Bankers who leave UK after Brexit can earn much more in Europe

13 November 2018

Bankers who work in the United Kingdom and start working elsewhere in Europe after Brexit can count on a substantial increase in income in many countries. European governments have, in fact, devoted all sorts of tax rebates to lure the highly skilled financial services talent to their countries. In France and Italy, 'millionaire bankers' can keep more or less an extra $200,000 per year of their salaries. 

On March 29 2019, at 11pm UK time, the United Kingdom is set to officially withdraw from the European Union. The move will complete a two-and-a-half year process triggered by a shock referendum result on June 23 2016, when 52% of British voters supported leaving the EU. Since then, British politics and business have worked to put all kinds of rules and procedures in place to ensure that after the ‘go-live’ of Brexit, all flows of goods, services and people continue as they were prior to the new situation.

One of the sectors most impacted by Brexit is the financial services industry. Global banks, insurance companies and other financial institutions based in London are reviewing their operations, and assessing whether or not after Brexit the city still can serve as their base for the European mainland. To date, around 5,000 London jobs have moved to other countries, and in the long-run, even more jobs may well move to the continent.

Barclays and Bank of America Merrill Lynch for instance have shifted to Dublin, Ireland. While Germany's Deutsche Bank is still reviewing plans, the bank said that it could move up to 4,000 jobs to Frankfurt, a path that Goldman Sachs has already taken. The US bank has so far moved over 200 employees from London's financial services hub, as well as moving client-facing staff to among others Milan and Madrid. HSBC is in the process of moving an estimated 1,000 jobs to Paris, while JP Morgan is expected to move hundreds of employees to Paris, Dublin, Frankfurt and Luxembourg. Other banking institutions sitting on post-Brexit plans include ING, RBS, Société Générale, Standard Chartered and UBS.

Bankers who leave UK after Brexit can earn much more in Europe

On the other side of the process, the governments of European countries are eager to attract Brexit-leavers to their territories. Financial services jobs are well-paid jobs, meaning that highly educated talent will either end up moving to the new locale or local talent will receive a higher income, boosting taxes. Because of this, Frankfurt, Paris, Dublin, Amsterdam and other European cities compete for the favour of banks, stock exchange companies and asset managers who seek accommodation on the continent after their departure from the United Kingdom.

According to analysis by consultants from PwC, the governments of Italy and France have put in place the most fiscal lures for Brexit-leavers, concerning companies, but also for individuals, in the form of a higher net salaries. Bankers who earn € 1 million in gross revenue in London after tax earn a net salary of €542,869. If they move to Italy, such taxes have been slashed, and they can take home more than two hundred thousand more (€772,805) for the first five years. In France there is an income improvement of €180,000 for British expats for eight years. City bankers who move to Amsterdam enjoy an income improvement of around €100,000 for the first five years. This increase is close to what they can expect in Spain and Ireland. Strikingly, Germany does nothing extra for tax purposes for individuals to attract employees from post-Brexit Britain.

However, the experts at PwC also found that after the fiscal schemes expire, usually after five years, then the scales tilt back in the favour of tax authorities. At this point, big earners are best off in Spain, followed by Luxembourg, Germany, Italy, or even remaining in the UK. The Netherlands is still behind France and Ireland, but the differences among regular tax regimes are much smaller than under the expat arrangements. In Spain, €561,247 remains from a gross salary of €1 million after taxes and social security contributions. In the Netherlands this is €488,970.

Government policy

The post Brexit fiscal approaches of governments are sparking much debate across European countries. In the Netherlands for instance, the government wants to shorten the expat scheme from 8 years to 5 years. In the country, during those years, expats in certain professions do not have to pay tax on the first 30% of their income. The rationale behind the scheme is an allowance for extra costs that foreigners often have when they come to work in the Netherlands, and as a result, the Dutch Minister of Finance is facing heavy criticism for his proposal to weaken it. Meanwhile, in Germany, the clamour is growing for the nation’s government to do more to attract bankers and other financial sector employees at an individual level too.

More news on

AI can improve operational risk management in banking

17 April 2019

Risk management is an integral part of banking. By taking financial risks, banks are able to generate the profits that are necessary to survive. Risk management aims to control this process by making potential losses more predictable. This makes the bank more robust to external fluctuations.

