I will present today the results of the work that the IMF has conducted with the Bank of Albania on the issue of de-euroization as part of its technical assistance provision.
Figure 1 represents the cycle of euroization as we have witnessed it, with small differences in several countries. It starts with phase zero, which is the natural level of euroization. Here, the benefits of euroization balance its costs, and the level of euroization is mainly driven by structural factors.
Then, because of external shocks or policy issues, significant macro-economic imbalances emerge, which usually lead to a large current account deficit and exchange rate depreciation. Expectations of exchange rate depreciation become entrenched. As a result, wages are indexed to the exchange rate and saving is done in foreign currency as a hedge against exchange rate risk.
Then the authorities step in and stabilize the macroeconomic context, which narrows the current deficit. As result, two-way exchange rate risk is reintroduced in the foreign exchange market; the risk of depreciation abates; the exchange may appreciate as well. It is more difficult to bet against the domestic currency.
However, the euroization usually does not disappear at that point, because the public is still attached to the insurance value of saving in foreign currency. While economic agents witness and are aware that in the short- to medium-term the exchange rate may move both ways, saving in foreign currency is still perceived as a cheap insurance against the risk of a recurrence of major macroeconomic shocks experienced in a non-distant past, which may engender an abrupt and significant exchange rate depreciation. So, what the policy makers can do in the short-term is to increase the “insurance” premium implicit in the foreign currency saving by keeping a minimum spread between the local currency and the foreign currency interest rate, widening it to the greatest possible extent via regulation while maintaining two-way risk in the foreign exchange market. At the same, by maintaining a stable macro-economic context, by strengthening institution and policy credibility, they should reduce the perceived risk of tail scenarios.
Then, you reach the third phase, where euroization is not due to hedging, is not due to insurance, but is the result of portfolio optimization, which means that the public saves in the currency that minimizes the variance of its saving portfolio. So, in that case, if you want to strengthen de-euroization of deposits, you must minimize the inflation compared to the exchange rate volatility to make saving in local currency the portfolio with the least variance. And it is there that the inflation targeting takes a more important role.
Phase 0: Optimal Level of Euroization
Let’s start from phase zero. What are the benefits of euroization? Financial euroization could contribute to financial deepening, by encouraging banks’ deposits rather than hoarding foreign currency cash, the integration in the international capital markets, and the access to financially attractive foreign funding sources.
What are the Costs?
The financial cost of euroization includes the loss of seigniorage. When the public uses foreign currency banknotes, it reduces the circulation of local currency banknotes, leading to a smaller liquidity deficit in the banking sector, which reduces the profit of the central bank. Because of the financial stability risk due to euro deposits, the central bank must also keep more foreign reserves than without euroization, which have an opportunity cost. This cost is usually non-negligible and could be considered a fiscal loss. For Albania, it is between 0.4 percent to 0.6 percent of GDP per year.
Another issue is the effectiveness of monetary policy. First, euroization directly reduces the scope of financial intermediation that the central bank can influence. Second, financial euroization impairs the monetary policy transmission, as the exchange rate transmission channel negatively interacts with the interest rate channel. In a non-euroized economy, a decrease in the policy rate leads to a depreciation of the exchange rate and a decrease in banks’ lending rates, which both support inflation. Factor in euroization, and the depreciation of the exchange rate due to a cut in the policy rate could lead to an increase in non-performing loans because of unhedged borrowers’ exposure to the exchange rate risk, which in turn deteriorates lending conditions as banks try to recoup the losses due to higher non-performing assets.
The last issue with euroization is related to financial stability. As mentioned earlier, unhedged borrowers are exposed to exchange rate depreciation. The depreciation of the exchange rate could also affect local currency borrowers, as a typical defense against exchange rate depreciation is to raise domestic interest rates, leading to an increase in local currency non-performing loans.
It is not possible to accurately measure the cost and benefits of euroization, except for the loss of seigniorage. Therefore, we decided to estimate an empirical euroization benchmark, which is defined as the level of euroization that could be predicted based on a country’s structural features. For the latter, we included the optimal currency area criterion, which determines the choice of an exchange rate arrangement and encompasses economic size, trade, and capital count openness, as well as remittances, which are also strongly correlated with euroization. We assume that the optimal level of euroization does not deviate much from its natural level.
Our estimate of structural euroization also includes policy variables. The objective is to neutralize them: we predict the level of euroization as if macroeconomic policies have not influenced euroization.
For Albania, we have estimated a natural level of euroization between 40 percent and 35 percent. It, therefore, deviates from the current level of 48 percent. It is not a very large deviation, and, relative to the other countries, Albania is not very far from its natural level. But the issue is that the deposits euroization increased recently, so it is deviating from this natural level, which justifies the recently announced de-euroization policies.
Phase 1: Maintaining Two-Way Risk in FX Market
The first step toward de-euroization is to introduce two-way risk in the market to fight one-way bets against the local currency. Figure 3 shows the odds of an exchange rate appreciation on an 80-day rolling basis. When the indicator is above one, the odds of an appreciation have been higher than the odds of a depreciation during the past 80 days. On the other hand, if the indicator is below one, the odds of a depreciation have been higher than the odds of a depreciation.
It should be noted that in Albania the odds have been close to one during the past five years, meaning that the public should have equal expectation of exchange rate appreciation or depreciation. This means that the public should perceive two-way risk in the foreign exchange market and, therefore, the hedging value of FX assets should have declined. As a result, the share of euro deposit increased when the odds of a depreciation were high, but stabilized when two-way risk was introduced in the market.
Phase 2: Why Would Euro Negative Interest Rates Matter for De-euroization?
So let’s move from phase 1 to phase 2.
