Macroprudential policy stance assessment: the case of Croatia

Tihana Škrinjarić*

Tihana Škrinjarić

Affiliation: Bank of England, London, UK

0000-0002-9310-6853

Correspondence
tihana.skrinjaric@bankofengland.co.uk

Article   |   Year:  2024   |   Pages:  421 - 463   |   Volume:  48   |   Issue:  4
Received:  April 9, 2024   |   Accepted:   July 10, 2024   |   Published online:  December 13, 2024

Download citation

https://doi.org/10.3326/pse.48.4.3

Abstract

1 Introduction

2021

2021;

2020

2021

2021

2019

2020

To measure the MP stance, literature² developed the Growth-at-Risk framework (GaR), as it relates current macro-financial conditions in the economy and MP measures to the entire distribution of the future GDP growth. One important aspect that is looked upon here is the so-called tail risk, i.e. what the worst possible outcomes of future real growth are. This is important, as financial stability risks lie in those tail risks, which we call systemic risk. Reducing these risks in the financial system could lead to a lower probability of a future crisis³ , and MP could increase the financial system’s resilience (2016). Thus, utilizing the GaR approach to measure MP stance could help in obtaining information on how this policy affects real growth, and macroprudential instruments could be tailored accordingly. GaR framework can be used to estimate the inter-temporal trade-offs of the application of MP instruments, as we can estimate the “benefits” of limiting the extreme negative outcomes of future growth and compare them to the potential “costs” of limiting mean future growth (2024). Several central banks already started publishing this concept regularly in their financial stability reports (2019; 2022; 2018; 2022; 2019), regular IMF reports (2017) for earliest applications, or IMF (2022a) for latest), ECB reports (2019), and regular risk identification (2021). Besides this specific characteristic of the MP stance assessment, the at-Risk approach also has other benefits, such as better forecasting capabilities compared to linear models, and in generating scenarios, as well as calibrating them in stress testing (2020; 2023). That is why empirical research⁴ that utilizes this framework has significantly increased in the last couple of years. 

The purpose of this paper is to construct a GaR framework and empirically evaluate MP effects and stance for the case of an active country with respect to this policy: Croatia. To evaluate the MP stance, we collected quarterly data on selected macrofinancial variables, and the MP index that captures all of the instruments and tools that have been used since the mid-90s to increase the resilience of the financial system. Then, we employ quantile regression as the methodological apparatus to estimate the GaR model for the Croatian case. There are two reasons why this single-country analysis could be interesting for an international audience. Firstly, GaR can be estimated either on a single-country basis, or in a panel quantile regression setting. The “one size fits all” approach in which countries are merged in a panel setting is not always the best⁵ . Certain specificities characterize individual countries and their experiences over time, and a such information could be lost in a panel setting. Ampudia et al. (2021) list some other drawbacks of panel settings, which include the high diversity in macroprudential measures across countries being truncated into simple indicators. Budnik et al. (2021) comment that panel GaR estimation could be biased if time-invariant country characteristics are omitted from the model. Moreover, research on the importance of structural⁶ differences among countries that affect the GaR results is growing (2022; 2022). Secondly, another reason is the unique experience of Croatia’s macroprudential policy in the last 20 years: even in the pre-GFC period, Croatia was among those countries that had a relatively active macroprudential policy (2016). A lot of measures were employed to tackle credit growth (2020), and higher activity of MP employing different measures started in 2003 (2011). This means that the MP stance assessment of Croatia includes an interesting period in which both tightening and loosening measures were included. Not many countries have such data luxury.

2 Literature review⁷

2018

2019

2019

2021

2.1 Initial contributions

2021

2010

2023

Giglio, Kelly and Pruitt (2015, 2016) examine the predictive power of many systemic risk measures (like Covar, MES, SRISK, etc.) for the case of the USA and selected advanced economies. The main focus was to obtain information on which measures are successful in forecasting future GaR, with in- and out-of-sample comparisons. Probably the most famous papers in this group are those of Adrian, Boyarchenko and Giannone (2016, 2019), who argue that most growth forecasts are point estimates and ignore the heterogeneity of different quantiles of future growth distribution. The authors show that lower quantiles of future US GDP growth have greater volatility when compared to the upper ones, which are fairly stable over time. Deteriorating financial conditions are related to a decrease in future average GDP growth, with low upside risks regardless of today’s financial conditions. This indicates a nonlinear relationship between financial conditions and future GDP growth distribution exists.

2.2 Extending the framework

2018

Other studies emerged afterwards, such as Plagborg-Møller et al. (2020), is an extensive study of future GDP growth distribution forecasting in 13 advanced economies, based on several nonparametric and parametric approaches, and a long list of GDP predictors for monthly and quarterly data from 1975⁸ to 2019. This study covers a wide selection of in- and out-of-sample forecasts and nowcasts. The authors found a great degree of heterogeneity of results, not just between the countries, but between different indicators used within the same category for a given country. Aikman et al. (2019a, 2019b) extend the Adrian, Boyarchenko and Giannone (2019) approach by looking at different variables of medium-term vulnerabilities in the financial system. The authors include credit growth information, house price growth, current account imbalances, etc. Another novelty is that they constructed a measure of banking sector leverage to see how the increase in capital requirements affects bank capital and growth-at-risk. As capital requirements were the main tool to build up the resilience of financial systems after the GFC, using an indicator that captures this could be observed as an indicator of macroprudential policy, whose effectiveness could be evaluated within the GaR setting. Here, the main results indicate that greater capital requirements led to a 0.9 p.p. cumulative improvement of GDP-at-risk over three years. This prompted the authors to make a CCyB (countercyclical capital buffer) simulation as an example of capital requirements being increased before the GFC hit. Not surprisingly, the results showed that such a requirement would offset the GDP-at-risk significantly.

2.3 Introducing the macroprudential policy indicator in the analysis

2014

2011

2016

2018

Two years later, the same authors (2020) published a paper with more details on their previous work. Here, the main results show that both policies reduce left tail risks by not affecting the median growth. This means that in the medium-term neither policy has significant negative effects on the real economy. Furthermore, if both policies are tighter, this can reduce the future tail risk by targeting credit growth, and if the monetary policy is loose, the impact of low interest rates on financial stability could be partially compensated with tighter MP. All of this implies that MP really was effective in the Canadian case. Finally, there is an interesting simulation made at the end of this study, where the authors show how much benefit would have been obtained if a tighter MP had been in place before 2018, leading to a reduction of central risk.

2.4 Recent findings on macroprudential policy effectiveness

2020

2020a

2020b

Other studies that could observe richer results are those of Brandao-Marques et al. (2020), in which the authors utilize the quantile regression approach to propose a cost-benefit approach to macroprudential policy. In a panel setting (period: 1990- 2016), they evaluate the effects of different types of policies on future GDP growth and inflation. The authors found evidence of MP trade-offs regarding lowering mean future growth and increasing the GaR growth (i.e. reducing the future losses). In particular, benefits were the results of BBM (borrower-based measures), whereas CBM (capital-based measures) were found to be better for building the system’s resilience. Cucic et al. (2022) is an empirical case study of Denmark’s GaR in the period 1982-2022. Both GaR and HaR (House price-at-Risk) were examined, and the effectiveness of the BBM and CBM measures was compared. The authors conclude that BBM measures shift the entire growth distribution right, whereas CBM measures have a trade-off between GaR and median growth.

Franta and Gambacorta (2020) goes into more details regarding BBM measures, as the authors apply the GaR approach to a panel dataset of 56 countries in the period 1980 to 2012 to evaluate the LTV (loan to value) and loan loss provisioning effects. The results show that LTV narrows the whole future distribution of the growth, whereas loan loss provisions only move the left tail of the distribution upward. Thus, LTV can be used to reduce the volatility of GDP growth, whereas loan loss provisioning decreases the GDP losses in the event of a crisis. Finally, as a last recent empirical work related to this research, we found Drenkovska and Volčjak (2022). It is a study of the Slovenian GaR case in the period 2003-2020. This paper is actually divided into two sections. In the first part, authors develop a financial stress indicator for the Slovenian case. Then, the second part uses this indicator alongside financial cycle and MP indicators to evaluate the MP stance. However, the authors did not find the MP indicator to be significant in the analysis. An explanation could be found in the utilisation of a single-country approach, with insufficient data being provided for the tail risk estimation part of the framework itself.

3 Challenges with the macroprudential policy index

2018

2020b

3.1 Definition and transformation of the MPI variable

First challenge is the definition of the MPI variable itself. One part of the research that uses the MP variable takes the MPI index (regardless of the form and transformation) and calls this the macroprudential policy stance. Examples include Akinci and Olmstead-Rumsey (2015): “... These cumulative variables sum the dummy variables (tightening net of easing) to get an idea of a country’s ʻmacroprudential policy stanceʼ in a given quarter...”; and Ćehajić and Košak, (2021): “we design our main macroprudential measures by summing all policy changes over time, both tightening and easing. This allows us to capture the overall macroprudential stance in a given country and time period.” Although this is not wrong, such an approach is not in line with the definitions of macroprudential stance as used in the GaR literature. However, it introduces more complexity in comparison to results across studies. Another issue is that some papers do not explicitly describe in which form the MPI indicator enters the analysis (net values, cumulative, etc.).

