Revue D’Économie Financière – Extrait du numéro 127 – Article Olivier Klein
Category: Economical policy
Is banking experiencing Uberisation? And if so, how do we deal with it? What assets do we have? There are certain analyses and lines of reasoning that we can put forward with a fair degree of certainty without looking into a crystal ball.
The word “Uberisation”, broadly speaking, can refer to the threat to an established model posed by a series of innovations and new players. We shall focus here on retail banking since it is, in principle, more greatly affected by this phenomenon than corporate banking. There are various types of digital innovations that may find an application in the banking sector: robotisation, the digitalisation of processes, of contracts and signatures, big data, artificial intelligence, payments, payments and many more. Clearly, these innovations are causing many profound changes and are creating a number of possibilities for revolutions that need to be analysed and incorporated into our strategies. Two types of profound change in particular are interesting to study.
The first approach involves asking the question of whether digitalisation could go as far as to do away with branches, or at least a significant reduction in the network bank model. This is a question that needs to be asked since there are far fewer customers visiting branches. We can therefore legitimately ask ourselves whether this will continue, whether this reduction is affecting all branch-based roles and whether we are heading towards an almost unique model of online banking or neo-banking.
The second approach involves studying the extent to which this technological revolution is allowing a number of start-ups, such as the fintechs, to grow and compete with commercial banks on one part or another of their value chain. Is there a possibility, in this case, that, in time, profitable segments will be lost? Could profitability fall without jeopardising the network banking model in the process?
These two questions are important and different, even if the answers one might give to them sometimes merge into one.
The first approach is fundamental: can we imagine a world where banks have no branches? This is what some analysts are claiming, speaking of banking’s “tomorrow’s Kodak” or, a little less radically, “the next iron and steel industry”. This question cannot be ignored. On the contrary, we need to get to the bottom of things. This subject merits answers founded on solid analyses. First of all, we need to differentiate the issue of digital from that of interest rates. We can see the convergence of these two phenomena but neither has anything to do with the other. On the one hand there is a very flat interest rate curve that is damaging the profitability of retail banks. We can reasonably expect that these rates will rise again, specifically with a sufficient gap between short- and long-term rates and a central bank which will gradually emerge from quantitative easing. We expect the ECB to make announcements in this vein at the end of October.
On the other hand, we have digital and its impact on profitability. I think we need to be careful not to answer the question of interest rates with digital by thinking that low rates structurally change the model. The interest rates as they are today do not change the model in itself but temporarily damage profitability, which is not exactly the same thing.
In my opinion, the line of reasoning to be followed involves going back to basics: What is the very essence of a retail bank? What is the essence of the banking relationship? Here we need to distinguish those non-variable contextual points which vary depending on current technology and how they are used by customers. In retail banking – the model of which can incidentally differ by country according to the customs and habits that are specific to each one – there are two major areas: everyday, transactional and relationship banking, that of “life plans” and advisory banking. These are two very distinct banking requirements, even if their paths often cross.
Clearly, everyday banking involves current transactions: picking up a cheque book, making a payment, making cash withdrawals or deposits etc. This type of banking practically no longer requires a network, even if there is still demand, albeit ever decreasing, to go the counter. The development of the internet, smartphones and automated machines means that this transactional banking practically no longer requires a network to carry out these common transactions. This reduction in visits to banks is very significant overall. This is very important because the demand for “cashiers” is becoming increasingly small and this must clearly be integrated.
On the other side is relationship banking, that of life plans and advisory services ̶ as well as being the banking called upon at those difficult times that can come upon everyone sooner or later – which essentially characterises the most profound relationship between private individuals and their bank, going far beyond managing means of payment. This is the long-term relationship with customers. This is crucial because this long time is linked to the fact that their life plans are being taken care of, both in their development and their deployment. These can be very important projects: financing one’s studies, your first start in the professional world, buying a home, preparing for retirement or planning for your heirs. They can also be small life plans that are linked, such as preparing for a trip or buying a car. This is all part of a world in which banks are entirely legitimate since they meet the needs of customers by offering them the necessary products, namely loans, savings and, of course, insurance (for property and people), which enables them to be protected. These plans require time to be for them to be prepared and brought to fruition. This creates a long and strong relationship of trust between customer and banker since it relates to the security of one’s property, person, family and, fundamentally, one’s well-being. The world of needs served by relationship banking is therefore a long-term one, just as loans, savings and insurance are long-term products.
It is a nomal and commonplace for thoughts to be polarised within society: at the end of the 1990s and throughout the 2000s, there was a lot of debate over whether mass retail would replace banking. At that time many people wrote knowledgeably that mass retail would remove whole swathes of banking. This is not what happened, however. Back in 2004, I wrote a long article in Les Échos on this subject. My thought process stemmed from the following point: mass retail deals with the short-term because what is bought through this medium is consumed almost immediately. If this does not satisfy us, it’s easy to change to a different brand or even a different outlet. In banking it is difficult to change quickly, depending on the consumer experience, because if someone takes out a loan, makes savings or takes out an insurance policy, this is generally a long-term affair. This is why bank advisors need to remain in their jobs for a sufficient length of time. This is also a strong demand from customers. Whereas in mass retail, by and large, there are no more “sales representatives” in store. I have never truly believed, therefore, that mass retail can take significant market share from banks, specifically because the fundamental analysis of what makes the very essence of what was the banking relationship led me to believe that no-one would buy savings in a pre-packaged form. The only meeting points between mass retail and banking are consumer credit and payment and loyalty cards, both of which are an exact extension of the act of purchase. There is therefore only one area in which, until now, there has truly been competition between mass retail and banks.