Here it is a question of foreign currency deposit as an insurance and what premium is paid. To encourage domestic currency intermediation, the intermediation spread in domestic currency should narrow vis-à-vis the intermediation spread in foreign currency so as to make domestic currency deposits and loans relatively more attractive financially vis-à-vis foreign currency deposits and loans. It is here that the negative euro interest rates are creating a problem from the perspective of leaning against and fighting excessive euroization. Why? First of all, euro loans are cheap because FX loans are mostly indexed to Euribor, and now Euribor is in a negative territory or very low. But, on the other hand, the negative rate is usually not passed onto euro deposits, especially on retail deposits.
Euro negative interest rates and de-euroization
Euro negative interest rates and de-euroization
Euro negative interest rates and de-euroization
What we noticed in the euro area is that negative rate transmitted very well to the money market, but not to retail funding. For Albania’s banking sector, which is mainly financed by retail deposits, zero is the floor and deposit rates did not follow the euro area money market rates. As a result, the foreign currency intermediation spread is compressed compared to the local currency intermediation spread, encouraging financial euroization.
Regulations preventing negative remuneration of customer deposits, – as well as the practice of many central banks in the Western Balkans to not pass the negative remuneration of their euro reserves on banks’ accounts with the central bank – contributed to the relative compression of the intermediation spread.
Phase 2: Increasing Euro Deposit Insurance Cost in the Context of Euro Negative Rates
So, what can we do? It is important to try to correct the relative mispricing of the foreign exchange deposits. The first priority is to leave it to the banks to decide whether and how to pass the negative rates to customers. What we have seen in the euro area is that banks themselves would not pass a negative rate if they felt they would be losing funding. Therefore, it is better to let them decide to which type of clients they can pass the negative rates.
The second important aspect is to pass the negative rate to bank accounts at the central bank, including the reserve requirements, to avoid subsidizing euro deposits in the banking sector. This is what the Bank of Albania did recently.
It is also important to recall what is the role of reserve requirement on foreign currency deposits. It is mainly prudential: the idea is to make sure that the banks keep a certain amount of high-quality liquid assets at the central bank always available to back up their euro deposits. This is important because the deposit guarantees will never have as many resources as the reserve requirements to back up euro deposits. It also forces banks to contribute to the reserve buildup at the central bank to provide emergency liquidity assistance in foreign currency, and it transfers the cost of the reserve build up to the banks, which in turn forces them to internalize the risk of foreign exchange intermediation. In that case, the central bank just acts as an intermediary collecting euros from the banking sector and investing them in the euro area money market in high-quality liquid assets.
That also justifies why the euro area money market interest rate should be transmitted to the reserve requirement remuneration. One other measure that is often applied is to discriminate in terms of the reserve requirement ratio between foreign currency and local currency to have an impact on banks’ cost of funding in favor of the local currency. It is important to have the right remuneration structure if you decide to differentiate between your reserve requirement ratio to avoid encouraging the collection of foreign currency deposit by remunerating the reserve requirement on foreign currency deposits above the market rate.
The last point to consider is adjusting the reserve requirement ratio depending on each bank contribution to euroization. This is also envisaged to be implemented by the Bank of Albania. For instance, a higher reserve requirement ratio could be applied to banks that have a percentage of foreign currency deposits higher than the rest of the industry, as they contribute more than others to financial euroization or, as planned in Albania, to apply a higher and more penalizing ratio on the amount of foreign currency deposits exceeding a certain critical threshold (for example, a higher ratio on FX deposits exceeding 50 percent or 40 percent of total deposits).
Phase 3: Portfolio Optimization
The last phase to address in the euroization life cycle is probably the most challenging for a small open economy. This phase is when euroization is driven by portfolio optimization considerations. The left-hand side of Figure 5, derived from the paper by Alan Ize & Levy Yeyati, shows the relation between the minimum variance portfolio and deposit euroization.
We note from the figure that there is a clearly positive relation between the two. The red dots represent inflation targeters with a flexible exchange arrangement in the Western Balkans. The blue dots are the countries with a fixed exchange rate arrangement. Normally, inflation targeters, like Poland, have a low minimum variance portfolio (MVP), and they also have a very low level of euroization. What is surprising is that, for instance, Albania has a relatively high minimum variance portfolio, which is largely because there is a high covariance between inflation and the exchange rate. So here we go back to the natural level of euroization: when you are a small open economy, you have a large pass-through, and then you create a link between the exchange rate and inflation. It is therefore more difficult to reduce the minimum variance portfolio. Another difficulty is that the pass-through is usually higher in countries that are highly euroized.
Phase 4: Pre-condition for More Exchange Rate Volatility
In this phase, the priority is to tackle the risk due to unhedged foreign exchange lending, which should reduce the fear of floating and, thus, facilitate the move toward inflation targeting and a flexible exchange rate.
The first order of business is to identify those borrowers who have a natural hedge against exchange rate swings, such as exporters, and those who are to a large extent unhedged. For this, a standardized and compulsory definition of unhedged borrowers for the banks is needed. Then, the first line of defense is to ask unhedged borrowers to have enough income relative to debt to be able to absorb potential increases in regular debt servicing installments as a consequence of exchange rate depreciation. This means imposing a limit to the maximum debt service-to-income ratio. The second line of defense is to ask the banks to have larger capital buffers to absorb the risk of non-performing loans in case of exchange rate depreciation due to unhedged borrowing. This may take place, for instance, by parametrizing the counter cyclical capital buffers on the basis of the share of unhedged borrowers in foreign currency to boost the loss-absorption capacity of banks with large indirect exposure to exchange rate risk. Alternatively the central bank could impose higher risk weight on unhedged FX lending to increase capital buffers, but also to discourage the provision of loans to unhedged borrowers.
