Advertisement

Brexit and macroprudential regulation: a DSGE perspective

  • Jürgen JergerEmail author
  • Jenny Körner
Original Paper
  • 80 Downloads

Abstract

This paper uses a small and simple theoretical DSGE model in order to conduct some exercises in comparative dynamics of shocks that can be associated with Brexit. We do so by comparing two policy environments, one where a flexible macroprudential regulation (FMR) is in place and one, where this is not the case. This enables us to evaluate whether and to what extent FMR helps to mitigate the Brexit related shocks. We conclude that FMR would indeed be helpful, although in quantitative terms only slightly so.

1 Introduction

The Brexit vote in June 2016 and the political negotiations and disputes that followed poses a multi-faceted challenge to policymakers as well as analysts on various levels. At the most basic level, no one can be sure, how Brexit will eventually look like. At the time of preparing this article for the workshop, it was not even clear, whether Brexit will happen by end of March 2019 according to the official timetable. And the final version of the paper was finished on the day after the narrow victory of Theresa May in the no-confidence vote that was initiated by her own party. This was also the day (Dec. 12, 2018) when the UK parliament should have voted on the “Brexit deal” that was negotiated between the British government and the EU. Since there is no political majority for any concrete form of the Brexit visible in the UK, it is anybody’s guess how absurd the political process will get and to which “solution” this will lead.

So even little more than three months before the official exit date, any option is on table, at least in principle. From simply and unilaterally cancelling the whole Brexit process to a hard or “disorderly” Brexit nothing can be ruled out. And there is, of course, still the possibility of a second referendum. Consequently, economists and other social scientists are confronted with the impossible task of predicting the effects of a rather elusive event. This would have also been true, however, if the “Brexit deal” mentioned above would smoothly pass the UK parliament, because this deal does not finally lay down the principles that will govern the future relationship between the UK and the EU but mainly stipulates a period of transition until the end of 2020 with the possibility of an extension until the end of 2022. From the experience so far, it seems rather clear that this option for postponement would be used before a final settlement might eventually be reached.

But even if we precisely knew the terms and conditions of Brexit, it is fair to say that the analysis of its consequences would still be hard. This is so because of two different kinds of complexity that are useful to distinguish.

First, the models at our disposal for empirical analyses are by far too simple in order to reliably map the thousands of legislative changes that would take place in any Brexit scenario. In empirical trade models, for instance, an obvious – but certainly less than perfect – way to model Brexit is simply to assume some increase of bilateral trade costs. And despite the fact that financial market issues have been taken into consideration much more seriously in macroeconomic models after the economic and financial crisis in 2008/9 – the literature on which this paper draws is part of this endeavour –, the models are nowhere near a full sketch of the complexities of the regulatory environment.

Second, empirical models can in many cases only be expected to deliver reasonable results for small changes of some exogenous variable. This might be the case because of the non-robustness of the empirical model with respect to policy changes (Lucas critique). But even if the Lucas critique does not apply, empirical models often rely on linear approximations of non-linear relationships – which again makes them unsuitable for the analysis of big exogenous shocks. And the magnitude of the Brexit related shocks might be rather big. In a recent study of different scenarios, the Bank of England estimates that a disorderly Brexit will cause a drop of GDP between 7.75 % and 10.5% by end-2023 relative to the May 2016 trend, cf. Bank of England (2018).

With these words of caution, we want to make clear that – strictly speaking – the task of delineating the effects of Brexit is impossible. But it is also a necessary (and therefore: good) tradition in social sciences to do the impossible. Hence, there are many studies that try to look at the effects of Brexit in many dimensions.1 These analyses range from strategic questions at the politcal level as in Patel and Reh (2016) to specific issues such as the impact on immigration (Wadsworth et al. 2016) and – clearly – international trade (Schoof et al. 2015); see Busch and Matthes (2016) for a meta-analysis of the literature. Of course, also this conference and the whole research project are also attempts of the scientific community to think about the likely consequences of Brexit.

In this paper, we aim to contribute to this endeavour by asking a very specific – and therefore limited – question. More specifically, we look at the possible contribution of a flexible macroprudential regulation (FMR) framework to mitigate the effects of exogenous shocks that can be associated with Brexit from a continental European point of view. It is important to stress the limitation of this contribution also in comparison to what most other papers in the literature on the Brexit are at least aiming at. Our exercise does not say anything about the effects of Brexit on levels or long-run growth rates of some macroeconomic variable – such as GDP, employment, trade volumes or migration flows. Rather, it treats Brexit as an event that leads to shocks the economy has to cope with and looks how FMR affects the shock absorption capacity. The question is still interesting, however, for at least two reasons. First, for all the long-run effects Brexit might have, it will be (and with regard to anticipation effects already is) a major economic shock. Second, macroprudential regulation has attracted a lot of attention recently and therefore is an increasingly important tool that complements the tool kit of both monetary policy makers and regulators of financial markets.