Furthermore, depending on the country, there are different mixtures between relationship and everyday banking models. A study carried out on behalf of the French Banking Federation in 2010 sought to discover which countries had a strong relationship model. France stood out as one of the strongest countries. This does not mean that French banks are not transactional but simply that they placed relatively more weight on the relationship aspect than many other countries do. Countries that have models which are far more transactional than relationship-focused therefore have a vested interest in closing large numbers of branches since branches have little new to offer. In relationship-focused banks, something else is going on.
Therefore, people are visiting their banks less and less for everyday matters, that is certain. However, will people’s appetite for relationship banking decline? For the last decade or so, relationships with banking networks have been evolving: physically and due to telephone, e-mail, visual communication and live chats etc. These are not however removing the need for banking advisors. Although banking advisors are needed as much as ever, if not more, it is therefore important to know where to place and accommodate them. Customers want to see their advisors face to face on a regular basis, for more complex subjects or for simple reassurance. Having branches closer to them is, therefore, not entirely incongruous, especially since branches, as a place where banking takes places, also act as reassurance for many private individuals and even professionals. Therefore, since we already have branches nearby, why deprive ourselves of this asset, especially when they offer us miles’ worth of advertising space in our branch windows that even online banks envy?
Therefore, relationship modes evolve and complement each other but they are not killing each other off. The essence of what they are is not changing because there is no fall in demand for banking related to life plans, far from it in fact. With the advent of the Internet, customers are ever more demanding when it comes to the quality of advice because they are adept at browsing the web to find out information and to compare and switch if they need to. They require their advisor to be even better, more responsive and more proactive than before.
In reality, the fact that fewer people are going to “counters” is an opportunity for banks and this is not a paradox. Firstly, digital is taking repetitive, unpaid tasks away from the counter, thereby reducing costs. We can therefore offer a lot more business time to customers who request more and, in doing so, make customer advisors out of those employees who used to be at the counters. Most of the time these are young people in their first job who don’t expect anything different, therefore making them easy to train. Thanks to digital, business time is developed and therefore avoids repetitive tasks for experienced sales staff.
The second argument, which follows from the first, is that we increase the productive business time of our own advisors, which increases our productivity. Our net banking profit is therefore increased by our ability to better serve and advise our customers and therefore meet their needs.
Thirdly, the customer experience is clearly made easier by digital since certain operations are much more easily handled remotely or by machines. Customer satisfaction is therefore increased because the bank becomes more practical.
Lastly, digital is also an opportunity because it makes it possible to improve the relationship model itself. Big data and artificial intelligence, which we are gradually trying to integrate, may allow us to better understand our customers and their needs, to better prepare for our meetings and therefore serve customers better. This therefore makes us much more efficient. This is about intelligent commercial productivity, which really satisfies customers since they will only be called regarding things that concern their true needs.
Increasing the practicality of our banks and the quality of the advice are therefore two fundamental keys to success. There are two axes that allow the banks to do this: training, the budget of which we have increased significantly at BRED, and digital itself!
We therefore need to ensure that online banking is just as practical. Of course, if we ensure it is just as practical but we don’t have correspondingly low prices, there needs to be something else that sets us apart: high quality advice. Without wanting to criticise them, because they are perfectly legitimate, purely digital banks do not have advisors.
Customers in France actually want both: highly practical everyday banking and an assigned advisor who can provide them with added value. They will only, therefore, seek to separate transactional and advisory banking or even to be satisfied with just one low-cost everyday bank if their usual bank does not excel at these two levels.
Retail banks therefore have a certain comparative advantage, provided that today, on the one hand, they continue to invest so they are as good as online banks when it comes to the practicality of everyday banking. Nothing impossible here. On the other, they also need to make sure that they can provide quality advice at the same time, the added value of which warrants remuneration. Significant investment in digital and training are therefore definitely two keys to success.
However, the agile organisation of each branch and of the network and optimising the use of resources to ascribe them to the most productive in terms of net banking profit is also crucial. In certain cases, banks may close branches because the need for transaction counters is disappearing. As a result, we effectively no longer need to have a branch every 200 metres in large cities, even if we still need branches to provide advice. Therefore, according to the configuration of banks today, the number of branches to be reduced may be quite different.
It should be added that, for the moment, online banking is not profitable precisely because it is having great trouble equipping customers. Furthermore, in order to gain customers, it incurs considerable cost corresponding to the need to advertise more than the other banks. Since they have no “shop windows”, online banks need to attract the customer before they spontaneously visit one of our branches. In the same vein, online banks need to offer many more gifts and free offers. For example, customers typically receive 80 euros for opening an account, but many students go to several banks one after another to collect these incentives in turn. Enhancing loyalty is not easy, therefore. As a result, online banking focuses mainly on transactional banking. It is therefore quite difficult to monetise these models and capitalise on customers unless they start to expand their offering and assign advisors, which is starting to happen here and there. If this were to develop, it would be very interesting since there could then be transactional banks that go digital and online banks which would start to play the traditional banks’ game by appointing advisors. Both types of bank would then start drawing nearer to one another in an interesting way.
However, this begs the question: can advisory services be digitalised? Can we do without humans therein? We could of course say the following: with quality big data and good automatic intelligence, automating “pushes” (text messages or e-mails) to customers would render the human advisor useless. The customer would receive intelligent suggestions, sometimes even more intelligent than those an advisor who has not been sufficiently trained or assisted would make. Why, therefore, would we need banking advisors in future since everything would be digital?
We are convinced that the opposite is true, even if it is impossible to predict with any certainty what will happen in ten or twenty years’ time.
