International Economics and Economic Policy

, Volume 16, Issue 1, pp 193–213 | Cite as

The UK’s banking FDI flows and Total British FDI: a dynamic BREXIT analysis

  • Fabian J. BaierEmail author
  • Paul J. J. Welfens
Original Paper


The City of London has been the global leader for the provision of international banking services since the 1980s when Thatcher-era deregulation, followed by the EU single market program, stimulated big international FDI inflows – mainly of US banks – into the UK. The “single passport” rule allowed international banks in the UK to serve the whole of the EU28 market from London whose supply-side dynamics contributed to economic growth in the UK and a rising output share of the UK banking system in British GDP. With the expected BREXIT, there are serious challenges for the City since the passporting of banks will end and the regulatory framework will be adjusted; EU equivalence rules for UK banks that might be valid after the implementation of BREXIT cannot be a substitute for passporting so that lower FDI inflows and higher FDI outflows in the banking sector should be expected; inflow dynamics should also be shaped by international M&A dynamics influenced by the real Pound depreciation in 2016, while the prospects of reduced EU market access post-BREXIT also became relevant in 2017/18 and should influence the FDI dynamics of the UK – a similar pattern might occur in the BREXIT implementation year (i.e. 2019) and the following adjustment period where the change in City banks’ access to the single market will matter; as regards the latter, quasi-tariff-jumping FDI outflows from the UK can be expected where the FDI of City of London banks could go primarily to the EU27/Eurozone or the US. The empirical findings confirm the expected FDI pattern for the UK banking sector – overall FDI inflows in the wake of the BREXIT referendum have increased, in line with the Froot-Stein effect, while FDI inflows to the UK banking sector have declined.


Brexit UK FDI Banking Financial markets 



United States


Foreign Direct Investment


United Kingdom


Gross Domestic Product


European Union


Mergers & Acquisitions


European Systemic Risk Board


European Banking Authority


Asian Infrastructure Investment Bank


European Securities and Market Authority


Organisation for Economic Co-operation and Development


European Central Bank


Bank for International Settlements


United Nations Conference on Trade and Development


Locational Banking Statistics


Coordinated Portfolio Investment Surveys


International Monetary Fund


Balance Sheet Items Statistics


Pseudo Poisson Maximum Likelihood


Loan to Value

JEL classification

F02 F4 F21 G1 G2 

1 Introduction

The Thatcher-era banking deregulation of the 1980s resulted in banking FDI flowing into London over a number of decades fuelling the subsequent expansion of the UK financial services sector which increasingly has been able to successfully serve EU27 (EU28 without the UK) clients from London’s financial center. In this context, exploiting economies of scope – locational advantages of the City of London – as well as economies of scale for specific banking transactions has contributed to the growth of the London banking center which, of course, has benefitted from the EU single market implemented in 1993 (with some years of delay for all financial services, including insurance). EU regulations have contributed to creating and restoring confidence in the London financial center with its many foreign banks, among them subsidiaries or branches of all leading banks from the US, Europe, Japan; and after 2000, also from China. The UK even quickly became a founding country of the multilateral Asian Infrastructure Investment Bank (AIIB) once China had agreed to set up the European AIIB subsidiary in London.

With BREXIT, the UK will face a serious challenge to the London banking system which is a global financial center and also the leading banking center in the EU28 (Coeuré 2017; Donnery 2017). The estimate for the market share of London, i.e. “City”, banks in the EU27 wholesale market has been close to 90% (Sapir et al. 2017). The UK has had a competitive banking services supply side; with some notable exceptions - as has emerged after the Transatlantic Banking Crisis where investigations have revealed that the Libor interest rate was rigged by the group of banks involved in calculating this important reference interest rate. Foreign direct investment (FDI) into the UK has contributed to making the City, London’s financial center, the biggest such center in the world and for decades the UK has been involved in the designing and implementation of EU financial services and banking regulations. These were revised and the relevant institutions modernized – including the creation of the European Systemic Risk Board (ESRB) in 2010 – after the banking crisis. EU institutions, including the European Commission and the European Banking Authority (EBA) have been part of the institutional setting under which foreign banks could be active in the UK, while national regulations have been complementary to the EU framework after the Transatlantic Banking Crisis of 2008/09, which affected many UK banks with activities in real estate in the US (the write-down of asset values in the US thus affected both US and British banks in London and contributed to higher spreads; see Born and Enders 2018).

While the government of Prime Minister May wishes to retain access to goods markets in the EU through an UK-EU free trade agreement, the policy stance in the field of financial services is to stay out of the EU single market and to rather rely on a new equivalence regime that could partly be a substitute for the current passporting of banks which is a system that allows any EU bank with a banking licence in one EU country to offer banking services in all EU countries, namely through a branch or through a subsidiary. For many financial services offered in the EU27, the ECB expects that City banks will set up a subsidiary in the EU27 if current business is to be carried on in full, while the ESMA (European Securities and Market Authority) could also require that specific activities have to be set up with adequate capitalization, banking infrastructure and staff in the EU27. Against this background, one may anticipate that there will be reduced banking FDI inflows to the UK in the context of BREXIT and more banking FDI inflows to the EU27. Gravity modeling on overall FDI in OECD countries has shown that EU membership as well as participation in the EU single market is a significant variable in terms of raised FDI inflows so that leaving the EU will reduce the inward FDI of the UK in the context of BREXIT (WELFENS/BAIER, 2018). Lower overall UK FDI inflows and, in particular, lower banking FDI inflows into the UK could have an effect on the UK’s overall portfolio capital inflows. Eichengreen (2018) has shown that the inward stock of portfolio investment is reduced by 12% for the UK post-BREXIT. This implies that London, as a financial center, could face lower liquidity and profits after BREXIT and foreign investors from the banking sector will anticipate these developments; other aspects could also affect inward banking FDI as well as overall FDI. However, in the subsequent analysis the key aspect is not so much true banking FDI, for which data are unavailable, but rather outward and inward deposits assets of banks in banking and non-banking sectors, respectively. The following analysis takes a closer look at the question of how outward banking activities are affected by BREXIT and also how inward banking activities can be expected to react with respect to BREXIT.

Looking effectively at the prospect for the UK banks’ outward FDI in the banking sector and the non-banking sector, one would have to anticipate the post-BREXIT trade and regulatory regime for the UK and the EU27. It seems clear that the worsening of the London City banks’ access to EU27 markets is a realistic case even if some equivalence rules will be implemented. Since the overall investment of UK banks in non-banking is somewhat smaller than in banking (in terms of foreign assets), the BREXIT will affect the relative international asset position of the banking sector to non-banking and thus the international risk exposure of UK banks; and this normally should lead to regulatory adjustments in the UK and in the EU27.

