Conduct of Monetary Policy with Financial Stability Issues and a Challenging Environment: The Case of Serbia
- International Monetary Fund
- Published Date:
- May 2017
I will cover the conduct of monetary policy in Serbia, with regard to our main financial stability concerns.
First, I will explain the inflation targeting monetary framework in Serbia and euroziation as a major financial stability concern. Then, I will touch on the challenging international environment. Afterward, I will try to explain the adjustments in conduct of monetary policy in Serbia and describe some challenges.
2. Inflation Targeting Framework in Serbia
Grounds for inflation targeting were introduced in Serbia in 2006, but only in 2009 did we officially run inflation targeting (IT) as a monetary framework, after a memorandum with the government was adopted. In the implementation of the IT framework, standard instruments are used. The key policy rate is our main policy instrument. The key policy rate is, currently, the maximum bid rate applied to the one-week liquidity absorbing repo operations. The key policy rate is complemented by a symmetric interest rate corridor to support the transmission of the key policy rate to money market rates and limit their volatility. We also use required reserves as an important auxiliary monetary and prudential instrument. Indeed, when the reserve requirement ratio is high as it is in Serbia, it can be used as a liquidity buffer in case of some external shocks, under certain stress scenarios. Also, we closely monitor FX market movements, and we are active in the FX market mainly to smooth short-term excessive volatility of the exchange rate. One more important supplementary instrument that we use are FX swap auctions. I will explain it later.
Generally, as you can see in this first graph, from 2013 onwards, we succeeded in lowering and maintaining inflation at low and stable levels. Currently, it stands at 3.6 percent and we expect through 2017 to remain within our target tolerance band set at 3 percent ±1.5 percent. On the right hand side of Figure 1, you can see our key policy rate changes and interest corridor movements together with money market rates. We have been in monetary policy easing cycle since May 2013. Up to now, we have decreased the key policy rate by 7.75 percentage points. The key policy rate is currently 4 percent in Serbia.
3. High Euroization as a Financial Stability Concern
High and persistent euroization is one of the major financial stability concerns in Serbia. Looking at the figure that shows the share of dinars in banking loans and banking deposits in the private sector, on the banking loans side, we can see that the degree of euroization has been steady at about 70 percent over a longer period of time. It moves sometimes up and down depending on the exchange rate movements and some government subsidy programs. On the deposit side, although euroization is even higher, you can see a favorable trend starting from 2012. We had an increase in dinar deposits in the private sector from 19.3 percent in 2012 to 28.8 percent at the end of 2016 (an increase of about 8 to 9 percentage points) that we attribute to a period of macro-economic stabilization in the country, low inflation, and relative stability of the exchange rate, which have contributed to strengthening the confidence in the national currency as a reliable store of value.
Reasons for the high euroization level in our country stem mainly from a legacy of protracted and acute macroeconomic instability. Serbia experienced hyperinflation during the 1990s that was, at that time, the second largest ever recorded in the world. Hyperinflation was followed by unfavorable geopolitical developments: disintegration of the ex-Yugoslavia, war, sanctions, etc. At that time, monetary policy and fiscal policy were not coordinated, so expansionary monetary policy together with soft budget constrains led to entrenched expectations of persistently high inflation. Very soon, as a consequence, an indexation mechanism emerged, with prices linked to exchange rate changes, and only formally published in dinars, a “black” currency market emerged, and trade was performed in foreign currency. Confidence in the national currency was eroded, so that it could hardly fulfill the traditional role of a currency: unit of account, means of exchange, and store of value. Generally, for a long period of time, following this period of acute instability, there was no unique, common approach to restore confidence in the national currency and reduce the use of foreign currency in the local economy. Instead of a common approach shared among and followed by all stakeholders – the central bank, government, and commercial banks – the National Bank of Serbia (NBS) was, in a way, left alone to implement its monetary policy in a highly euroized environment.
