A Positive View Of The Future Of Cash!

Dr. Johannes Beermann, Member of the Executive Board of the Deutsche Bundesbank spoke about the future of cash at the Payment Asia Summit. Shenzen, China.

As the member of the Executive Board of the Deutsche Bundesbank responsible for cash management, I arguably very much represent what many of you may consider the “”old world of payments””. A world in which there is limited space for innovation and progress. A world that is generally high in risk but low in reward. That is what is often claimed, at least.

In giving you my European and my German perspective, in particular, let me tell you: this case is not as straightforward as it may seem. In Germany and the euro area at large, the circulation of cash remains on the rise. The Bundesbank has issued more than half of the value of euro banknotes currently in circulation. Handling and distributing cash is a major operational task performed by national central banks in the euro area – particularly the Bundesbank. This also means that we need to continue investing in our cash infrastructure.

Cash serves various economic functions – making payments is just one of them. Our estimates suggest that roughly one out of ten banknotes issued by the Bundesbank is used for making payments in Germany. This limits the size of the pie that is up for grabs by the various non-cash payment alternatives.

Let us focus on cash as a payment instrument nonetheless. Usage of cash as a means of payment is declining – this is true both internationally and in Germany. But the level of cash usage is still high in many countries – and especially so in Germany. There may be less cash around, but we are far from being cashless. So why is it that, as of yet, physical cash has not disappeared beneath the waves in the vast ocean of digital payment methods?

2. Cash as an independent means of payment

In my view, this has to do with the special features that cash offers. We regularly monitor payment behaviour in Germany to understand households’ motives for using particular forms of payment over others. Protection against financial loss, personal privacy and a clear overview of spending are crucial features that households expect from payment instruments. Cash scores favourably in all of these areas, according to our surveys. My interpretation of these results: German households value independence – and physical cash offers three unique forms of independence, which distinguishes it from digital payment systems.

First, independence from one’s socio-economic background. Cash is tactile and does not require any technical equipment. The use of cash is easily understood across the generational divide. It is this haptic nature of cash, which, in my view, is an important element of strengthening financial inclusion. Ensuring access to cash may be particularly relevant in rural areas with insufficient banking or technological infrastructures. Cash is, in that sense, also a means of safeguarding social cohesion.

Second, independence from technological ecosystems. Given the still fragmented payments landscape in Europe, cash currently remains the one truly universal means of payment when it comes to P2P transactions in the euro area. Fintech companies are shaking up the traditional banking system in Europe. These companies can often leverage their global reach and huge customer base. This may bring benefits for consumers, for instance regarding cross-border payments. But it also means that customers are becoming locked into particular payment ecosystems. Cash offers an easy way out, at least for certain transactions.

Third, independence from social control and data collection. As legal tender, cash is fully backed by the domestic central bank. Cash is the obvious choice of payment method when it comes to personal privacy. This strengthens individual freedom. At the end of the day, digital payment systems work by using personal data. Collecting data is not harmful per se. But in the age of Big Data, collecting detailed data means obtaining valuable information which, in turn, makes it possible to construct patterns of individual behaviour. From a consumer protection point of view, the question arises as to how much information is necessary to carry out a particular transaction. From an economic point of view, personal data may be seen as an additional source of transaction costs to be factored in when comparing the underlying cost structures of different payment methods.

3. Retailing – the source of future transformations?

Payment methods tend to evolve in stages. For example, the adoption of mobile payment solutions is typically preceded by the widespread use of credit and debit cards. This is the case in Germany, where contactless payments have just started to catch on. China, on the other hand, seems to be a case in its own right. A comparative study in China and Germany supports this. The evidence reported there for the year 2017 suggests that cash and debit card payments account for the bulk of German retailers’ revenue. Mobile-based payment solutions did not play a noticeable role at that time. The reverse picture emerges for Chinese consumers in major cities. Third-party mobile payment providers clearly dominate here, having leapfrogged debit and credit card payments.

Payment habits in China are still in a state of flux as payment technologies continue to evolve. Seamless payment methods are on the rise. These methods essentially try to counter the “pain of paying” with a physical smile. To what extent similar shopping experiences are becoming popular in Germany remains to be seen. There are serious concerns surrounding data protection, and these would need to be alleviated first. In my view, the transition towards a society with less cash has to be driven by the user and not the supplier. It appears that, at least in Germany, consumers value the existing diversity of payment options. Cash continues to be an important part of this. In the bank-centred financial system in Germany, commercial banks are a major actor in the provision of a payment infrastructure that can cater for both cash and its digital alternatives.

Retailing in Germany is transforming, too. On the one hand, German retailers are increasingly turning to Chinese providers of mobile payment solutions, with a particular view to increasing sales to Chinese tourists. On the other hand, retailers have also increased the scope of their activities by closing the cash cycle in Germany. Nowadays, more and more shops are providing basic banking services for their customers such as cash withdrawals and deposits at the counters. To me, this shows that the transformation of the payments landscape is anything but complete.

4 CBDC as a cash substitute?

In the digital era, it should not be surprising that central banks, too, are discussing the potential merits and drawbacks of digital forms of a central bank currency (CBDC). There are currently many operational issues relating to CBDC that remain unresolved. This pertains, for example, to the technology implemented. Blockchains and the underlying distributed ledger technology seem promising, and central banks are open to them in principle. There are several potential use cases in settlement and payment systems, for instance, which are worth exploring further. But handling and safely storing vast amounts of data does not necessarily require distributed ledgers. We need to understand the underlying technologies better in terms of operational risk.

Also, the exact set-up of a CBDC needs to be thought through as the specifications may determine the potential effects. Broadly speaking, there are two conceivable variants of a CBDC. The wholesale type restricts access to CBDC to selected financial market participants for a specific purpose. The retail type, on the other hand, could grant domestic or even non-domestic non-banks access to CBDC on a wide scale.

The wholesale variant may be seen as an improvement on existing structures in terms of processing securities trading and foreign exchange transactions, but it would have little or no effect on monetary policy. The retail variant, however, could potentially mean a paradigm shift in the economic relationships between households, commercial banks and central banks that have evolved to date. uch a fundamental shift is not free of risks, and it requires careful consideration.

There is also the question of how strong households’ appetite for such a form of CBDC would actually be. This user perspective should not be left out in the discussion.

We need to see matters in perspective. After all, many of these debates have been fuelled by the plans announced by the Libra consortium. To me, what this shows, first and foremost, is the need to offer fast and cost-efficient systems for cross-border payments. We should go one step at a time. There are already several innovative market solutions that have the potential to be transformed into an efficient pan-European digital payment solution. In addition SEPA instant credit transfers could serve as a basis for pan-European payment solutions. We should develop these systems further before contemplating further, more radical steps.

5. Conclusion

The old world of payments versus the new world. This story is not new. At the turn of the millennium, there was a strong admiration for what was referred to as the new economy in Germany. New economy was a term used to describe internet start-ups which often relied on little physical capital to generate, at times, staggering market valuations. This was in contrast to the old economy. Think of brick-and-mortar car plants with, in some cases, considerable overheads. At this point, we can say that “”the new has become a bit old and the old has become a bit new””. Economic structures have integrated. The basic market forces still apply: the companies that survive are those that are competitive and offer a unique product. I view the world of payments in very much that spirit. To me, digital payments offer exciting prospects. But that does not necessarily imply the extinction of existing payment methods. It may very well actually increase the diversity of payment methods. Cash offers these unique forms of independence from social and electronic networks, which suggests to me that it will continue to enjoy great popularity in the euro area.

German Bank Stress Test Hit By Low Interest Rates

Moody says that on 23 September, German bank regulator Bundesanstalt fuer Finanzdienstleistungsaufsicht (BaFin) and the German central bank, the Deutsche Bundesbank, published the results of their biennial stress test applied to 1,412 small and midsize German banks that fall under BaFin’s direct oversight. The tested banks represent 38% of German banking system assets. The stress test scenario was designed well ahead of the recent decision of the European Central Bank (ECB) to lower the rate on its deposit facility to a negative 0.50%.

Instead, the scenario tests the banks for a severe economic downturn combined with rising interest rates and credit spreads. The capital buffers of the tested German banks remained robust under stress with a Common Equity Tier 1 ratio of 13.0% on average, down 3.5 percentage points from the year-end 2018 starting point of 16.5%. The drawback of this scenario is that it does not address the more plausible future environment of weakening economic growth combined with extended low interest rates.

A return on equity survey that accompanied the stress test, however, was more revealing. The survey required the banks to contrast their five-year base-case assumptions for return on equity evolution (from 2018 to 2023) with five defined interest-rate scenarios.

The closest interest rate scenario to current market expectations simulates the present ultra-low interest rates lasting throughout the five years. In terms of potential to impair the banks’ profitability, it is also the most severe. Based on the assumption of a static balance sheet,2 banks would see their profitability falling by more than 50%. The high probability of this interest rate scenario unfolding suggests that the banks need to materially increase their focus on cost management to protect their credit profiles.

Unchanged rate environment is the most severe scenario. The survey results show improved forecasts for base-case profitability at the end of the five-year horizon against the same survey two years ago (2016-21). In part, this reflects moderate progress in trimming high cost bases. The German regulators cautioned, however, that the improvement is substantially driven by the fact that about half of the participating banks (Group B) assumed a rise in interest rates over the five-year horizon from year-end 2018 levels.