Based on a theoretical analysis one can analyze the UK’s banking FDI outward prospects to EU27 and other countries in an empirical framework and then also draw policy conclusions for the UK, the EU27 and other countries. The key insights will be the asymmetric FDI dynamics in the UK’s banking and non-banking FDI – and, more broadly speaking, the internationalization of banks’ assets and of non-banking firms’ assets.

2 Theoretical aspects

As regards the economic logic of banking FDI, the general approach of combining ownership-specific advantages, locational advantages and internalization advantages has played a role with regard to the cumulated FDI inflows (Dunning 1998, 2001). Economies of scale/density and networking effects also played a particular role for London as a banking center (Gehrig 1998). As regards international mergers & acquisitions (M&As) one can clearly point to the role of the real exchange rate as emphasized by (Froot and Stein 1991): A real devaluation will lead to higher overall FDI inflows. One may, however, consider sector-specific aspects and to the extent that the focus is on banking FDI, the changes in post-BREXIT City of London access to the EU single market suggests a strongly declining attractiveness of the UK for banking FDI; hence FDI banking inflows to the UK should decline. To the extent that US banks in the City would relocate to New York, in doing so they would benefit from existing equivalence agreements between the US and the EU which largely maintains that the regulatory quality in the US is equivalent to that in the EU so that banks located in the US could offer services from the US to EU27 clients within the level playing field of a transatlantic financial market.

As regards the role of rather low corporate tax rates, one can also argue that there is a positive link to FDI inflows in OECD countries as the empirical gravity equation of (Welfens and Baier 2018) has shown. The Bank for International Settlements has looked into some of the key aspects of the internationalization of the banking sector (Brei and Von Peter 2017).

To the extent that British banks could develop specific ownership advantages, this would contribute to the UK’s financial services current account surplus on the one hand, on the other hand such banks also gain an improved basis for becoming successful with outward FDI. Until the pro-BREXIT majority in the EU referendum of 2016, there was, however, no big incentive for UK-based banks to set up major activities in the EU27. This has broadly changed after 2016 since the European Central Bank (ECB) and European Banking Authority (EBA) have signaled that London banks interested in offering the full range of standard banking activities would have to create subsidiaries in the Eurozone so that branches of primarily London-based banks would have to become subsidiaries with a distinct source capitalization.

Overall, foreign direct investment inflows into the UK have increased in the British EU referendum year of 2016, but FDI inflows to the banking sector and the financial services sector should generally fall if banks are not anticipating that the EU passport solution for banks would hold after the UK’s leaving of the European Union; and also not anticipating that equivalence rules would be broad enough to be an effective substitute for British and non-British banks in London to continue their pre-BREXIT level of banking and financial services provided to EU27 clients from the UK. To the extent that neither continued passporting nor broad equivalence is possible from a UK perspective, BREXIT brings the following analytical outlooks:
  • The short-term impact of BREXIT will be a strong nominal Pound depreciation which will facilitate international M&As in all sectors initiated by foreign investors in the UK; this is an argument, based on (Froot and Stein 1991), to expect an expansion of UK FDI inflows, namely in a world of imperfect capital markets;

  • as most simulations for BREXIT imply a long-run reduction of economic growth in the UK, this dynamic negative market size effect will dampen the appetite of foreign investors to invest strongly in the UK. The net impact of the above real exchange rate effect and the market size-dampening effect of BREXIT should be additional FDI inflows. One should note, however, that uncertainty about the reduction of medium-term UK economic growth will depend on the outcome of the UK-EU negotiations in 2019 – the broader the future British market access to the EU27 single market, the smaller the growth-dampening effect of BREXIT will be. Only in the worst case scenario of a No-deal BREXIT could one expect that the market size effect would quickly dominate the real exchange rate effect;

  • as regards the sectoral banking perspective, there are two impulses for relocating UK banking activities abroad, namely partly to the EU27 and partly to the US (New York or other US financial centers). US banks with London subsidiaries will often have a tendency to relocate activities back to New York which for certain financial transactions would represent a second-best economies-of-scale solution;

  • one particular aspect of banking FDI in some fields could be oligopolistic interdependency which has been emphasized from a theoretical perspective – and with a focus on many sectors with oligopolistic structures – by Knickerbocker (1973). This could imply that once a major US investment bank in London decides to relocate activities i back to the US and relocate banking activities j to the EU27, other US banks would follow suit. Moreover, as soon as one big British bank relocates to the EU27, other big UK banks will likely follow.

The subsequent empirical analysis will focus on the key issue of to what extent BREXIT has affected FDI – or rather general total investment flows into banking in the UK, on the one hand, and British outflows on the other. The analysis naturally considers the development of the nominal exchange rate for the British Pound over the period 1977–2018, while the nominal interest rate vis-à-vis the US$ will be of particular relevance.

3 Empirical analysis

The first problem we face when analyzing the structure of foreign investments and assets in the UK banking and non-banking sectors is that a broader databank for bilateral sectoral FDI data does not exist; while the UNCTAD and OECD databanks provide bilateral FDI data, they do not do so on a sectoral level. Although both provide sectoral data, this data is not on a bilateral but rather an aggregated level. Therefore, our first challenge is to find alternative data which describes investment patterns in the UK banking sector and the UK non-banking sector, which is our primary contribution to economic science.

On a national level, some countries do collect micro-data on their firms, including banks, and their investments abroad, such as the MiDi databank of the Deutsche Bundesbank. However, this is done on an individual country level and consolidated statistics or databanks do not exist up to this point. Data on international banking activity (structured with loans/assets/deposits) is also collected on a country level, and then reported to the Bank for International Settlements (BIS). The BIS collects this information relating to the banks of 44 reporting countries and publishes a quarter-wise dataset, the “Locational Banking Statistics” (LBS). Following the bilateral and sectoral structure of the dataset, it becomes evident that this can be analyzed using gravity models quite well.

3.1 Gravity models for FDI and banking

The term gravity model is used in general to describe models in which the economic size of two trading countries, measured in terms of GDP, is a trade stimulating factor, whereas the distance between those two countries restricts or dampens trade between them (Tinbergen 1962). A solid theory evolved around this, at first purely intuitive, model with (Anderson and Van Wincoop 2003), who provided a solid econometrical basis considering the consumption- and production-side of countries and transportation costs. In order to capture multilateral in- and outward resistance,1 the fixed effects panel data estimation method has established itself as useful, see Anderson (2011) and (Head and Mayer 2014). Those models were extended with dyadic fixed effects by Bruno et al. (2016) for FDI BREXIT analyses, and refined by (Welfens and Baier 2018) with respect to zero and negative FDI flows and missing values, getting an even more precise model. The latter is considered as the current state-of-the-art model when analyzing multilateral panel (dyadic) fixed effects gravity FDI.

