Stability of a Currency Board: The case
of Bulgaria – hypothetical short-run
consequences of a crisis
Index
1. Introduction
1
2. Literature Review
2
2.1 History, features and financial system of the Bulgarian
Currency Board (CB)
2
2.2 Advantages of CB
7
2.3 Disadvantages of CB
10
3. Testing the stability of the Currency Board- applied sudden
stop in FDI in a crisis period
3.1
Empirical analysis
16
3.2
Results, evaluation and some limitations
18
4. Conclusion
21
Appendix
22
1
1. Introduction
Bulgaria is a small country in Eastern Europe and has been in the European Union since January 2007. The country introduced a currency board (CB) monetary system in 1997 following a major financial crisis which escalated in a hyperinflationary three-month period in 1996-1997.
In the current global crisis and political disturbances in Bulgaria it is more and more questioned whether the country should continue operating under the monetary policy regime of a currency board or abandon it. There are obvious benefits and costs. Three of the greatest advantages distinguished are economic credibility, maximum transparency and an automatic adjustment mechanism (Williamson, 1995). The last one, especially, (symmetric correction of the balance of payments and the money stock) maintains the currency board in the long run. The stability of the Bulgarian currency ( the lev) gives certainty for investing in the country. Another benefit is the easier fulfillment of the criteria for accession to the Eurozone in small open countries (as in the case of Bulgaria) (Mundell, 1997).
On the contrary, advocates against the currency board highlight the impact the global financial crisis has on the country and the even worsening effect of the board. The Bulgarian National Bank (BNB) loses autonomy in conducting monetary policy and thus cannot print money in case of a crisis and its role as a lender of last resort (LLR) to banks facing solvency problems is limited. Consequently, the only way money supply could grow is by accumulating euros in the Bulgarian National Bank’s currency reserve (held in a foreign bank like the ECB). The two ways in which BNB can obtain euro is from exports or foreign direct investments. The first has been mostly negative throughout the years following the introduction of the CB and the latter has been almost always positive.
Today, in the presence of political and economic disturbances in Bulgaria which is on its way to Eurozone accession, a deeper analysis of the economic indicators connected to the national currency and the CB stability in terms of its automatic balance of payments adjustment mechanism is needed. Consequently, the goal of the present paper can be summarized by the following research question:
2
How stable is the Bulgarian Currency Board in a hypothetical short-run scenario of a three-month crisis situation in terms of its Balance of Payments automatic adjustment mechanism? What is the effect on the economy of a negative shock to FDI and the whole Financial Account falling to 0?
In this paper, part two provides the literature review. In this paper, the history, the functioning and the basic characteristics of the currency board are explained. In the third part, a regression analysis is performed and the liquidity of the board of a hypothetical three month crisis period is examined. The empirical test is based on a linear regression including a time-series monthly data from 2000-2013 of the variables used. The conclusion is presented in the last part.
2. Literature Review
2.1
Currency board monetary system in Bulgaria –brief history, basic
characteristics and financial system
The currency board arrangement (CBA) was established in Bulgaria after a severe financial crisis which culminated in a hyperinflationary period in 1996-1997.
The harsh twin (banking and currency) crisis was the major reason for the adoption of the CBA. The subsequent events describe the Bulgarian economy throughout the period of financial turmoil which led the country to introduce the currency board monetary regime (Dobrev, 1999).
After the fall of the Communist wall in 1989 the Bulgarian economy was highly unstable and a major constraint was its high amount of financial heritage, namely, its large foreign debt amounting to 125 percent of GDP by the end of 1995 (Pavlov, 1999). To finance this high budget deficit (comprising of government deficit and a deficit of state-owned companies), the Central Bank of Bulgaria was printing Bulgarian levs and thus increasing the money supply. This resulted in high monthly inflation which escalated in hypeinflationary rates (inflation exceeding 50 percent for at least three months) of 240
3
percent in the beginning of 1997 (Barlemann & Nenovsky, 2004). Furthermore, budget deficit increased substantially reaching values between 4.4 -10.9 percent of GDP throughout 1992-1997 as a consequence of the increased government expenditure combined with less tax revenues (Pavlov, 1999). Another central difficulty for the Bulgarian economy during the above discussed period was the enormous losses of the state-owned enterprises and their financing through ‘bad’ credit provided by commercial banks. The banks were forced to subsidize those uncollectible loans by the government in an attempt to prevent the unemployment which would have resulted from the closure of the state-owned companies that were going bankrupt. Subsequently, in mid-1997 most of the commercial banks were becoming illiquid with 18 out of 46 banks being declared bankrupt already in 1996 (Pavlov, 1999). This led to the start of the banking crisis. Moreover, the bad-loan portfolios were followed by worsening of the monetary conditions. The Bulgarian exchange rate was highly volatile and the Bulgarian Central Bank took actions to stabilize it. In the short run, the Bulgarian National Bank was quite successful. However, when foreign currency reserves dropped to some 500 million USD in 1994-1995, several exchange rate devaluations took place, therefore inducing a rise in international competitiveness (Barlemann & Nenovsky, 2004). Reserves increased to 1500 million USD in late 1995. Nevertheless, the effect was only temporary as in 1996 foreign exchange reserves quickly started declining again to some critical amounts of 600 million USD prompting the start of the currency crisis. In 1997, the Bulgarian Central Bank engaged in a number of devaluations reaching a level of 230 percent with respect to the US dollar, reserves dropped to 300 million USD, output declined sharply by 7 percent and unemployment rose to 14 percent (Barlemann & Nenovsky, 2004).