When collecting data from established databases, one needs to assign numerical outcomes. This is because all measures in such databases are given in a descriptive form, whereas MPI is defined as a simple binary variable. Positive unit value indicates a tightening measure that took place in a given quarter and negative unit value indicates a loosening one. Finally, a zero value is assigned to ambiguous and/ or absent measures (see formula 1). Descriptions on how the researcher can assign these values to different tools can be found in Cerutti, Claessens and Laeven (2017), Budnik and Kleibl (2018), Garcia Revelo, Lucotte and Pradines-Jobet (2020), etc. In a given quarter, we can define a measure  :

Usually, values in (1) are counted for a group or all MP tools. Thus, formula (2) can present the cumulative value of all MP actions. This is a starting point in a lot of research, reflecting a general policy direction. Based on Garcia Revelo, Lucotte and Pradines-Jobet (2020), several other variants of MPI can be defined as follows. The first variant represents quarterly MPI with three values: -1, 0, and 1, based on tightening, loosening, or absence of all measures:

$$$$

(3)

This means that regardless of the overall sum in each quarter being +1 or more, it will be rescaled to +1, and something similar is true for negative values. Thus, this transformation looks only at the information if the macroprudential policy is tightening or loosening, regardless of the number of measures. A second measure is to divide the original mpi_i,t by the number of measures in each quarter. This is suitable for cross-country analyses. Several other specifications on a single-country analysis are found in Ćehajić and Košak (2021) as follows:

$$ $$

(4)

and

$$\stackrel{}{\stackrel{}{{\mathrm{<i><br></i>}}_{}}}\begin{array}{c}\end{array}$$

(5)

As many countries have a lot of zero values in the above MPI specifications, some authors (2015) try to overcome this by using a cumulative MPI index as defined in (2). However, Plagborg-Møller et al. (2020) and McCracken and Ng (2016) comment that it is better to utilize stationary variables if possible. MPI defined in (2) is often not stationary, especially in MP active countries. That is why more research is looking at year-on-year changes in the cumulative index:

$${\mathrm{<i>d\_MPI</i>}}_{\mathrm{<i>i</i>}<i>,</i>\mathrm{<i>t</i>}}={\mathrm{<i>MPI</i>}}_{\mathrm{<i>i</i>}<i>,</i>\mathrm{<i>t</i>}}-{\mathrm{<i>MPI</i>}}_{\mathrm{<i>i</i>}<i>,</i>\mathrm{<i>t</i>}<i>-</i>\mathrm{<i>4</i>}}$$

(6)

as found in Galán (2020a, 2020b), Vandenbussche, Vogel and Detragiache (2015), Cerutti, Claessens and Laeven (2017), Alam et al. (2019). Finally, the ESRB Report (2021) utilized a 20-quarter change of the cumulative MPI. This could be especially problematic, as quantile regressions as methodological apparatus for GaR modelling are estimated for growth up to 16 quarters ahead. This makes it difficult to interpret measures that were put into force 20 quarters ago, and their effects 16 quarters in the future (i.e., 9 years in total). 

3.2 Intensity issues

A couple of papers emerged in the last couple of years that try to adjust the intensity of MPIs. Eller et al. (2020), Vandenbussche, Vogel and Detragiache (2015), and Richter et al. (2018, 2019) have been working on this. Fernandez-Gallardo and Paya (2020) follow Meuleman and Vander Vennet (2020), and in the first step assign a positive value for tightening actions, the opposite for loosening, and zero for ambiguous impact or no measures. For the intensity adjusting part, policy actions that are activated for the first time receive the highest weights, a lower value for changes in the level, even lower for changes in the scope, and maintenance of both level and scope is given the lowest weight. When a measure is deactivated, the cumulative index gets value zero. Galán (2020b) besides the usual analysis (see the literature review section), includes one section of robustness checking, in which the author looks at the intensity of the LTV ratio and its effect on GaR. Since the iMaPP (Integrated Macroprudential Policy) database that author uses in the study includes mean regulatory LTV ratios, the author opted to test its effectiveness during the upswing and downturn of the financial cycle, and obtained results that are consistent to the main ones in the first part of the study: in the expansion phase of the cycle, tightening of LTV limits would improve GaR by approximately 0.5 p.p. Negative effects on the median growth are also significant but negative.

Unfortunately, a consensus on how to solve this important challenge has not yet been found because research states that “we assign a higher weight to policy actions we consider to be more important”, as in Meuleman and Vander Vennet (2020). This is something future research needs to work on, trying to find an objective way to define such adjustments, and some solutions are summarized in Škrinjarić (2024).

3.3 Endogeneity issues

2018

2018

1999

2016

2019

2018

2019

3.3.1 Obtaining non-systematic policy shocks

1999

1972

2018

2022a

Some examples of using this within MP stance assessment are the following. Brandao-Marques et al. (2020) utilized ordered probit regression to obtain the non-systematic MPI shock¹⁰. Previous quarter credit-to-GDP gap, house price gap and previous year cumulative value of MPI itself are used as explanatory variables in the regression. There was a similar approach in Gelos et al. (2022). Galán (2020a) utilizes regression to extract residuals from a model that includes all other regressors that enter the GaR equation. Here, however, all of the variables are assumed to affect MPI in the same quarter. This could be questionable, due to knowledge that in practice, data collection and decision making takes some time. Lagged values of explanatory variables should be considered in explaining macroprudential decisions today. However, the results in this study showed little difference in the results when the macroprudential policy variable was purged from specific effects from macro-financial variables to the results without any “cleaning” of the data. Reasoning could be as already mentioned challenge of regressing the MPI on variables from the same quarter. Duprey and Ueberfeldt (2020) also have an interesting approach, due to the MPI variable taking values -1, 0 and 1. The authors opted to use propensity score matching, such that the probability of MPI variable taking positive or negative values depends on the lags of other variables that are considered in the GaR part of the analysis.

3.3.2 Considering dates of announcements and enforcements of measures

2019

2021

2020

3.3.3 Lagging values in models

2017

2022

2022

Furthermore, there are also some papers that lag both the MPI indicator and the rest of the macro-financials that are used in the GaR setting, such as Eller et al. (2020). Authors decide on the lag selection based on the BIC (Bayes information criterion). This is a purely statistical approach to the definition of the best model and should be complemented with MP knowledge from practice. It is still not completely clear which of these approaches are correct, as sometimes they do the opposite things (lagging MPI versus lagging variables that affect MPI), which leads to different economic interpretations, alongside having econometric consequences. As endogeneity has been examined for monetary and fiscal policy for some time now, future work that focuses on macroprudential policy should try to compare these approaches to see what should done next.

3.4 Different sources of MPI data issues

2018

2022a

4 Methodology and data description

4.1 Quantile regression¹³

where $y_t$ is the dependent variable, $\theta$ is the quantile, betas are parameters that need to be estimated at each quantile, $x_{t,k}$ are conditional variables, ${\epsilon }_t$ is the error term. To estimate model (1), for every quantile $Q_{\theta }(y|X)$ , $0<\theta <1$ , a minimization problem is solved:

$$\underset{{\beta }_k}{\arg \min }\sum _{t:y_t\ge {\hat{y}}_t}^{}\underset{}{}\theta |y_t-{\beta }_0\mathrm{(\theta )}-\sum _{\mathrm{k=1}}^K{\mathrm{}}^{}{x}_{t,k}{\beta }_k\mathrm{(\theta )}|+\sum _{t:y_t<{\hat{y}}_t}^{}(1-\theta )|y_t-{\beta }_0(\theta )-\sum _{k=1}^K{x}_{t,k}{\beta }_k(\theta )|$$

(8)

where $\hat{y}$ is the estimated value of $y$ . For the case of evaluating MP effects on future real growth $y$ , the real growth is defined $t_{+}$ quarters ahead:

$$y_{t+h}=100\%\times (\frac{{\mathrm{r\_GDP}}_{t+h}}{{\mathrm{r\_GDP}}_t}-1)/\frac{h}{4}$$

(9)

where usually h = 1, ... , 16. A basic specification of a QR model that describes y_t+h could be the following:

$$y_{t+h}(\theta )={\beta }_0(\theta )+{\beta }_1(\theta )\mathrm{MPI}{\mathrm{_t}}_{}+{\beta }_2(\theta )y_t+{\beta }_3(\theta ){\mathrm{<i>Stress</i>}}_t+{\beta }_4(\theta ){\mathrm{FV}}_t+{\epsilon }_t(\theta ),$$ $$\theta =0.05,\dots ,0.95$$

(10)

where MPI is the macroprudential policy indicator, Stress denotes indicator of financial stress, and FV is a financial vulnerabilities variable. Mentioned variables are an example of which variables are usually used in this modelling. In other words, the GaR model in (10) usually includes the autoregressive term y, the MPI indicator in some form as commented on in section 3.1., and then a variable that corresponds to financial conditions, stress or systemic risk indicators, with a final variable of financial vulnerabilities or financial cycle indicator. This work will look into financial stress and several financial cycle indicators in the empirical section. To evaluate the values of coefficients at different forecast horizons h, the dependent variable on the left-hand side of equation (10) is changed, such that the h quarter growth from equation (9) is regressed on the same set of variables in equation (10). This is the local projection approach, as often used in macroeconomic literature and the plotted reaction of GDP growth to changes in explanatory variables is considered as the impulse response (2023).

Goodness of fit of a QR model can be measured with pseudo-R squared, evaluated at each quantile θ:

$${R_{\theta }}^2=1-\frac{{\mathrm{<i>RASW</i>}}_{\theta }}{{\mathrm{<i>TASW</i>}}_{\theta }}$$

(11)

where RASW_θ is the residual absolute sum of weighted deviations of real values to the estimated ones, and TASW_θ is the total absolute sum of weighted deviations.

Rationales on why the MP effectiveness changes across the growth distribution and alongside different coefficients are in place both for financial conditions/stress and financial cycle indicators is as follows. The early warning methodology literature (2018) on which variables forecast financial crises has shown that financial cycle indicators have good predictive power of not only financial crises, but also subsequent economic downturns. This means that the OLS estimates cannot capture those tail risks. If one used the threshold regression approach, the usual assumptions that hold for OLS should also hold here, which would not capture the heteroskedasticity in the data that is present, alongside other assumptions regarding the error term. Moreover, by applying threshold regression, one would need to sub select the dataset into more regions for which the model is estimated above and below a certain threshold, which would result in an insufficient number of datapoints for the tail risk. Moreover, one would also need to test for more than one threshold, which is not feasible for such a short time series. Furthermore, if one opted for time varying parameters, same assumptions on the error term need to be imposed as for OLS, which is again not satisfied in practice. Quantile regression does not rely on any assumption on the distribution of the datasets that are used, or about the error term. Next, the seminal contribution of Adrian, Boyarchenko and Giannone (2019) has shown that there is almost no effect on mean growth when the usual set of variables that are used is complemented with a financial stress indicator. On the opposite side, there is a significant negative effect on the downside growth risk. This was empirically tested on other economies in the following research, thus it made a stepping stone by which to move from linear models to the quantile regression approach. From the economic point of view, the reason on why the MPI would have different effects on GDP growth at different parts of the distribution is that if done properly, increases of MPI value (which means that the policy is tightening) in good times should increase the resilience of the financial system, which would prepare it to deal with a future crisis better by avoiding huge losses. However, increasing the resilience does come at a cost, by, e.g. banks having greater capital requirements would perhaps decide to reduce lending in some capacity, which would decrease the investment in the economy and thus, the mean growth would be reduced as a consequence.

