criteria taken into account by the GMT index which will all lead to a high economic independence score (see Segalatto et al. (2006)).
This makes us exclude from the sample the European countries using the Euro, Saint Kitts and Nevis, the African countries members of the CFA (Colonies Francaises d’Afrique) Franc zone, and the countries belonging to the Common Monetary Area in the South African region (South Africa, Lesotho, Swaziland and Namibia).
I remove the observations which can naturally be considered as outliers and bias the estimates and the conclusions: the hyperinflationary observation for Belarus in 1999, where inflation reached 293%; observations for “Dominican Republic” and “Costa Rica”, as these 2 central banks suffered from unusually severe problems (Stella and Lonnberg, 2008) and have CBFS values in the sample sometimes going below −100%; Serbia is ignored because the country split during the sample period. I chose to keep a panel as balanced as possible in order to have a number of observations roughly similar when splitting the sample, therefore I don’t keep the countries for which only few observations (less than 5) are available.
Similar to Perera et al. (2013) the control variables are lagged one year in order to limit potential endogeneity issues.
As Roodman (2009) explains, the usual Sargan statistic is inconsistent when one suspects nonsphericity in the errors, thus making the Hansen overidentification test superior to the Sargan test in such cases. Also, Roodman (2009) notes that, in the context of System GMM, conventional significance levels of 0.05–0.10 are “liberal” and states that a p-value of 0.25 “should be viewed with concern”. On the other hand, a p-value close to 1 is said to suggest that the number of instruments is too high, making the test unreliable. Accordingly when possible I select the number of lags so that the p-value of the Hansen test lies between 0.3 and 0.9 (threshold arbitrarily fixed).
Neither the independence variable is included since it is time-invariant.
The results are available on request.
The absence of significance could also be due to the inclusion of fixed effects which could capture the effect of independence, or to the reasons mentioned here-after.
See Segalotto et al. (2006) for a comparison of these two indicators.
Separately I also performed regressions including an interaction term between the government surplus / deficit and CBFS1, to see whether the relationship I obtained would reflect a broader context of integrated budget constraint of the central bank and the Treasury (see Del Negro and Sims (2015) for a discussion). Interaction terms were not significant while not affecting the results, suggesting that the effect of CBFS I obtained here is distinct from any potential consolidated budget consideration.
Our sample does not contain any advanced economy with low fiscal space, which is why we focus only on emerging and developing economies. While many specification choices and robustness checks performed in the previous sections were intended to make sure that the effect obtained does not reflect certain country-specific characteristics, it would still be interesting in further research to apply the analysis for periods in which advanced economies have limited fiscal space (with appropriate CBFS indicators) to guarantee that the conclusions of the paper can firmly be transposed to advanced economies. The comparison between emerging and developing countries though still allows us to alleviate the potential concerns that the conclusions would be driven by countries with levels of income at the tail of the distribution.
I compute the turnover index on 14 years (1996–2009), corresponding to the length of two seven-years mandates, from the database of Dreher et al. (2010).
The same holds with a log-specification and with a one-step estimator (available on request).
References
Adler G, Castro P, Tovar CE (2016) Does central bank capital matter for monetary policy? Open Econ Rev 27:183–205
Afonso A, Gomes PM, Rother P (2011) Short and long-run determinants of sovereign debt credit ratings. Int J Financ Econ 16:1–15
Arellano M, Bond SR (1991) Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations. Review of Economic Studies 58:277–297
Benecká S, Holub T, Kadlčáková NL, Kubicová I (2012) Does central bank financial strength matter for inflation? An empirical analysis. In: Czech National Bank Working Papers. Czech National Bank, series 3
Blundell R, Bond S (1998) Initial conditions and moment restrictions in dynamic panel data models. J Econ 87:11–143
Buiter WH (2008) Can central banks go broke? In: CEPR Working Papers. Centre for Economic Policy Research, London, No. 24
Calderón C, Schmidt-Hebbel K (2008) What drives inflation in the world? In: Central Bank of Chile Working Papers. Central Bank of Chile, No. 491
Campillo M, Miron JA (1996) Why does inflation differ across countries? In: NBER Working Paper Series. National Bureau of Economic Research, No. 5540
Cargill TF (2005) Is the bank of Japan's financial structure an obstacle to policy ? IMF Staff Papers, vol 52 International Monetary Fund, Washington, DC, pp. 311–334
Collyns C, Kincaid GR (2003) Managing Financial Crises; Recent Experience and Lessons for Latin America. In: IMF Occasional Papers. International Monetary Fund, Washington, No. 217
Cukierman A, Webb SB, Bilin N (1992) Measuring the independence of central banks and its effect on policy outcomes. World Bank Econ Rev 6:353–398
Del Negro M, Sims CA (2015) When does a central bank’s balance sheet require fiscal support? J Monet Econ 73:1–19
Dreher A, Sturm JE, De Haan J (2010) When is a central bank governor replaced? Evidence based on a new data set. J Macroecon 32:766–781
Dziobek CH, Dalton JW (2005) Central bank losses and experiences in selected countries. In: IMF Working Paper Series International Monetary Fund, Washington, DC, No 05/72
Gariga AC (2016) Central bank independence in the world: a new data set. Int Interact 42:849–868