We know that machines can beat humans in many areas. We also know that humans and machines acting in concert beat the machine alone. We need to remain very modest, however, because who knows today what artificial intelligence will be capable of tomorrow? Artificial intelligence experts themselves remain very cautious. There are, however, some key elements to be borne in mind.
The first element is that trust is a key part of the banking relationship for a very simple reason: people are trusting us with their money and helping them to construct their life plans. We enter into the inner world and security of people and their families. Having an interpersonal relationship allows us today to cultivate infinitely more trust than is possible with a robot, even an “intelligent” robot. The youngest among us, who are extraordinarily accustomed to digital, for example, need our banks to have assigned customer advisors, even if they are visiting our branches less often. At BRED, we have also experimented with sending commercial proposals to groups of customers who are in identical situations via text message or e-mail. As is always the case with mailings, there was a positive response rate of around 2% to 3%. We then sent out e-mails or texts again to other people with identical characteristics and then had the customer advisors ring them regarding the same subject. Our success rate then multiplied ten-fold. This very modest experiment therefore provides some hope for the human relationship.
The second challenge hinges on the fact that the basis for trust also resides with the institution’s reputation, which is an added value and an asset for banks.
Moreover, cognitive science currently shows that, to be able to make a decision, rational intelligence but also emotional intelligence is needed. Studies present certain cases of people who have been injured and lost the use of part of their brain used for emotional intelligence. They are therefore entirely incapable of making decisions even though their capacity for reasoning and analysis remains intact. Advances in cognitive science therefore demonstrate that, in order to make a good decision, we need to have an inkling of the solution as well as good analysis. Therefore, the human relationship can be a powerful aid in the decision-making process. In the same vein, econometric studies have recently updated the belief that a lesson could be learned more successfully in a “classroom setting” with a teacher, than through massive open online courses (MOOC). This in no way calls into question the extraordinary interest MOOC have in disseminating knowledge and their ability to reach many more students. However, this does not mean the classroom teacher has no future.
Finally, every day we receive more diverse and varied requests from people or organisations we do not know, in the form of “push” e-mails or text messages. A few years from now, we will all be saturated with these requests, if we aren’t already. What will make the difference at that point are humans capable of calling, in addition to these “pushes”, providing additional added value. This differentiation will surely prove decisive.
For all these reasons, retail banks will probably not be threatened with extinction by what seems to be a possible “Uberisation”.
There is a second approach that involves asking ourselves whether there is a possibility that profitable market segments will be lost due to external players like fintechs.
Let’s take fintechs which are flourishing. We are seeing more and more development of services proposed on certifications and authentications, biometrics, budgetary management, electronic safes, aggregators, payments, blockchain etc. The question therefore is: is there a risk of becoming disintermediated, as a bank, from portions of the banking chain that would be profitable?
The example that could be of legitimate concern is that of external aggregators which are now capable of accessing data, suggesting bank transfers and therefore initiating payments. On top of these, they can also add budgetary management services and suggest best-priced banking products. What could actually stop them, in future, from analysing our customers’ data and offering them the banking services they need? Such players could, for example, offer consumer credit by using brokers and suggest the cheapest provider – though not necessarily the most suitable one – which may not be the customer’s traditional bank. This hypothesis of banks being partially disintermediated is perfectly foreseeable.
I think – perhaps wrongly – that these fears may be exaggerated.
Firstly, many fintechs will not have access to customer data, such companies being, for example, those that offer budgetary management software. These companies will have difficulty in taking market share away from certain segments that are currently operated by banks. There are therefore two solutions open to them: either they cooperate with specific banks, the former being bought out by the latter or by forming a more or less exclusive partnership, or they form cooperative platforms with several banks in order to offer services that can be shared. Thus, they invite and integrate themselves into the banks’ value chain without disrupting their model. They would even help to enrich it since banks would become even stronger in the global model with their customers by expanding their services. For example, in the BCPE group, we sought out fintechs to offer professional customers CRM solutions linked to payments. Therefore, either the banks have the IT investment capacity to enhance their services themselves or they can look to subcontract. In fact, the reality is, of course, a mixture of the two. What does change, however, is the fact that banks were wont to do everything themselves, whereas in the future this is also likely to be an assembly profession, not just a complete, fully integrated profession. There’s nothing wrong with assembling as long as that allows us to expand our overall relationship base and our revenue.
The second case, which could of course pose a problem, is the one in which fintechs will have access to some of the data. Web scratching – which may soon be prohibited or at least highly regulated – and, more generally, DSP2 and APIs, pose the question of opening up banks’ data and customer accounts. Currently, all banks have built their aggregators to try to make sure that their customers don’t need to leave the bank’s environment to access their accounts at other banks and this works well. In future, by regulation, accessing data will be accompanied by strong authorisations which will certainly, and legitimately, make it much more difficult for external players other than the customer’s bank to process data (whether or not these players are banks).
The discussion is also currently focused on knowing which data can be accessed. We can say that there is growing awareness amongst customers, and the population in general, of the danger of allowing uncontrolled use of their data. This trend is likely to accelerate, being particularly noticeable among young people. This awareness will most likely put the brakes on any intrusion.
Furthermore, a new data protection regulation will be implemented next year (the General Data Protection Regulation – GDPR – which will apply from May 2018), which will reiterate the fact that the data belongs to the customers and that any use of such data must be approved by the customer. This will apply not only to banks but also to all data users. This is probably good news since, even if it is difficult for banks to justify everything, this regulation will make it possible to slow down the arrival of third parties that wish to use this data in a cavalier fashion, which will again put the brakes on intrusion. Banks must remain this trusted third party which processes people’s personal data and they clearly must not be allowed to disclose this information without the customer being aware of this and giving their consent.
