In my presentation, I would like to emphasize some risks to financial stability that are hidden if one observes only the aggregate data. In a sense, I am going to play the devil’s advocate. One of the things that I noticed in the reports of the IMF Spring Meetings, and that goes with the general wisdom in the world, is that regarding Central European and Southeastern European countries, their credit cycle position is in an expansionary phase. Several facts are mentioned to underpin this conclusion: credit growth is increasing, bad debt recovery is increasing, NPL ratios are decreasing, asset prices are increasing, and bank profitability is increasing. All correct, with a lot of “BUTs.”
I am going to treat some of those “buts” in my presentation, with the general disclaimer: “If I say something that you strongly disagree with, it is not the Central Bank of Bosnia and Herzegovina. It is my fault.”
In the first chart, I present financial assets and liabilities, without capital, of the banking sector of Bosnia and Herzegovina and the determinants of net interest income (Figure 1).
The determinants of net interest income in BH
Source: CBBH.Notes: Depicted are financial assets and liabilities, net of capital. Non-interest bearing assets up to 2013 include cash in vaults; as of 2014, reserves accounts with the CBBH are included. As of 2016, non-interest bearing assets exclude excess reserves, as the negative remuneration rate on excess reserves has been introduced. For this reason, these are on the liabilities side, although, technically, these are banks’ assets.The determinants of net interest income in BH
Source: CBBH.Notes: Depicted are financial assets and liabilities, net of capital. Non-interest bearing assets up to 2013 include cash in vaults; as of 2014, reserves accounts with the CBBH are included. As of 2016, non-interest bearing assets exclude excess reserves, as the negative remuneration rate on excess reserves has been introduced. For this reason, these are on the liabilities side, although, technically, these are banks’ assets.The determinants of net interest income in BH
Source: CBBH.Notes: Depicted are financial assets and liabilities, net of capital. Non-interest bearing assets up to 2013 include cash in vaults; as of 2014, reserves accounts with the CBBH are included. As of 2016, non-interest bearing assets exclude excess reserves, as the negative remuneration rate on excess reserves has been introduced. For this reason, these are on the liabilities side, although, technically, these are banks’ assets.In Figure 1, I showed 2008, which is the peak year, and then I showed all years from 2012 onward. The slide shows what was happening in terms of the banking sector and the changes in balance sheet positions. As one sees, there is a very mild credit growth from 2013 onward. “Nostro accounts” have been fairly stable. Non-interest bearing assets until 2013 are represented exclusively by banknotes and coins. After 2013, they also include required reserve accounts with the central bank for the reason that we lowered the remuneration rate at 0, while we lowered it below zero for the excess reserves as of 2014.
On the liability side, there is a strong deleveraging process as evidenced in the chart. Foreign liabilities dropped from 5.5 billion convertible marks in 2008 to approximately 2.2 billion currently. Regarding liabilities, domestic deposits represent the main source of interest expenses for commercial banks, together with excess reserves since 2016. In fact, although excess reserves are commercial banks’ assets, they engender an interest rate cost for commercial banks, as banks get charged a negative rate on excess reserves. The chart shows that as of 2014, the interest generating assets are fairly constant while the interest-bearing liabilities of commercial banks are increasing. Required reserves are roughly at the level of 4 billion and there are a few reasons for that. But the main reason, historically, was the regulatory requirements regarding the liquidity risk management. Regarding the network composition of commercial banks, in Bosnia and Herzegovina, banking supervision is not the responsibility of the central bank, but of two separate, sub-state level banking agencies that operate under very harmonized rules. Despite the harmonized rules, this structure poses a challenge from the perspective of consolidated reporting to the same foreign banking group. I am going to touch upon domestic deposits later on during my presentation, but they currently represent quite a significant fraction of liabilities of BH commercial banks. Such funding structure might possibly limit the credit growth in the future, especially if the EU regulation on cross-border exposure of banking groups is taken into account. Currently, commercial banks from the EU have a 100 percent risk weight on their exposure to Bosnia and Herzegovina even for a central bank’s reserves. Such a requirement weighs on banks’ profitability and, partially, explains the deleveraging trend.