So, what is the relationship between euroization and financial stability concerns? In the absence of domestic currency sources of funding, banks started to use foreign currency-denominated sources of funding (domestic or imported) to extend loans to their clients, and because they didn’t want to carry FX exposure in their balance sheets, they extended loans in foreign currency or in domestic currency with a FX clause indexing the payment amounts to the exchange rate, thereby, de facto, directly transferring the exchange rate risk to their clients. The clients’ exchange rate risk, however, represented a credit risk for the banks because of debt servicing capabilities of unhedged borrowers; corporates and households started to be very sensitive to sharp exchange rate movements. Over time, after a few episodes of sharp currency depreciation, the problem with non-performing loans (NPLs) has arisen. Recently, NPLs have been treated as a major financial stability concern, and that’s why two years ago the government and the NBS together adopted a NPLs resolution strategy as a set of measures to tackle the problems of NPLs. Currently, NPLs are about 17.4 percent of total loans, lower by approximately 4 percentage points compared to 12 months before. Therefore, although still high, NPLs are on a decreasing path.
The NBS runs a “managed floating” exchange rate regime, which should provide adjustment to the economy via the exchange rate in case of some external shocks, so that the burden of adjustment can be transferred across the economy. Such potential advantage of the managed floating exchange rate regime is undermined in the context of high euroization rates as it cannot be used in full capacity due to financial stability concerns. That accentuates the importance of exchange rate movements from a central bank’s perspective. It requires a more active role of the central bank in the FX market in order to minimize excessive exchange rate fluctuations and provide relative exchange rate stability within narrow boundaries.
Regarding monetary policy implications, because the share of domestic currency in the financial system was just about 30 percent, it forced us to keep policy rates very high over a longer period of time, and have major changes in its level to be effective enough to transfer to the money market interest rates. Otherwise, the interest rate channel would not have functioned efficiently. The effectiveness of the interest rate transmission channel was, therefore, partially undermined by the degree of euroization and had to take into account any possible indirect effect of changes of interest rates on the exchange rate and, eventually, on the debt servicing costs of unhedged borrowers. Changes in the key policy rate are primarily determined by the changes of interest rates in the euro area. During periods of downward pressure on the dinar exchange rate, monetary policy had to take into account the risk of possible increase in NPLs, and use FX interventions to prevent significant depreciation.
Even in tranquil times, much like now (low inflation and weak economic growth), the possibility of using some central bank instruments to promote credit activity may be limited because of high euroization. For example, stimulating credit activity by diminishing required reserves ratios may not be so efficient if banks are already burdened with high levels of NPLs. That would prevent them from further credit activities against risky borrowers, such as corporates and households, and point them to safe assets.
4. Challenging Environment
Small, open economies with a still significant pass-through of exchange rate changes to inflation, such as Serbia, are very careful about global capital movements. Among various things, we particularly follow actions of the central bank of the largest economic areas, the ECB and the US Fed. And I will just comment that the Fed quantitative easing (QE) exit and more restricting monetary policy have had significant effects in portfolio outflows from Serbia, mainly from our domestic currency government bonds, because US funds were major investors in these bonds. Actually, Serbia attracted these investors in the Serbian government debt market during the Fed QE programs. When these flows were reversed, we had to cope with the exit of some investors and the resulting depreciation pressures on the currency.
On the other hand, the ECB negative rates and the asset purchase programs – that put plenty of euro area liquidity on markets – have not had significant portfolio inflow effects on our country. We have not seen so visible and direct exchange rate effects, although significant incentive to lending emerged because of low interest rates, which also significantly contributed to ease debt servicing costs and to put NPLs on a downward trajectory.
I will shortly mention some spillovers of euro area negative rates to our economy. Regarding financial stability, we didn’t see excessive risk-taking from banks in Serbia. So far, we have not seen even desirable risk-taking, because banks are risk-averse, and hesitant to invest in the private sector because of high NPLs. They prefer investing in government bonds that are risk-free and attractive due to higher interest rates. We have not seen cash substitution either (withdrawing savings from banking sector), despite very low interest rates on FX, even though, we tend to have a high propensity for keeping money “under the “mattress.”
Serbia has continued enjoying growth in FX savings since the introduction of negative euro area interest rates and the launch of the asset purchase programs. We have not experienced currency substitution on the deposit side either (FX currency savings being converted to dinar currency), although that would be desirable from the deeuroization perspective.