They added that even the flat interest rate assumption employed by the other half (Group A) based on year-end 2018 interest rates has become an optimistic scenario. Exhibit 2 shows that long-term interest rates have in fact declined by around 100 basis points since the end of 2018, broadly in line with Scenario 4 (the most severe scenario), with an even more pronounced decline in longer-term rates.

The banks’ simulated results are better under the assumption of dynamic balance sheets. This is because active management intervention to counteract interest rate-driven income declines remains an important lever to reduce the profitability pressure on German banks. Close to half the bank managers surveyed indicated that under Scenario 4, an interest rate drop of 100 basis points, they would consider applying negative interest rates to the deposits of retail clients. Less than one third excluded this for retail and corporate clients under this scenario.

Managers of banks that participated in this year’s stress test expect a combination of weaker deposit margins and a reduced contribution from maturity transformation business to outweigh moderately improving lending margins. This was the case even in Scenario 1 (based on year-end 2018 institutions’ plans with a dynamic balance sheet). It illustrates that banks with lower dependence on maturity transformation results (i.e. with higher fees and commissions income) and with more flexible funding options are better positioned to defend their profitability during a period of extended low interest rates.

As in prior years, the stress test exercise excluded the 21 large German banking groups that are under the direct supervision of the ECB.

We believe that, on aggregate, these larger banks benefit from their access to more diversified funding sources.

The low interest rate environment has particularly compressed the net interest margin on newly originated loans if these new loans have been financed with retail deposits. This is because banks have so far felt unable to charge for retail deposits and these deposits have effectively been floored at 0% interest, even as lending rates have continued to fall.

In contrast, the use of secured or unsecured market funding sources (where costs have continued to fall) available to the larger banks, enables them to substantially offset the decline in interest rates earned on their newly originated loans. This is illustrated in Exhibit 3 using residential mortgage loans as an example. On the other hand, despite economies of scale, large banks’ cost efficiency materially lags the efficiency of smaller German banks – and in turn the efficiency of small German banks lags international peers.

Advanced Economies and Financial Shocks

Vice-President of the Deutsche Bundesbank Prof. Claudia Buch spoke on “Have the main advanced economies become more resilient to real and financial shocks? and makes three telling points. First, favorable economic prospects may lead to an underestimation of risks to financial stability. Second resilience should be assessed against the ability of the financial system to deal with unexpected events. Third there is the risk of a roll back of reforms.

The global economy is currently in good shape, and valuations on financial markets are high. Global GDP has been growing for nine consecutive years now, and output has surpassed the levels prior to the financial crisis (IMF 2017). Growth has not only been sustained for a number of years, it is also relatively broad based. And, according to recent estimates by the International Monetary Fund (IMF), growth is expected to continue in the coming years (IMF 2018). Projections suggest that the world economy will expand by 3.9% in 2018, compared with 3.7% in 2017. These favorable economic outlooks, together with inflation and interest rates that remain below their historical averages, contribute to high valuations on financial markets.

As regards the regulation of financial markets, much progress has been made since the financial crisis. G20 reforms that have been agreed upon post crisis aim at enhancing the resilience of the financial system, ending too-big-to-fail, reforming derivatives markets, and transforming shadow banking into a resilient source of finance. Many of the reforms are well under way, and a first assessment of their effects becomes feasible.

Does this mean that all is well and that markets are resilient with regard to future shocks? I will argue in the following that it is not the time to be complacent:

  • Favorable economic prospects may lead to an underestimation of risks to financial stability.
  • Resilience should be assessed against the ability of the financial system to deal with unexpected events.
  • There is the risk of a roll back of reforms.

1 Favorable economic prospects may lead to an underestimation of risks to financial stability

Market participants are currently quite optimistic. This reflects that the European economy is in good shape and that the global economy, too, is expanding at a brisk pace. Market participants, bolstered by the current growth forecasts, are expecting interest rates to slowly start picking up again. And not only are market participants optimistic, the range of expectations has also narrowed down. Yet, this optimism and convergence of expectations may harbor risks for financial stability.

Assuming that the current outlook will be sustained, risks to financial stability will be limited. A gradual upturn in interest rates would strengthen the stability of the financial system. Banks may see their interest margins recover – in particular if interest rates stay out of negative territory. Life insurers and pension institutions would find it easier to generate returns and to honour promised payments to customers.

But how would markets respond to an unforeseen economic slowdown? What if interest rates stay low for much longer? What if political risks materialize and risk premiums increase abruptly? Such unexpected events may affect many market participants at the same time – thus potentially threatening the functioning of the entire financial system.

A resilient financial system needs to be in a position to weather also unexpected, but by no means unrealistic, scenarios. During the extended spell of low interest rates, risk has been growing in the global financial system. Low interest rates and strong growth might cause risks to be underestimated. These risks can be mutually reinforcing in the financial system. Therefore, the resilience of the financial system might be overstated.

The good economic conditions might lead markets participants to ignore the risks of scenarios involving heavy losses. Gazing into the rear view mirror for too long increases the risk of overlooking hazards on the road ahead. The longer booms persist, the greater the inclination to extrapolate current trends, and the weaker are incentives to take precautions against unforeseen events. If the real economy takes a worse path than expected, that would also drive up credit risk.

In Germany, for example, the number of insolvencies has almost halved in the past years, dropping from just over 39,300 in the year 2003 to a little more than 20,000 in the year 2017 (Federal Statistical Office of Germany 2018). To put that number into perspective: There are just over 11,000 municipalities in Germany. Hence, there has been an average of less than two insolvencies per municipality. Reflecting these numbers, credit risk provisioning for German banks is currently at record lows. Moreover, we know from empirical studies that entry and exit of firms is an important mechanism behind innovation and, ultimately, economic growth (Foster, Grim, Haltiwanger, and Wolf 2018).

Yet, low default rates on loans are a backward-looking indicator and do not provide a good proxy for future risks. Larger banks use their own risk models to assess risks as a basis for their capital requirements. These banks might underestimate the risks which might arise if economic activity unexpectedly worsens. This might, in turn, also lead to an underestimation of how much capital is needed to provide sufficient buffers against losses.

There are, in fact, scenarios which could hit the financial system hard.One risk scenario is a faster-than-expected upturn in interest rates. For instance, risk premiums in international financial markets might increase unexpectedly. A rapid increase in interest rates would drive up short-term funding costs and hit particularly those financial institutions that have invested into low rate, fixed term assets. In Germany, for instance, 44% of housing loans are fixed-rate mortgages with a maturity of more than ten years, this number being up from 26% at the beginning of 2010. An abrupt increase in interest rates would put pressure on banks – their funding costs would go up, and their interest income would, initially, rise at a slower pace. An abrupt rise in interest rates would also send valuations down from their current high levels and thus cause losses.

A second risk scenario is that of persistently low interest rates which would incentivize a search for yield. Empirical results for the US show that risk-taking by banks intensifies if interest rates are low for long (Buch, Eickmeier, and Prieto 2014; Dell’Ariccia, Laeven, and Suarez 2017). Also, life insurers and pension institutions would find it increasingly difficult to generate sufficient income to cover the returns that some of them have promised to pay.

2 Resilience should be assessed against the ability of the financial system to deal with unexpected events

The resilience of a financial system has two key components – buffers against risks and the risks themselves. Notwithstanding difficulties with the definition of capital and liquidity buffers, these can be measured with a reasonable degree of certainty. On average, banks’ capital ratios have increased since the crisis. Within the euro area, the Tier 1 capital ratio, which is measured in relation to banks’ risk-weighted assets, increased from 8.8% in the year 2008 to 14.7% in 2016 (CGFS 2018). This raise has been achieved by a decline in total assets, a decrease in average risk-weights, and by a strengthening of banks’ capital positions. Measured relative to banks’ total assets, bank capital has increased to a lesser extent, from 3.7% in the year 2008 to 5.8% in 2016.

Banks have also built up buffers against liquidity shocks. By the end of 2016, the Liquidity Coverage Ratio (LCR) was 130% on average in an international sample of around 90 large banks (CGFS 2018). More than 90% of those banks had already met the regulatory requirement of the LCR set at 100%. Banks’ long-term resilience against liquidity shocks has also substantially improved. In particular, the Net Stable Funding Ratio (NSFR) has increased significantly from 43% in 2012 to 115% in 2016 (CGFS 2018).

Notwithstanding higher levels of capital in the banking sector, global debt levels remain elevated. At 144% of GDP, private non-financial sector debt has been higher in late 2017 than its pre-crisis level of 125% in 2007 (BIS 2018). This trend is explained, not least, by increased leverage in the corporate sector, particularly in emerging market economies. Complacency with regard to the resilience of the financial system is thus a risk, and assessment of debt sustainability might be overly positive.

Moreover, measuring risks and the associated losses when these risks materialize is  difficult. Risks to global financial stability are genuine uncertainties, rather than well-defined risk scenarios with estimates of probabilities and losses-given-default attached to them. Writing contracts that describe all relevant contingencies and that condition the responses of the contractual parties on these outcomes is typically not feasible.