Hall RE, Reis R (2015) Maintaining central bank solvency under new-style central banking. In: NBER Working Paper Series. National Bureau of Economic Research, No. 21173
Heller PS (2005) Understanding fiscal space. In: IMF Policy Discussion Papers. International Monetary Fund, Washington, DC, No. 05/4
International Monetary Fund (2000) Debt and reserve-related indicators of external vulnerability. Public Information Notice No. 00/37, International Monetary Fund, Washington, DC
International Monetary Fund (2010) Recovery, risk and rebalancing. World Economic Outlook, International Monetary Fund, Washington, DC
Ize A (2005) Capitalizing central banks: a net worth approach IMF Staff Papers, vol 58 International Monetary Fund, Washington, DC, pp. 289–310
Jácome LI, Parrado E (2007) The quest for price stability in Central America and the Dominican Republic. In: IMF Working Papers. International Monetary Fund, Washington, DC, No. 07/54
Jácome LI, Vázquez FF (2008) Any link between legal central bank independence and inflation? Evidence from Latin America and the Caribbean. Eur J Polit Econ 24:788–801
Jeanne O, Svensson LEO (2007) Credible commitment to optimal escape from a liquidity trap: the role of the balance sheet of an independent central bank. Am Econ Rev 97:474–490
Klüh U, Stella P (2008) Central bank financial strength and policy performance: an econometric evaluation. In: IMF Working Papers. International Monetary Fund, Washington, DC, No. 08/17
Loungani P, Swagel P (2001) Sources of inflation in developing countries. In: IMF Working Papers. International Monetary Fund, Washington, DC, No. 01/198
Nickell SJ (1981) Biases in dynamic models with fixed effects. Econometrica 49:1417–1426
Perera A, Ralston D, Wickramanayake J (2013) Central bank financial strength and inflation: is there a robust link? J Financ Stab 9:399–414
Rajan RG, Zingales L (1998) Financial dependence and growth. Am Econ Rev 88:559–586
Reis R (2013) The mystique surrounding the central bank's balance sheet applied to the European crisis. Am Econ Rev 103:135–140
Reuters News (2005) Bundesbank Blows Hole in Eichel’s 2005 Budget. March 15
Roodman D (2009) How to do xtabond2: An introduction to difference and system GMM in Stata. Stata J 9:86–136
Segalotto JF, Arnone M, Laurens BJ (2006) Measures of central bank autonomy: empirical evidence for OECD, developing, and emerging market economies. In: IMF Working Papers. International Monetary Fund, Washington, DC, No. 06/228
Sims CA (2013) Paper money. Am Econ Rev 103:563–584
Stella P (2005) Central bank financial strength, transparency, and policy credibility, IMF Staff Papers, vol 52 International Monetary Fund, Washington, DC, pp. 335–365
Stella P (2008) Central bank financial strength, policy constraints and inflation. In: IMF Working Papers. International Monetary Fund, Washington, DC, No. 08/49
Stella P, Lönnberg A (2008) Issues in central bank finance and independence. In: IMF Working Papers. International Monetary Fund, Washington, DC, No. 08/37
Stock J, Watson M (2006) Heteroskedasticity-robust standard errors for fixed effects panel data regression. Econometrica 76:155–174
Valencia F, Laeven L (2012) Systemic banking crises database: an update. In: IMF Working Papers. International Monetary Fund, Washington, DC, No. 12/163
Williamson DS (2014) The balance sheet and the future of FeD policy. Open Econ Rev 25:163–170
Windmeijer F (2005) A finite sample correction for the variance of linear efficient two-step GMM estimators. J Econ 126:25–51
World Bank Group (2015) Global economic prospects, January 2015: having fiscal space and using it. World Bank Publications. World Bank, Washington, DC
Acknowledgements
The author is grateful to the referees for careful reading of the paper and valuable suggestions and comments. The author is also thankful first and foremost to Massimo Giuliodori, from whom the paper benefited from very useful feedback, but also to Christian Bordes, Peter Stella, Franc Klaassen, Oana Furtuna, Artures Juodis, Naomi Leefmans, Marc Pourroy, Cédric Tille, Belma Colakovic, Christian Stoltenberg, Sophie Brana, Anil Perera, Sona Benecká, Bernd Hayo, Sinéad Shannon, Kenneth De Beckker and Quentin Vandeweyer for their help, comments or insightful discussions they provided me with at different stages of this contribution. I also thank the University of Amsterdam for the exceptionally good conditions they provided me with during the production of this paper.
Author information
Authors and Affiliations
Corresponding author
Electronic supplementary material
ESM 1
(DOCX 29 kb)
Appendix 1
Appendix 1
1.1 Descriptive statistics and correlations for countries categorized as no de facto fiscal support countries
1.2 Appendix 2: Other regression results
In Table 13, columns (1) (2) (3) and (4), I repeat the approach of Perera et al. (2013) to test for a link between CBFS and inflation for countries with a fixed exchange rate regime and countries where the central bank enjoys low independence. That is, I select the countries with fixed and managed-floating regimes based on the IMF de jure exchange rate regime classification, and the countries where central bank independence, as measured by the turnover index,Footnote 37 is lower than the sample average. This leads us to select 32 countries in the first case and 31 in the second case (against 13 and 9 respectively for Perera et al. (2013)). As Table 9 shows, I do not find any robust link in the resulting sample.Footnote 38
Tables 13 and 14 also show the robustness tests results discussed in section 5.
Rights and permissions
About this article
Cite this article
Pinter, J. Does Central Bank Financial Strength Really Matter for Inflation? The Key Role of the Fiscal Support. Open Econ Rev 29, 911–952 (2018). https://doi.org/10.1007/s11079-018-9496-x
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11079-018-9496-x
Keywords
- Central bank financial strength
- Central bank capital
- Central bank balance sheet
- Inflation
- Fiscal space
- Central bank law