The rest of the paper is structured as follows. In the next section, we briefly review the renewed interest in macroprudential regulation. Section 3 introduces the model that is a slight modification of Jerger and Körner (2018). In Section 4, we present the results of a couple of exercises as outlined above. Some tentative conclusions are offered in Section 5.

2 Macroprudential regulation

After the economic and financial crisis a decade ago, it (again) became more than obvious that the stability of financial markets in general and of banks in particular is of very high importance for the real economy.

In fact, real output fell by about 5% globally in 2009 – the by far largest slump since the Great Recession in the 1920/30s. Although empirical and theoretical accounts of frequent and deep financial crises existed early on (see e.g. Aliber and Kindleberger 2017, the first edition of which was published by Charles E. Kindleberger in 1978 or Minsky 1977), there was a general perception that the economics profession was surprised by the crisis in 2008.2 The common pattern that is shared by financial crises is a credit boom that eventually leads to the inability of debtors to service their debt. Therefore, a sound crisis prevention policy has to focus on the prevention of unsustainable debt levels. It is clear, however, that the threshold between sustainable and unsustainable debt can not be known for sure ex ante.

Macroprudential regulation aims at conditioning the credit policy of the bank sector to the macroeconomic situation thereby making an unsustainable credit boom less likely. It has been recognized early on that this instrument – or rather set of instruments – is especially useful when inflation and nominal interest rates are low (Borio and Shim 2008). Dehmej and Couppey-Soubeyran (2017, 06) emphasize the aspect that macroprudential regulation can also serve as a remedy for economic and financial imbalances between members of the Euro area. Unlike the usual instruments of monetary policy, macroprudential regulation can take into account the macroeconomic situation at the national level as well as national differences in the financial systems.

The ECB clearly documented its perception of the importance of macroprudential regulation, when it introduced the bi-annual ECB Macroprudential Bulletin in March 2016. This interest in the field is also echoed by other central banks, e.g. in China, cf. Klingelhöfer and Sun (2017).

One of the consequences of the economic and financial crises a decade ago is the fact that policy interest rates are at or close to zero. Hence, the zero lower bound (ZLB) binds. In this situation, adverse shocks might easily lead to negative policy rates if monetary policy is charaterized by a Taylor rule or some modification of it. Although different varieties of quantitative easing can be interpreted as finding a way around the ZLB, nominal policy rates (for liquidity providing transactions) can not be negative. Hence, a model has to deal with a binding ZLB. Farhi and Werning (2016) and Korinek and Simsek (2016) also look at the effects of macroprudential regulation when monetary policy is restricted by the ZLB. Quite close to our model is the paper by Rubio and Yao (2017) who also looks at a FMR in a low interest-rate environment. They do so however, by contrasting the low-interest rate scenario with an alternative scenario in which the steady state interest rate is higher.

3 The model

In this paper, we use a slightly modified version of the model used in Jerger and Körner (2018), which in turn draws heavily on Rubio and Carrasco-Gallego (2015).