Net interest margin (Figure 2) over the past five years has been fairly constant. I emphasize the fact that interest bearing assets were fairly constant, while liabilities were increasing. Net interest income minus interest expenses is, again, fairly constant. Relating to profitability of commercial banks in Bosnia, one must notice that business and direct expenses are those that drive the profitability of the banking sector in Bosnia and Herzegovina, especially the provision for the losses in the previous period.
Profitability of Bosnia and Hercegovina banks
Sources: Banking agencies, own calculations. “Net interest margin” defined as end-of-year net interest income to total assets.Profitability of Bosnia and Hercegovina banks
Sources: Banking agencies, own calculations. “Net interest margin” defined as end-of-year net interest income to total assets.Profitability of Bosnia and Hercegovina banks
Sources: Banking agencies, own calculations. “Net interest margin” defined as end-of-year net interest income to total assets.In 2015, the provisions were 524 million convertible marks, while in 2016 they amounted to 270 million. Very often we have claims by the media that the banking sector is profitable, despite relatively low interest rates and low credit growth. But one must have in mind what drives the profitability of commercial banks in Bosnia and Herzegovina, in 2016, the higher profits were driven by lower provisions.
About the NPLs issue, the ratio at the end of 2016 was 11.8 percent. This figure is very good compared to the rest of the region. However, I must emphasize that such a comfortable ratio has been reached after a long restructuring period, as much as the legal framework allowed, since we do not have a law regulating the transfer of NPLs and impaired assets to special purpose vehicles. There are persisting uncertainties concerning the tax treatment of such transfers and the disposal of impaired loans that the tax authorities have not clarified yet. That is one of the most important reasons the NPL ratio in Bosnia and Herzegovina is still relatively high, despite that the banking license of some entities has been revoked, so that their impaired assets do not count any longer as NPLs, and the non-performing assets of some of the remaining banks have been written off.
In terms of what is happening with interest rates (see Figure 3), let us compare the interest rates in Bosnia with those in the EU, as published on the ECB website.
Interest rates on loans to non-financial companies
Sources: CBBH, ECB, own calculations.Series: Firms, loans up to and including 1 million EUR, up to 1 year, 3 months moving average.Interest rates on loans to non-financial companies
Sources: CBBH, ECB, own calculations.Series: Firms, loans up to and including 1 million EUR, up to 1 year, 3 months moving average.Interest rates on loans to non-financial companies
Sources: CBBH, ECB, own calculations.Series: Firms, loans up to and including 1 million EUR, up to 1 year, 3 months moving average.The chart shows moving averages of interest rates in EU and Bosnia for loans to firms for an amount up to EUR 1 milion and with a maturity non exceeding one year. One can see that they are pretty much moving in the same direction as interest rates in the EU and that there is a direct pass-through of changes in the EU interest rates to interest rates in Bosnia and Herzegovina for both large banking groups – the top three in terms of assets represented in the chart by a dashed line – and the banking system as a whole.
However, if one looks at the interest rates on loans to households (see Figure 4), the weighted average interest rates charged by the top three groups, which are coming from the EU, in Bosnia are rather different from both the EU and overall Bosnian average.
Interest rates on loans to households
Sources: CBBH, ECB, own calculations.Series: Households, general consumption loans with maturity up to 1 year, 3 months moving average.Interest rates on loans to households
Sources: CBBH, ECB, own calculations.Series: Households, general consumption loans with maturity up to 1 year, 3 months moving average.Interest rates on loans to households
Sources: CBBH, ECB, own calculations.Series: Households, general consumption loans with maturity up to 1 year, 3 months moving average.As of 2014, the interest rates have been increasing. In my view, this increase is quite interesting. If the average interest rates for the banking system as a whole are fairly unchanged, and the top three groups’ rates are increasing, it must be the smaller banks charging much less, thus affecting the average interest rates in Bosnia and Herzegovina.
Therefore, another financial stability issue might be: which banks are driving the current credit growth?
Which banks extend loans?
Note: Figures in red indicate by how much bank’s balance sheet had to expand in order to experience reported annual credit growth.Sources: CBBH, own calculations based on December figures.Which banks extend loans?