Whereas profits can be made by accepting certain financial risks, operational risk is intrinsically different. Operational risks only cause losses – financially in terms of bottom-line impact and non-financially in the form of for instance reputational damage. The consequences of operational risk events can have a large impact on an organisation and the financial system as a whole as experienced during the last financial crisis. It is not surprising therefore that operational risk is receiving more attention within the financial sector, with banks trying to minimise the operational risks they take, given the resources available while keeping in mind the strategic goals of the organisation.

In the past decade, there has been major progress in the development of artificial intelligence (AI). AI algorithms excel at data analysis and have evolved to the point where they surpass human performance for a wide variety of tasks. More and more businesses exploit these technological advances to optimise different kinds of processes such as marketing, sales and e-commerce, manufacturing and logistics. In today’s growing data-driven world, this trend is expected to continue on the back of widening opportunities for use cases, including in the area of operational risk management.AI can improve operational risk management in banking

Challenges in operational risk management

In June 2011 the Basel committee published the Principles for the Sound Management of Operational Risk (BCBS), which provides a framework for the development of proper operational risk management. Three years later, a survey was conducted to measure to what extent banks complied with these principles. One of these principles states that banks should write a risk appetite and tolerance statement. Banks reported that this is more challenging for operational risk than for other risk categories and attributed this to the nature and pervasiveness of operational risk. The banks that did comply with this principle often reported the use of backward-looking metrics of operational risk, such as operational losses as a percentage of gross revenue.

The above example underlines the challenges which banks face in the management and measurement of operational risk. Compared to financial risk, operational risk is a more qualitative field of study. Whereas financial risk management has been the main priority of banking for a longer time, operational risk management is much younger resulting in less extensive historical data. Predictive modeling becomes more of a challenge in this situation.

On top of that, the events in operational risk are much more diverse in scope. The Basel committee defines operational risk as the “risk of loss resulting from inadequate or failed internal processes, people and systems or from external events”. Internal fraud, data leakage and reputational damage are very different problems, yet can be very closely related as well. A lot of the data in operational risk consists of textual input which contain qualitative information. The qualitative nature of operational risk is reflected in the Basel framework, which encompasses guidelines for organisational structures, culture and awareness, and qualitative reporting.

The computer as a reader

Artificial intelligence could play a valuable role in upgrading operational risk practices, with in particular machine learning – the field of self-learning computer algorithms – showing promise. Machine learning algorithms can make predictions based on data fed to the algorithm. Recently, major progress was made in the field of natural language processing (NLP). NLP focuses on using textual data for predictions. As an example, an algorithm can learn to rate hotel reviews. By processing large amounts of reviews together with their given ratings, the algorithm can learn to give a rating to a new review it has never seen before.

“There are many opportunities for operational risk management to exploit AI and other related technological advances.”
– Lars de Ruiter and Matthias Geerse, Solid Professionals

In operational risk, many textual reports are written regarding specific risks or risk appetite of the organisation as a whole. These reports are generally written by risk managers – experts in their field. Assuming that risk reports hold information which is absent in historical loss data, this information could be extracted and used for predictive purposes. So why should computers perform this task instead of risk experts? The answer lies within the physical limitations of human being – as well as the natural biases they have to cope with. While humans may struggle to remember a piece of text, a computer algorithm easily processes thousands of books and finds structure within using statistical methods with near-perfect precision.

Towards prediction of operational risk

The performance of NLP algorithms using word embeddings has increased tremendously of late. Using the embedding as a starting point, these algorithms learn the meaning of full sentences and use this information for further predictions. This is more generally called sentiment analysis. A popular application in the financial setting is predicting stock price movement from news articles, or twitter feeds. A recent paper by Denmark’s national bank predicts corporate distress of firms by analysing their annual reports. Taking textual data under scrutiny can make a contribution to quantitative predictive modelling, even in the financial sector where the use of words is very different to day-to-day language.

In summary, there are many opportunities for operational risk management to exploit AI and other related technological advances. One possibility is the classification of risk events. As an example, a computer algorithm can read risk descriptions written by risk managers and classify them according to their impact and frequency. Combining loss data with risk reports, improved prediction of risk events might lead to more accurate prediction of future losses. A different application is the measurement of more abstract concepts such as the financial health or maybe the cultural aspects of a company.

An article by Lars de Ruiter and Matthias Geerse. Both are finance & risk consultants at Solid Professionals, a consultancy from the Netherlands.