In addition, risks are highly endogenous and depend on the ability of the financial system to adjust to shocks. Seemingly small shocks can propagate in the financial system and become contagious (Allen and Gale 2000). Contagion effects can arise in the financial system whenever market participants are contractually highly interconnected or if investment strategies are very similar. In that situation, a relatively minor shock can affect the functioning of the entire financial system – with negative implications for the real economy.

The financial system needs to have sufficient buffers to be able to absorb even unexpected events, which can become mutually reinforcing in the system. Two examples can illustrate such systemic events:

First, risks stemming from an interest rate hike, revaluations in markets and increased credit losses could materialize simultaneously. Asset values might plunge; write-downs would erode equity. And, in particular, credit risk could increase if economic activity unexpectedly declines.

Second, common exposures to the same risk factor can trigger systemic risks. Take one example based on the German financial system. One of the German banking system’s strengths is its large number of smaller credit institutions operating directly within the regions in which they are based. Yet, the bulk of Germany’s smaller credit institutions – its credit cooperatives and savings banks – are highly exposed to interest rate risk. This might undermine stability precisely when interest rates climb more briskly and more strongly than anticipated. Small though these institutions may be individually, together they account for a significant share – roughly 50% – of lending to domestic enterprises and households (Deutsche Bundesbank 2018). If those institutions were to run into difficulties, the repercussions for the economy as a whole could be severe.

Generally, resilience with regard to such systemic events should be measured with regard to the mechanisms that are in place to deal with losses. Equity capital, for example, provides an ex ante insurance mechanism. Whenever risks materialize, the value of equity adjusts, and dividend payments can be suspended. Thereby, equity investors bear upside and downside risks. Standard debt contracts, in contrast, are insensitive to the borrower’s situation. Risks are not shared unless the debtor enters insolvency proceedings and unless risk sharing occurs through haircuts. Insolvency proceedings, however, are often inefficient, may create distortions if investments are postponed, and may lead to the liquidation of viable parts of businesses.

Market participant needs to hedge against negative scenarios – by making sure that they put enough capital into any investment and basing expectations on the most realistic scenarios possible. But what happens when risk materializes that the individual cannot readily grasp, when risks become mutually reinforcing in the system?

At the system level, the credibility of regimes for the recovery and resolution of financial institutions is a crucial element of resilience. In 2011, the Financial Stability Board published an international standard for sound resolution regimes: the “key attributes of effective resolution schemes for financial institutions” (FSB 2011). Their overall aim is to establish a framework that allows for systemically important financial institutions to exit the market without endangering financial stability. In particular, the burden of losses should be shifted from taxpayers to shareholders and creditors. This reduces moral hazard and funding advantages due to public bail-outs.

While the “key attributes” are applicable to all kind of financial institutions, their implementation has so far advanced mostly for banks. In the European Union, the attributes have been transposed via the Bank Recovery and Resolution Directive (BRRD) for all EU members as of 2015. For euro area countries, the Single Resolution Mechanism was established in 2016 with the Single Resolution Board as a common resolution authority. The new rules were applied for the first time in 2017. These first applications highlight shortcomings that need to be addressed (Deutsche Bundesbank 2017). For example, the decision that failing banks have reached their point of non-viability should be based on more specific criteria in order to avoid (costly) delays of the “failing or likely to fail” decision. Moreover, the first applications of the BRRD revealed discrepancies in bail in rules according to the European resolution framework, state aid rules, and national insolvency laws. Finally, contagion risks that might emerge from banks’ cross-holdings of each other’s liabilities can be mitigated by implementing holding restrictions.

3 There is the risk of a roll back of reforms

Fading memory of the crisis and a favorable economic environment intensify pressure to relax financial regulations. The risk of a roll back of reforms is particularly acute when business models of incumbent financial institutions are under pressure as a combination of excess capacity in some markets, changing patterns of globalization, and technological change.

Since the financial crisis, reforms have been initiated – the impact of which is now gradually making itself felt in the markets. Higher capital and resilience of the financial system are a stated objective of these reforms. We can now begin evaluating the effects of the reforms. Have they achieved their objectives? Do the reforms bring with them unintended side-effects?

Under the German presidency in 2017, the G20 have agreed on a framework for a structured evaluation of the reforms (FSB 2017). A structured evaluation is needed to gauge the costs and benefits of reforms. Such an analysis needs to take the perspective of society as a whole because not all the costs being discussed in the public arena are in fact costs to society. The costs of failures in the financial system should be borne by those who cause them – the shareholders and potentially also the creditors of financial institutions – rather than by the taxpayer. Technically speaking, the reforms aim at cutting implicit subsidies for systemic institutions. For the private sector, funding costs are thus going up. Yet, society benefits because costs of financial distress are shifted from society to the originators. Also, if financial crises occur less often and are less severe, reforms carry a lower economic and social price tag. Finally, benefits of the reforms can only be reaped over the longer term, whereas higher funding costs tend to be recognized immediately.

Evaluation, therefore, means making the costs and rewards of the reforms for society as a whole more transparent and disclosing any unintended side-effects. So far, there are no indications that the reforms have impaired the ability of the financial system to provide services and to lend to the real economy. The evaluation of the reforms should not be used as a pretext to water them down or weaken the resilience of the financial system. Ultimately, a stronger, better capitalized, and more resilient financial system promises a “double dividend” in terms of growth and stability.

Artificial intelligence (AI) in finance: Six warnings from a central banker

Prof. Joachim Wuermeling Member of the Executive Board of the Deutsche Bundesbank spoke about AI.  Consumers may be rated by AI when applying for a mortgage. Pooling data points from internal transactions, social networks and other sources provides a more meaningful picture of banks’ borrowers. But if too much trust is put in “intelligent” systems, the stability of financial markets may be at stake.

1.  Don’t miss out on the opportunities of AI in finance …

AI in finance could impact on the functioning of our financial system in a profound way. Some suggest that AI is enhancing the power of the human brain in the same way that electricity enhanced the power of the body 150 years ago. Hence, it could become a big thing in finance.

Artificial intelligence and big data are currently the strongest and most vivid innovation factors in the financial sector. Using AI in finance may trigger dramatic improvements in many businesses. AI elevates the role of data as a key commodity. Used wisely, big data make outcomes more reliable and may improve financial mediation. Process chains can be organised in new ways. “The scope and nature of banks’ risks and activities are rapidly changing,” as a recent Basel Committee analysis puts it.

This evolution towards increased use of non-human intelligence is not something that has just occurred in the last few years. The first invention of neural networks, a central pillar of most AI systems, dates back to the year 1943.

Until a few years ago, the main users of big data and AI in the area of finance were certain hedge funds and high-frequency trading firms. In recent times, the application of AI in finance has begun to spread widely, via “normal” banks, FinTechs and other financial service providers, to the general public.

Since 2011, HFT has accounted for about 45–50 % of all trading in US equities. The figures for the main European indices are in the same region (with about 40 % for German DAX futures). Taken together with all other “normal” algorithmic trading activities, we currently estimate the amount of algorithmic trading to be in the realm of 80–90 % of the entire trading volume for equities (and somewhat less but still very high in other market segments).

A single normal trading day generates about 3–6 million data points about prices, order deletions and modifications in DAX futures alone. No human can analyse these amounts of data simply by looking at them in an Excel spreadsheet. More sophisticated and sometimes also AI-driven techniques are necessary to do the job.

AI profoundly changes the functioning of our financial system in at least three areas: products, processes and analysis. This is true for both front office functions (eg customer business, trading) and back office functions (eg executing trades, risk management, market research). Special-purpose AI can solve specific problems, eg in customer engagement, financial management or cybersecurity.

Applications focused on market operations cover various core areas eg trading, portfolio composition, backtesting and validation of models, market impact analysis, modelling trading of large positions and stress testing. Dynamic portfolio adjustment, depending on the macro environment, may be strengthened by AI.

With the help of AI, various human shortcomings in dealing with finance can be mitigated. As behavioural finance has taught us, biases, inertia and ignorance lead to the malfunctioning of markets. AI allows humans to reach out beyond their intellectual limits or simply avoid mistakes.

2 … but beware of the risks

But opportunities are always accompanied by risks. As regards the financial system, if too much trust is put in “intelligent” systems, the stability of financial markets may be at stake. The workings of AI can be a mystery; it can trigger loss of control, make fatal errors, and have a procyclical effect due to its mechanistic functions. Pattern recognition has its limits. This can be dangerous particularly in crisis scenarios. An autopilot would never have been able to land a jet on the Hudson River. Nor can algorithms stabilise in periods of financial stress.

Looking at the recent turbulence in equities and the market for VIX-related financial products, it can be concluded that the events of 5 February share many similarities with a “flash crash”. Unfortunately, as with the original flash crash of May 2010, we have only limited knowledge about the direct drivers that triggered the event. It can be assumed that algorithmic market participants were quite active during the relevant period. But as to which strategies were applied and to what effect, we have no knowledge so far. The rise in volatility in the S&P 500 then nearly instantly affected the VIX industry, making it not the cause but more the first victim of this market event, with losses up to 95 % on assets. We do not expect this phenomenon to disappear in the future. On the contrary, more of these flash events are to come.

AI is still in its infancy. Continuous processes for the entire AI lifecycle still have to be defined and scaled for business needs. That means that AI must be embedded in the process of acquiring and organising data, modelling, analysis and delivering analytics. The skills gap, particularly with regard to data science and machine learning expertise, is the foremost challenge. At this stage, non-human intelligence is far from replacing the human brain in any respect. Computers are like school pupils dividing numbers mechanically without having understood what they were doing.