Once again, therefore, banks’ ability to provide for and improve their overall relationship model with their private customers will be a deciding factor in resisting Uberisation while integrating and offering new services that meet the needs of the customers served by banks.
Therefore, if banks invest heavily in training and digital and carry out the essential changes their organisations require, there is no reason to believe that the retail banking model is dead. However, as with all enduring models, it no longer needs to be a chemically pure model. On the contrary, that model now needs to be intimately combined with digital. Today we are seeing in all areas of distribution that purely digital models are having a hard time surviving and that purely physical distribution models are dying out. The future will involve this mixture, with the right replies being found in understanding what the very essence of the banking relationship is. It seems to me that this is possible.
The risk of Uberisation will also, as in other sectors, have provoked strong competition, which is essential in a sector that is highly regulated and poorly suited to swift changes. We must be lucid, however: this stimulating effect which, overall, has led to an improvement in the banking model and benefited its customers, is not the only factor. A reduction in profitability, all other things being equal, caused by new players entering the market and putting pressure on prices, is likely. However, other sectors can no doubt be developed in parallel, connected with the recurring activity of commercial banks, which will see their revenue grow.
The risk of Uberisation, to quote the title of the conference, must be assessed in depth in the light of the assets that can be mobilised by commercial banks. Coming out on top, therefore, seems possible. But this is contingent on correctly appreciating the changes that need to be made and the adoption of a deliberately offensive strategy by these same commercial banks.
On 26 October, the ECB (European Central Bank) will probably announce how it is planning to recalibrate its current highly accommodative monetary policy. Primarily consisting of massive purchases by the ECB of sovereign and corporate bonds and the introduction of negative interest rates, this policy has proven its usefulness in combatting the risk of deflation and the disintegration of the eurozone. It has, therefore, been effective.
Gradual withdrawal of the policy now appears necessary. Deflationary fears are now behind us, growth in the eurozone is confirmed and the unemployment rate has fallen considerably. Although we are experiencing stubbornly low inflation, continuation of the policy entails significant risks.
Through a policy of very low and even negative interest rates, below the nominal growth rate, the ECB, by supporting borrowers, impacts the remuneration of savers and lenders. Germany, a country of declining demographics and thus more sensitive to this situation, reminds the ECB regularly of this. Furthermore, and whether or not they are contractually required to deliver minimum yields, institutional investors (insurance companies, pension fund managers, etc.) may therefore be inclined to extend the duration of their investments and accept higher counterparty risks in exchange for higher remuneration. Should it continue beyond its necessary duration, this policy could cause future financial instability.
Additionally, such a policy may encourage speculative behaviour, a cause of bubbles, consisting of borrowing at low rates in order to buy risky assets (equities or real estate) in order to benefit from the yield differential. Yet, although such bubbles had not clearly been seen until recently, certain assets appear to have been experiencing quite rapid price hikes over the last few months, both on the US equities markets, for example, and on real estate markets of a number of large American and European cities (including in Germany).
By seeking to position long-term interest rates at very low levels, it destroys the differential between banks’ lending rates and the rates applicable to their sources of funds, while savers’ bank deposit rates cannot fall below zero. But this interest margin constitutes a fundamental building block of retail banking income. In the case of France, for example, since 2016 this negative effect has not been offset by higher lending volumes and a lower cost of credit risk, due to the same very low interest rates. Yet at the same time, results from their other activities (investment banking, international, insurance, etc.) have enabled them to generate very good overall results. Consequently, sooner or later the lower income from retail banking in domestic markets runs the risk of impeding their ability to support lending growth alongside resurgent economic growth, at a time when the solvency ratio demanded under prudential rules continues to rise.
For all these reasons in particular, the start of normalisation of the ECB’s monetary policy has now become necessary. It would also enable the institution to re-establish vital room for manoeuvre to combat any future cycle reversal, particularly as the budgetary policy of many European governments currently has little room for manoeuvre given their levels of public debt.
To implement this turnaround, from 2014 the US Federal Reserve commenced a gradual tapering and subsequently ended its asset purchase programme, and finally gradually increased its key rates (short-term rates). The ECB will probably announce its own tapering plans on 26 October. By deciding to unwind its asset purchase programme very gradually, and by first of all stabilising its stocks, it could trigger a very prudent rise in long-term rates over the coming years. At the same time it could also raise negative rates towards zero, a situation that can only exist in very exceptional circumstances. Key rates would only be raised after this first step.
The rates rise will be managed very prudently, as it also involves significant risks. It could cause major market shocks if it is very sudden and poorly anticipated. Similarly, in view of the high levels of sovereign, corporate and household debt, it can only be implemented very gradually. The euro has already risen sharply against the dollar. With the rise of our currency clearly having an effect which could counteract anticipated inflation growth following higher economic growth in the zone, the ECB cannot, however, run the risk of accelerating the revaluation of the euro while it is seeking to get inflation back up to around 2%.
The policy implemented by the ECB has, in practice, been designed to buy time for the eurozone, to enable its states to carry out structural reforms and to make the necessary modifications to the institutional and organisational framework of the monetary zone itself. As the policy cannot last for ever, it is becoming all the more imperative for the countries concerned to implement such reforms in order to enhance their competitiveness (quality/price) and sustain their growth potential. And consequently, in the absence of austerity policies, to reduce public deficits, including welfare, and structural current account deficits. The objective must be to create the foundations of a strengthened eurozone, through better coordination, with greater solidarity and where all members will be able to improve their growth potential.