3.1 Consumers/workers and borrowing constraint

There are two groups (of equal size) of representative agents – savers and borrowers – with group specific discount rates βi, i ∈{s,b}. The assumption βb < βs distinguishes the two groups. Both groups consume some consumption good and housing and supply labour. Their behaviour is described by the solution of the following intertemporal maximization problem
$$\max\limits_{c_{i,t},h_{i,t},n_{i,t}, b_{i,t}} E \sum\limits_{t = 0}^{\infty} {\beta^{t}_{i}} \kappa_{t} \left[\ln \left( c_{i,t}\right) + j \ln\left( h_{i,t}\right) - \frac{n_{i,t}^{\eta}}{\eta}\right], $$
where ci,t, hi,t, ni,t and j denote consumption, housing services, working hours and the weight of housing, respectively. η − 1 ≥ 0 is the inverse of the Frisch labour supply elasticity. κt denotes an intertemporal aggregate demand shock, specified as
$$ \ln(\kappa_{t}) = \rho_{\kappa} \ln(\kappa_{t-1})+\varepsilon_{\kappa,t},\\ $$
(1)
where 0 < ρκ < 1 and \(\varepsilon _{\kappa ,t}\sim N(0,\sigma ^{2}_{\varepsilon _{\kappa }})\).
Savers and borrowers face the budget constraints
$$ c_{s,t} + b_{s,t} + q_{t} h_{s,t} = b_{s,t-1} \frac{r_{t-1}}{\pi_{t}} + w_{s,t} n_{s,t}+q_{t} h_{s,t-1} + x_{t} $$
(2)
and
$$ c_{b,t} + \frac{r_{t-1}}{\pi_{t}} b_{b, t-1}+q_{t} h_{b,t} = q_{t} h_{b,t-1} + b_{b, t}+w_{b,t} n_{b,t}, $$
(3)
respectively, where bs,t,qt,rt− 1,ws,t,(wb,t) and πt denote lending, the price of housing, both in units of consumption, the gross nominal interest rate in t − 1, the real wage rate earned by savers (borrowers) and overall inflation rate, respectively. xt are dividend payments from the production sector that are assumed to accrue to savers.
Macroprudential regulation is introduced by limiting borrowing by a maximum loan-to-value (LTV ) ratio as follows:
$$ l \geq E_{t} \left[\frac{r_{t} b_{b, t}}{\pi_{t} q_{t + 1} h_{b,t}}\right]. $$
(4)

We will come back to the specification of a flexible macroprudential regulation below.

3.2 Production and price setting

The production sector is modelled in a very basic way. Most importantly, we neglect any factor of production other than the labour supply of savers and borrowers. More specifically, final output yt is assembled from a continuum of intermediate goods yt(z) according to
$$ y_{t}={{\int}_{0}^{1}} \left( y_{t} (z)^{\frac{\varepsilon -1}{\varepsilon}} dz \right)^{\frac{\varepsilon}{\varepsilon-1}}. $$
(5)
The intermediate goods are produced by means of the CRS technology
$$ y_{t}(z)= \gamma_{t} n_{s,t}(z)^{\alpha} n_{b,t}(z)^{1-\alpha}. $$
(6)

\(\gamma _{t} \sim N(1, \sigma ^{2}_{\gamma })\) denotes a productivity shock. The specification in Eq. 6 implies that borrowers’ and savers’ labour services are different and hence receive different wageswb,t and ws,t, respectively. See Iacoviello and Neri (2010) for possible justifications of this modelling strategy.

Prices Pt(z) are the result of profit-maximizing firms at the intermediate goods level who have to observe Rotemberg price adjustment cost of \(\frac {{\Phi }_{p}}{y_{t}}=\left (\frac {\phi _{p}}{2}\right ) \left (\frac {P_{t}}{\pi P_{t-1}} -1\right )^{2}\) and the demand function for their products from the final goods sector \(y_{t}(z) = \left [\frac {P_{t}(z)}{P_{t}} \right ]^{-\varepsilon }y_{t}\).

For symmetric intermediate goods firms, the supply side of the economy can be summarized by the familiar New Keynesian Phillips curve:
$$ \hat{\pi}_{t} = \frac{(\varepsilon -1)}{\phi_{p}}\hat{\xi}_{t} + \beta_{s} \kappa_{t} E_{t} \hat{\pi}_{t + 1} $$
(7)
where \(\pi _{t}=\frac {P_{t}}{P_{t-1}}\) and ξt denote gross inflation and marginal cost, respectively. As usual, hats denote deviations from the steady state.

3.3 Monetary policy and flexible macroprudential policy

Monetary policy is modelled with the modified Taylor rule
$$ r_{t}=\max\left[ 1 , r^{1-\rho_{r}} r_{t-1}^{ \rho_{r}} \left( \left( \frac{\pi_{t}}{\pi}\right)^{\omega_{\pi}}\left( \frac{y_{t}}{y}\right)^{\omega_{y}}\right)^{1-\rho_{r}} \right]. $$
(8)

This rule reflects the ZLB in an obvious way. When the ZLB is not binding, interest rates follow the second expression in the maximum function of Eq. 8.

The introduction of flexible macroprudential regulation (FMR) follows Lambertini et al. (2013) by making the maximum LTV ratio a function of the level of debt relative to its steady state value according to
$$ l_{t}=l_{t-1}^{ \rho_{l}}\left( l\left( \frac{b_{t}}{b}\right)^{\chi_{l}}\right)^{(1-\rho_{l})}. $$
(9)

l is the steady state value of the LTV ratio. 0 < ρl < 1 and χl capture the persistence and the reaction parameter of the rule, respectively. Clearly, χl < 0 implies a countercyclical macroprudential regulation which is what we want to look at.