3. Consumers should take care: they remain the risk-takers

What makes this development so significant is the fact that it is not just occurring at the level of systemic institutions, markets and stock exchanges. With robo advisers, for example, AI can directly influence and control the daily financial decisions of customers and ultimately their personal wellbeing. Society has barely begun to understand the economic, ethical and social implications of AI.

While client interaction is made more convenient by mobile banking, chatbots or virtual customer assistants, banks can find out more about customer habits and provide them with tailor-made financing.

Consumers may be rated by AI when applying for a mortgage. Pooling data points from internal transactions, social networks and other sources provides a more meaningful picture of banks’ borrowers. But denials may be hard to understand. It may become even harder to challenge a decision made by algorithms.

The proper functioning of the applications is not a given. Simple flaws, cyberattacks and criminal behaviour render the systems extremely vulnerable. Consumers should be cautious. They need to be protected. Laws may have to be modified to cover new threats. Responsibility and liability in the case of malfunctioning machines have to be clarified.

4. FinTechs should not ignore the legitimate concerns of society and supervisors

Agile tech companies are driven by an admirable energy and inspiration. By nature, they take risks. They create an idea, build a prototype and try it out immediately in the real world. Regulation, supervision, obligations and requirements must make them extremely nervous.

But the wellbeing of society depends on rules. The public demands cybersecurity, data privacy, consumer protection and financial stability. FinTechs should not brush aside the concerns of their stakeholders. Business can only flourish if it is broadly accepted by citizens.

FinTechs usually pick up specific elements of the work chain of finance or create new features. Using technology, they modularise and customise products as a third party or standalone provider.

FinTechs are part of the finance sector but are not necessarily supervised. As long as they carry out tasks for supervised entities, these institutions are responsible for the behaviour of the FinTech.

5.  AI needs new forms of supervision

“Artificial intelligence” may sound glamorous from a technological perspective, but in banking supervision, the well-established principle of “same business, same risk, same rules” has so far proved to be a sound standard for innovations. Whether they employ AI themselves or outsource it to FinTechs, from the supervisors’ point of view responsibility remains entirely with the bank.

For German supervisors, IT governance and information security nowadays are equally as important as capital and liquidity requirements.

All financial institutions should address the risks posed by new technologies. Banks have to implement effective control environments needed to properly support key innovations. This includes the requirement to have appropriate processes for due diligence, risk assessment and ongoing monitoring of any operations outsourced to a third party.

The European MiFID II includes the requirement that firms applying algorithmic models based on AI and machine learning should have a robust development process in place. Firms need to ensure that potential risks are considered at every stage of the process.

Regulators increasingly have to apply AI-supported analytical methods themselves to recognise vulnerability patterns, scan lengthy reports or analyse incoming data.

In any case, we must strike a balance between financial stability and avoiding barriers for potential new entrants, products and business models. Alongside technological progress, regulators have to constantly reassess the current legal framework, supervisory models and resources.

6. Central banks should embrace AI

Central banks have access to huge amounts of very valuable data stemming from market operations, supervision, payments and statistics. They are well positioned to tap the benefits of AI so they can enhance their ability to fulfil their mandate for price stability and the stability of the financial system.

Machine learning is already being used at the Bundesbank in different narrow segments. The experiences of all users have been good without exception. While monitoring the technical progress, we are currently discovering further use cases and defining our AI foundation, strategy, organisation and processes.

Here is a list of examples, which is by no means exhaustive:

In risk management, neural networks assess and evaluate the financial soundness of the markets. Market research is supported by adopting web mining techniques and machine learning in content analysis, topic modelling and clustering of relevant articles. In statistics, machine learning enables new methods for data quality management, eg in the context of securities holdings or the classification of company data. Furthermore, the informational content of seasonality tests is assessed by a random forest machine learning technique. For our IT user help desk, the handling of routine requests via automated chatbot responses could be a useful support measure. We use social media data to detect trends, turning points or sentiments. Machine learning methods can be applied for variable selection purposes in econometric models.

ANNEX: Use case – monitoring of real estate markets

An interesting data source is internet platforms. For example, some rental and housing platforms have the potential to improve the analysis and monitoring of real estate markets via the provision of information such as list prices and structural and locational characteristics of the property market at a disaggregated level.

This is mainly based on the assumption that these data contain information on the expectations and interests of economic agents with respect to future decisions. In such contexts, a wide range of topics or “search strings” are often potentially relevant. This can result in many different, highly correlated time series.

Furthermore, the “textual analysis” method is increasingly applied in research, as large amounts of “unstructured” information on businesses and the economy are available electronically on the internet. In order to operationalise textual data for econometric analysis, machine learning algorithms can be helpful. Learning methods can be applied to classify textual documents into different categories which can then be used to draw statistical inferences.

 

A Change in Interest Rates Could Hit Parts of the [German] Banking System Hard

Dr Andreas Dombret Member of the Executive Board of the Deutsche Bundesbank spoke on What’s the state of play in Germany’s banking sector? He makes the point that current profitability of banks is low and links it to low interest rates.

Included in the speech was a fascinating passage on rate mismatch, which is a rising environment could have a strong negative impact. Worth reflecting more widely on this phenomenon.

German banks have expanded their maturity transformation in recent years. In order to stabilise profits in times of very low interest rates, they have increasingly extended the lives and the interest rate lock-in periods of their loans. For instance, the percentage share of longer-term loans and advances – that is to say claims with a maturity of more than five years – at German banks has risen from 60% in 2007 to just under 70%. The ratio among institutions in the savings bank and cooperative bank sector is especially high. At 83%, it is significantly higher than the 47% ratio for commercial banks.
Thus, we are in a situation in which banks and savings banks are holding many long-term, low-yielding investments in their books. Moreover, valuations for many investments are extremely high. By contrast, risk provisioning in the German banking system is very low, at 0.6% of total assets. This makes banks vulnerable to unexpected macroeconomic developments, such as an abrupt hike in interest rates or an unforeseen deterioration in economic activity.

At the same time, they have shortened the maturities of their liabilities. The ratio of overnight deposits to total liabilities towards non-banks has risen within ten years from 36% to around 60%. An important aspect in this context is that customers are parking their funds in deposits because interest rates on investments are so low. It is difficult to predict how these funds will be shifted as soon as more attractive investments become available. Historical experience gives us a rough idea, but in view of the extreme situation of the current low-interest-rate environment it is of only limited use as a guide for the future.

The bottom line here is that we see an increased vulnerability of banks to changing interest rates. This is why we have been keeping a very close eye on the topic of interest rate risk.

Our low-interest-rate survey 2017 focused on this very issue, simulating the implications of possible shocks for small and medium-sized banks and savings banks. One scenario entailing an abrupt rise in the yield curve by 200 basis points highlights banks’ short-term vulnerability. In such a scenario, profits would initially plummet by around 55% before staging a recovery in the medium term. This means that the speed at which interest rates are raised is crucial.

The stress test carried out as part of our low-interest-rate survey combined several risks at once. Besides an abrupt rise in the yield curve of 200 basis points, it simulated a simultaneous increase in credit and market risk. On aggregate, in such a case the tier 1 capital ratio would drop from around 16% to around 13%, which is to say by some three percentage points. Yet a more detailed look reveals the positive impact of substantially improved capital levels. Small and medium-sized institutions prove largely resilient to a simultaneous rise in the three types of risk.

Thus the stress test scenario presents a mixed picture overall. On the whole, German banks and savings banks are robust and in good shape. But this should not blind us to the major challenges and the associated need for adjustments that banks face. Elevated interest rate risk and the low level of risk provisioning increase vulnerability to shocks. I would therefore call on banks to focus their risk management operations primarily on the issues of maturity transformation and interest rate risk.

Is a Central bank-issued digital currency a realistic prospect?

Interesting speech from Carl-Ludwig Thiele, Member of the Executive Board of the Deutsche Bundesbankentitled “From Bitcoin to digital central bank money – still a long way to go“.

He says the Bundesbank actively shapes the ongoing conversation about distributed ledger technology (DLT) by contributing insights of its own, not least because as a central bank, trust is its most precious asset. The stability and efficiency of systems alone is their primary concern.

They wish to neither hype up a “hot topic” nor hinder the development of highly promising innovations.  But, healthy scepticism, coupled with curiosity and critical analysis, is warranted when it comes to both DLT and central bank-issued digital currency. He concludes that a Central bank-issued digital currency, is currently an unrealistic prospect.

“The road to a digital central bank – assuming there would be any benefits in the first place – would be a very lengthy one. At present, there is not even a recognised basic blockchain. Major consortiums are developing different types of basic blockchains, each with their own particular features. Not all of them can be used in the financial sector”.

The original promise of Bitcoin was to forge a “trustless” payment system – that is, one that required no trust. I quote from Satoshi Nakamoto’s paper from 2008 (Bitcoin: A Peer-to-Peer Electronic Cash System): “What is needed is an electronic payment system based on cryptographic proof instead of trust, allowing any two willing parties to transact directly with each other without the need for a trusted third party.”