(Brei and Von Peter 2017) use the locational banking statistics (LBS) data of the BIS to measure the distance effect in banking, and base their methodology on (Head and Mayer 2014). They provide a rather short, but up-to-date literature review on gravity models in international banking and finance focusing on transaction and monitoring costs, which seem to be linear with respect to the distance between an investing bank in country i and the subject of the investment in country j. This proxies relative frictions limiting the volume of transactions between countries. Frictions in international banking can arise due to issues such as risk assessment,2 information frictions and asymmetries,3 including search costs. (Lane and Milesi-Ferretti 2008) analyze bilateral factors driving portfolio equity holdings across countries, using the Coordinated Portfolio Investment Surveys (CPIS) of the IMF. This dataset was also used by (Okawa and Van Wincoop 2012) to examine asset trade and the cross-border financial frictions which underlie them. The authors also provide a broad literature review on papers which use BIS data in a gravity framework in order to analyze external claims by banks.4 This confirms the structural convenience of the dataset for gravity estimations with particular respect to precision and quality of the data it contains. The dataset will be described in more detail below.

As identified above, it is not only distance-related frictions which arise in the field of international investment and banking, where traditionally we do not have transportation costs or other similar costs. In analyzing significant structural policy changes, such as BREXIT, where the distance between partners does not change in terms of the number of miles or kilometers between them, we rather focus on frictions such as the ending of the “single passport” rule, possible exclusion from the EU single market and exchange rate dynamics, all of which are important for M&A, but also greenfield investment decisions.

3.2 BREXIT and the effect on investments in banking and industry

On the one hand, global investment bankers face higher risks when investing in UK banks, as there has been no structured BREXIT plan delivered up to this point and investors have been left in uncertainty for the past 2 years (i.e., from 23 June 2016 to December 2018). While there have been discussions with the EU27 on maintaining free trade in goods, but not in services, aspects such as FDI restrictions have been quite neglected in the negotiation process, the omission of the “single passport” rule, which allows international banks in the UK to serve the EU27 market from London, increases uncertainty about future investment in the UK banking sector. The risk of a No-deal BREXIT followed by considerably increasing transaction barriers exists. Even in the event of a soft BREXIT, the international market share of UK banking services will decrease. Decreasing growth in the UK economy is dampening the appetite of investors to decide to invest in the UK banking sector. Business is currently being transferred to New York, Dublin, Amsterdam and Paris, not only due to increasing cross-border frictions but also for reasons of oligopolistic interdependencies.

On the other hand, the Pound depreciation in the last 2 years can lead to increasing brownfield investment, as international investors can “buy out” UK plants more easily and more cheaply. Increasing trade barriers can also be an additional reason for increasing investments in the producing sector: Following Dunning’s (1979) eclectic paradigm, especially location-specific advantages such as production-to-market and supply-chain maintenance for UK industry are arguments for increasing investments especially during the run-up to the implementation of BREXIT. Considering the effects of corporate tax rates on FDI inflows (Welfens and Baier 2018), the announcements in September 2018 of Prime Minister Theresa May with regard to cutting the UK’s corporate tax rate to the lowest in G20 is likely to foster foreign investments in the UK’s industrial sector.

Vice versa, when we look at UK investments in the EU and the world, one expects that especially investments in the banking sector are spreading, while investments in the industry are stagnating.

3.3 Data and model specification

We use data on the assets, loans and deposits of global banks in the UK, and of UK banks in the world, respectively, provided by the CBS from the BIS, which describe cross-border banking. In order to understand how this data is compiled in detail, we look at the raw data format which is reported to the BIS by individual countries:
  • Data is delivered to the BIS on a monthly basis, on the external positions recorded on the balance sheets of domestic banks.

  • The format of the data is prescribed by the ECB’s Balance Sheet Items Statistics (BSI); all domestic banks (including the domestic offices of foreign banks) are required to report the aforementioned data according to the ECB format.

  • The data is used to observe the global activities of domestic banks; they serve as an input for monetary and balance-sheet statistic aggregates and are the basis for the LBS

  • Assets and liabilities include indebtedness certificates, bonds and securities, investment asset pools, special purpose entities used for the purposes of asset securitization, firm derivatives, banknotes and coins.5

Therefore, in the present analysis we rather look at bilateral financial positions, and not at fixed assets or real FDI. The currently available data ranges from 1977 to the first quarter of 2018, reported on a quarterly basis (as of September 23, 2018). In general, the CBS has global data, but availability can depend on the willingness of countries’ central banks to report.6 We use that data in a country-to-country format where we match data reported by the UK with the data of the counterpart country, following (Brei and Von Peter 2017), and find that the bilateral data is of an overall good quality compared to bilateral FDI data from OECD or UNCTAD (with few differences in counterparty data, where deviation does occur these are relatively small in dimension). We always use the highest number reported. The UK reports data for only 30 counterpart countries (and offshore centers) who are investing in UK banks and industries, resulting in 4860 observations.7 The most important partners, including the US, France, Germany, Netherlands, Japan etc., are covered, also covered are many countries often regarded as tax havens such as Switzerland, Hong Kong, and Luxembourg. We are therefore confident that our dataset presents a representative and meaningful sample size.

Figures 1 and 2 show the financial claims of global banks in the UK and vice versa. It cannot be determined definitively what effect the BREXIT vote in particular has had, as many factors, which would need to be controlled for in an extended statistical analysis, influence (especially short-term) investment positions. The generally increasing interdependence of global investments from the 1980s on and the 2008 financial and banking crisis, which started with the collapse of the Lehman Brothers investment bank in New York in September of that year, can be seen quite clearly.
Fig. 1

Financial Claims of Global Banks in UK, by Sector. Source: Own calculations; combining the assets/liabilities of UK partner countries, based on data reported by BIS (timeframe 1977–2018)

Fig. 2

Financial Claims of UK Banks in the World, by Sector. Source: Own calculations; combining the assets/liabilities of UK partner countries, based on data reported by BIS (timeframe 1977–2018)

Note that while all claims (and assets as counterpart) are reported in current USD, they have originally been recorded in countries’ own currencies and British Pounds, respectively, and therefore are affected by a different set of exchange rates to USD. When controlling for exchange rates to USD later, via time fixed effects and a control variable, this negatively affects the integrity our data to a small extent, as the consolidated data relates to many different exchange rates and not only those relating to the USD. Particularly affected in this regard are holdings of coins and banknotes as direct cash reserves, which might lead to a distortion of our results. Therefore it is of interest to determine what magnitude holdings of coins and banknotes are included; as we have no access to internal Bank of England data, we check data on German cash reserves in foreign currencies total (as only aggregates are accessible) and find that they account for only a marginal fraction of total foreign assets (about 0.01%).8 Taking this as a benchmark and even assuming that UK banks would hold more foreign coins and cash due to dimensional differences (significance of Euro(zone)-Pound transactions), we feel confident in neglecting this aspect in our further analyses.