4.2 Fitting the conditional distribution of estimated growth

2003

by matching the quantiles of the skewed t-distribution to the empirical quantiles obtained from the estimation. The empirical quantiles are usually the 5^th, 25^th, 75^th and 95^th. Some exceptions can be made to the 10^th and 90^th, when dealing with fewer data. Although the QR model obtains more estimated percentiles, Adrian, Boyarchenko and Giannone (2019) opt to have fewer quantiles for (13) to avoid over-parametrization. Another approach is found in Lloyd, Manuel and Panchev (2022) and Mitchell, Poon and Zhu (2021), where a non-parametric approach is used: conditional quantiles are mapped to conditional density, and interpolations across adjacent quantiles are made to smooth out the density. These papers use a measure of model accuracy, the unconditional coverage (UC) test of Kupiec (1995), usually called a back-testing technique, to evaluate the quality of the model. The UC test null hypothesis assumes that, on average, the conditional quantile is a correct coverage of the selected percentile of the forecasted distribution.

4.3 Distance to tail concept

2021

2022

4.4 Data description and stylized facts

2023

2015

2016

Another relevant variable in the GaR approach is financial vulnerability. This variable is also problematic as well, as different authors utilize a wide range of variables that capture credit dynamics, house price dynamics, credit institution vulnerabilities, etc. ESRB (2021) uses d-SRI (domestic systemic risk indicator) as it is based on panel estimation and is more comparable across countries. Individual studies focusing on a single-country analysis often substitute this indicator for one better suited for that country. This study will do the same. Besides the usual¹⁵ credit dynamics variables and d-SRI, this study observes ICSR (Indicator of Cyclical Systemic Risks) as a Croatian version¹⁶ of the composite indicator.

Table 1 gives a brief overview of financial vulnerability measures tested in this study, with graph 5 showing their dynamics (2022; 2022b, 2023c). Dynamics in graph 5 shows that the credit growth was substantial in the 2000s due to financial deepening. Value of composite indicators of cyclical systemic risks increased during the 2000s, reflecting not only the rising credit dynamics but other relevant categories of financial vulnerabilities, such as house price dynamics, external imbalances, private sector debt burden, mispricing of risk, etc. (2023a). Indicators reached their maximal values in 2007 and dropped fast when the crisis hit. The prolonged recession lasted for a few years, and in 2017 a mild recovery started. Finally, something to keep in mind is the problem of the non-stationarity of the data. White, Kim and Manganelli (2015a, 2015b) assume that data for this analysis are stationary. The best option would be to have all variables transformed in such a way that they are stationary, as the predictability of time series is mostly based on the stationarity of the series itself. The d-SRI indicator is the variable commonly used, but it does not satisfy this assumption. All of the specifications in table 1 and graph 5 will be tested in section 5.2.1. to find which variable definition is the best in terms of model performance.

The financial stress indicator for the Croatian case is somewhat problematic, as the ECB (2023) version has a very different dynamic compared to the one that the CNB developed (the period for this variable is Q1 1990 to 2Q 2022). Graph 6 contrasts the two indicators, where it is seen that during the GFC and the sovereign bond crises, the ECB version does not capture the stress, whereas the COVID-19 crisis and war in Ukraine are not much reflected in the CNB version. CLIFS is constructed based on three markets: equity, bond and foreign exchange market (see Duprey et al., 2015), whereas HIFS is the tweaked version of CISS (2012), in which some minor things are changed based on data unavailability, and includes all five markets (besides the aforementioned ones in HIFS, money and bank markets).

5 Empirical application: Croatian case

5.1 Trying to reduce endogeneity of macroprudential policy index

As previously remarked, the macroprudential policy cannot immediately react to macro-financial surroundings due to data lags, legislation bounds, etc. That is why we compare several model specifications to obtain the non-systematic policy shock in the first step, in which lagged values of real growth, financial stress (HIFS, the HR version), and financial vulnerabilities (y-o-y change of credit-to-GDP ratio) are used. We are interested in talking about causal effects, and since there exists a bulk of literature on monetary and fiscal policy that takes this approach in extracting policy shocks to talk about these effects, currently we opted to take this approach (2016).

5.2 Quantile regression results

– Real GDP growth, forecasting horizons h = 4 and 12 to contrast short- and medium-term results in models. 

– Residuals from models (M₃) in table 2 for MPI and . 

– All nine financial vulnerabilities indicators from graph 5. 

– Original values of MPI and for models where lagged variables are included.

5.2.1 Selecting best variables for each indicator category

5.2.2 Selected models’ results

– Model (1): y-o-y change of credit-to-GDP ratio, HIFS, residuals of MPI 

– Model (2): 2-year change of credit-to-GDP ratio, HIFS, residuals of MPI 

– Model (3): ICSR, HIFS, residuals of MPI.

When comparing the effects of MPI variables in models (1) to (3), graph 9¹⁹ shows the estimated coefficients for h = 4 for the QR case (dotted curve), which is contrasted to the OLS results (red dashed line). In all three cases, the MPI QR estimates differ over quantiles and are different compared to the OLS lines. At first glance, the effects on the lower tail of the growth distribution are positive and greater than the median (central) value. This is in line with previous research showing that tighter macroprudential policy positively affects the future lower tail of GDP growth distribution (2020a, 2020b; 2020). However, positive, albeit almost nonsignificant, results regarding the effect on the median could also be explained. When times are good, in terms of economic growth and the upward phase of the financial cycle, imposing higher reasonable macroprudential requirements cannot hurt future average growth, especially when credit institutions have fairly high own voluntary buffers.

Moreover, beta values for the median growth case are constant over all observed horizons, around 1%, but not significant. This is in line with Suarez (2021), where MP stance and effects on future growth are explained from a theoretical point of view: if no effects on median growth are found, this could indicate that there are still some non-linear MP effects on growth that are not captured in the GaR setting. Another explanation is that there exists a natural limit to some MP tools that being applied in practice did not have negative effects on economic growth. As the Croatian macroprudential policy is fairly active, the latter could be true. The Croatian banking system has been one of the highest capitalized in the EU since the capital requirements were introduced after the GFC. Thus, this affects the results as well, due to the frequency of increases made over time, and this is in line with Aikman et al. (2018) who found positive effects of higher bank capitalization on GaR, with no significant reduction of median growth. Although not significant, the beta coefficients for the 10^th percentile start with the highest value for h = 4, and for each subsequent horizon decline. When compared to previous related literature (see literature review section), the signs of the estimated parameters are in line with related studies. However, the insignificance of the results could be also explained by the information that Croatia does not have borrower-based measures, which were found to be more effective in this analysis in previous related papers (2022).

5.2.3 Distribution fitting results

5.2.4 Macroprudential stance measure (distance to tail)

6 Discussion and conclusion

2021

2021

The inconclusive results of this study are in line with the comments of Reichlin, Ricco and Hasenzagl (2020), who agree that the relationship between financial and real variables is difficult to model. This is proven in this paper, where it is shown that the results vary with respect to variable definition, and model specification. One conclusion could be entirely different, if some of the changes are made. Such findings indicate that more work needs to be done before fully operationalizing the GaR approach to evaluate the MP stance on a regular basis. This is one of many papers that find GaR modelling challenging. This means that so far potential usage of this framework to tailor MP tools could be questionable. Alessandri and Di Cesare (2021) warn about the empirical problems and that MP instrument calibration based on such an approach should be very cautious. From this, we conclude that so far, MP tools should not be tailored according to the results obtained in this study. The initial results obtained from this study need to be taken with some caution. Results, although still insignificant, show that the effect of tighter macroprudential policy is positive on the lower tail of the future GDP growth, without effects on the median growth when focusing on a one-year-ahead forecast. This means that the distribution is reshaped to lower the magnitude of downside risks. Tightening in normal times does not disturb future average growth, but it could have greater positive effects on reducing the downside risks when they materialize.

When comparing the approach and results to the ESRB (2021) report, one needs to take into consideration a couple of things. Although it seems that the results in the report are stable and more usable than what was obtained here, there are caveats and challenges found in the ESRB approach. In the mentioned report, the transformation of the MPI indicator is rather questionable. All MP tools are converted into +1 or -1, with the final MPI variable calculated as a cumulative value of all tools that were observed over time for a given country. Then, a 5-year change is the basis for the transformation of the cumulative indicator. This 5-year change is used as the main policy indicator in the GaR model where the forecasts are made up until 4 years ahead. This would mean that the effects of macroprudential policy could take place over a 9-year period, which is too long. Furthermore, a 5-year change has a long memory and often is not stationary for countries like Croatia. The second challenge is the failure to solve the endogeneity issue of the policy variable. So, the results in the mentioned report are not comparable to those obtained here.

Some shortcomings of the approach taken in this research include the single equation approach. Future GDP growth is the only variable in this setting that is assumed to be endogenous. However, the interaction between the financial system and the real economy is two-sided. There has been some work on this that started emerging in the recent period: the ECB has been working on the quantile vector autoregression approach to deal with endogeneity of macro-financial variables and the feedback loops between the real and financial sector (2020). In such a framework, one does not need to have a two-step procedure of purging the MP variable from macroeconomic shocks. Rather, the nature of vector autoregressive models can capture all interactions in one go.