I feel that too little attention is being paid to Nakamoto’s primary goal of constructing a groundbreaking, trustless electronic payment system which, like cash, would facilitate peer-to-peer (P2P) transactions. At the same time, Nakamoto was looking to create a currency which was not based on trust. This aspect – forging a new currency that does away with central banks – has become a major talking point in the current debate. I have come here today to explain why a trustless currency is not feasible, and I will also argue that the merits of blockchain can be harnessed more readily with trustworthy institutions than without.

To get a grasp of Bitcoin, we need to put our minds to the essence of money. There are two types of money. Money as a commodity, and money as a claim.

Money as a commodity, that could be a commonly used consumer good which is mostly non-perishable. Cigarettes, for instance, were used as a money substitute in Germany after the Second World War.

But equally, money could be a durable good – gold being the most prominent example of this. Gold is extraordinarily durable, and it has an intrinsic value as a sought after industrial metal, say, or as jewellery. Indeed, for centuries, delivering gold was regarded as the ultimate form of settling a claim.

Consumer and durable goods which can be used as money substitutes both have an intrinsic, consumption or utility value.

Virtual currencies, meanwhile, which are transferred much like goods, are a fabrication. That is not to consign them straight to the category of “fraud”. Yet they have no intrinsic value, just an exchange value. You can’t consume or use them, only exchange them.

On the other hand, there is money as a claim. The bulk of our money – central bank money and commercial bank money – is a claim on either the central bank or a commercial bank.

Every euro in cash and every euro in credit balances in TARGET2 represents a liability for the Eurosystem. And the euro is backed by the Eurosystem with its constituent central banks, one of which is the Bundesbank.

Unlike consumer or durable goods, central bank money does not have any consumption or utility value. And the issuing central bank’s credit quality and integrity is reflected in the value of its currency. The value of a currency, then, hinges on trust in the central bank.

Not just that: the issuer – so in the euro’s case, the Eurosystem – takes collateral from its monetary policy counterparties as a “deposit” for providing euro currency. That indirectly anchors the euro in the real economy.

Virtual currencies, by contrast, have no issuer, no footing in the real economy. No one has to redeem them. They are a fabrication and propagate according to a fictitious set-up in virtual systems which, in some cases, can be altered or newly created at the whim of a small group of participants. What is more, their governance regime is opaque, if not to say obscure – not to mention the fact that the identity of the participant or participants – no one knows for sure how many there are – behind the pseudonym Satoshi Nakamoto remains shrouded in mystery.

Virtual currencies are exchanged in the same way as goods, but they have no intrinsic value of their own. That is undoubtedly one reason why their value is highly volatile. Over the long term, that naturally also exposes Bitcoin holders to the risk of total loss. For us, Bitcoin is not money, it is a speculative plaything. The great number of sometimes dubious initial coin offerings is a clear indication that Bitcoin is more of a funding instrument.

To repeat: it is more of a speculative plaything than a form of payment. Hence my repeated warnings against investing in virtual currencies. We are witnessing a remarkable increase in the value of some virtual currencies. But that does not alter the risk of total loss.

2 Blockchain/DLT in the world of payments

For us, Bitcoin’s most important contribution is the underlying blockchain technology, or to put it more broadly, distributed ledger technology (DLT). This technology could help boost efficiency in payment and settlement processes.

That is why we have been looking at this technology from three different perspectives. First, the Bundesbank develops and runs major payment and settlement systems, often in conjunction with other central banks, and in this context we explore innovative technical capabilities which can contribute to their stability and efficiency.

Second, the Bundesbank acts as a catalyst to forge improvements in payment operations and settlement structures. The better the Bundesbank grasps the practical implications of technologies or processes, the more forcefully it will be able to present its arguments, which always aim to preserve the stability and enhance the efficiency of payment and settlement systems.

Third, the Bundesbank monitors the stability of systems and tools used in the field of payments and settlement. Being able to gauge the relative merits of state-of-the-art technology is a key skill in this regard. That is why the Bundesbank – much like other central banks worldwide – has been putting a great deal of thought into DLT, even though this technology is still very much in its infancy.

Potentially, distributed data storage means that DLT can simplify reconciliation processes associated with complex work-sharing value added chains. DLT is seen as having disruptive potential since it generally allows transactions to be carried out directly – that is, without intermediaries.

Developed originally for the virtual currency Bitcoin, DLT will nonetheless require extensive modifications if it is to be adapted to the needs of the financial sector. For one thing, the legal framework as it stands requires participants to be identifiable, transactions to be kept secret from third parties, and transactions to be settled with finality.

For another, transaction throughput needs to be high. That said, some of the consensus mechanisms, as they are known, absorb so much time and energy that efficient settlement seems barely possible. Furthermore, they require substantial additional data transfers, which adds to the costs.

For comparison purposes, the Bitcoin network, at its peak, settles roughly 350,000 transactions worldwide every day, and given its current configuration, appears to be running at almost full capacity. The German payment system alone, meanwhile, processes more than 75 million transactions on average every business day, according to the data for 2016.

The traditional answer to the problem of mounting complexity in the interactions of a multitude of independent participants has been to use a central bank – an institution which centralises the settlement of payment transactions. Hence the name: Central. Bank. This arrangement channels the many different bilateral payment flows and order books into larger flows which are then routed via or by the central bank and posted in a central bank account. That was a huge step towards greater stability and efficiency in the world of payments.

As a matter of fact, that is why we are seeing a trend towards centralisation and hierarchical structures in the development of basic blockchains as well. There are multiple reasons why a pure P2P settlement arrangement does not appear viable.

A pure P2P world appears unfeasible without trusted institutions. I call this factor the lack of a real reference framework. Bitcoins, you see, are merely virtual, and they change hands between virtual participants. They never leave the Bitcoin blockchain, and they will never have a real point of reference until they are exchanged for real currency, which takes place outside the blockchain.

Once real transactions come into play, a real point of reference is needed. You can trade a house on the blockchain in the form of a virtual token. But on the blockchain, that tells you nothing about whether the house even exists, whether it has the features it is said to possess, and whether it belongs to the seller in the first place. To verify all those things, there needs to be a trustworthy outside third party.

The basic matter of a participant’s personal identity needs to be verifiable outside the blockchain. Only then can we conduct real transactions with that participant.

That is why I feel that the purported goal of settling transactions without trustworthy third parties is a pie in the sky proposition.

All in all, we are highly sceptical about the extent to which DLT can be put to use in the financial sector. Given the current state of the art, it is somewhat unlikely that DLT will become a widely used application in individual and retail payments.

In the field of securities settlement, though, the shrinking processing times and reconciliation costs might prove to be a more important factor and suggest that DLT does have its uses.

The Deutsche Bundesbank is analysing the pros and cons of DLT in a project it is running with Deutsche Börse. While this project indicates that DLT does indeed have its functional merits, it is still unclear how far DLT also has the edge over today’s technology in terms of security, efficiency, costs and speed.

3 Central bank-issued digital currency

When using DLT, the question might arise in future as to whether central bank-issued digital currency could be provided for the safe settlement of larger transactions.

Central bank-issued digital currency would rank alongside cash and credit balances with the central bank as another form of central bank money, and it would also need to be posted as a liability on the central bank’s balance sheet.

There are several technical options in terms of the form this would take. Transfers could be value-based (like cash) or account-based (like deposits), anonymous or registered, its use could be restricted – in terms of amount or payment purpose, say – and it could be remunerated or, like cash, earn no interest.

The specific design dictates not just how far the supposed benefits of DLT-based central bank-issued digital currency will come into play, but also the macroeconomic repercussions, which also need to be factored into any overall verdict on its merits.

Arguably, the most important question here concerns who exactly should be allowed to use central bank-issued digital currency, or, to be more specific, whether central bank-issued digital currency should be issued to non-banks as well. Because if that were the case, we would probably see substitution effects between the different forms of money. Confining its use to the settlement of transactions among banks, on the other hand, would not involve any substantial changes over the status quo.

In particular, non-banks could convert their sight deposits at banks into central bank-issued digital currency if storage as an entry on the distributed ledger appears more secure and more convenient than hoarding it as cash.

Significant parts of non-banks’ sight deposits being shifted into a blockchain, however, and no longer being available­ to the credit institutions as virtually unremunerated funding ­might have considerable repercussions for the interest margin, the scale of lending ­as well as the business models in the banking system and the banking system’s structure.

Moreover, simply expanding the monetary base accompanied by sight deposits being shifted into central bank-issued digital currency would require a larger amount of collateral and would thus have a significant impact on the structure and risk profile of the central banks’ balance sheets.

There is a wide variety of potential monetary policy and stability policy implications. And these are currently being investigated by a number of central banks. As things stand, the likely consequences remain to be seen.

In a nutshell, the title of my speech today: “From Bitcoin to digital central bank money – still a long way to go” sums up the status quo of our considerations.

The road to a digital central bank – assuming there would be any benefits in the first place – would be a very lengthy one. At present, there is not even a recognised basic blockchain. Major consortiums are developing different types of basic blockchains, each with their own particular features. Not all of them can be used in the financial sector.

At the same time, applications for payment and settlement systems are being developed on these shifting sands. There is a lot going on in this field. Technology has been advancing at a pace unseen in the past decades.