It should be noted that asset position changes could reflect both FDI flows as well as portfolio flows. FDI stocks are expected to play a rather strong role in the financial sector – on obvious reputational grounds which require a strategic investor and often strong control from the equity side; FDI stocks are also expected to play a strong role in technology-intensive sectors where foreign investors would typically seek 100% ownership in high-technology sectors (Jungmittag and Welfens 2016). The available database does not allow to make a distinction in terms of FDI versus portfolio capital flows in the respective sectors considered in the case of the UK.

In addition, as we see no possibility to control for exchange rate splits (Euro, USD, Yen etc.) described above, due to the nature of the collected data, and considering that previous studies using said data neglected this effect completely, we see the need to point this aspect out but do not correct for it in our analysis. To show that this effect should be quite small, consider the main currencies’ exchange rates individually.

Figure 3 and Table 1 show the split of investments in the UK of the seven biggest investors in the UK; one can see that the US is traditionally the biggest partner of the UK, followed by France which started catching up with Germany after the financial crisis of 2008. Switzerland traditionally had a very large share of total investments in the UK in relative terms, but did not experience a rise of investment as for example the US did, especially since the turn of the millennium, as we can see from the absolute numbers in Table 1.
Fig. 3

Financial Claims in the UK of the UK’s Top Seven Partners. Source: Own calculations; combining assets/liabilities of UK partner countries, data reported by BIS reported data (timeframe 1977–2018)

Table 1

Financial Claims in the UK of the UK’s Top Seven Partners, in bn USD







US:United States










































Own calculations; combining assets/liabilities of UK partner countries, data reported by BIS (timeframe 1977–2018)

Our study uses these data as dependent variables in a gravity setting, resulting in six different models. Models (1)–(3) use global investments in the UK in total and by sector (banking and non-banking), while models (10)–(12) use global UK investments (total, by sector).9 The GDPs of countries traditionally serve as control variables for the economic size of the respective countries and sectors, while we control via country fixed effects for a broad range of individual unobservable factors which are time-invariant (such as language, culture, history, whether countries share a border or have access to the sea, but also for example the World Bank’s “doing business” indicator which is time non-varying for many countries). The panel structure controls for years and quarters, and therefore also for time-variant USD exchange rates, assumed it evolves relatively constant.10 We therefore compile the following model:
$$ \ln POSITIO{N}_d={\alpha}_0+{\alpha}_1\mathit{\ln}{GDP}_{ot}+{\alpha}_2\mathit{\ln}{GDP}_{dt}+{v}_d+{\delta}_o+{\tau}_t+{e}_{odt} $$

α0 regression constant,

GDPot time-variant characteristics of the origin country such as GDP,

GDPdt time-variant characteristics of the destiny country such as GDP,

vd destination event BREXIT vote,

δo country fixed effects for the origin country,

τt time fixed effects,

eodt error term.

We additionally control in models (10)–(12) for whether the origin/destination country is an EU27 member, in order to check whether EU27 countries have different investment patterns when it comes to their decision to invest in the UK. We capture the BREXIT effect with a dummy variable which switches from 0 to 1 as of the third quarter of 2016, i.e. 8 days after UK citizens voted to leave the EU, following (Friewald and Subrahmanyam 2012) methodology. A poisson pseudo maximum likelihood (PPML) setting is utilized (see Welfens and Baier 2018 for a review).
Compared to traditional gravity modeling, we therefore use a single-country model and deviate from (Anderson and Van Wincoop 2003) regarding multilateral resistance, losing out on the explanatory power of for example ‘how do German investments in UK change if the US-German investment relationship changes’ to use an example of two very important partners of the UK. However, via country and time fixed effects, we control for a broad part of this. What we gain by using this design is that we can measure UK-specific effects, as we only look at inflows and investment into the UK as well as UK outflows. With a general gravity model, country specific measurements become more difficult, usually quantitative results have to be applied with qualitative arguments to mask certain countries of interest. Moreover, predictions are somewhat hard to discern and defend with this model; our goal is to show whether and how investment patterns have changed due to the BREXIT vote. We want to find answers to our main hypotheses:
  • Hypothesis 1: Mid- and long-term banking investments in the UK (total) will drop due to increasing frictions and less economic growth. UK investments abroad will drop for the same reasons.

  • Hypothesis 2: The depreciation of the Pound after the BREXIT vote fosters brownfield investment and therefore international financial claims will rise particularly in the non-banking sector. Mirror effect: The UK is more likely to invest less in the non-banking sector abroad.

  • Hypothesis 3: EU27 regulatory pressure on London City banks to relocate EU wholesale banking activities to the EU/Eurozone; increasing risks and frictions, a smaller market share of the UK banking industry (in the UK as an EU27 wholesale market) and oligopolistic interdependencies in the banking sector will reduce international financial claims in the banking sector. The mirror effect will be that the UK is more likely to invest more in the banking sector abroad.

The subsequent analysis presents empirical evidence on these hypotheses – with the already mentioned caveat in terms of available data.

3.4 Empirical findings

The findings for models (1)–(3), which analyze monetary inflow, investments and claims of foreign banks in the UK, are presented in Table 2. All dependent variables were first tested for stationarity.11
Table 2

Claims of Foreign Banks in the UK, in all sectors (Non-banking and Banking), 1977–2018




















































Standard errors in parentheses, *** p < 0.01, ** p < 0.05, * p < 0.1

The partner_eu variable indicates whether the home country of the investing bank is an EU27 member, brexitvote is a dummy variable which switches to 1 at the third quarter of 2016. Time fixed effects and country fixed effects were applied in all models, but not displayed for ease of interpretation of the results. The relatively high R-squared in all models show that our multiple fixed effects models have good explanatory power. Model (1) shows the total claims of foreign banks in the UK in all sectors, while model (2) and model (3) distinguish between the non-banking sector (2) and the banking sector (3). Countries’ GDPs are positive and significant in all models, which is in line with the usual gravity theory.