Some authors warn that the empirical research relies on quantile regressions too heavily and found that GARCH (generalized autoregressive conditional heteroskedasticity) models outperform the QR one (2021). Thus, future work could consider other methodological directions, like the MIDAS-QR (mixed data sampling), where higher frequency data could be used for forecasting purposes (2022). Some authors are starting to focus on the DSGE (dynamic stochastic general equilibrium) modelling approach (2018). However, others criticize this framework for not capturing tail risks (2016), so an opportunity may exist to extend DSGE to GaR analysis. It is expected that the GaR framework will become more prevalent in climate change analysis. Bayoumi, Quayyum and Das (2021) and Kiley (2021) already provide an introduction. As climate disasters are becoming more frequent, it would not be surprising to see more and more applications capable of visualising the effects on financial stability.

Notes

2017

2023d

2024

2009

2012

2013

2013

2013

2020

2012

2012

2011

2021

2019

2020

2020

2020

2021

2022

2020

2021

2022

2022a

2023a

2024

2015

2019

2018

2005

2013

1978

2022

2022a

2023b

2015

2020

^* The author states that the views expressed in this paper do not represent the views of the Bank of England. This paper was written at time the author was employed at Croatian National Bank. The author would like to thank two reviewers for their helpful comments and advice.

¹ However, there is still no consensus on this, as found in Arslan and Upper (2017).

² Besides the papers mentioned below, see, e.g. Škrinjarić (2023d, 2024).

³ Also, financial crises are costly. Reinhart and Rogoff (2009) estimate that crisis episodes are related to significant increases in government spending, as government debt increases on average by 86% for three years following a banking crisis. Laeven and Valencia (2012) estimated that the cumulative cost of banking crises is about 23% of GDP during the first four years of their duration, and fiscal costs amount to about 6.8% of GDP (Laeven and Valencia, 2013). Jordá, Schularick and Taylor (2013) found that financial crises are costlier than other recessions, as after five years, the real GDP per capita is lower by 5% compared to other “normal” recessions. Recoveries from financial crises are slower when compared to other types of crises, as found in Kannan, Scott and Terrones (2013). Other relevant findings about the costs of financial crises can be found in Koh et al. (2020), Jordá, Schularick and Taylor (2012), Claessens, Kose and Terrones (2012), and Papell and Prudan (2011). To avoid the costs caused by financial crises, systemic risk should be reduced, in which endeavour macroprudential policy plays an important role.

⁴ Other interesting and important variables have been analysed, such as inflation-at-risk (López-Salido and Loria, 2021), bank capital-at-risk (Lang and Forletta, 2019; 2020), house-price-at-risk (Deghi et al., 2020), unemployment (Adams et al., 2020), or capital flows (Eguren-Martin et al., 2021; Gelos et al., 2022).

⁵ E.g., Plagborg-Møller et al. (2020) found significant country heterogeneity in their results. After a battery of forecasts and estimations made, the authors found a couple of significant mean growth predictors, and fewer still for the volatility of growth, alongside different signs of results, and great cross-country heterogeneity in the results, which prompted the authors to conclude that some caution needs to be taken when one tries to build a theoretical model on empirical results.

⁶ As an example, a credit-to-GNI (Gross National Income) gap and y-o-y GNI growth rates are used for the case of Ireland instead of GDP (Gross Domestic Product), as GNI is a better representation for this case, as well as ICSI (Irish Composite Stress Index) instead of CLIFS (Country-Level Index of Financial Stress) in O’Brien and Wosser (2021).

⁷ Besides the works that are examined below, it is worth mentioning other preceding research that links financial conditions and financial vulnerabilities to the real economy. A comprehensive overview is given in Boyarchenko, Favara and Schularick (2022) and Škrinjarić (2022a; 2023a).

⁸ For the US case, and 1980 for other countries.

⁹ As explained in Škrinjarić (2024), another example of different effects could be that when comparing the countercyclical buffer (CCyB) actions, one country could immediately introduce a value of 2% to take place in a given banking system. This country would get +1 value in the quarter when this action is taken. If the country does not change CCyB in the next few quarters, the value remains at +1. On the other hand, if another country decides to introduce 0.5% CCyB value, and in each subsequent quarter increase it by 0.5 p.p., it will get positive unit value every quarter and could have a cumulative indicator greater than the first country.

¹⁰ As it takes values {-2, -1, 0, 1, 2}.

¹¹ Others include the ESRB database for EU countries: [CrossRef], or Cerutti, Claessens and Laeven (2015). Some other sources are listed and commented in Alam et al. (2019), in appendix I, table 4.

¹² This number comes from counting the total number of all tools in the Excel sheet provided in Budnik and Kleibl (2018).

¹³ Introduction to quantile regression, alongside advantages to other approaches, such as being robust to outliers, heteroskedasticity, non-normality, etc., can be found in Koenker (2005), Davino, Furno and Vistocco (2013), or Koenker and Bassett (1978).

¹⁴ GDP deflator was used to deflate the original nominal GDP series.

¹⁵ See Škrinjarić and Bukovšak (2022) for Croatia’s best individual credit dynamics indicators.

¹⁶ See Škrinjarić (2022a; 2023b) for the composite indicator for Croatia.

¹⁷ SIC values resulted in the same ordering. As these are just ordinary regressions, the idea is to see the tradeoff between the explanatory power of the model versus the number of parameters included in the model. Information criteria give us this information.

¹⁸ I.e., we compare four models M_i, where i stands for how many lags of all variables on the right-hand side (RHS) of the ordered probit equation symbol were included. The explanatory variables included lagged value of the real growth itself, as it is usually put in GaR modelling, and the other variables included were: HIFS and YoY change of the credit-to-GDP ratio. E.g., M_3 means that all variables on the RHS were included with lags 1, 2 and 3 to regress the MPI dynamics on.

¹⁹ Although we are mostly interested in comparing the median and the GaR value, i.e. calculating distance to tail, we are showing all quantiles in graph 9 to get a better picture on the stability of the beta coefficients.

²⁰ We use MPI “shocks” from a variable that was defined with values being equal to +1 or -1 (before purging out effects of other variables), and as we are not using moving sums as some studies do, there is no reason to believe that a measure that takes +1 value in, e.g. 1Q 2019 should have effects 12Q ahead, as this is way overstretching the effects that macroprudential policy has. Using moving sums or similar transformations would include a greater autocorrelation, i.e. memory of the variable and thus it would make sense to look at a longer time horizon. But to observe effects 3 years in future is really overstretching and we don’t hear central banks saying that the macroprudential measures have significant effects on the GDP growth or in general the real economy so long in the future.

²¹ Here, we depict the median value of DTT as just a statistical value. It does not replace the theoretical “optimal” DTT value.

²² Some initial steps have been done by Vandenbussche, Vogel and Detragiache (2015) and followed by Eller et al. (2020).

Disclosure statement

References

1. Acemoglu, D., Ozdaglar, A. and Tahbaz-Salehi, A., 2015. Microeconomic origins of macroeconomic tail risks. NBER Working Paper, No. 20865 [CrossRef]


2. Adams, P. A. [et al.], 2020. Forecasting Macroeconomic Risks. FRBNY Staff Reports, No. 914 [CrossRef]


3. Adrian, T., Boyarchenko, N. and Giannone, D., 2016. Vulnerable growth. FRBNY Staff Report, No. 794.


4. Adrian, T., Boyarchenko, N. and Giannone, D., 2019. Vulnerable growth. American Economic Review, 109(4), pp. 1263-1289 [CrossRef]


5. Aikman, D. [et al.], 2018. Measuring risks to UK financial stability. BoE Staff Working Paper, No. 738.


6. Aikman, D. [et al.], 2019a. Credit, capital and crisis: a GDP-at-risk approach. Staff Working Paper, No. 824.


7. Aikman, D. [et al.], 2019b. How do financial vulnerabilities and bank resilience affect medium term macroeconomic tail risk? 


8. Akinci, O. and Olmstead-Rumsey, J., 2015. How Effective are Macroprudential Policies? An Empirical Investigation. International Finance Discussion Papers, No. 1136 [CrossRef]


9. Akinci, O. and Olmstead-Rumsey, J., 2018. How effective are macroprudential policies? An empirical investigation. Journal of Financial Intermediation, 33, pp. 33-57 [CrossRef]


10. Alam, Z. [et al.], 2019. Digging Deeper – Evidence on the Effects of Macroprudential Policies from a New Database. IMF Working Paper, No. 2019/066. 


11. Alessandri, P. and Di Cesare, A., 2021. Growth-at-risk in Italy during the covid-19 pandemic. Notes on Financial Stability and Supervision, No. 24.


12. Ampudia, M. [et al.], 2021. On the effectiveness of macroprudential policy. ECB Discussion Paper, No. 2559 [CrossRef]


13. Arslan, Y. and Upper, C., 2017. Macroprudential frameworks: implementation and effectiveness. BIS Papers, No. 94.


14. Azzalini, A. and Capitanio, A., 2003. Distributions Generated by Perturbation of Symmetry with Emphasis on a Multivariate Skew t-Distribution. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 65(2), pp. 367-389 [CrossRef]


15. Bambulović, M. and Valdec, M., 2020. Testing the characteristics of macroprudential policies’ differential impact on foreign and domestic banks’ lending in Croatia. Public Sector Economics, 44(2), pp. 221-249 [CrossRef]


16. Banco de España, 2021. How do central banks identify risks? A survey of indicators. Documentos Ocasionales, No. 2125.


17. Bank of Japan, 2018. Financial Stability Review. Tokio: Bank of Japan.


18. Bank of Lithuania, 2019. Financial Stability Review. Vilnius: Bank of Lithuania.


19. Banque centrale du Luxembourg, 2022. Financial Stability Review. Luxembourg: Banque centrale du Luxembourg.


20. Bayoumi, T., Quayyum, S. and Das, S., 2021. Growth at Risk from Natural Disasters. IMF Working Paper, No. 2021/234. 


21. Blanchard, O., 2016. Do DSGE models have a future? Peterson Institute for International Economics Policy Brief, PG 16-11. 