Risk In A Low Interest Rate Environment

The German central bank has published the results of a survey of smaller financial institution across Germany, examining the impact of current low interest rates.  Of note is their commentary on home loan lending. Bundesbank calculations show they are experiencing price hikes in major towns and cities of 15% to 30% above the level that is justified by the fundamentals. They call out the extended risks in the sector, thanks to large loans being made on the back of more affordable repayments. They warn of systemic risks, a hit to CET1 ratios, and propose actions to manage this scenario. All highly relevant to the Australian context, where our regulators remain coy about the risks in the system.

Over the past few months, we asked 1,555 institutions about their profitability and resilience – in numerical terms, that’s 88% of all German institutions. With aggregate total assets of around €3,000 billion, they represent roughly 41% of the entire German banking sector.

Thanks to the results of the survey, we can provide an exclusive insight into the current and future risk situation facing German banks and savings banks. The survey was based both on assumptions made by the institutions and on stress scenarios defined by the supervisors.

The check-up covers three areas. First, it analyses profitability on the basis of business figures – here, we look not only at the institutions’ planning and forecasts up until 2021, but also at how the annual results would change under the assumption of different interest rate levels in the future. We supplemented these quantitative analyses with qualitative questions, whereby the banks and savings banks gave a uniform assessment of the future of their institution and the banking sector.

The second area is stress testing. In this context, we analysed what would happen to the capital base of the institutions if they had to cope with particular external stress events. For the first time, in addition to the classic stress factors of interest rate, market and credit risk, we also stress tested residential real estate.

Residential real estate also formed part of the third area of analysis, in which we examined other key risks. Besides risks in the residential real estate sector, we performed a detailed analysis of lending standards and the position of building and loan associations.

Our examination centres on the question of how the low-interest-rate environment is affecting the resilience of banks and savings banks. In order to grasp the effects of possible future interest rate levels on credit institutions, we calculated how the profitability of the institutions would change if the interest rate level were to move in one direction or the other. What were the underlying scenarios we used?

For the first area, the analysis of profitability, we first asked the institutions to provide their three-years plans using a uniform template and requested forecasts for a further two years. In addition to this, we examined supervisory interest rate scenarios: a sustained low-interest-rate environment, a positive interest rate shock of 200 basis points, a negative interest rate shock of 100 basis points, and a turn in the yield curve of +200 basis points at the short end and -60 basis points at the long end, each at the beginning of this year. The institutions’ balance sheets could be altered for some scenarios, and not for others.

For the stress test, the impact of a positive interest rate shock of 200 basis points was compared with that of a continued low-interest-rate environment. The scenario also assumes a 200 basis point increase in the probability of default and a 20% increase in the loss given default. Moreover, for interest-bearing securities, risk premiums were assumed to rise by between 30 and 1,500 basis points, depending on credit quality. A 20% loss in value was assumed for other securities, such as equities.

Ladies and gentlemen, when doctors start explaining all their findings down to the very last detail, some patients might think, “Don’t beat around the bush, just tell me what I have”. So here are the key diagnoses, to start with.

  • Small and medium-sized German credit institutions expect their profits to continue shrinking between 2016 and 2021, according to their planning. In concrete terms, they expect their pre-tax profit to dwindle by 9% and their total return on capital to fall by 16%.
  • In the same period, the aggregate common equity tier 1 (CET1) ratio is expected to rise from 15.9% to 16.5%.
  • Further cuts in interest rates would reduce overall pre-tax profitability by up to 60%. All told, however, these effects are less drastic than in the 2015 survey.
  • Under the conditions prevailing in our stress test, around 4.5% of the institutions would fail to meet the prudential requirements set out in pillars I and II plus the capital conservation buffer, taking into account hidden reserves.
  • One thing that increases in the low-interest-rate environment is competition: over 70% of institutions expect competition from other banks and savings banks to pick up – and as many as 85% see fintechs as a source of mounting rivalry.
  • On this score, nearly every second institution can see a prospect of mergers and takeovers in the medium term.
  • One set of findings I am sure you will all be interested in is from the residential real estate market. This much I can tell you already: we are seeing the unsecured portion of housing loans increasing at one in three institutions, but there is no sign of a worrying easing of credit standards.
  • And the good news is that a simulated extreme drop in housing prices in Germany would shave just one percentage point, or thereabouts, off institutions’ CET1 ratio.

2  Ongoing decline in profitability forcing banks to fight back

Let us now take a closer look at the details, starting with earnings. In this part of the survey, we asked credit institutions for their budgeted figures for the period until 2021. You can see straight away that the trend does not look good. Smaller and medium-sized German credit institutions are expecting results – measured in terms of their pre-tax total return on capital – to shrink by an average of 16% by 2021. The 2015 survey had even projected a decline by one-fourth.

What is behind this drop in profitability? This chart shows the aggregate decline in results over the 2016-21 period, broken down by type of result. The heaviest losses are projected to come from the 0.27% pre-tax drop in net interest income, corresponding to a contraction of more than 3 billion euro in absolute terms, and from the increase in loss provisions – the latter including positions such as expected credit losses. Here, the decline in annual earnings to the tune of 0.43 percentage points is equivalent to future loss provisions amounting to more than €5 billion in absolute terms.

The overall decline, however, will first be offset by an improvement in net commission income in the amount of 0.24 percentage points, corresponding to a figure of almost €3 billion in absolute terms. The second dampening factor at play here is the reduction by 0.5 percentage points (equivalent to more than 6 billion euro) in additional reserves – these factors will keep the decline in profitability in check at a negative 1 billion euro.

On aggregate, then, the banking sector is expecting to see a steady decline in profitability in the years ahead – however, there are growing signs that institutions are beginning to fight back. But these steps still don’t go far enough – further, more decisive action will be needed to turn things around.

Let’s now move on and look at what happens when interest rates change. To illuminate this point, we specified a number of uniform interest rate scenarios and asked banks and savings banks to calculate how their business figures until 2021 would change if interest rates remained static, increased or declined. The blue bar highlights the 16% decline in total return on capital I have just described.

If the interest rate level stayed low or even shrank further, their results would slump, as you can see from the dark blue line (-41%) and especially the solid red line (-60%). Assuming a dynamic balance sheet, portfolio adjustments can cushion this impact accordingly, as the red dashed red line (again -41%) illustrates. A rise in interest rates would be a different story. To begin with, the short-term burden of interest rate risk would materialise, hitting bank profitability. But on a more cheery note, over the medium to long term, results would even move back above the current figure from 2016 (+7%).

But banks and savings banks are not projecting such an upbeat scenario as this interest rate scenario is not regarded as being likely to happen.

Institutions’ earnings, then, are under pressure. This might lead them to take on greater risks, which are normally rewarded by higher returns. If that does not work out, institutions will end up taking excessive risk on board. For this reason, supervisors need to focus on the resilience of the institutions.

And as you can see from the chart, one in three institutions are expecting their CET1 ratio to contract. The left-hand side of the black bar shows institutions whose ratios are declining. Another thing the survey responses tell us is that as many as two out of three institutions are projecting a drop in their total capital ratio. However, that’s not a point we should overdramatise because the average outcome across all institutions is that while the total capital ratio looks set to shrink from 18.3% to 17.8%, the CET1 ratio is projected to climb from 15.9% to 16.5%.

The main question, though, is what the one in three institutions which are expecting the CET1 ratio to drop are planning to do. The institutions in this group, which account for a handsome 32% of participants, are planning to increase their total assets and exposures, but not to step up their equity capital to the same extent – on aggregate, this will slightly reduce the capital ratio.

These are all early warning signals of a heightened propensity among credit institutions to take risks, and we are monitoring developments very closely indeed.

3 Continued intense competition a catalyst for merger plans

But why are institutions taking on greater risks? One likely reason is that there aren’t any superior straightforward alternatives. Efficiency gains can only be achieved through costly optimisation measures. And low-risk investments are hotly contested in Germany’s banking sector, plus their returns have been depressed by the low-interest-rate environment.

In this setting, speculation has long been rife over further consolidation, and not just in the German banking sector. Our survey now delivers clear indications not only that competition remains as fierce as ever, but also that institutions are even expecting it to intensify – and not only because of fintechs, but also due to other credit institutions, especially regional ones.

It is hardly surprising that consolidation continues to make headway under these conditions. But what we did not expect were the figures on future mergers: Around every tenth institution is already in the process of implementing a merger or has specific merger plans. What’s more, almost half of all banks can envisage a merger in the next five years. However, considerably more banks see themselves as the acquiring institution rather than as the institution to be acquired. That’s another reason why I expect we will ultimately see fewer mergers than the responses might initially suggest.

The number of German institutions is likely to continue falling in the years ahead, too. In our role as bank supervisors, we only want to warn banks that not all mergers are sustainable. In this respect, too, banks would do well to carry out a comprehensive check-up to identify avoidable problems in good time.

4 Elevated risks through housing loans

And now we come to a new element of our check-up: the housing market. Fear of a housing market bubble and rising real estate risks in banks’ balance sheets has been the subject of heated debate for some months now, not least because of constantly rising property prices. Bundesbank calculations show that we are experiencing price hikes in major towns and cities of 15% to 30% above the level that is justified by the fundamentals. Credit growth in Germany has likewise gathered momentum of late, notably at the smaller institutions. But this needs to be put into perspective: growth is still comparatively moderate compared with the euro area in the early 2000s.