For overall investments in the UK, we find a negative coefficient for the BREXIT vote which is not significant. This indicates that total investments in the UK did not decrease significantly since mid-2016. However, when splitting investments into sectors, we find a positive and significant effect of brexitvote in the non-banking sector model (2) (0.207**, standard error 0.094), while we find a negative significant effect for the banking sector model (3) (−0.521***, standard error 0.112), indicating that since the BREXIT referendum, banks have increased their investments in non-banking operations, but decreased investments in the banking sector. For robustness, we estimated the same model including annual average nominal exchange rates for the USD to British pound in Table 4 (see Appendix),12 where we find that we lose out on the significance of the BREXIT vote with respect to non-banking investments (model 5), but gain significance for a negative overall impact on investments in UK (model 4). This underlines Hypothesis 2 where we argue that due to the BREXIT vote and the subsequent Pound depreciation, we find increasing investments in the non-banking sector, and supports the (Barrell and Pain 1997) theoretical model. The coefficients of all other control variables remain unaffected, adding to the robustness of the analyses. When controlling for correlation in Table 6 (in Appendix), we find the expected negative correlation for brexitvote and dollar_pound_rate, which supports our arguments.

When the ‘nationality’ of the investing bank is that of an EU27 member country, it only matters significantly for investments in the non-banking sector model (2) (0.204***, standard error 0.055): EU27 countries invest significantly more in the UK non-banking sector.13 When checking for the Eurozone instead of EU27 membership, we find the same results for brexitvote in models (7)–(9), but the Eurozone holds significantly more claims in the UK – especially in UK banks – than the rest of the world (Table 5, in Appendix). Subsequently we find that non-Eurozone EU27 countries are more likely invest in UK industry, while Eurozone EU27 countries are more likely to invest in EU banks.

Table 3 shows the findings for models (10)–(12), where we analyze the investment pattern of UK banks outside their home country.
Table 3

Claims of UK Banks in the World, in all sectors (Non-banking and Banking), 1977–2018




















































Standard errors in parentheses, *** p < 0.01, ** p < 0.05, * p < 0.1

Table 4

Claims of Foreign Banks in the UK, Additionally Controlling for Average Annual Exchange Rates, 1977–2018



























































Standard errors in parentheses, *** p < 0.01, ** p < 0.05, * p < 0.1

Table 5

Claims of Foreign Banks in the UK, Controlling for Partner in the Eurozone, 1977–2018




















































Standard errors in parentheses, *** p < 0.01, ** p < 0.05, * p < 0.1

Table 6

Correlation Matrix of Investment Variables in the UK, BREXIT vote and USD-GBP exchange rates































Once again, GDPs show highly significant results; larger countries in terms of GDP attract more investments, which adds to the robustness of the model in general. The variable brexitvote shows no significant results in all models (10), (11) and (12), indicating that UK banks’ investment choices have not been affected by the BREXIT vote on 23 June 2016. Note that this does not mean that UK investment patterns did not change in the time frame since then, but changes were controlled for via fixed effects and were not caused by preparations for a future BREXIT, which would have been captured by a significant brexitvote variable.

EU membership however has significant signs for the total asset model (10), the non-banking sector (11) and the banking sector (12); UK banks invest significantly more in non-EU27 countries than in EU27 countries in total (−0.240***, standard error 0.041), a finding that in large part is explained by the role of the US, where the UK holds assets worth $1275 billion in Q1 2018 which is about the amount of France, Germany and the Netherlands combined. The UK also has holdings in Japan worth $390 billion, followed by Australia, Switzerland and Hong Kong, which are all non-EU27 countries. This, however, is mainly driven by the non-banking sector, as we can tell by model (11), whereas the UK is significantly more likely to invest in the EU27 banking sector than in the banking sectors of non-EU27 countries, even though only slightly (0.096*, standard error 0.056). This mirrors the merge of the UK and EU27 banking industries in the last decades, and thus far has not been affected by the BREXIT vote.

The fact that the sectoral effects of the partner_eu variable are different in sign could, however, also indicate an indirect BREXIT effect: The partner_eu variable has a positive effect and shows that in general there are positive incentives for the banking sector to invest in EU partner countries which could also reflect the particular aspect of the banking sector being part of the services sector which for practical purposes often requires a complementary local presence, for example in investment banking. By contrast, the non-banking sector, read the manufacturing industry, will often consider outward asset stocks (e.g. outward FDI stocks in EU partner countries) to be a substitute for trade which could explain the negative sign for the partner_eu variable outside the banking sector. The foreign presence of the home country bank typically will encourage non-banking FDI outflows into the host country selected by the respective bank (for empirical evidence, see Poelhekke 2014). The implication is that British non-banking FDI outflows to the EU27 could increase on the basis of this mechanism to some extent; however, the net effect on non-banking FDI still should be negative according to our regression results.

Overall, we see a one-sided effect: While global banks shifted their UK investments from the banking sector to the non-banking sector after the BREXIT vote, and EU27 investments in the UK non-banking sector additionally increased in general, UK banks’ outward investments have not been affected by the UK decision to leave the EU up to this point.

4 Policy conclusions

The full implementation of BREXIT would be a stronger signal than the EU referendum decision which left unclear for most investors how future EU-UK trade and investment relations – and cooperation in the banking sector – would look after March 29, 2019. The results presented have shown a distinct pattern of asset accumulation abroad in the banking sector versus the non-banking sector.

The UK should expect reduced capital inflows into the non-banking sector post-BREXIT while the prospects for inward capital flows into the banking sector are rather unclear as regards the pure BREXIT effect. As, however, most simulations of macro models show a negative long-run GDP effect for the UK post-BREXIT – and very much so in the case of a No-Deal BREXIT (see, e.g., HM GOVT. (2016) a pre-referendum study by the Treasury) – one should expect that overall capital inflows into the UK (both in the banking sector and in the non-banking sector) would decline post-BREXIT. Thus, for the UK government there will be a strong incentive to stimulate output growth in the UK by various policy measures, in the fields of monetary policy, broadly defined, fiscal policy as well as in regulatory policy.