Co-written with Thibault Dubreuil, Finance Major at HEC
Whereas the Euro Zone is at a crossroads and the governments of the member states are taking a close look at the way it works, Olivier Klein takes a look here at the possible evolution of monetary and interest rate policy in this context.
The possible evolution of monetary policy and interest rate policy and the consequences on the economy and the banks.
Are interest rates going to rise and what are the fundamentals which would justify it? What are the effects of the low interest rates on the economy and on the capacity to finance the banks?
An initial, rather general but fundamental thought: the very low interest rates are not a phenomenon due exclusively to the central banks. It can be seen that growth has started again, both in the United States and in Europe. This is good news. This will probably drive internet rates up. But since the financial crisis we have seen a period of very low growth with a global overproduction crisis, which has led to very sluggish supply and demand. Clearly low demand and a high level of saving combined with low investment are the manifestation of global overproduction. Similarly, low gains in productivity and declining global demographics – apart from in India and in Africa – are keeping supply sluggish. The simultaneity of sluggish demand and supply has resulted in very low growth rates and naturally reduces interest rates.
If the markets were working well and if the economy spontaneously self-regulated, interest rates would return to so-called natural levels. In economic theory these natural rates are those which equalise saving and investment at a level of full employment. It just happens that this is a rate which cannot be observed, it is a rate calculated by economists. And these calculations lead to extremely low interest rates. Sometimes it is even calculated that these rates should have been negative to equalise demand and supply, i.e. savings and investment, at a level of full employment. Particularly in the eurozone.
It can thus clearly be seen that it is not only the central bank which is pushing interest rates down.
This leads us to think that not so long ago we were in a period of secular stagnation, the effects and reasons of which we all know, in other words, at least since the crisis, a classic period of debt reduction with very low growth, which occurs after all the major over-indebtedness crises, such as that we witnessed from 2007 onwards. Both are possible at the same time, at least temporarily. In both cases this justifies extremely low interest rates.
To these considerations must be added the fact that the brutal financial crisis we have witnessed has brought us into a period of major deflagration with a very high risk of deflation. To fight against this deflation the central banks have pursued extremely aggressive but necessary policies. All the major central banks have acted by reducing interest rates even further than the markets pushed them to, in other words below neutral levels. As we know, the neutral level is equal to the actual growth rate plus the inflation rate, thus to the nominal growth rate. When interest rates are pushed below neutral rates this is done because we want to reinvigorate growth by pushing inflation up again – and thus avoid deflation – and, of course, in order to limit the risks of an over-indebtedness crisis to prevent the “snowball” effects on debt due to nominal interest rates above the nominal growth rate.
When there is extremely low growth and virtual nil inflation we face the zero lower bound. Monetary policy and self-regulation of the economy are thus constrained in principle by the impossibility of bringing interest rates below zero, whereas they should be to rebalance supply and demand at full employment level. In France, for example, the banks cannot propose deposits at negative interest rates, except to major institutional clients. We are limited to this zero rate. It can clearly be seen that this may be a trap for a lasting situation of under-balance. If interest rates cannot fall sufficiently low, the consequence may thus be to remain in a situation of under-balance, of under-production, with lasting under-employment and with a persistent risk of deflation. With interest rates which, although very low, no longer have enough spring to rise back, because they should be even lower.
If we now return to the effects on the economy of the very low rates, they are well known. In principle they drive growth back up by an in initial effect, the stimulation of consumption and investment and the reduction of the attractiveness of saving.
The second effect is the wealth effect. The drop in interest rates drives the price of wealth assets up, be it property or shares, which in turn bolsters consumption and investment for both households and companies.
In 2007 debt levels of private agents in the most advanced countries reached extremely high levels. This over-indebtedness crisis, which is the fundamental reason behind the financial crisis we witnessed in 2007-2009 led to an over-indebtedness crisis for countries. From 2008-2009 onwards countries ran up high levels of debt to meet this financial and economic crisis. This led a certain number of countries to situations of over-indebtedness, thus joining the situation of the private players.
This naturally led, as always in financial history, to great periods of painful debt reduction which potentially asphyxiate growth.
The effect of very low interest rates, lower than the neutral interest rates, enables these periods of debt reduction to be facilitated. As previously mentioned, this enables the well-known “snowball” crises to be avoided. If interest rates are above neutral rates and if there is a high level of debt, debt snowballs because the debt interest must be financed by increasing the debt itself. Conversely, if we have interest rates below neutral rates debt can be reduced less painfully. Obviously this is what the central banks have done by greatly reducing their short interest rates, down to zero. This is usual in monetary policy. The new feature by the central banks was to bring certain short interest rates below zero to avoid the zero lower bound. The ECB initiated a policy of negative interest rates on bank deposits at the central bank. We are at -0.40% today. If the ECB did it, it is probably because the natural interest rate in the eurozone is negative. It is also obviously a way of encouraging the banks not to hold on to cash reserves at the central bank but rather to use them to grant more loans. This is indeed what happened, moreover. The banks considered that it was better to grant loans, even at 1.50%, rather than lose -0.40% leaving cash in the central bank. This meant a differential of 1.90%. All the banks were therefore encouraged to grant more loans. And this pushed interest rates down again since the credit offer increased and competition between the banks was thus fiercer. So the banks lent more, within an intelligent central bank policy, even if it is not very usual and even if, obviously, it involved risks.