Note: Figures in red indicate by how much bank’s balance sheet had to expand in order to experience reported annual credit growth.Sources: CBBH, own calculations based on December figures.Which banks extend loans?
Note: Figures in red indicate by how much bank’s balance sheet had to expand in order to experience reported annual credit growth.Sources: CBBH, own calculations based on December figures.The horizontal axis represents the contribution to total credit growth in 2016, with respect to 2015, of each commercial bank in Bosnia and Herzegovina. The vertical axis shows the share of individual bank in total assets in 2016. Clearly, there might be vulnerability forming. A single bank, the one in the far top right-hand corner, has accounted for one third of credit growth of the banking sector in one year. Another serious issue is the contribution of small banks, represented by red dots. Very small banks, those with a share in total assets of the banking sector below 3 percent, accounted for a quarter of credit growth in 2016 compared to 2015. A very small one, with the share of banking sector assets under 1 percent, accounted for about 10 percent of increase in stock of loans in 2016, compared to 2015.
The red figures show how much their balance sheets had to expand in order to experience this large credit growth. In short, there is a group of very small banks expanding rapidly. This is something, especially in the case of households, that we witnessed in 2008 when large players started getting rid of bad clients. Some small banks were picking them up, thereby enjoying rapid credit growth. These same banks experienced serious problems with the NPLs a few years later; some of them are still struggling.
Regarding the interest rates on deposits to households (see Figure 6), one notices a peculiar thing.
Interest rates on deposits of households
Sources: CBBH, ECB, own calculations.Note: New business coverage.The average interest rates on deposits of households over one and up to two years and over two years in Bosnia and Herzegovina are higher than those in EU. There are two things that we have to remember from earlier slides. I mentioned deleveraging of commercial banks in Bosnia and Herzegovina, which is the reason banks increasingly rely on deposits of domestic sectors. These are clearly savings deposits.
The vast majority of households’ deposits are transaction accounts and sight deposits. Transaction accounts are high around payday, generally at the end of the month, which is also the reporting period. So, even if households deposits look fairly stable and high, it does not mean that there is a great lending potential. Finally, which households’ deposits are growing?
In Figure 7, we have the cumulative number of deposits on the x axis and cumulative value of deposits on the y axis. The deposits up to a maximum of 50,000 convertible marks (KM) are all insured by the Deposit Insurance Agency (DIA). In Bosnia, everybody is quite aware of the significance of the DIA, as some banks have recently had their licences revoked and the prompt reaction of the DIA was very important in maintaining the perception of security of banks’ deposits, thereby preventing deposit outflows elsewhere. The chart shows that over 98.5 percent of the total number of deposits are those up to 50,000 KM. The vast majority are, actually, up to 30,000 KM. The cumulative value of deposits in this group was about 50 percent, which represents a 10 percentage point decrease with respect to 2010. Now, look at large individual deposits. Less than 0.15 percent of total number of deposits is over 20 percent of total value of deposits. From the chart it is evident that the number of large deposits is increasing over time.
Which households’ deposits are growing?
Sources: Deposit insurance agency, own calculations.Therefore, this perception of the overall households’ deposits growth does not mean that the average person saves more. There is a possibility that “cherry-picked clients” of commercial banks, who have accounts with the same banking group abroad as well, got more favorable interest rates for saving in Bosnia and Herzegovina. By transferring their personal, or their company, deposits from the headquarters, for example, in Austria, to personal bank accounts in Bosnia and Herzegovina, household deposits in Bosnia grow, while banking groups themselves would reduce intragroup exposure to Bosnia and Herzegovina – assuming that the extra domestic deposits replace intra-group lending.
To conclude, if the credit growth is as sluggish as it is right now, relying on the activity of a handful of very small banks, and if domestic deposits, as they are, become the main sources of financing, there is a concern that we are possibly moving toward an environment in which long-term investments get increasingly financed via short-term loans. This is similar to the environment already experienced in Bosnia in the early 2000s. This leads to similar maturity mismatch and liquidity risk problems as we have experienced in the past.