22. BoE, 2021. How do we monitor UK financial system? Bank of England, 23 April 2021. London: Bank of England.


23. Boyarchenko, N., Favara, G. and Schularick, M., 2022. Financial Stability Considerations for Monetary Policy: Empirical Evidence and Challenges. FRB of New York Staff Report, No. 1003. 


24. Brandao-Marques, L. [et al.], 2020. Leaning Against the Wind: A Cost-Benefit Analysis for an Integrated Policy Framework. IMF Working Paper, WP/20/123 [CrossRef]


25. Brownlees, C. and Souza, A. B. M., 2021. Backtesting global Growth-at-Risk. Journal of Monetary Economics, 118, pp. 312-330 [CrossRef]


26. Buch, C. M., Vogel, E. and Weigert, B., 2018. Evaluating macroprudential policies. Working Paper Series, No 76 [CrossRef]


27. Budnik, K. [et al.], 2021. The growth-at-risk perspective on the systemwide impact of Basel III finalisation in the euro area. ECB Occasional Paper Series, No 258 [CrossRef]


28. Budnik, K. and Kleibl, J., 2018. Macroprudential regulation in the European Union in 1995-2014: introducing a new data set on policy actions of a macroprudential nature. ECB Working Paper Series, No. 2123 [CrossRef]


29. Carstens, A., 2021. The role of macroprudential policies during economic crises. Speech by Agustín Carstens, General Manager, Bank for International Settlements, at the 45th regular session of the Council of Arab Central Banks and Monetary Authorities' Governors, Abu Dhabi, 19 September 2021.


30. Cecchetti, S. and Suarez, J., 2021. On the stance of macroprudential policy. Reports of the Advisory Scientific Committee. ESRB: Advisory Scientific Committee Reports, 2021/11 [CrossRef]


31. Ćehajić, A. and Košak, M., 2021. Tightening and Loosening of Macroprudential Policy, Its Effects on Credit Growth and Implications for the COVID-19 Crisis. Economic and Business Review, 23(4), pp. 207-233 [CrossRef]


32. Central Bank of Ireland, 2022. Financial Stability Review I and II. Dublin: Central Bank of Ireland


33. Cerutti, E. Claessens, S. and Laeven, L., 2015. Database. Washington: International Monetary Fund. 


34. Cerutti, E., Claessens, S. and Laeven, L., 2017. The use and effectiveness of macroprudential policies: New evidence. Journal of Financial Stability, 28(c), pp. 203-224 [CrossRef]


35. Chavleishvili, S. and Kremer, M., 2023. Measuring systemic financial stress and its risks for growth. ECB Working Paper Series, No. 2842 [CrossRef]


36. Chavleishvili, S. and Manganelli, S., 2020. Forecasting and stress testing with quantile vector autoregression. ECB Working Paper Series, No. 2330 [CrossRef]


37. Christiano, L. J., Eichenbaum, M. and Evans, C. L., 1999. Chapter 2 Monetary policy shocks: What have we learned and to what end? In: Handbook of Macroeconomics, Vol. 1, Part A, pp. 65-148 [CrossRef]


38. Claessens, S., 2014. An Overview of Macroprudential Policy Tools. IMF working paper, No. WP/14/24 [CrossRef]


39. Claessens, S., Kose, M. A. and Terrones, M., 2012. How do Business and Financial Cycles Interact? Journal of International Economics, 87(1), pp. 178-90 [CrossRef]


40. CNB, 2023. Internal database of HIFS – Croatian index of financial stress. Zagreb: Croatian National Bank.


41. CNB, 2023. Online database Croatian National Bank. Zagreb: Croatian National Bank.


42. Cucic, D. [et al.], 2022. Evaluating the macroprudential stance in a growth-at-risk framework. Economic Memo, No. 14.


43. Ćehajić, A. and Košak, M., 2019. Tightening and loosening of macroprudential policy, its effects on credit growth and implications for the COVID-19 Crisis. Economic and Business Review, 23(4), pp. 207-233 [CrossRef]


44. Davino, C., Furno, M. and Vistocco, D., 2013. Quantile regression: theory and applications. New York: John Wiley & Sons [CrossRef]


45. De Schryder, S. and Opitz, F., 2019. Macroprudential policy and its impact on the credit cycle. Ghent University, Department of Economics, Working Paper, 2019/990.


46. De Schryder, S., Opitz, F., 2021. Macroprudential policy and its impact on the credit cycle. Journal of Financial Stability, 53, 100818 [CrossRef]


47. Deghi, A. [et al.], 2020. Predicting Downside Risks to House Prices and Macro-Financial Stability. IMF Working Paper, WP/20/11 [CrossRef]


48. Deutsche Bundesbank, 2018. Financial Stability Review. Deutsche Bundesbank.


49. Deutsche Bundesbank, 2021. Macroprudential policy and growth-at-risk. Monthly Report, July 2021, pp. 65-83.


50. Drenkovska, M. and Volčjak, R., 2022. Growth-at-risk and Financial Stability: Concept and Application for Slovenia. Discussion Papers, No. 5.


51. Duprey, T. and Ueberfeldt, A., 2018. How to Manage Macroeconomic and Financial Stability Risks: A New Framework. Staff Analytical Note, 2018-11.


52. Duprey, T. and Ueberfeldt, A., 2020. Managing GDP tail risk. Bank of Canada Working Paper, No. 2020-03.


53. ECB, 2010. New quantitative measures of systemic risk. IV special features. Financial Stability Review, December 2010, pp. 147-153. Frankfurt am Main: European Central Bank.


54. ECB, 2018. MacroPrudential Policies Evaluation Database. Frankfurt am Main: European Central Bank.


55. ECB, 2019. Financial Stability Report. European Central Bank. Frankfurt am Main: European Central Bank.


56. ECB, 2023. CLIFS data for Croatia. Frankfurt am Main: European Central Bank.


57. Eguren-Martin, F. [et al.], 2021. Capital Flows-at-Risk: Push, Pull and the Role of Policy. ECB Working Paper, No. 2021/2538 [CrossRef]


58. Eller, M. [et al.], 2020. Macroprudential policies in CESEE – an intensity-adjusted approach. Focus on European Economic Integration, Oesterreichische Nationalbank (Austrian Central Bank), Q2/20, pp. 65-81.


59. ESRB, 2019. Features of a macroprudential stance: initial considerations. Frankfurt am Main: European Systemic Risk Board.


60. ESRB, 2021. A framework for assessing macroprudential stance. Report of the Expert Group on Macroprudential Stance – Phase II (implementation). Frankfurt am Main: European Systemic Risk Board.


61. Fernandez-Gallardo, A. and Paya, I., 2020. Macroprudential Policy in the Euro Area. Economic Working Paper Series, No. 2020/020. 


62. Ferrara, L., Mogliani, M. and Sahuc, J-G., 2022. High-frequency monitoring of growth at risk. International Journal of Forecasting, 38(2), pp. 582-595 [CrossRef]


63. Flament, G., Hurlin, C. and Lajaunie, Q., 2023. The at-Risk approach: a new tool for stress tests and overlays [CrossRef]


64. Franta, M. and Gambacorta, L., 2020. On the effects of macroprudential policies on Growth at Risk. Economics Letters, 196, 109501 [CrossRef]


65. Gächter, M., Geiger, M. and Hasler, E., 2022. On the structural determinants of growth-at-risk. eeecon Working Papers in Economics and Statistics, No. 2022-06.


66. Galán, J. E. and Rodríguez-Moreno, M., 2020. At-risk measures and financial stability. Revista de Estabilidad Financiera, (39), pp. 69-96.


67. Galán, J. E., 2020a. The benefits are at the tail: uncovering the impact of macroprudential policy on growth-at-risk. Documentos de Trabajo, No. 2007 [CrossRef]


68. Galán, J. E., 2020b. The benefits are at the tail: uncovering the impact of macroprudential policy on growth-at-risk. Journal of Financial Stability, 100831 [CrossRef]


69. Garcia Revelo, J. D., Lucotte, Y. and Pradines-Jobet, F., 2020. Macroprudential and monetary policies: The need to dance the Tango in harmony. Journal of International Money and Finance, 108, pp. 102-156 [CrossRef]


70. Gelos, G. [et al.], 2022. Capital flows at risk: taming the ebbs and flows. Journal of International Economics, 134, 103555 [CrossRef]


71. Giglio, S., Kelly, B. and Pruitt, S., 2015. Systemic risk and the macroeconomy: An empirical evaluation. NBER working paper, No. 20963 [CrossRef]


72. Giglio, S., Kelly, B. and Pruitt, S., 2016. Systemic risk and the macroeconomy: An empirical evaluation. Journal of Financial Economics, 119(3), pp. 457-471 [CrossRef]


73. Holló, D., Kremer, M. and Lo Duca, M., 2012. CISS – A Composite Indicator of Systemic Stress in the Financial System. ECB Working Paper Series, No. 1426 [CrossRef]


74. IMF, 2017. Global Financial Stability Report. Washington: International Monetary Fund.


75. IMF, 2020. Stress Testing at the IMF. Washington: International Monetary Fund.


76. IMF, 2022. Chapter 1 - Financial Stability in the New High-Inflation Environment. In: IMF Global Financial Stability Report. Washington: International Monetary Fund.


77. IMF, 2022. Integrated Macroprudential Policy (iMaPP) Database. Washington: International Monetary Fund.


78. IMF, 2022. The Macroprudential Policy Survey. Washington: International Monetary Fund.


79. Jordá, O., 2023. Local Projections for Applied Economics. Federal Reserve Bank of San Francisco Working Paper, No. 2023-16 [CrossRef]


80. Jordá, O., Schularick, M. and Taylor, A. M., 2012. When Credit Bites Back: Leverage, Business Cycles, and Crises. NBER Working Paper, No. 2011-27 [CrossRef]


81. Jordá, O., Schularick, M. and Taylor, A. M., 2013. When Credit Bites Back. Journal of Money, Credit and Banking, 45(s2), pp. 3-28 [CrossRef]


82. Kannan, P., Scott, A. and Terrones, M. E., 2013. From Recession to Recovery: How Soon and How Strong. In: S. Claessens [et al.], eds. Financial Crises, Consequences, and Policy Responses, pp. 239-274.