If the UK would want to avoid a negative effect of BREXIT on banking FDI inflows, one could consider three basic policy options:
  • To deregulate banking in the UK, by signaling, for example, that the loan-to-value (LTV) ratio policy will be rather generous so that the prospects for an expansion of loans in the UK would be rather favorable. This in turn might attract complementary FDI inflows from outside the banking sector, say in the construction sector. However, any deregulation that pushes for a soft LTV policy stance also runs a risk, namely to the extent that monetary policy is shaped by a Taylor rule (for the case of the US see e.g. Bachmann and Ruth 2017). As the Taylor rule suggests (with r standing for the normal real interest rate, π for the inflation rate and π’ for the central bank’s inflation target; Y is output and Y′ normal output; H and H′ are positive parameters) to set the central bank interest rate iCB = r + H(π - π’) + H′(Y-Y′), a soft LTV policy that would raise output in the construction sector - and thus raise Y - could lead to a generally more restrictive monetary policy since the Bank of England would have to react to the more frequent positive output gaps (Y-Y′). It is not clear how this interaction would ultimately affect the stability of the UK’s banking system and of the British economy at large.

  • One natural policy option for attracting higher overall FDI into the UK could be further reductions of the corporate tax rate as emphasized in (Welfens and Baier 2018). Whether or not this is politically feasible and how the EU countries would react to such a strategic reduction of corporate tax rates is unclear.

  • Promotion of Fintech activities of UK banks could be useful, as higher expected profits of banks would stimulate FDI inflows in the UK.

The Eurozone and the EU27, respectively, could benefit from higher banking FDI inflows in the medium term and the long run, not least since mainly big banks will relocate to some Eurozone countries. Big banks already have subsidiaries or branches in the Eurozone so that legal adjustment is sometimes needed in the short term, while the hiring of more staff and the implementing of complementary asset accumulation could be a gradual stock adjustment process. The stability of the Eurozone/EU27 banking system could be reinforced by the inflow of higher banking FDI from London’s City banks. However, there is one caveat, namely that the ratio of banking value-added to the GDP of the respective host country should not be raised toward a higher critical level – if there were any future national or international banking crisis, the governments of host countries must come up with bridging financing and possibly the ability and willingness to recapitalize ailing banks. From this perspective, the Eurozone has no interest in witnessing the attracting of high banking FDI inflows into countries such as Italy, Greece, Belgium or Cyprus, all of which are countries which currently face a high public debt-GDP ratio. A particular challenge for the Eurozone and the EU27 could emerge if a considerable share of London banking activities would be relocated to the US where new banking deregulation and other economic policy initiatives, including tax reforms, would raise the US relative stock market valuation so that US-based banks could more easily take over foreign rivals, for example banks in the Eurozone/EU27. A similar argument holds, of course, if UK banking deregulation would be adopted. Any broad banking deregulation in the US and in the UK would thus put new pressure on the EU27 to also deregulate banking; if this is done in an excessive way, the seeds of the next international banking crisis would have been sown. From this perspective, it is quite important that EU27 countries would coordinate their international activities more strongly, say at the IMF, the G20 and at the Bank for International Settlements. To the extent that the UK government takes sides with the Trump Administration in multilateral organizations, the EU27 could be facing a rather difficult challenge in the future – post-BREXIT.

If BREXIT should trigger partial instability in the Eurozone – for example in the context of induced conflicts between a majority of Eurozone countries and Italy over the latter’s proposed fiscal policy – there could be additional effects to be considered. Outside the banking sector, changes in asset positions and FDI stocks, respectively, could also be influenced by changes in relative unit labor costs. Whether or not British trade unions’ wage policy, for example, will change after BREXIT remains an open question. If the EU27 should face a higher concentration ratio in banking post-BREXIT, this could also affect relative labor costs in that sector. These are questions for future research. One may wish that the governments of the EU28 countries would finally consider publishing the available FDI data in the banking sector in a transparent and timely fashion – this would help economists and others to conduct research in a more precise way which, in turn, would generate potential benefits for policymakers who could get a better understanding of FDI dynamics in key sectors of the economy. Once the UK leaves the EU it is likely to adopt broad banking deregulation so that OECD regulatory indices for the UK and the EU27 countries can be expected to differ more strongly post-BREXIT, which represents yet another field for future research.


  1. 1.

    Exports from country i to country j depend on trade costs across all possible export markets (outward resistance); imports into country i from country j depend on trade costs across all possible import markets (inward resistance).

  2. 2.

    The farther away an investment is, the more difficult it is to anticipate correctly changes and developments in investments; this might be linked to economic and cultural differences, but also, for example, to something as seemingly innocuous as different time zones.

  3. 3.

    Collecting information about clients and customers is increasingly costly over greater distances.

  4. 4.

    A substantial number of papers also use data on external claims by banks from the BIS. Some recent papers that have estimated empirical gravity equations for equity, bond and bank holdings include Ahearne et al. (2004), Aviat and Coeurdacier (2007), Balli (2008), Balli et al. (2008), Balta and Delgado (2008), Berkel (2007), Bertaut and Kole (2004), Buch (2005, 2002), Chan et al. (2005), Coeurdacier and Martin (2009), Coeurdacier and Guibaud (2005), Daude and Fratzscher (2008), de Santis and Gerard (2009), Eichengreen and Luengnaruemitchai (2006), Faruqee et al. (2004), Forbes (2008), Gande et al. (2009), Garcia-Herrero et al. (2009), Gelos and Wei (2005), Ghosh and Wolf (2000), Hahm and Shin (2009), Jeanneau and Micu (2002), Kim et al. (2006), Kim et al. (2007), Lane and Milesi-Ferretti (2005, 2008), Lane (2005), Martin and Rey (2004), Pendle (2007), Portes and Rey (2005), Portes et al. (2001), Rose and Spiegel (2004), Salins and Benassy-Quere (2006), Vlachos (2004) and Yu (2009).

  5. 5.

    A detailed listing and definition of assets/liabilities/claims can be found on the website of the Deutsche Bundesbank: pp. 263–268

  6. 6.

    Some central banks are reticent for reasons of banking secrecy; mainly due to tax avoidance strategies and competitive advantages between countries and their banking sectors. The Bank of England, for example, is one of the central banks who do not report their total linkages to all foreign countries and banks.

  7. 7.

    Country list of UK partners: Austria, Australia, Belgium, Brazil, Canada, Switzerland, Chile, Germany, Denmark, Spain, Finland, France, Guernsey, Greece, Hong Kong SAR, Ireland, Isle of Man, Italy, Jersey, Japan, Korea, Luxembourg, Macao SAR, Mexico, Netherlands, Philippines, Sweden, Chinese Taipei, United States, South Africa.

  8. 8.