And our survey currently sounds something of an “all-clear” for Germany. We see no far-reaching loosening of credit standards or conditions. If these were significantly softened, this would point to the emergence of a housing bubble capable of threatening financial stability. But we have found no sign of that.

What we certainly do see, though, is a growing tendency among institutions to incur greater risks. These movements have been minor so far – but we need to be especially alert to them.

What exactly do we see? First, in the current low-interest-rate environment, there is an increase in mortgage loans in banks’ balance sheets – both the overall volume and the average loan size have risen distinctly. Customers seem to be taking advantage of the low interest rates to offset part of the price increases – and because rates are so low, they can also finance their purchases without any additional costs. Moreover, the rate fixation period is simultaneously being extended. On top of that, institutions also seem to be willing to grant loans against less collateral. The sum result of these factors is increased risk-taking on the part of banks.

Parallel to this, the interest rate margins, that is the interest they demand minus their funding costs, have contracted significantly over the last two years. One reason for this appears to be the fierce competition for mortgage business, which remains a safe and therefore attractive business segment for banks.

Let us move on now to the housing stress test. How well would banks and savings banks cope with a bust in the housing sector? To put it in no uncertain terms: we do not see any real estate bubble that should give us cause for concern. Nevertheless, we do need to be on our guard. That is why we took the precaution of performing the housing stress test. To this end, we simulated a decline in housing prices and examined how such an occurrence would impact on banks in terms of losses and their capital ratio.

Therefore, we first needed to simulate a macroeconomic setting appropriate for the hypothetical house price developments, using a suitable model. This then enabled us to determine both the impact on default probabilities and on loss ratios for housing loans. Based on the changes in these parameters, we were able to derive the hypothetical increase in impairment charges as well as the losses in interest income, both of which diminish the capital base. Furthermore, using the standard approach, the banks’ risk-weighted assets expand on account of the reduction in the value of eligible collateral. These partial effects cause the CET1 of the banks in question to shrink.

From the upper chart you can see that we simulated very pronounced price corrections which, however, experience in other countries has shown to be plausible in a crisis situation. The dashed lines show the development of housing market crises in other countries. The dark blue line represents our extreme scenario, which simulates how a drop in prices similar to that experienced during the Spanish housing crisis from 2011 onwards would have affected the banks in our survey. In this simulation, prices plunge by 30%. The light blue line represents the less extreme, yet still severe, scenario – which we call “adverse”. In this case, prices fall by 20%, which is not exactly a small margin either.

The credit institutions would sustain heavy losses under the extreme scenario: the result would be a drop in their CET1 ratio of 0.9 percentage points. In this case, the small and medium-sized German banks and savings banks would have to raise additional capital of around €12 billion in order to lift their CET1 ratio back to its original level. And even under the somewhat less dramatic, adverse scenario, the CET1 ratio would still fall by 0.5 percentage points. Here, capital would have to be topped up to the tune of €5.6 billion.

So you see, the risks stemming from the residential mortgage market are relevant for banks. What is more, taking contagion into account would intensify the impact considerably. We see signs of growing competition to secure mortgage loan business in the low-interest rate environment. Nevertheless, the stress test shows that banks need to look closely at how well they would be able to cope with the associated risks in the event of a crisis.

The message to banks and savings banks, then, is that they ought to make provisions if they want to stay healthy in the long term – that is and remains, without a doubt, the best medicine.

Catalysing A Financial Crisis

Interesting article from Deutsche Bundesbank discussing the causes of a financial crisis. High private sector debt is significant – Australia, please note – “Asset price booms are particularly harmful if they are debt-financed”!

Financial crises are costly – output and financial wealth are lost, unemployment increases, and social gaps widen. The costs may be prolonged, and they may become chronic. The global financial crisis that began ten years ago with the liquidity squeeze on global financial markets in August 2007 is still casting long shadows.

Global debt levels remain elevated. Debt levels of the non-financial sector relative to GDP stood at 220% by the end of 2016 compared with 179% a decade earlier (BIS 2017). In the aftermath of the financial crisis, risks were shifted from the private to the public sector (Figure 1). In the euro area, government debt due to the support for financial institutions went up by €488 billion, or 4.5% of GDP, between 2007 and 2016 (Eurostat 2017a). Today, in the euro area, government debt relative to GDP is about 24 percentage points higher than it was prior to the crisis (Eurostat 2017b).

The global financial crisis has had a significant impact on economic growth and unemployment. The estimated median loss varies between 4% and 9 % (Ball 2014, Mourougane 2017, Ollivaud and Turner 2014). Such output losses have also had social consequences. In the euro area, the unemployment rate went up from 9.2% in 2005 to 11.2% in 2015 (Eurostat 2017c).
Answering the question of how economies can be protected from financial crisis is thus a key challenge for policymakers. Complete “protection” against fluctuations on financial markets is not possible and would impair critical functions of markets in terms of the allocation of resources. But reducing excessive risk-taking, making crises less likely and reducing their costs should be the ambition of policymakers. In this note, I want to highlight three elements of a strategy for making future progress.

First, agreed financial sector reforms need to be implemented. Enhancing the resilience of the financial system and improving buffers against unexpected shocks has been a key goal of post-crisis financial sector reforms. High levels of debt can increase the fragility of finance, make financial crisis more likely, and be an impediment to growth. In response to the global financial crisis, governments have thus set out to tackle the underlying causes of the kind of financial distress that can seriously harm the economy. Regulations have been amended in order to strengthen the financial system’s capacity to buffer shocks and to promote strong, sustainable, balanced, and inclusive growth.

Second, complementary reforms can make the reform agenda fully effective. In Europe, the Capital Markets Union is such a complementary project. Implementing the Capital Markets Union can represent a major step forward towards achieving a more resilient financial system and putting in place improved mechanisms of cross-border risk sharing in Europe.

Third, effects of post-crisis reforms need to be evaluated. Full implementation of post-crisis financial sector reforms should be followed by a structured evaluation of the effects of reforms. A structured evaluation is needed in order to assess the impact and the effectiveness of the reforms implemented and to study potential unintended consequences.

1 What are the drivers and costs of financial crises?

Financial crises have been a recurrent theme in economic history. Reinhart and Rogoff (2008b) have put together a large historical database covering eight centuries and 66 countries in Africa, Asia, Europe, Latin America, North America, and Oceania. These countries represent about 90% of world GDP. The database comprises information on a large set of economic indicators as well as indicators of crisis episodes (0/1 indicators), including external and domestic defaults, banking crises, currency crashes, and inflation outbursts. Analysing these data, the authors conclude:

“Capital flow/default cycles have been around since at least 1800 – if not before. Technology has changed, the height of humans has changed, and fashions have changed. Yet the ability of governments and investors to delude themselves, giving rise to periodic bouts of euphoria that usually end in tears, seems to have remained a constant” (Reinhardt and Rogoff 2008b, p. 53).

While fluctuations on financial markets are part of regular market processes, making crises less likely and less costly should be a key goal of economic policy. After the crisis, macroprudential policy has thus become established as a new policy area. A stable financial system fulfils its core macroeconomic functions smoothly and at all times. These functions include the efficient allocation of financial resources, the provision of risk-sharing mechanisms, and the provision of an efficient and secure financial infrastructure, including the payments system.

Yet, financial stability can be threatened if the distress of one institution or a group of financial institutions can “infect” the entire system. Channels of infection can be direct contagion through financial linkages or indirect contagion through asymmetries of information, panics, or fire sales. Through such channels, decisions by individual market participants can have external effects on the functioning of the financial system. Such “externalities” are all the greater, depending on how pronounced the risk-taking incentives are, how high the leverage of individual market participants is, how large the  institutions are (“too big to fail”), how connected they are (“too connected to fail”), and how high common exposures to similar risks are (“too many to fail”). The real economy can be affected through a credit crunch when banks are forced to reduce their lending activities in response to the crisis (Brunnermeier 2009, Brunnermeier and Oehmke 2013).

One key factor that affects the stability of the financial system is the structure of finance (Bernanke, Gertler andGilchrist 1996; Gambacorta, Yang, and Tsatsaronis 2014). The larger the share of debt finance is, the larger the “financial accelerator” effects can be – seemingly small shocks can then have large and systemic implications. Economic fluctuations may become magnified and threaten the stability of the entire financial system. The channel of transmission between debt and output fluctuations can run through consumption or investment (Cecchetti, Mohanty, and Zampolli 2011, Sutherland and Hoeller 2012):

  • High levels of household debt can affect the stability of the real economy through the adjustment of consumption. Evidence for the United States shows that, during the crisis, households with high levels of real estate debt cut down consumption in response to shocks to asset prices, thus amplifying the cycle (King 1994; Mian and Sufi 2014, Jordà, Schularick and Taylor 2015; Mian, Sufi and Verner forthcoming). Similar effects have been documented for other countries: The loss in consumption during the financial crisis was particularly severe in economies that experienced a large run-up of household debt prior to the crisis. And these same countries experienced the fastest increases in house prices in the pre-crisis period (Glick and Lansing 2010, Leigh et al. 2012).
  • High levels of debt may also impair the ability of firms to smooth employment and investment when an adverse shock hits. High leverage has, for example, been shown to have negative effects on the performance of firms as a consequence of industry downturns (González 2013).
  • High levels of public sector debt can be destabilising. Strained government finances may, for example, weaken the ability to ensure financial stability (Das et al. 2010, Davies and Ng 2011). Furthermore, high levels of public sector debt may amplify under some conditions the effects of cyclical shocks due to raising sovereign risk (Corsetti et al. 2013).
  • Given the importance of the banking sector for the allocation of resources across all sectors of the economy, excessive leverage in the financial sector can be particularly harmful for the real economy. An insufficiently capitalised financial sector or banking system is thus a threat to financial stability. Adverse shocks can then set in motion a downward spiral of asset valuations and prices that ultimately threatens the solvency of financial institutions.