Following this serious financial destabilization, replacement of the floating exchange rate regime with a new monetary system was considered necessary. Furthermore, the new system should meet three important requirements – to restore confidence in the financial system, to revive the economy and to bring stability in the long-run. Since reputation is the base of any inflation-targeting system and consequently ineffective in the short-run, Bulgaria introduced a currency board monetary system on July the 1st 1997 (Minea & Rault, 2011).
4
A currency board arrangement (CBA) is a monetary rule-based system in which the domestic currency is linked to a foreign currency and exchanged at a fixed rate that is guaranteed by law obligation (Balino & Enoch, 1997). The Bulgarian CB has been adapted to the country’s specific requirements of its political and economic reality and thus significantly differs from an Orthodox typical CB whose characteristics are defined according to Hanke and Schuler (1994, pp. 2 -10). The CBA in Bulgaria exhibits the following common features with an orthodox one (Dobrev, 1999):
It maintains full foreign exchange cover for its note, coin and deposit liabilities as set by law
It maintains a truly fixed exchange rate with respect to the reserve currency as set by law
It cannot finance domestic government spending as set by law The main differences are listed below (Dobrev, 1999):
An Orthodox CB Bulgarian CB
Usually supplies only notes and coins Supplies notes, coins and deposits Does not regulate commercial banks Regulates commercial banks Is not a Lender of Last Resort Strictly limited LOLR
Bulgarian Financial System
The Bulgarian financial system is divided in three main components, namely: the Issue Department, the Banking Department and the BNB Banking Supervision Department. The key for understanding the functioning of the currency board is the Issue Department which performs all operations related to the issue of Bulgarian levs and providing their unlimited exchange with reserve currency with the public, government and commercial banks (Dobrev, 1999). Its liabilities, consisting of notes and coins, reserves of commercial banks, government deposit and net worth expressed by the Banking Department deposit, are fully backed by the assets side comprising of the country’s foreign exchange reserves. The Issue Department’s Balance Sheet includes:
5
FX - foreign exchange reserves; C - currency in circulation;
R - reserves of commercial banks;
G - the government deposit with the Issue Department;
B - the Banking Department deposit with the Issue Department – a buffer absorbing shocks caused by asset operations; a positive amount permits the conduction of some discretionary monetary policy and provides an opportunity for exercising a strictly limited Lender of Last Resort (LOLR) function in case of a crisis.
Issue Department FX C R G B
The reason for describing and focusing on the Issue Department is that, in practice, its balance sheet plays the role of the currency board. In case of a crisis, the above mentioned department is the one to ensure and maintain the stability of the economy. In the past five years, the world is experiencign a severe financial crisis, which, in the third section of this paper, is hypothetically assumed to escalate and leave the Bulgarian economy without any financial inflows from abroad, which are a leading factor for the Bulgarian economic growth. Whenever some disturbances in the balance of payments and its automatic adjustment mechanism occur leading to a lasting deficit, the BNB, which has very limited functions, can engage in some attempts to restore BoP equilibrium. The government cannot conduct monetary policy and therefore, whenever there is an abrupt fall of the foreign exchange reserves of a significant amount, the country can finance this decline with the currency in circulation, gold reserves, cash from maturing securities and Banking Department deposits, if sufficient amounts are present in the BNB balance sheet.