On the other hand, negative euro area interest rates posed serious challenges to NBS FX reserve management. The traditional principle “security comes first” led to negative yields from the investment of NBS FX reserves in euro-denominated assets, and placement of euro reserves operationally became much more demanding because of preset timelines, new rules for charging penalty rates by first-class international counterparties, etc. We experienced a period when banks tried to transfer the cost of negative rates to our central bank by allocating euro funds over required FX reserves in our system. In response to that the NBS introduced a penalty rate on banks, FX holdings beyond mandatory reserve requirements, and offered banks the opportunity to place funds with central bank accounts under a contract that would enable them to get the same yield for the funds that the NBS gets in international markets. Also, there is a question of volatility of our FX reserves because our intervention currency is the euro and under our required reserve instrument, full averaging is allowed. That could put FX reserve management under pressure to hold more euros on sight deposits and induce more costs.
5. Adjusted Conduct of Monetary Policy
Financial stability and a changing market environment had significant effects in designing and adjusting monetary policy instruments in Serbia. Here, I am going to concentrate on a few adjusted instruments designed specifically to promote domestic currency.
First, in collateral policy, only securities issued in domestic currency can be used for open market operations and credit facilities. Currently, only government bonds and some international bonds issued in dinars are available.
Second, required reserve rates are set in such a way as to stimulate both longer term and dinar sources of funding and to discourage shorter-term and FX sources of funding. For example, currently the required reserve rate for short term FX sources of funding in our country is 20 percent and 0 percent for longer term and dinar sources of funds. Also, remuneration in required reserve policy is used to stimulate domestic sources of banks’ funding: the domestic currency required reserve is remunerated and FX required reserve is not. Colleagues from the IMF correctly noted that even a 0 percent remuneration rate for the FX (euro required reserve) in the international environment means subsidizing banks because of the negative euros area interest rates. As mentioned before, banks in Serbia were trying to deposit euro at the NBS in excess of the reserve requirements, so, in 2014, we introduced a penalty rate for banks holding FX funds in excess of the required reserves.
Third, as I explained before, the NBS had to be more involved in the FX market to lower short-term exchange rate oscillations. That is why the NBS invested a lot of effort in market-monitoring in previous periods. This investment in market-monitoring capabilities allow us to provide quality information to decision-makers to increase the efficiency of our FX interventions, that is, to be able “to achieve more with less.” The NBS interventions are performed on both sides of the market. We are not price makers but price takers, and we believe that, in this phase of our development, it is important to provide together low inflation and relative exchange rate stability with the aim of promoting deposit dinarization, which would in turn enable and encourage loan dinarization.
Finally, the NBS also uses FX swap auctions as a supporting tool to provide banks with an additional liquidity management instrument, to support the development of the interbank swap market, and to support banks to overcome internal trading limits with other banks.
I will try to conclude by providing some lessons that we learned and possible challenges in Serbia.
The NBS prefers using market incentives rather than the adoption of administrative measures. The region has already suffered a lot from Swiss franc mortgage loans and, in some countries, FX loans to households are forbidden, and some central banks provided conversion in other currencies, while the NBS did not use any administrative measure. We still prefer using market incentives in the first place.
Unilateral euroization is not an option for Serbia. In light of our economic competitiveness, internal and external imbalances, and financial system vulnerabilities, adopting the euro at the present juncture could be rather detrimental for the economy, and not beneficial. The aim of Serbia is to adopt the euro ultimately, as a member of the EU and after all convergence criteria are met.
In our opinion, the de-euroization process is a kind of a marathon, not a sprint race. We see it as a long road that needs to be crossed with a series of small steps. In this phase, we consider that “deposit de-euroization” is the key. Higher deposits in domestic currency would later provide ground for higher lending activity in domestic currency.
We insisted on improving macroeconomic fundamentals in good times and doing structural reforms to build buffers against external shocks in bad times. The government put a lot of effort into the past three or four years to lower the fiscal deficit and to implement structural economic reforms that contributed to a more favorable business environment in Serbia, upgrades by major rating agencies, and improving external imbalances. Also, it was beneficial for implementing monetary policy in Serbia. The challenges for the future are to be persistent in providing good macroeconomic results, to persevere with our policies to deliver good results over a long period of time.
Last, but certainly not least, there is a need to invest more effort in general financial literacy and education of the public. Specifically regarding market risks (exchange rates and interest rate risks), to which the most sensitive parts of the economy – households and corporates – are exposed, there is more to be done to raise public awareness and understanding. Only properly understanding these risks can enable protecting against them. Education is an important part of financial market development, because the public needs first to understand the main market risks, and only then would people and corporations start thinking about the hedging against them.