    In September 2018, German banks reported holding €233 million in foreign coins and banknotes in contrast to €1,855,669 million in other assets. Source: Balance sheet statistics of German banks, September 2018, Deutsche Bundesbank.

  9. 9.

    Note that models (4)–(9) represent robustness checks for asset inflow to the UK, including exchange rates and controlling for the Euro Area instead of EU membership

  10. 10.

    For robustness, we check the model with exchange rates on an annual (not quarterly) basis in order not to interfere with time FE; we use annual data for the period 1977–2016 from the Federal Reserve Bank of St. Louis, for 2017/2018 averaged daily data from the Bank of England, where for 2018 we average the first quarter only; see Table 3 (in Appendix).

  11. 11.

    The Breitung unit root test, Hadri Lagrange multiplier stationarity test, and Im-Pesaran-Shin unit root test show significant P-values indicating stationarity, while Levin-Lin-Chu and Fisher unit root tests show insignificant values; we follow Fidrmuc (2008) and transfer his findings on OLS to PPML stating that our fixed effects take into account potential non-stationarity.

  12. 12.

    We use the FRED real foreign exchange rate index; decreasing numbers mean weaker BPS and stronger USD.

  13. 13.

    Adding exchange rates in model (5) leaves results unchanged.



This paper is part of EIIW research funded by the Deutsche Bundesbank. While the authors gratefully acknowledge funding from the Deutsche Bundesbank within the project “The Influence of Brexit on the EU28: Banking and Capital Market Adjustments as well as Direct Investment Dynamics in the Eurozone and other EU Countries”, opinions expressed within represent those of the authors and do not reflect the views of the Deutsche Bundesbank or its staff. We gratefully acknowledge research assistance by Christina Peussner and editorial assistance by Kennet Stave and David Hanrahan (EIIW), as well as comments from the participants of the 4th FDI workshop November 2019 in Mainz; and the EIIW project workshop 2018 at HIS Markit, Frankfurt. The usual disclaimer applies.