In order to close the model, we have to introduce market clearing conditions. We do so by fixing housing supply at unity (hs,t + hb,t = 1). Goods market clearing is given by \(y_{t} = c_{b,t} + c_{s,t}+ \left (\frac {\phi _{p}}{2}\right ) \left (\frac {P_{t}}{\pi P_{t-1}} -1\right )^{2} y_{t} \), whereas bond market clearing requires bs,t = bb,t. Finally, labour market clearing is given by \({{\int }_{0}^{1}} n_{i,t}(z)dz = n_{i,t}\), i ∈{s,b}.

3.4 Calibration and simulation of the model

One unusual feature of the simulation exercise below is the fact that we take seriously the ZLB. Formally, this is achieved by the definition of the modified Taylor rule in Eq. 8. Technically, this is handled by using the Occbin toolbox developed by Guerrieri and Iacoviello (2015). For the calibration of the benchmark model we use the parameter values displayed in Table 1. These values are commonly used in the relevant literature with the aim of matching data from the Eurozone.
Table 1

Calibration parameters

Parameter

Description

Value

β s

discount factor of savers

0.99

β b

discount factor of borrowers

0.975

η

parameter associated with labour elasticity

2

j

weight of housing in utility function

0.1

l

steady state LTV ratio

0.7

α

labor share of savers

0.64

ϕ p

price adjustment cost parameter

58

ε

price elasticity of demand

6

ρ r

interest rate smoothing parameter in Taylor rule

0.8

ω y

output parameter in Taylor rule

0.1

ω π

inflation parameter in Taylor rule

2

ρ l

smoothing parameter in LTV rule

0.2

χ l

reaction parameter in LTV rule

-2

ρ κ

persistence of preference shock

0.9

σ κ

standard deviation of preference shock

0.02

4 Brexit shocks: some illustrative exercises

4.1 Brexit as an adverse demand shock

The probably most obvious implication of Brexit is that it reduces demand for goods and services in the Eurozone at least temporarily. This can be motivated by shrinking incomes in the UK, a depreciation of the British Pound relative to the Euro but also by a more sceptical – if not outright hostile – attitude towards continental EU countries that some British policymakers are clearly trying to evoke. In Fig. 1 we therefore look at the effects (impulse response functions) of a temporary preference shock. In terms of the model outlined in Section 3, this can be achieved by a temporary reduction of κt, which makes – relative to the steady state – both savers and borrowers less willing to consume and therefore leads to a postponement of consumption and thus to lower aggregate demand. More precisely, we assume a 2% reduction of κt in t = 1 which later on evolves according to Eq. 1. The solid green line depicts the situation with a FMR in place, whereas the dotted red line assumes a constant LTV ratio. Note that in the logic of the model the LTV (whether flexible or not) is always binding due to the different discount rates of savers and borrowers.
Fig. 1

Impact of an adverse demand shock with ZLB (baseline scenario)

This shock is strong enough to make the ZLB binding for a sustained period of time. Since inflation goes down by more than nominal interest rates, the real interest rate increases. This explains the decrease of indebtedness. Perhaps somewhat counterintuitively, house prices increase due to the adverse demand shock. But this is easily explained by the fact that housing is the only aggregate saving vehicle. The countercyclical increase of the LTV ratio in the FMR scenario almost completely stabilizes debt. However, FMR, also stablizes output and inflation to some extent. In Jerger and Körner (2018) we show that this stabilizing role of FMR for output and inflation vanishes if the ZLB is assumed not to bind. In this case, monetary policy is sufficiently powerful to bring down real interest rates after an adverse demand shock. Even in this case, however, FMR stabilizes the level of indebtedness greatly.