The destabilising effects of debt arise from its contractual features. Standard debt contracts are insensitive to the borrower’s situation. An adjustment to idiosyncratic shocks can occur only through new lending or through haircuts on existing loans after risks have materialised. In contrast, the value of equity adjusts if the borrower’s situation changes. In this sense, equity provides an ex ante risk-sharing mechanism. In other words, equity as a claim on real assets has stabilising features compared with debt as a claim on nominal assets.

Empirical studies do indeed show that excessive private (and public) sector indebtedness, asset price misalignments, international linkages of banks, and high external imbalances are drivers of financial crisis (Borio and Drehmann 2009). Furthermore, banking crises have often been preceded by real estate price booms (Reinhart and Rogoff 2008a, 2008b). This was the case in the 2008 global financial crisis, but it is also true of earlier crises in the 1970s to 1990s in, for example, Spain, Sweden, Norway, and Finland. Asset price booms are particularly harmful if they are debt-financed (Jordà, Schularick and Taylor 2015). Consequently, measures of private sector indebtedness, such as credit relative to GDP, have been identified as predictors of banking crises (Detken et al. 2014, Drehmann and Juselius 2014, Laeven and Valencia 2012). Moreover, asset price booms may be fuelled by increasing capital inflows from abroad with the potential for severe negative consequences when these capital flows stop (Kaminsky and Reinhart 1999).

G20 Data Gaps Initiative – Real Estate Data Hole

Interesting speech from Prof Claudia Buch Vice-President of the Deutsche Bundesbank on “Data needs and statistics compilation for macroprudential analysis.” Availability of data on real estate markets does not match the importance of these markets for financial stability. The lack of data is profound.

Surveillance of risks to financial stability requires good data and information. The second phase of the G20 Data Gaps Initiative plays an important role for improvements in the statistical infrastructure. Apart from providing a conceptual framework for the collection of data, implementation of new concepts nationally and internationally will be crucial.

First, with its framework for the evaluation of financial sector reforms post-implementation, the FSB has started an ambitious project. The success of this project will depend crucially on the timely and comprehensive availability of granular data. Now is the time to start developing protocols defining how statistical and policy evaluation work can be integrated more closely.

The real estate sector plays an important role for the real economy and the financial system. Monitoring developments in real estate markets is, therefore, key to an early identification of vulnerabilities.

  • More than two-thirds of all Europeans own the homes they live in. Residential property typically forms the largest component of homeowners’ wealth.
  • The majority of households borrow to finance a home purchase. In many places, housing assets can be used as collateral to access funding.  Mortgage debt is thus the main financial liability of the household sector.
  • Mortgage loans are also a major asset of the financial system. In advanced economies, about 60 percent of banks’ total lending portfolios are held in the form of mortgage loans.

Given this large exposure of financial institutions, risks to financial stability can occur if a strong rise in house prices coincides with a strong expansion in mortgage loans and an easing of credit standards.

Risks can build up if market participants form overly positive expectations regarding future developments in debt sustainability. They may not give due consideration to the possibility that asset prices may fall and that interest rates may rise. If property prices subsequently decline, and if this is coupled with a simultaneous increase in default rates, banks may not be able to offset losses from mortgage lending.

The bursting of credit-driven real estate price booms does significant and long-lasting damage to the real economy.  A fall in house prices may also affect financial institutions more directly through their specific investments in residential real estate assets.

The availability of data on real estate markets does not match the importance of these markets for financial stability. The European Systemic Risk Board (ESRB 2016) has thus recommended “closing real estate data gaps”. Much work needs to be done to improve data on real estate in terms of coverage as well as of comparability across countries.

The lack of data is profound. For Germany, indicators are available only for (aggregated) prices and credits. Information on credit standards is insufficient for monitoring financial stability. Information is limited to the Eurosystem’s quarterly Bank Lending Survey (BLS). But this survey includes only qualitative information, and it is constrained to a sample of 139 large banks. As regards markets for commercial real estate, reliable indicators on both price and lending volumes are lacking.

The G20 Data Gaps Initiative aims at improving the availability of Residential Property Price Indices (RPPI) (IMF and FSB 2016). By the year 2021, G20 economies are to provide nationally available data on Commercial Property Price Indices to the BIS. In September 2016, the BIS had already published such data, including information on coverage and methodologies, for a number of countries.

Second, we have made much progress in the surveillance of non-bank finance or “shadow banking”. Assessing risks in this area requires drilling down further, using the infrastructure that we have in terms of data and methodologies. But it also requires further developing our analytical tools, especially in order to strengthen our understanding of shock transmission channels and the relevance of common exposures and inter-sectoral linkages for the latter, including those that extend across borders.

Third, international capital flows have many positive effects – but can also propagate shocks across borders. To address this concern, timely and granular data are needed for policy use. An improved sharing of and accessibility to sufficiently granular data is crucial for monitoring systemic risk. This implies the use of common identifiers in order to allow a better linking of different micro datasets and a more refined analysis of channels of propagation.

The financial crisis, ten years on – what have we learned?

In a long, but well worth reading speech, Dr Jens Weidmann President of the Deutsche Bundesbank paints an interesting picture of what happened, and why, and what has, and needs to still be done.

I will pick out just five sections, which to me at least, resonate with the current Australian situation.

Walter Eucken, founder of the Freiburg school and a pioneer of the social market economy, condensed the liability principle into a simple formula: Whoever reaps the benefits must also bear the liability.

This tenet was so dear to him that he declared it a constitutive principle of our economic order – for, in his view, economic agents will make responsible decisions only if the liability principle is enforced.

For instance, when banks become so big that their failure could bring the entire financial system to its knees, they can rely on politicians to throw them a lifeline if they run into difficulties. Thanks to this implicit insurance policy against the risk of insolvency, the banks benefit from a funding advantage even in normal times, as investors perceive the risk of default to be lower, and the capital market, deeming them too big to fail, therefore cuts them a certain amount of slack.

Furthermore, complex financial market products and confusing market structures had caused a fog to descend on the financial markets, with the resulting lack of transparency likewise serving to help drive the mispricing of risk. As a result, many banks were therefore undercapitalised in terms of their balance sheet risk.
But the rules that apply to enterprises, banks and investors must ultimately apply to governments, too. Their purse strings also tend to be loosened if they are absolved, either in part or in full, from bearing the financial consequences of their projects. In a monetary union, the impact of one country’s debt – felt in the form of rising interest rates – becomes more widespread across all of the other member states, not only because of the single capital market but also, similarly to the response to the too-big-to-fail problem, it makes sense for member states to come to each other’s rescue during times of crisis. To this extent then, too, there is a greater incentive to run up debt.

The interest rate environment led to a “search for yield”. What’s more, thanks to the low interest rates, low-income households were also able to shoulder the debt. At the same time, homeowners’ debt levels were falling as a result of ever-rising property prices, irrespective of their mortgage repayments. In the United States, many homeowners used this opportunity to refinance and take out additional mortgages. Borrowers became all the more vulnerable to rising interest rates and falling property prices, culminating in the subprime crisis of 2007.

To make matters worse, the bursting of the property price bubble in countries such as Ireland and Spain shook the banking system, which had helped to finance the construction boom on a large scale. Negative feedback effects of government support measures included a drastic deterioration in public finances, which intensified the banking crises in these countries even further. This was because banks held sizeable amounts of bonds from their own countries.

Just like so many others, Queen Elizabeth II also asked the simple, yet not so easy to answer question during a visit to the London School of Economics in the spring of 2009: Why did no one see it coming

The reputation of economists has undoubtedly suffered as a result of the crisis.

In his book “The Art of Thinking Clearly”, the Swiss author Rolf Dobelli writes: in 2007, economic experts painted a rosy picture for the coming years. However, twelve months later, the financial markets imploded. Asked about the crisis, the same experts enumerated its causes: monetary expansion under Greenspan, lax validation of mortgages, corrupt rating agencies, low capital requirements, and so forth. In hindsight, the reasons for the crash seem painfully obvious, and yet not a single economist (…) predicted how exactly it would unfold. On the contrary: rarely have we seen such a high incidence of hindsight experts.

The hindsight bias Dobelli goes on to write, is one of the most prevailing fallacies of all. We can aptly describe it as the ‘I told you so’ phenomenon.

And so the recent financial crisis will not have been the last crisis that we encounter. This is assured by the “This time is different” syndrome, as described by Carmen Reinhart and Kenneth Rogoff. Its core consists of the firm belief that financial crises only happen to other people in other countries; a crisis cannot occur here and now in our country. We are doing things better, we are smarter, we have learned from past mistakes.
But even if we do not fall for the “This time is different” trap, even the best economists in the world are not exactly sure what will trigger the next crisis.