Moreover, as in the United States, but later, the ECB took more radical measures in the shape of quantitative easing, in other words developing its own bank balance by directly buying public and private debts. In reality this involved giving itself the means to control long interest rates, too, whereas the traditional practice of the central banks is to control the short rates. It had to control long rates to bring them to rates which were compatible, in particular, with the budgetary solvency trajectory of the nations. Between 2010 and 2012 we entered a major crisis in the eurozone, a highly risky period since a contagious defiance had set in where we had the following catastrophic dynamic: fear as to the solvency of the public debt which pushed interest rates on public debt upwards, which in turn reinforced the insolvency risk. The extraordinarily welcome policy of Mario Draghi was to initiate quantitative easing to reduce countries’ long interest rates, ending the vicious circles. Without this the eurozone would probably have exploded. His famous “whatever it takes” was a salvation.
We well know that quantitative easing has consequences on foreign exchange. The foreign exchange level may, however, under certain conditions, help increase the level of growth. There have been attempts to reduce the dollar or reduce the euro, etc. by the central banks concerned, via quantitative easing policies.
The question we are asking ourselves here, today, is how long can the very low, or even negative, rates continue and have we definitively moved into a phase of increased interest rates? And if the answer is positive, at what speed will this increase occur?
At the peak of the crisis in 2008-2009 I was persuaded that the very low long interest rates would be lasting. I wasn’t totally wrong as we are in 2017 and they are still extremely low. For me they were lasting due to the context and the reasons I just explained. Why might we change our paradigm now and think that the rates may rise again?
I said it earlier. Firstly, because there is a return to growth and nominal interest rates are quite strongly determined by the nominal growth rate. With the nominal growth rate rising all over the world, this is a good reason to think that interest rates must rise.
In other words there are brakes and problems that may ensure extremely low or even negative interest rates may last a long time. The first brake is that the very low long rates policy may not work. It is not enough to reduce interest rates to encourage companies and households to borrow. This was indeed the case in France in 2014, when interest rates fell sharply. Lending did not return to growth straight away, and this was not attributable to the banks, which would have liked to grant more loans. There was simply a problem of demand for credit, because everyone was in a sort of depression where nobody wanted to borrow more. Finally, at the end of 2014 and in 2015 growth in the loan mass was seen in France, due to the very low interest rate policy. This brake no longer exists, therefore, since we have demand for credit which is, in our opinion, insufficient, but in any case at a good level.
In parallel, in principle, very low interest rates discourage saving. But interest rates must be compared to inflation. We have very low interest rates but also very low inflation. Overall savers have not been badly treated, at any rate less than in the years where interest rates were much higher than today but lower than inflation rates. But there is a psychological effect to having very low interest rates. Many households consider that they are not managing to constitute gradually the savings they would like to have when they retire because interest is not high enough to capitalise at a sufficient level to reach these amounts; they will possibly save more and consume less to ensure themselves the levels they want later. In this case, the effect of very low interest rates may be exactly the contrary of what is expected according to traditional economic theory. Today the effect is not clear-cut. We can clearly see that the very low interest rates have not significantly discouraged saving. But this effect may occur sooner or later.
The third brake on very low interest rate monetary policy is that the wealth effects, which are strong in the United States, are less strong in Europe, notably because the composition of household savings financial portfolios is not the same at all. It is based far less on shares. It is composed more of money market products and property, etc. Thus, the wealth effects are much less evident econometrically across Europe.
Two risks must also be taken into consideration. The first is to see the return of speculative bubbles. As interest rates are very low, it is easy to borrow to become a buyer on the wealth assets market. This could lead to the development of bubbles.
Today, and at least a few months ago, we could not really see signs of a bubble. No property bubble is apparent in Europe. Nor is there a clear bubble on the stock market, at least in Europe, even if, in the United States, I am not at all sure that certain sectors are not already overvalued. This risk, although not yet proven, nonetheless exists, especially if such a situation regarding interest rates were to continue further.
It can also be seen that institutional players are having difficulty in meeting the yield obligations they may have, whether it be pension funds, health mutual insurers or investment funds. The buying of much more risky assets than those made previously is also starting to be seen. As everything which presents little risk has a virtually negative return, a trend towards much more risky assets can be seen. The next change in the economic environment and market may result in loan and bond defaults. In short, more fragile balance sheets.
There is also a risk on the banks. We well understand that they need an interest rate slope to ensure profit margins. Why? Because, basically, they borrow money from depositors at interest rates linked to short interest rates and they predominantly lend at long fixed rates. A drop in interest rates during the transition period is not, in general, good for the banks. But after the transition the banks should be able to restore their margins. If we were, before the transition phase, at a 5% credit rate on average on stock and at a 2.50% rate on deposits, for example, and if we return, after transition, to 2.50% and zero respectively, the bank’s margin rate is indeed reconstituted. Since today all the long and short rates are all around zero, the credit rates on stock are falling incessantly in banks’ assets and the deposit interest rates can practically fall no more since they are practically at zero and they cannot become negative. We are faced with the zero lower bound phenomenon. And this is leading the commercial banks in France to see their margin rates, and thus their income, fall inexorably.
But there are other effects that the central bank highlights, and rightly so. According to it, because interest rates are low, the credit volume may bounce back up. That’s true. As I said, from the end of 2014-2015 onwards, that’s what has happened in France. We have seen a positive volume effect on loans which has enabled the negative interest rate on commercial bank net interest margins to be compensated. This was exactly the case in 2015. In 2015 half the commercial banks in France had a net interest margin which fell slightly, the other half which increased slightly and, in total, the banks saw an aggregate net interest margin which was unchanged. This was no longer the case at all in 2016. In commercial banking in France in 2015, Net Banking Income did rise by 1.8 %, because the volume effect compensated the interest rate effect, as we have just seen, and commission rose slightly. But in 2016 NBI fell by 4% on average because the volume effect was less than the cumulative interest rate effect, despite the increase in commission. Even if interest rates are now stagnating, or even if they rise very slightly, the drop in the stock interest rate due to the natural repayment of old loans or renegotiations, or early repayments, would lead to a falling interest rate on stock.