Whereas our results so far show at least a possible stabilizing role for FMR, we would like to note that this result does not generally hold. If we assume a slightly lower volatility for the preference shock (σk = 0,015 instead of the baseline value of σk = 0,02), our scenario still leads to a situation where the ZLB binds. Real interest rates, however, now decrease; this in turn leads to an increase of debt but also to less pronounced reactions of inflation and output. In this situation, FMR thus primarily stabilizes the debt level. See Fig. 2 for the full set of impulse response functions.
Fig. 2

Impact of an adverse demand shock (σk = 0,015)

4.2 Brexit as an adverse supply shock

Whereas the interpretation of Brexit as a demand shock is rather straightforward, it might also be thought of as an adverse supply shock. Again, a couple of channels suggest themselves. First, Brexit might render some capacities rendundant that before Brexit are/were used to serve the UK market; second, the more and more sophisticated transboundary division of labour – trade in intermediate goods or trade in tasks (Grossman and Rossi-Hansberg 2008) – will be disrupted to some extent, leading to a decline in overall productivity; third, trade costs and most plausibly also other kinds of transaction costs will increase. Again, our model does not lend itself to look at different steady state levels of the macroeconomic variables. But it is interesting enough to simulate an adverse suppy shock. We do so by assuming a temporary productivity decrease of 3%, i.e. setting γt = 0,97 for one period.3 The corresponding impulse response functions can be seen in Fig. 3.
Fig. 3

Impact of an adverse supply shock

The adverse supply shock drives up prices and lowers output as one would expect. By depressing real income and making consumption more expensive, it also leads to a decrease of house prices. A countercyclical LTV ratio helps to stabilize the level of debt but does not help to stabilize output and inflation.

4.3 Brexit as a shock to ECB policy

The ECB has a clear mandate to prioritize price stability. Nevertheless, policy makers look at all aspects of the macroeconomic environment – and are even obliged to do so as long as the goal of price stability is not compromised. Since Brexit poses a major real economic shock (whether from the demand or the supply side), the ECB might consider to give a relative bigger weight to its output goal – and by implication a lower weight to its inflation goal. We depict this situation by looking at the effects of an increase of the weight ωy in the Taylor rule (8) from its benchmark weight of 0.1 to 0.3. Note that this still implies a much higher weight of the inflation goal (ωπ = 2). Figures 4 and 5 show the effects of the benchmark demand and supply shock. Therefore, they can be directly compared to Figs. 1 and 3, respectively.
Fig. 4

Impact of a demand shock when the central bank cares more about the output goal

Fig. 5

Impact of a supply shock when the central bank cares more about the output goal

It is worth noting that the stabilizing role of FMR for output and inflation after a demand shock is no longer present. This can be interpreted such that FMR is a substitute for a higher weight of inflation in the policy function. But as in all other scenarios, a FMR is very effective with regard to the stabilization of debt.

4.4 Brexit as a shock to price adjustment costs

Price adjustment costs in macroeconomic models serve the purpose of introducing some sand in the wheels of instantaneous market clearing. And even if there is no straightforward link between Brexit and price adjustment costs in a literal sense, Brexit certainly throws additional and cruder sand in those wheels by imposing additional costs to firms. One straightforward channel for this is the fact that firms will find it harder to evaluate/judge their competitive enviroment after Brexit and therefore will be more reluctant to change prices. Hence, for this exercise, we look at a higher value of ϕp which is assumed to increase from its benchmark value of 58 to 100. Again, this change can only be evaluated by looking at specific shocks – the demand and supply shocks (see Figs. 6 and 7, respectively) we analyzed before.
Fig. 6

Impact of a demand shock (ϕp = 100)

Fig. 7

Impact of a supply shock (ϕp = 100)

As expected, the additional amount of price inertia leads to a less pronounced reaction of all variables relative to the benchmark of Figs. 1 and 3, respectively. As in the previous scenarios, FMR looses its stabilizing properties with regard to inflation and output after a demand shock to quite some extent. Hence FMR is less useful – or less needed – for stabilizing the macroeconomy if prices are stickier. However, the role of FMR for the stabilization of debt after the shocks remains intact.

5 Conclusions

This paper applies the DSGE framework of Rubio and Carrasco-Gallego (2015) and Jerger and Körner (2018) to a set of exercises that are intended to capture some aspects of Brexit. Unlike in most other studies on the possible effects of Brexit, we focus on the differences in the dynamic adjustment paths after some shocks. We specifically look at the potential of flexible macroprudential regulation (FMR) to serve as an additional stabilization device.

Whereas our results are always in the affirmative concerning the positive role of FMR for the stabilization of debt, this is much less (or not at all) the case for the stabilization of output and inflation. FMR thus remains a good idea. A decisive value added for the mitigation of the Brexit shocks is only visible for a severe demand shock in the benchmark scenario.

Footnotes

  1. 1.

    In the arguably most restrictive database for scientific publications in economics a search for Brexit AND effect (OR some synonyma) returns 172 papers; the same exercise in Google Scholar leads to 37.900 (sic!) hits.

  2. 2.