6
Money Supply
For the purpose of this paper analysis, it is important to describe the Bulgarian money supply and its reaction to changes in the foreign exchange reserves since the Bulgarian currency board is particularly sensitive to capital inflows and outflows and to the balance of payments surplus/deficits, which ultimately cause money supply expansion or contraction. Furthermore, operating under the monetary regime of a currency board, sterilization of the money supply cannot occur when affected by capital flows (Avramov, 1999).
Some notations:
Mb – monetary base; M – money supply; m – money multiplier; c – cash to deposits ratio; r – reserves to deposits ratio; D – deposits.
Mb = C + R = FX – G – B M = C + D= m.Mb m = 1+c/1+r
M = m. [FX – G – B] - this equation determines the money supply in Bulgaria under the CB monetary system.
Subsequently, from this equation the following deductions can be made as summarized by Dobrev (1999, p.20):
- ‘’ceteris paribus, changes in FX, which are not matched by an equal change in B or G,
proportionally cause identical changes in the money supply. Such changes arise from exchange of domestic currency against reserve currency by the public or the commercial banks;’’
- ‘’ceteris paribus, changes in B, which are not related to equal changes in FX, proportionally cause inverse changes in the money supply. Such changes arise from covering the BNB operating and investment costs and possible exercise of the Lender of Last Resort function of the Banking Department;’’
7
- ‘’ceteris paribus, changes in G, which are not related to equal changes in FX, proportionally cause inverse changes in the money supply (see also Nenovsky and Hristov, 1998, pp. 16 - 18). Such changes arise from issuance and redemption of securities, tax collection, payment of salaries, pensions and subsidies from the state budget and all other types of budgetary operations executed in the domestic market.’’
An extension can be made to the previous subsection description of the Bulgarian financial system. The first deduction which follows from the money supply equation shows that in a case of a decrease of the foreign exchange reserves, the money supply also declines with the same amount. Consequently, the picture from the current and previous subsections can be summarized as follows: a decline in capital inflows or increase in current account deficits cause BoP deficit and disequilibrium. This, in turn, results into an identical fall in the money supply. If the decline in FX is extremely sharp and the balance of payments does not automatically adjust, the BNB Issue Department comes into action. The foreign exchange reserves on the asset side are falling and so the Bulgarian Central Bank can help endure the critical situation by using cash and notes, commercial banks reserves or the Banking Department deposit until GDP and the current account can react and restore equilibrium.
2.2 Advantages of a Currency Board
Four of the greatest advantages of a currency board as discussed byWilliamson (1995, p.13) are:
‘’that they assure convertibility; that they instill macroeconomic discipline; that they provide a guaranteed payments adjustment mechanism; that because of those three features they create confidence in the monetary system and therefore promote trade, investment and growth.’’
One of the main features of the currency board is the assurance of convertibility of the currency at a fixed exchange rate. In order for this guarantee to be taking place, there must be certain amount of reserves to meet any demand for foreign currency (Miller, 1999). Accordingly, only then the currency board credibility will be sustained and citizens would be able to exchange currency freely and in unlimited quantities.
8
The second advantage of a currency board as discussed by Williamson (1995) is that it instills microeconomic discipline. As it was already discussed in the first subsection, the economic situation which led to the adoption of the currency board in Bulgaria was highly unstable. After the introduction of the new monetary regime in 1997, the currency board was very successful in bringing an end to the preceding troublesome period. Inflation dropped down from values of 240 percent in the beginning of 2007 to 1.0 percent in 1998 on an annual basis. The annual average inflation measured by the harmonized index of consumer prices for August 2012- July 2013 has been 2.1 percent (WEO, 2013). Furthermore, government debt which had a value of 70 percent of GDP in 1996 has been declining ever since the introduction of the currency board and its amount for 2012 is 18.5 percent of GDP. Accordingly, major accomplishments have been made and the currency board can be given high marks for the overall Bulgarian macroeconomic discipline in the current global crisis.
The last advantage of a currency board mentioned is the creation of confidence and promotion of trade and growth as discussed by Williamson (1995). Before the adoption of this monetary regime, the exchange rate was highly volatile and after the currency board was introduced confidence was restored as shown by the base interest rate movements – they declined dramatically from level of 97.5 percent in 1996 to rates below 10 percent ever since. Consequently, the exchange rate was stabilized and the country has been attracting foreign capital although trade of goods has been negative and growth has been slow as a result of the ongoing world crisis.
In this paper, only the third advantage of providing a guaranteed balance of payments adjustment mechanism is discussed in greater detail as it is most relevant for the topic in question.