The commercial banks are thus entering very troubled waters. The European Central Bank replies, rightly, that thanks to the fact that interest rates are very low, which has also reinvigorated the economy and French and European growth somewhat, the cost of the credit risk has also fallen. It is right. In 2015, commercial banks in France saw the credit risk cost fall by 12.2%. In 2016, it fell by 14.2%. So, if I now take the NBI variation less the credit cost variation to analyse the overall effect, what do we see? For all the commercial banks in France, in 2015, a net positive effect of +3%; but in 2016, a negative effect of -3.3%. In other words, the drop in the cost of loans in 2016 was not sufficient to compensate the fall in NBI caused by the interest rate effect.
Also, the fall in the cost of risk cannot be lasting. The effect of the falling credit stock interest rate is lasting, however. The fall in the cost of risk is not lasting, in fact, since a mere slowdown in the economy would drive this cost upwards. We cannot wager on that compensating lower NBIs in the long term.
The question which can be raised is, fundamentally, whether this very low or even negative interest rate policy in Europe is desirable or not. Certain economists say that it was and is very dangerous. I do not share this judgement. I think it was perfectly desirable and that the favourable effects, as the ECB rightly says, have far outweighed the risks taken. The risks it took had to be taken, because the risks which would have existed if it had not carried out this monetary policy would have been far greater: deflation, prolonged stagnation, etc.
Where does that leave us today? Firstly, it must be acknowledged that the return to growth in many parts of the world legitimises an increase in interest rates, as we said earlier. The Fed is pushing them slowly upwards, but with fresh uncertainty over Donald Trump’s policy and its possible consequences for the American economy. The dollar rose but is now falling again. It can clearly be seen that the markets are uncertain regarding the success of the Trump policy or, on the other hand, the problems it may cause. And then the Fed is increasingly sensitive, rightly in my opinion, to the effect of an increase in interest rates in the United States on emerging countries. To a certain extent, it is the Fed which defines the monetary policy of emerging countries which very often have currencies linked to the dollar. It can clearly be seen that if the Fed increases its interest rates too quickly, it will partially cut the financing of emerging countries. It is a classic effect which means that when interest rates are very low in the United States, stakeholders borrow dollars there to invest them in emerging countries which have far higher growth rates and thus far higher interest rates, thus benefitting from a highly favourable transfer. If interest rates rise in the United States the money will be withdrawn from the emerging countries and return to the United States. This can, then, create profound crises in the emerging countries, as we saw a little over a year ago, when the Fed increased or threatened to increase its interest rates. So the Fed will be prudent in increasing its interest rates, I am certain, being very conscious of these two phenomena.
As for the ECB, I think it understands the challenges very well and that it has manifestly acted well until now. It must nevertheless face up to several new phenomena. The first is that the effect of the low interest rates is starting to fade and even become dangerous, as stated earlier. I took the example of France but it is also true elsewhere and the French banks are among the most solvent. They are in excellent health compared with Germany or Italy. But you can see that even the French banks are affected in their commercial banking business in France.
At the same time, let’s not forget that we are asking the banks to increase their solvency ratios very significantly. In general, since Basel III, they have been asked to double their so-called “hard” equity. It is difficult, however, to ask the banks to greatly increase their solvency ratios, and thus their equity, at the same time as reducing their profits. Growth must not, for example, take off again more strongly while the banks are caught in a trap, unable to sufficiently follow the excess credit demand which would result from this.
Fundamentally we clearly understand that the ECB thus initiated this policy – even if it does not say so – to facilitate the budgetary solvency trajectories of the various eurozone countries, as we have observed. In reality it has bought some time. The ECB “deal” is clear. It is carrying out an extremely low interest rate policy pending two things from the nations. The first is that they carry out the structural reforms necessary to increase their growth potential and reduce their structural deficit, thus facilitating their future budgetary trajectory by protecting their solvency. The second is that they constitute the institutional conditions of a viable eurozone. We know today that the incompleteness of the eurozone is manifest in terms of institutional arrangements, in other words the ability to make the zone work without it necessarily always being up to the countries doing least well to incur the cost of the necessary adjustment, with the consequences on votes that we know. The ECB is saying to the member states: “Quickly organise the eurozone a little better”.
The problem we face today and which leads me to question the increase in interest rates in the eurozone, is that the countries which should have done so have not done this work. The structural policies have virtually not been carried through where they were necessary. This started in Italy but was stopped following the failed referendum. In France we have not done much. There will be no capacity for exiting the dangerous solvency areas of the countries concerned if there are not, on the one hand, these structural policy efforts and, on the other, the completion of a more complete, better regulated eurozone which works better, in other words less asymmetrically.
The Germans, however, severely criticise the ECB’s monetary policy which is not necessarily favourable to them. They have a higher growth rate; they therefore do not need such low interest rates. Furthermore, these rates reduce the return on Germans’ savings who, as we know, have a much older population. They thus need a higher return on their savings, even more so for institutional investors who had, in the past, sold annuities at fixed rates.
The head of the ECB is not wavering from his policy. As for the Germans, they are obsessed with the question of the moral hazard, insofar as they do not want the solidarity factors necessary to the eurozone. They refuse, and this can be understood, to be the only ones to pay for everyone if the others do not make their structural reforms, thus finding themselves sooner or later in a situation of being dependent on Germany in the long term.