  1. Ahearne AG, Griever WL, Warnock FE (2004) Information costs and home bias: an analysis of US holdings of foreign equities. J Int Econ, Elsevier 62(2):313–336CrossRefGoogle Scholar
  2. Anderson JE (2011) The gravity model. Annu Rev Econ 3:133–160CrossRefGoogle Scholar
  3. Anderson JE, Van Wincoop E (2003) Gravity with gravitas: a solution to the border puzzle. Am Econ Rev 93:170–192CrossRefGoogle Scholar
  4. Aviat, A., Coeurdacier, N. (2007), The geography of trade in goods and asset holdings, Journal of International Economics Vol. 71, Issue 1, 8 March 2007, Pages 22–51Google Scholar
  5. Bachmann R, Ruth S (2017) Systematic monetary policy and the macroeconomic effects of shifts in loan-to-value ratios, contribution at the conference of the macroeconomic committee. Université du Luxembourg 30 June - 1 July 2017Google Scholar
  6. Balli F (2008) New Patterns of International Portoflio Allocation and Income Smoothing. Working Paper, Massey UniversityGoogle Scholar
  7. Balli F, Louis RJ, Osman M (2008) International portfolio inflows to GCC markets. Are there any general patters? MPRA Paper 10158Google Scholar
  8. Balta N, Delgado J (2008) Home bias and market integration in the EU. CESifo Economic Studies 55(1):110–144CrossRefGoogle Scholar
  9. Barrell R, Pain N (1997) Foreign direct investment, technological change, and economic growth within Europe. Econ J 107:1770–1786CrossRefGoogle Scholar
  10. Berkel B (2007) Institutional determinants of international equity portfolios – a country level analysis. BEJM 7(1):34CrossRefGoogle Scholar
  11. Bertaut, C., Kole, L.S. (2004), What Makes Investors Over or Underweight? Explaining International Appetites for Foreign Equities, International Finance Discussion Paper No. 819, Board of Governors of the Federal Reserve SystemGoogle Scholar
  12. Born, A., Enders, Z. (2018), Global banking, trade, and the international transmission of the great recession, CESifo Working Paper No 6912, Access date 7 Jan 2019
  13. Brei, M., Von Peter, G. (2017), The Distance Effect in Banking and Trade BIS, Working Papers No. 658, Bank for International SettlementsGoogle Scholar
  14. Bruno R, Campos N, Estrin S, TIAN M (2016) Technical Appendix to ‘The Impact of Brexit on Foreign Investment in the UK’. Gravitating Towards Europe: An Econometric Analysis of the FDI Effects of EU Membership. Center for Economic Performance, London School of Economics and Political Science, LondonGoogle Scholar
  15. Buch CM (2002) Are banks different? Evidence from international data? Int Financ 5(1):97–114CrossRefGoogle Scholar
  16. Buch CM (2005) Distance and international banking. Rev Int Econ 13(4):787–804CrossRefGoogle Scholar
  17. Chan K, Covrig V, Ng L (2005) What determines domestic Bias and foreign Bias? Evidence from mutual fund equity allocations worldwide, The. J Financ 60(3):1495–1534CrossRefGoogle Scholar
  18. Coeurdacier N, Guibaud S (2005) International equity holdings and stock returns correlations: does diversification matter at all for portfolio choice? Working paper, London business schoolGoogle Scholar
  19. Coeurdacier N, Martin P (2009) The geography of asset trade and the euro: insiders and outsiders. Journal of Japanese and International Economies 23(2):90–113Google Scholar
  20. Coeuré, B. (2017), European CCPs after Brexit, speech to the global financial markets association, Frankfurt (20 June 2017), Access date 7 Jan 2019
  21. Daude C, Fratzscher M (2008) The pecking order of cross-border investment. J Int Econ 74:94–119CrossRefGoogle Scholar
  22. De Santis RA, Gerard B (2009) International portfolio reallocation: diversification benefits and European monetary union. Eur Econ Rev, 2009 53(8):1010–1027CrossRefGoogle Scholar
  23. Donnery, S. (2017), Remarks by Ms. Sharon Donnery, deputy governor (central banking) central Bank of Ireland at Chatham house, London, (23 March 2017),
  24. Dunning JH (1979) Explaining changing patterns of international production: in defence of the eclectic theory. Oxf Bull Econ Stat 41(4):269–295CrossRefGoogle Scholar
  25. Dunning JH (1998) Location and the multinational Enterprise – a neglected factor? J Int Bus Stud 29(1):45–66CrossRefGoogle Scholar
  26. Dunning JH (2001) The eclectic (OLI) paradigm of international production: past, present and future. Int J Econ Bus 8(2):173–190CrossRefGoogle Scholar
  27. Eichengreen, B. (2018), The international financial implications of Brexit, University of California at Berkeley, Forthcoming in International Economics and Economic Policy Google Scholar
  28. Eichengreen B, Luengnaruemitchai P (2006) Bond Markets as Conduits for Capital Flows: How Does Asia Compare? NBER Working Paper No.12408. National Bureau of Economic Research, CambridgeCrossRefGoogle Scholar
  29. Faruqee H, Li S, Yan I (2004) The determinants of international portfolio holdings and home Bias, IMF working paper no.04/34. International Monetary Fund, Washington D.C.Google Scholar
  30. Fidrmuc J (2008) Gravity models in integrated panels. Empir Econ 37(2):435–446CrossRefGoogle Scholar
  31. Forbes KJ (2008) Why do Foreigners Invest in the United States? NBER Working Paper No.13908. National Bureau of Economic Research, CambridgeCrossRefGoogle Scholar
  32. Friewald NJ, Subrahmanyam R (2012) M.G. (2012), illiquidity or credit deterioration: a study of liquidity in the US corporate bond market during financial crises. J Financ Econ 105(1):18–36CrossRefGoogle Scholar
  33. Froot K, Stein J (1991) Exchange rates and foreign direct investments. Q J Econ 106(4):1191–1217CrossRefGoogle Scholar
  34. Gande, A., Parsely, D., Wei, S.J. (2009), Does source country corporate governance affect international investment patterns? Working paper, Columbia business schoolGoogle Scholar
  35. Garcia-Herrero A, Wooldridge P, Yang DY (2009) Why Don't Asians invest in Asia? The determinants of cross-border portfolio holdings. AEP 8(3):228–256Google Scholar
  36. Gehrig T (1998) Cities and the geography of financial Centres, CEPR discussion paper no. 1894, LondonGoogle Scholar
  37. Gelos GR, Wei SJ (2005) Transparency and international portfolio holdings. J Financ 60(6):2987–3020CrossRefGoogle Scholar
  38. Ghosh S, Wolf H (2000) Is there a curse of location? Spatial determinants of capital flows to emerging markets. In: Edwards S (ed) Capital flows and the emerging economies. University of Chicago Press, Chicago, pp 137–158Google Scholar
  39. Hahm JH, Shin K (2009) Complementarity among international asset holdings. IJFS 23:37–55Google Scholar
  40. Head, K., Mayer, T. (2014), Gravity equations: workhorse, toolkit and cookbook. In Handbook of International Economics. Edited by Gene M. Grossman and Kenneth S. Rogoff. New York: Elsevier, Vol. 4, Chap 3, pp. 131–195Google Scholar
  41. HM GOVT (2016) HM Treasury analysis: the long-term economic impact of EU membership and the alternatives, London April 2016Google Scholar
  42. Jeanneau, S., Micu, M. (2002), Determinants of International Bank Lending to Emerging Market Countries, BIS Working Paper No. 112, Bank for International SettlementsGoogle Scholar
  43. Jungmittag A, Welfens PJJ (2016) Beyond EU-US trade dynamics: TTIP effects related to foreign direct investment and innovation. EIIW Paper 212Google Scholar
  44. Kim, S., Lee, J.W., Shin K. (2006), Regional and Global Financial Integration in East Asia, Working Paper, Tufts UniversityGoogle Scholar
  45. Kim W, Sung T, Wei SJ (2007) Does Corporate Government Risk at Home Affect Investment Choices Abroad? NBER Working Paper No.13721. National Bureau of Economic Research, CambridgeGoogle Scholar
  46. Knickerbocker FT (1973) Oligopolistic reaction and multinational enterprises. Harvard University Press, Oligopolistic reaction and multinational enterprise, CambridgeCrossRefGoogle Scholar
  47. Lane, P. (2005), Global bond portfolios and EMU, European Central Bank Working Paper No 553Google Scholar
  48. Lane, P., Milesi-Ferretti, G.M (2005), The International Equity Holdings of Euro Area Investors, Working Paper, Trinity College and IMFGoogle Scholar
  49. Lane P, Milesi-Ferretti GM (2008) International investment patterns. Rev Econ Stat 90(3):538–549CrossRefGoogle Scholar
  50. Martin P, Rey H (2004) Financial super-markets: size matters for asset trade. J Int Econ 64:335–361CrossRefGoogle Scholar
  51. Okawa Y, Van Wincoop E (2012) Gravity in international finance. J Int Econ 87:205–215CrossRefGoogle Scholar
  52. Pendle, L. (2007), What Determines Australia's Foreign Equity Investment?, Working Paper, University of SydneyGoogle Scholar
  53. Poelhekke, S. (2014), Do global banks facilitate foreign investment?, working paper, VU University Amsterdam and Dutch central BankGoogle Scholar
  54. Portes R, Rey H (2005) The determinants of cross-border equity flows. J Financ Econ 65:269–296Google Scholar
  55. Portes R, Rey H, Oh Y (2001) Information and capital flows: the determinants of transactions in financial assets. Eur Econ Rev 45:783–796CrossRefGoogle Scholar
  56. Rose AK, Spiegel MM (2004) A gravity model of sovereign lending: trade, vol 51. Default and Credit, IMF Staff Papers, pp 50–63Google Scholar
  57. Salins V, Benassy-Quere A (2006) Institutions and bilateral asset holdings, vol 2006-19. CEPII Working PaperGoogle Scholar
  58. Sapir, A., Schoenmaker, D., Veron, N. (2017), Making the best of Brexit for the EU27 financial system, POLICYBRIEF issue 1, February 2017.
  59. Tinbergen J (1962) Shaping the world economy: suggestions for an international economic policy. The Twentieth Century Fund, New YorkGoogle Scholar
  60. Vlachos J (2004) Do regulatory similarities increase bilateral portfolio holdings? CEPR discussion paper no.4417. In: Centre for economic policy researchGoogle Scholar
  61. Welfens PJJ, Baier F (2018) BREXIT and foreign direct investment: key issues and new empirical findings. International Journal of Financial Studies 6(2):46. CrossRefGoogle Scholar
  62. Yu G (2009) Accounting standard and international portfolio holdings: analysis of cross-border holdings following mandatory adoption of IFRS, working paper. University of Michigan Ross School of BusinessGoogle Scholar

Copyright information

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

Authors and Affiliations

  1. 1.EIIW at the University of Wuppertal and Schumpeter School of Business and EconomicsWuppertalGermany
  2. 2.Johns Hopkins UniversityWashingtonUSA

Personalised recommendations