    Queen Elizabeth – whose private wealth was also severly affected by the crisis – famously asked why nobody saw the crisis coming during her first visit ever to the London School of Economics in late 2008. It would be interesting to discuss the nature and depth of the professional ignorance. But the profession should be very clear about the impossibility to forecast specific and rare events and their timing. Here it might suffice to note that there have been quite a few voices around who pointed to patterns in the sense of von Hayek (2007) that were destabilizing financial markets. This is also true with respect to the recent de-regulation of financial markets in the U.S. by the Trump administration.

  3. 3.

    We do not assume an autoregressive process for γt here, because the effects we get would not be qualitatively different, albeit more pronounced.

Notes

Acknowledgments

We are grateful for comments of participants of the workshop “The Influence of Brexit on the EU28: Banking and Capital Market Adjustments plus Macro Perspectives” on October 12, 2018 in Frankfurt/Main.

References

  1. Aliber RZ, Kindleberger CP (2017) Manias, panics, and crashes: a history of financial crises, 7th ednGoogle Scholar
  2. Bank of England (2018) EU withdrawal scenarios and monetary and financial stability A response to the House of Commons Treasury Committee. (November), 1–88Google Scholar
  3. Borio CEV, Shim I (2008) What can (macro-) prudential policy do to support monetary policy?Google Scholar
  4. Busch B, Matthes J (2016) Brexit-the economic impact: a meta-analysis. Technical report, IW-ReportGoogle Scholar
  5. Dehmej S, Couppey-Soubeyran J (2017, 06) The role of macro-prudential policy in the prevention and correction of imbalances in the euro area. Technical reportGoogle Scholar
  6. Farhi E, Werning I (2016) A theory of macroprudential policies in the presence of nominal rigidities. Econometrica 84(5):1645–1704CrossRefGoogle Scholar
  7. Grossman GM, Rossi-Hansberg E (2008) Trading tasks: a simple theory of offshoring. American Economic ReviewGoogle Scholar
  8. Guerrieri L, Iacoviello M (2015) OccBin: a toolkit for solving dynamic models with occasionally binding constraints easily. J Monet Econ 70:22–38CrossRefGoogle Scholar
  9. Iacoviello M, Neri S (2010) Housing market spillovers: evidence from an estimated DSGE Model. Amer Econ J Macroecon 2:125–164CrossRefGoogle Scholar
  10. Jerger J, Körner J (2018) Assessing macroprudential regulation: the role of the zero lower bound. Appl Econ Lett 25(9):580–583CrossRefGoogle Scholar
  11. Klingelhöfer J, Sun R (2017) Macroprudential policy, central banks and financial stability: evidence from ChinaGoogle Scholar
  12. Korinek A, Simsek A (2016) Liquidity trap and excessive leverage. Amer Econ Rev 106(3):699–738CrossRefGoogle Scholar
  13. Lambertini L, Mendicino C, Punzi MT (2013) Leaning against boom-bust cycles in credit and housing prices. J Econ Dyn Control 37(8):1500–1522CrossRefGoogle Scholar
  14. Minsky HP (1977) The financial instability hypothesis: an interpretation of Keynes and an alternative to “standard” theory. ChallengeGoogle Scholar
  15. Patel O, Reh C (2016) Brexit: the consequences for the EU’s political system. UCL Constitution Unit Briefing Paper 1–5Google Scholar
  16. Rubio M, Carrasco-Gallego JA (2015) Macroprudential and monetary policy rules: a welfare analysis. Manch Sch 83(2):127–152CrossRefGoogle Scholar
  17. Rubio M, Yao F (2017) Macroprudential policies in a low interest-rate environment. Working Paper (April 2017)Google Scholar
  18. Schoof U, Petersen T, Aichele R, Felbermayr G (2015) Brexit ? potential economic consequences if the UK exits the EU. Bertelsmann StiftungGoogle Scholar
  19. von Hayek FA (2007) Die Theorie komplexer Phänomene. In: Gesammelte Schriften, A 1, Wirtschaftstheorie und WissenGoogle Scholar
  20. Wadsworth J, Dhingra S, Ottaviano G, Van Reenen J (2016) Brexit and the impact of immigration on the UK. CEP Brexit Anal 5:34–53Google Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2019

Authors and Affiliations

  1. 1.University of RegensburgRegensburgGermany
  2. 2.Leibniz Institute for East and Southeast European StudiesRegensburgGermany
  3. 3.European Central BankFrankfurtGermany

Personalised recommendations