83. Kiley, M. T., 2021. Growth at Risk from Climate Change (August, 2021). FEDS Working Paper, No. 2021-54 [CrossRef]


84. Koenker, R. W. and Bassett, G., 1978. Regression Quantiles. Econometrica, 46(1), pp. 33-50 [CrossRef]


85. Koenker, R., 2005. Quantile regression. Cambridge: Cambridge University Press [CrossRef]


86. Koh, W. C. [et al.], 2020. Debt and Financial Crisis. Policy Research Working Paper, No. 9116 [CrossRef]


87. Kraft, E. and Galac, T., 2011. Macroprudential Regulation of Credit Booms and Busts, The Case of Croatia. Policy Research Working Paper, No. 5772 [CrossRef]


88. Kupiec, P., 1995. Techniques for verifying the accuracy of risk measurement models. Journal of Derivatives, 2, pp. 173-184 [CrossRef]


89. Laeven, L. and Valencia, F., 2012. Systemic Banking Crises Database: An Update. IMF Working Paper, No. WP/12/163 [CrossRef]


90. Laeven, L. and Valencia, F., 2013. Systemic Banking Crises Database. IMF Economic Review, 61, pp. 225-270 [CrossRef]


91. Lang, J. H. [et al.], 2019. Anticipating the bust: a new cyclical systemic risk indicator to assess the likelihood and severity of financial crises. Occasional Paper Series, No 219 [CrossRef]


92. Lang, J. H. and Forletta, M., 2019. Bank capital-at-risk: measuring the impact of cyclical systemic risk on future bank losses. ECB Macroprudential Bulletin, 9.


93. Lang, J. H. and Forletta, M., 2020. Cyclical systemic risk and downside risks to bank profitability. ECB Working Paper Series, No. 2405 [CrossRef]


94. Lim, C. [et al.], 2011. Macroprudential Policy: What Instruments and How to Use Them? IMF working paper, No. WP/11/238 [CrossRef]


95. Lloyd, S., Manuel, E. and Panchev, K., 2022. Foreign vulnerabilities, domestic risks: the global drivers of GDP-at-Risk. BoE Staff Working Paper, No. 940 [CrossRef]


96. López-Salido, D. and Loria F., 2021. Inflation at Risk. FEDS Working Paper, No. 2020-013 [CrossRef]


97. Lucas, R. E., 1972. Expectations and the Neutrality of Money. Journal of Economic Theory, 4(2), pp. 103-124 [CrossRef]


98. McCallum, B. T., 1999. Analysis of the Monetary Transmission Mechanism: Methodological Issues. NBER Working Paper, No. 7395 [CrossRef]


99. McCracken, M. W. and Ng, S., 2016. FRED-MD: A Monthly Database for Macroeconomic Research. Journal of Business and Economic Statistic, 34(4), pp. 574- 89 [CrossRef]


100. Meuleman, E. and Vander Vennet, R., 2020. Macroprudential policy and bank systemic risk. Working Papers of Faculty of Economics and Business Administration, No. 19-971 [CrossRef]


101. Mitchell, J., Poon, A. and Zhu, D., 2021. Multimodality in Macroeconomic Dynamics: Constructing Density Forecasts from Quantile Regressions. Working paper.


102. O'Brien, M. and Wosser, M., 2021. Growth at Risk & Financial Stability. Financial stability notes, 2021(2).


103. O'Brien, M. nad Wosser, M., 2022. Assessing Structure-Related Systemic Risk in Advanced Economies. Research Technical Paper, 2022(3).


104. Ossandon Busch, M. [et al.], 2022. Growth at risk: Methodology and applications in an open-source platform. Latin American Journal of Central Banking, 3(3), 100068 [CrossRef]


105. Papell, D. H. and Prudan, R., 2011. The Statistical Behavior of GDP after Financial Crises and Severe Recessions. Paper prepared for the Federal Reserve Bank of Boston conference on “Long-Term Effects of the Great Recession”, October 18-19 [CrossRef]


106. Plagborg-Møller, M. [et al.], 2020. When is Growth at Risk? Brookings Papers on Economic Activity, 2020(1), pp. 167-229 [CrossRef]


107. Portes, R. [et al.], 2020. The global dimensions of macroprudential policy. ESRB: Advisory Scientific Committee Reports, 2020/10 [CrossRef]


108. Prasad, A. [et al.], 2019. Growth at Risk: Concept and Application in IMF Country Surveillance. IMF Working Paper, No. 36 [CrossRef]


109. Ramey, V. A., 2016. Chapter 2 - Macroeconomic Shocks and Their Propagation. In: Handbook of Macroeconomics, Vol. 2, pp 71-162 [CrossRef]


110. Reichlin, L., Ricco, G. and Hasenzagl, T., 2020. Financial variables as predictors of real growth vulnerability. Discussion Papers, 05/2020 [CrossRef]


111. Reinhart, C. M. and Rogoff, K. S., 2009. This Time is Different: Eight Centuries of Financial Folly. Princeton Press.


112. Richter, B., Schularik, M. and Shim, I., 2018. The costs of macroprudential policy. NBER Working paper, No. 24989 [CrossRef]


113. Richter, B., Schularik, M. and Shim, I., 2019. The costs of macroprudential policy. Journal of International Economics, 118, pp. 263-282 [CrossRef]


114. Sánchez, A. C. and Röhn, O., 2016. How do Policies Influence GDP Tail Risks? OECD Economics Department Working Papers, No. 1339.


115. Škrinjarić, T. and Bukovšak, M., 2022. New Indicators of Credit Gap in Croatia: Improving the Calibration of the Countercyclical Capital Buffer. CNB Working Papers, No. W-69<


116. Škrinjarić, T., 2022a. Introduction of the composite indicator of cyclical systemic risk in Croatia: possibilities and limitations. CNB Working Papers, No. W-68.


117. Škrinjarić, T., 2022b. Do we get cold feet when deciding on countercyclical capital buffer? Ways to deal with uncertainty in estimating credit to GDP gap in real time. The 28th Dubrovnik Economic Conference. Dubrovnik: Croatian National Bank, 2022. 


118. Škrinjarić, T., 2023a. Introducing a composite indicator of cyclical systemic risk in Croatia: possibilities and limitations. Public Sector Economics, 47(1), pp. 1-39 [CrossRef]


119. Škrinjarić, T., 2023b. Leading indicators of financial stress in Croatia: a regime switching approach. Public Sector Economics, 47(2), pp. 205-232 [CrossRef]


120. Škrinjarić, T., 2023c. Credit-to-GDP gap estimates in real time: A stable indicator for macroprudential policy making in Croatia. Comparative Economic Studies, 64(3), pp. 1-33 [CrossRef]


121. Škrinjarić, T., 2023d. Macroprudential stance assessment: problems of measurement, literature review and some comments for the case of Croatia. CNB Working Papers, No. W-72.


122. Škrinjarić, T., 2024. Growth-at-Risk for macroprudential policy stance assessment: A Survey. BoE Staff Working Papers, No. 1075.


123. Suarez, J., 2021. Growth-at-risk and macroprudential policy design. ESRB Occasional Paper Series, No. 19. 


124. Swedish authorities, 2022. Response by the Swedish authorities to the European Commission´s public consultation on improving the EU’s macroprudential framework for the banking sector. 


125. Tölö, E., Laakkonen, H. and Kalatie, S., 2018. Evaluating Indicators for Use in Setting the Countercyclical Capital Buffer. International Journal of Central Banking, 14(2), pp. 51-111.


126. Vandenbussche, J., Vogel, U. and Detragiache, E., 2015. Macroprudential policies and housing prices: A new database and empirical evidence for Central, Eastern, and Southeastern Europe. Journal of Money, Credit and Banking, 47(S1), pp. 343-377 [CrossRef]


127. Vujčić, B. and Dumičić, M., 2016. Managing Systemic Risks in the Croatian Economy. BIS Paper, No. 86l.


128. White, H., Kim, T-H. and Manganelli, S., 2015a. VAR for VaR: measuring tail dependence using multivariate regression quantiles. ECB Working Paper Series, No. 1814 [CrossRef]


129. White, H., Kim, T-H. and Manganelli, S., 2015b. VAR for VaR: Measuring tail dependence using multivariate regression quantiles. Journal of Econometrics, 187, pp. 169-188 [CrossRef]

Acemoglu, D., Ozdaglar, A. and Tahbaz-Salehi, A., 2015. Microeconomic origins of macroeconomic tail risks. NBER Working Paper, No. 20865 [CrossRef]

Adams, P. A. [et al.], 2020. Forecasting Macroeconomic Risks. FRBNY Staff Reports, No. 914 [CrossRef]

Adrian, T., Boyarchenko, N. and Giannone, D., 2016. Vulnerable growth. FRBNY Staff Report, No. 794.

Adrian, T., Boyarchenko, N. and Giannone, D., 2019. Vulnerable growth. American Economic Review, 109(4), pp. 1263-1289 [CrossRef]

Aikman, D. [et al.], 2018. Measuring risks to UK financial stability. BoE Staff Working Paper, No. 738.

Aikman, D. [et al.], 2019a. Credit, capital and crisis: a GDP-at-risk approach. Staff Working Paper, No. 824.

Aikman, D. [et al.], 2019b. How do financial vulnerabilities and bank resilience affect medium term macroeconomic tail risk? 