Exiting low interest rate policies is therefore conditioned by the fact that France in particular is carrying out structural reforms which reassure the Germans who would thus accept a much better institutional arrangement for the running of the eurozone, with intra-zone solidarity factors so that the cost of the adjustments does not weigh on the weakest countries.
This is where we currently stand. The European Central Bank, it seems to me, is going to exit negative short rates sooner or later because this position is becoming difficult to maintain today. But exiting a very low interest rate situation will depend fundamentally on the capacity of countries to carry out their own reforms and simultaneously integrate the reforms of the eurozone necessary for its future. In 2019 when Mario Draghi’s term of office ends, everything will depend on the relative strength of the countries and on their respective abilities to be heard, in other words to have triggered the structural policies sufficient to be credible. This credibility of the major countries conditions the possibility of increasing the viability of the eurozone, by developing several federalist elements, such as the mutualisation of part of the sovereign debts or tax transfer elements, as exists between states in the United States. To support temporarily those going through asymmetric troubles, without asking them to act only by austerity measures.
Which would enable interest rates to be increased much more easily. If we increase them significantly without having done that the intrinsic risk of the eurozone is increased. If we don’t raise them, the risks of a very low interest rate policy described earlier will become increasingly strong, whereas growth is beginning to rise again, along with, to a limited extent, inflation.
Without thinking of a significant rise, what is most likely, in my opinion, is that we shall, at least, see a moderate increase in interest rates from the end of 2017 or early 2018. Short interest rates could come back from their negative territory towards zero. And long interest rates could be managed towards neutrality by the central bank, in other words between 2 and 2.5%. This would be compatible with the level of growth and inflation that we can currently anticipate. This moderate increase will stop facilitating countries’ debt reduction, without, however, propelling them into a snowball effect.
Monetary policy of low, even negative, short-term and long-term rates unquestionably made it possible to avoid a catastrophic risk in 2007–2009, and then in the eurozone from 2010 to the present day. And subsequently to revive growth, if only at a low level, by stimulating the demand for credit and sustaining consumption and investment.
In the period of low growth we are currently experiencing, the monetary policy implemented by the ECB is helping the public and private sector to reduce indebtedness by guaranteeing nominal interest rates lower than the nominal growth rate, or at least at the same level. This was an essential move, lest we forget that the crisis of 2007-2009 occurred on the heels of a cycle of household and corporate debt accumulation that gradually became unsustainable. This major financial and economic crisis led in turn to a very sharp rise in public debt. Causing a drastic fall in long-term rates by buying government bonds, Mario Draghi succeeded in halting the infernal cycle based on the contagion of mistrust vis-à-vis the public debt of certain European countries. This mistrust led to a speculative hike in their interest rates, which in turn worsened both their public deficit and consequently national debt, causing a further loss of confidence.
This policy of very low, and even negative, rates was also intended to sustain global demand for credit. In principle, interest rates lower than the growth rate will sooner or later encourage lower savings and higher consumption and investment and, ultimately, stimulate growth. The current rates for property loans are a perfect illustration of this, with historically low levels. Finally, by increasing asset values (real estate, equity, etc.), lower rates also give rise to a favourable wealth effect on consumption and investment.
But if confidence fails to follow, the demand for credit can remain sluggish in spite of lower rates. In 2014 in France for example, demand remained below banks’ expectations in terms of the projects they wished to finance. Conversely, between late 2014 and early 2015, French companies rediscovered the taste for investment with a strengthening in the demand for credit.
What has been the impact on the banks? Very low rates unquestionably eat into the profitability of the banks. A bank’s net interest margin corresponds to the interest received on its outstanding loans less the interest paid on deposits. If the margin rate falls, coming up against the impossibility of lowering remuneration on deposits (which is virtually impossible to move into negative territory) relative to that received on loans, banks’ income will fall. The current challenge for the banks is therefore to offset this loss due to the rate effect by a positive volume effect. If total demand for credit increases, notably due to the lower rates caused by the central bank, every bank can take advantage. But should demand fail to grow sufficiently, the sector contracts.
The total volume of credits in France in 2015 increased sufficiently to offset the negative rate effect. But this volume effect tailed off in the first half of 2016.
However, the lower interest margin was offset during this period by the lower cost of risk. By supporting the economy, lower rates will mechanically lower the cost of credit risk. Since 2014, with the lower cost of risk having gathered pace, the banks have been able to offset the negative rate effect and inadequate volume effect. But we are now coming up against an impasse. Were the rate effect to persist, the cost of risk could not fall indefinitely and continue to produce its offsetting effect.
By sharply reducing banks’ future profitability, very low rates would ultimately restrict the supply of credit, at a time when banking regulation is demanding significantly higher solvency ratios with the corresponding strengthening of capital. All the more so as it is impossible to easily make capital increases due to the profitability of the banks falling below their cost of capital. The continuation of such a policy could therefore ultimately have a negative impact on growth. It should be noted that, unlike in the USA where the markets provide the majority of financing requirements, in Europe the situation is the opposite. Maintained at such low levels, sooner or later interest rates could also cause a property or equity bubble. Finally, they also undermine life insurance companies and pension funds.
The policy of very low rates has been essential. Which other monetary policy could have been implemented without taking even higher risks? It has also bought a certain amount of time, notably in the eurozone, providing room for governments to implement the structural reforms required to lift their growth potential and to make vital institutional changes within the monetary zone (genuine coordination of economic policy, certain elements of public debt pooling, etc.). But it is unclear if this time has been spent wisely. And time is of the essence.