Akinci, O. and Olmstead-Rumsey, J., 2015. How Effective are Macroprudential Policies? An Empirical Investigation. International Finance Discussion Papers, No. 1136 [CrossRef]

Akinci, O. and Olmstead-Rumsey, J., 2018. How effective are macroprudential policies? An empirical investigation. Journal of Financial Intermediation, 33, pp. 33-57 [CrossRef]

Alam, Z. [et al.], 2019. Digging Deeper – Evidence on the Effects of Macroprudential Policies from a New Database. IMF Working Paper, No. 2019/066. 

Alessandri, P. and Di Cesare, A., 2021. Growth-at-risk in Italy during the covid-19 pandemic. Notes on Financial Stability and Supervision, No. 24.

Ampudia, M. [et al.], 2021. On the effectiveness of macroprudential policy. ECB Discussion Paper, No. 2559 [CrossRef]

Arslan, Y. and Upper, C., 2017. Macroprudential frameworks: implementation and effectiveness. BIS Papers, No. 94.

Azzalini, A. and Capitanio, A., 2003. Distributions Generated by Perturbation of Symmetry with Emphasis on a Multivariate Skew t-Distribution. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 65(2), pp. 367-389 [CrossRef]

Bambulović, M. and Valdec, M., 2020. Testing the characteristics of macroprudential policies’ differential impact on foreign and domestic banks’ lending in Croatia. Public Sector Economics, 44(2), pp. 221-249 [CrossRef]

Banco de España, 2021. How do central banks identify risks? A survey of indicators. Documentos Ocasionales, No. 2125.

Bank of Japan, 2018. Financial Stability Review. Tokio: Bank of Japan.

Bank of Lithuania, 2019. Financial Stability Review. Vilnius: Bank of Lithuania.

Banque centrale du Luxembourg, 2022. Financial Stability Review. Luxembourg: Banque centrale du Luxembourg.

Bayoumi, T., Quayyum, S. and Das, S., 2021. Growth at Risk from Natural Disasters. IMF Working Paper, No. 2021/234. 

Blanchard, O., 2016. Do DSGE models have a future? Peterson Institute for International Economics Policy Brief, PG 16-11. 

BoE, 2021. How do we monitor UK financial system? Bank of England, 23 April 2021. London: Bank of England.

Boyarchenko, N., Favara, G. and Schularick, M., 2022. Financial Stability Considerations for Monetary Policy: Empirical Evidence and Challenges. FRB of New York Staff Report, No. 1003. 

Brandao-Marques, L. [et al.], 2020. Leaning Against the Wind: A Cost-Benefit Analysis for an Integrated Policy Framework. IMF Working Paper, WP/20/123 [CrossRef]

Brownlees, C. and Souza, A. B. M., 2021. Backtesting global Growth-at-Risk. Journal of Monetary Economics, 118, pp. 312-330 [CrossRef]

Buch, C. M., Vogel, E. and Weigert, B., 2018. Evaluating macroprudential policies. Working Paper Series, No 76 [CrossRef]

Budnik, K. [et al.], 2021. The growth-at-risk perspective on the systemwide impact of Basel III finalisation in the euro area. ECB Occasional Paper Series, No 258 [CrossRef]

Budnik, K. and Kleibl, J., 2018. Macroprudential regulation in the European Union in 1995-2014: introducing a new data set on policy actions of a macroprudential nature. ECB Working Paper Series, No. 2123 [CrossRef]

Carstens, A., 2021. The role of macroprudential policies during economic crises. Speech by Agustín Carstens, General Manager, Bank for International Settlements, at the 45th regular session of the Council of Arab Central Banks and Monetary Authorities' Governors, Abu Dhabi, 19 September 2021.

Cecchetti, S. and Suarez, J., 2021. On the stance of macroprudential policy. Reports of the Advisory Scientific Committee. ESRB: Advisory Scientific Committee Reports, 2021/11 [CrossRef]

Ćehajić, A. and Košak, M., 2021. Tightening and Loosening of Macroprudential Policy, Its Effects on Credit Growth and Implications for the COVID-19 Crisis. Economic and Business Review, 23(4), pp. 207-233 [CrossRef]

Central Bank of Ireland, 2022. Financial Stability Review I and II. Dublin: Central Bank of Ireland

Cerutti, E. Claessens, S. and Laeven, L., 2015. Database. Washington: International Monetary Fund. 

Cerutti, E., Claessens, S. and Laeven, L., 2017. The use and effectiveness of macroprudential policies: New evidence. Journal of Financial Stability, 28(c), pp. 203-224 [CrossRef]

Chavleishvili, S. and Kremer, M., 2023. Measuring systemic financial stress and its risks for growth. ECB Working Paper Series, No. 2842 [CrossRef]

Chavleishvili, S. and Manganelli, S., 2020. Forecasting and stress testing with quantile vector autoregression. ECB Working Paper Series, No. 2330 [CrossRef]

Christiano, L. J., Eichenbaum, M. and Evans, C. L., 1999. Chapter 2 Monetary policy shocks: What have we learned and to what end? In: Handbook of Macroeconomics, Vol. 1, Part A, pp. 65-148 [CrossRef]

Claessens, S., 2014. An Overview of Macroprudential Policy Tools. IMF working paper, No. WP/14/24 [CrossRef]

Claessens, S., Kose, M. A. and Terrones, M., 2012. How do Business and Financial Cycles Interact? Journal of International Economics, 87(1), pp. 178-90 [CrossRef]

CNB, 2023. Internal database of HIFS – Croatian index of financial stress. Zagreb: Croatian National Bank.

CNB, 2023. Online database Croatian National Bank. Zagreb: Croatian National Bank.

Cucic, D. [et al.], 2022. Evaluating the macroprudential stance in a growth-at-risk framework. Economic Memo, No. 14.

Ćehajić, A. and Košak, M., 2019. Tightening and loosening of macroprudential policy, its effects on credit growth and implications for the COVID-19 Crisis. Economic and Business Review, 23(4), pp. 207-233 [CrossRef]

Davino, C., Furno, M. and Vistocco, D., 2013. Quantile regression: theory and applications. New York: John Wiley & Sons [CrossRef]

De Schryder, S. and Opitz, F., 2019. Macroprudential policy and its impact on the credit cycle. Ghent University, Department of Economics, Working Paper, 2019/990.

De Schryder, S., Opitz, F., 2021. Macroprudential policy and its impact on the credit cycle. Journal of Financial Stability, 53, 100818 [CrossRef]

Deghi, A. [et al.], 2020. Predicting Downside Risks to House Prices and Macro-Financial Stability. IMF Working Paper, WP/20/11 [CrossRef]

Deutsche Bundesbank, 2018. Financial Stability Review. Deutsche Bundesbank.

Deutsche Bundesbank, 2021. Macroprudential policy and growth-at-risk. Monthly Report, July 2021, pp. 65-83.

Drenkovska, M. and Volčjak, R., 2022. Growth-at-risk and Financial Stability: Concept and Application for Slovenia. Discussion Papers, No. 5.

Duprey, T. and Ueberfeldt, A., 2018. How to Manage Macroeconomic and Financial Stability Risks: A New Framework. Staff Analytical Note, 2018-11.

Duprey, T. and Ueberfeldt, A., 2020. Managing GDP tail risk. Bank of Canada Working Paper, No. 2020-03.

ECB, 2010. New quantitative measures of systemic risk. IV special features. Financial Stability Review, December 2010, pp. 147-153. Frankfurt am Main: European Central Bank.

ECB, 2018. MacroPrudential Policies Evaluation Database. Frankfurt am Main: European Central Bank.

ECB, 2019. Financial Stability Report. European Central Bank. Frankfurt am Main: European Central Bank.

ECB, 2023. CLIFS data for Croatia. Frankfurt am Main: European Central Bank.

Eguren-Martin, F. [et al.], 2021. Capital Flows-at-Risk: Push, Pull and the Role of Policy. ECB Working Paper, No. 2021/2538 [CrossRef]

Eller, M. [et al.], 2020. Macroprudential policies in CESEE – an intensity-adjusted approach. Focus on European Economic Integration, Oesterreichische Nationalbank (Austrian Central Bank), Q2/20, pp. 65-81.

ESRB, 2019. Features of a macroprudential stance: initial considerations. Frankfurt am Main: European Systemic Risk Board.

ESRB, 2021. A framework for assessing macroprudential stance. Report of the Expert Group on Macroprudential Stance – Phase II (implementation). Frankfurt am Main: European Systemic Risk Board.

Fernandez-Gallardo, A. and Paya, I., 2020. Macroprudential Policy in the Euro Area. Economic Working Paper Series, No. 2020/020. 

Ferrara, L., Mogliani, M. and Sahuc, J-G., 2022. High-frequency monitoring of growth at risk. International Journal of Forecasting, 38(2), pp. 582-595 [CrossRef]

Flament, G., Hurlin, C. and Lajaunie, Q., 2023. The at-Risk approach: a new tool for stress tests and overlays [CrossRef]

Franta, M. and Gambacorta, L., 2020. On the effects of macroprudential policies on Growth at Risk. Economics Letters, 196, 109501 [CrossRef]

Gächter, M., Geiger, M. and Hasler, E., 2022. On the structural determinants of growth-at-risk. eeecon Working Papers in Economics and Statistics, No. 2022-06.

Galán, J. E. and Rodríguez-Moreno, M., 2020. At-risk measures and financial stability. Revista de Estabilidad Financiera, (39), pp. 69-96.

Galán, J. E., 2020a. The benefits are at the tail: uncovering the impact of macroprudential policy on growth-at-risk. Documentos de Trabajo, No. 2007 [CrossRef]

Galán, J. E., 2020b. The benefits are at the tail: uncovering the impact of macroprudential policy on growth-at-risk. Journal of Financial Stability, 100831 [CrossRef]

Garcia Revelo, J. D., Lucotte, Y. and Pradines-Jobet, F., 2020. Macroprudential and monetary policies: The need to dance the Tango in harmony. Journal of International Money and Finance, 108, pp. 102-156 [CrossRef]

Gelos, G. [et al.], 2022. Capital flows at risk: taming the ebbs and flows. Journal of International Economics, 134, 103555 [CrossRef]

Giglio, S., Kelly, B. and Pruitt, S., 2015. Systemic risk and the macroeconomy: An empirical evaluation. NBER working paper, No. 20963 [CrossRef]

Giglio, S., Kelly, B. and Pruitt, S., 2016. Systemic risk and the macroeconomy: An empirical evaluation. Journal of Financial Economics, 119(3), pp. 457-471 [CrossRef]

Holló, D., Kremer, M. and Lo Duca, M., 2012. CISS – A Composite Indicator of Systemic Stress in the Financial System. ECB Working Paper Series, No. 1426 [CrossRef]

IMF, 2017. Global Financial Stability Report. Washington: International Monetary Fund.

IMF, 2020. Stress Testing at the IMF. Washington: International Monetary Fund.

IMF, 2022. Chapter 1 - Financial Stability in the New High-Inflation Environment. In: IMF Global Financial Stability Report. Washington: International Monetary Fund.

IMF, 2022. Integrated Macroprudential Policy (iMaPP) Database. Washington: International Monetary Fund.

IMF, 2022. The Macroprudential Policy Survey. Washington: International Monetary Fund.

Jordá, O., 2023. Local Projections for Applied Economics. Federal Reserve Bank of San Francisco Working Paper, No. 2023-16 [CrossRef]

Jordá, O., Schularick, M. and Taylor, A. M., 2012. When Credit Bites Back: Leverage, Business Cycles, and Crises. NBER Working Paper, No. 2011-27 [CrossRef]

Jordá, O., Schularick, M. and Taylor, A. M., 2013. When Credit Bites Back. Journal of Money, Credit and Banking, 45(s2), pp. 3-28 [CrossRef]

Kannan, P., Scott, A. and Terrones, M. E., 2013. From Recession to Recovery: How Soon and How Strong. In: S. Claessens [et al.], eds. Financial Crises, Consequences, and Policy Responses, pp. 239-274.

Kiley, M. T., 2021. Growth at Risk from Climate Change (August, 2021). FEDS Working Paper, No. 2021-54 [CrossRef]

Koenker, R. W. and Bassett, G., 1978. Regression Quantiles. Econometrica, 46(1), pp. 33-50 [CrossRef]

Koenker, R., 2005. Quantile regression. Cambridge: Cambridge University Press [CrossRef]

Koh, W. C. [et al.], 2020. Debt and Financial Crisis. Policy Research Working Paper, No. 9116 [CrossRef]

Kraft, E. and Galac, T., 2011. Macroprudential Regulation of Credit Booms and Busts, The Case of Croatia. Policy Research Working Paper, No. 5772 [CrossRef]

Kupiec, P., 1995. Techniques for verifying the accuracy of risk measurement models. Journal of Derivatives, 2, pp. 173-184 [CrossRef]

Laeven, L. and Valencia, F., 2012. Systemic Banking Crises Database: An Update. IMF Working Paper, No. WP/12/163 [CrossRef]

Laeven, L. and Valencia, F., 2013. Systemic Banking Crises Database. IMF Economic Review, 61, pp. 225-270 [CrossRef]

Lang, J. H. [et al.], 2019. Anticipating the bust: a new cyclical systemic risk indicator to assess the likelihood and severity of financial crises. Occasional Paper Series, No 219 [CrossRef]

Lang, J. H. and Forletta, M., 2019. Bank capital-at-risk: measuring the impact of cyclical systemic risk on future bank losses. ECB Macroprudential Bulletin, 9.

Lang, J. H. and Forletta, M., 2020. Cyclical systemic risk and downside risks to bank profitability. ECB Working Paper Series, No. 2405 [CrossRef]

Lim, C. [et al.], 2011. Macroprudential Policy: What Instruments and How to Use Them? IMF working paper, No. WP/11/238 [CrossRef]

Lloyd, S., Manuel, E. and Panchev, K., 2022. Foreign vulnerabilities, domestic risks: the global drivers of GDP-at-Risk. BoE Staff Working Paper, No. 940 [CrossRef]

López-Salido, D. and Loria F., 2021. Inflation at Risk. FEDS Working Paper, No. 2020-013 [CrossRef]

Lucas, R. E., 1972. Expectations and the Neutrality of Money. Journal of Economic Theory, 4(2), pp. 103-124 [CrossRef]

McCallum, B. T., 1999. Analysis of the Monetary Transmission Mechanism: Methodological Issues. NBER Working Paper, No. 7395 [CrossRef]

McCracken, M. W. and Ng, S., 2016. FRED-MD: A Monthly Database for Macroeconomic Research. Journal of Business and Economic Statistic, 34(4), pp. 574- 89 [CrossRef]

Meuleman, E. and Vander Vennet, R., 2020. Macroprudential policy and bank systemic risk. Working Papers of Faculty of Economics and Business Administration, No. 19-971 [CrossRef]

Mitchell, J., Poon, A. and Zhu, D., 2021. Multimodality in Macroeconomic Dynamics: Constructing Density Forecasts from Quantile Regressions. Working paper.

O'Brien, M. and Wosser, M., 2021. Growth at Risk & Financial Stability. Financial stability notes, 2021(2).

O'Brien, M. nad Wosser, M., 2022. Assessing Structure-Related Systemic Risk in Advanced Economies. Research Technical Paper, 2022(3).

Ossandon Busch, M. [et al.], 2022. Growth at risk: Methodology and applications in an open-source platform. Latin American Journal of Central Banking, 3(3), 100068 [CrossRef]

Papell, D. H. and Prudan, R., 2011. The Statistical Behavior of GDP after Financial Crises and Severe Recessions. Paper prepared for the Federal Reserve Bank of Boston conference on “Long-Term Effects of the Great Recession”, October 18-19 [CrossRef]

Plagborg-Møller, M. [et al.], 2020. When is Growth at Risk? Brookings Papers on Economic Activity, 2020(1), pp. 167-229 [CrossRef]

Portes, R. [et al.], 2020. The global dimensions of macroprudential policy. ESRB: Advisory Scientific Committee Reports, 2020/10 [CrossRef]

Prasad, A. [et al.], 2019. Growth at Risk: Concept and Application in IMF Country Surveillance. IMF Working Paper, No. 36 [CrossRef]

Ramey, V. A., 2016. Chapter 2 - Macroeconomic Shocks and Their Propagation. In: Handbook of Macroeconomics, Vol. 2, pp 71-162 [CrossRef]

Reichlin, L., Ricco, G. and Hasenzagl, T., 2020. Financial variables as predictors of real growth vulnerability. Discussion Papers, 05/2020 [CrossRef]

Reinhart, C. M. and Rogoff, K. S., 2009. This Time is Different: Eight Centuries of Financial Folly. Princeton Press.

Richter, B., Schularik, M. and Shim, I., 2018. The costs of macroprudential policy. NBER Working paper, No. 24989 [CrossRef]

Richter, B., Schularik, M. and Shim, I., 2019. The costs of macroprudential policy. Journal of International Economics, 118, pp. 263-282 [CrossRef]

Sánchez, A. C. and Röhn, O., 2016. How do Policies Influence GDP Tail Risks? OECD Economics Department Working Papers, No. 1339.

Škrinjarić, T. and Bukovšak, M., 2022. New Indicators of Credit Gap in Croatia: Improving the Calibration of the Countercyclical Capital Buffer. CNB Working Papers, No. W-69<

Škrinjarić, T., 2022a. Introduction of the composite indicator of cyclical systemic risk in Croatia: possibilities and limitations. CNB Working Papers, No. W-68.

Škrinjarić, T., 2022b. Do we get cold feet when deciding on countercyclical capital buffer? Ways to deal with uncertainty in estimating credit to GDP gap in real time. The 28th Dubrovnik Economic Conference. Dubrovnik: Croatian National Bank, 2022. 

Škrinjarić, T., 2023a. Introducing a composite indicator of cyclical systemic risk in Croatia: possibilities and limitations. Public Sector Economics, 47(1), pp. 1-39 [CrossRef]

Škrinjarić, T., 2023b. Leading indicators of financial stress in Croatia: a regime switching approach. Public Sector Economics, 47(2), pp. 205-232 [CrossRef]

Škrinjarić, T., 2023c. Credit-to-GDP gap estimates in real time: A stable indicator for macroprudential policy making in Croatia. Comparative Economic Studies, 64(3), pp. 1-33 [CrossRef]

Škrinjarić, T., 2023d. Macroprudential stance assessment: problems of measurement, literature review and some comments for the case of Croatia. CNB Working Papers, No. W-72.

Škrinjarić, T., 2024. Growth-at-Risk for macroprudential policy stance assessment: A Survey. BoE Staff Working Papers, No. 1075.

Suarez, J., 2021. Growth-at-risk and macroprudential policy design. ESRB Occasional Paper Series, No. 19. 

Swedish authorities, 2022. Response by the Swedish authorities to the European Commission´s public consultation on improving the EU’s macroprudential framework for the banking sector. 

Tölö, E., Laakkonen, H. and Kalatie, S., 2018. Evaluating Indicators for Use in Setting the Countercyclical Capital Buffer. International Journal of Central Banking, 14(2), pp. 51-111.

Vandenbussche, J., Vogel, U. and Detragiache, E., 2015. Macroprudential policies and housing prices: A new database and empirical evidence for Central, Eastern, and Southeastern Europe. Journal of Money, Credit and Banking, 47(S1), pp. 343-377 [CrossRef]

Vujčić, B. and Dumičić, M., 2016. Managing Systemic Risks in the Croatian Economy. BIS Paper, No. 86l.

White, H., Kim, T-H. and Manganelli, S., 2015a. VAR for VaR: measuring tail dependence using multivariate regression quantiles. ECB Working Paper Series, No. 1814 [CrossRef]

White, H., Kim, T-H. and Manganelli, S., 2015b. VAR for VaR: Measuring tail dependence using multivariate regression quantiles. Journal of Econometrics, 187, pp. 169-188 [CrossRef]