Guaranteed Adjustment Mechanism
The third great advantage according to Williamson (1995) is that:
‘’it builds in a payments adjustment mechanism. This is none other than what has often been identified as the gold-standard adjustment mechanism, otherwise known as the
price-9
specie-flow mechanism, based on perhaps the oldest model in economics, that of David Hume. (1995, p. 17, emphasis in original)’’
The money supply movements based on this automatic adjustment mechanism follows the change in FX causing a change in the Monetary Base and thus altering the broad money supply (Gedeon, 2009). Similar to the gold standard, the monetary base will decline as a result of a current account deficit and BoP deficit since foreign exchange reserves, FX, decrease as a result of current account deficits. Subsequently, a money supply reduction occur pushing interest rates upward and the economy contracts. As a consequence, aggregate demand and the price level decrease. The increased interest rates results in capital inflows. The fall in output decreases import demand and exports become more competitive because of the price level reduction. All three effects trigger equilibrium restoration. In case of BoP surpluses the complete opposite happens (Balino & Enoch, 1997; Desquilbet & and Nenovsky, 2004; Dornbusch & Giavazzi, 1999; Hanke & Sekerke, 2003; Miller, 1999).
Most of this literature implies that trade imbalances are the cause for money supply expansion or contraction which is inaccurate. The Balance of Payments (BoP) which also includes capital flows and investments determines the gross FX and subsequently the monetary base. Moreover, a change in the BoP can inversely affect the money supply as a consequence of increasing amount of foreign direct investments (FDIs) or private capital (Gedeon, 2009).
The above explained automatic adjustment mechanism maintains the long-term sustainability of the currency board. It depends on a strong connection between money supply and foreign exchange reserves (Miller, 1999).
The Bulgarian CB has its advantages but the flexibility generated between the foreign exchange reserves and the money supply could hamper the mechanism of automatic adjustment which, alternatively, provides for the BoP equilibrium (Miller, 1999).
10
2.3 Disadvantages of the Bulgarian CB
Williamson (1995) also discusses currency board disadvantages. He lists several drawbacks out of which the most applicable to Bulgaria are: the transition problem, the adjustment problem and the political problem.
The transition problem is about whether a country is able to reduce inflation promptly after the introduction of the currency board. Stabilizing and pegging the exchange rate should cause a fall in inflation but the latter could have a momentum making the real exchange rate be overvalued. The resulting BoP imbalance will ultimately be corrected for by the gold standard mechanism but it might be sluggish and painful (Miller, 1999). In Bulgaria, inflation did have a momentum but for a very short period of time – from high inflationary levels in the beginning of 2007 it increased 16 percent throughout the last months of the year and only 1 percent in 1998. The real exchange rate did not appreciate against the US dollar back in the days of the currency board introduction but the current account went to a deficit from a surplus. Therefore, a transition problem did occur and it could be seen mostly through the trade balance account while inflation did not play a central role. The value of the current account surplus in 1997 was 426 USD millions while just in the following year it moved to a deficit of 272 USD millions and it has remained in a deficit up until today with slight exceptions in some months.
The last disadvantage according to Williamson (1995) is the political problem. His main concern relates to the degree at which the fiscal authorities will be controlled by the currency board operation. So far, the Bulgarian currency board has been able to maintain an environment where budget deficits are being controlled by the government (Miller, 1999).
In this paper a detailed focus is given on the disadvantage which is most relevant for the examined topic and this being the adjustment problem that arises as a result of BoP disequilibria.
Adjustment problem
Williamson describes the adjustment problem as the possibility of the economy to contract when current account deficits persist over time (Williamson, 1995). In the presence of BoP
11
deficits, a contraction in the monetary base and a reduction in the money supply and aggregate demand set into motion the adjustment course of actions. The current account imbalances could be reduced through a decline in output or prices. Frequently, when exports decrease the economy grows at a slower pace. When a country is operating under a currency board monetary system, not many actions could be taken to compensate for these contractionary pressures. Bulgaria is missing adjustment instruments of monetary policy and floating exchange rate and requires greater flexibility to be able to deal with it.
Another channel through which current account imbalances could be offset is through increases in FDIs. In this case, the foreign investments will cause a fall in trade deficit, enhance economic growth and ensure fiscal balances. Bulgaria is highly dependent on foreign investments and its policy aim has been attracting those FDIs since the introduction of the currency board in 1997. The financing of the current account is still delicate (exports are very low compared to imports as shown in Figure 1) and the BoP equilibrium relies heavily upon the capital account balance, meaning that FDIs and other capital inflows into Bulgaria are uncertain in the current ongoing global crisis (Figure 2).
The sensitivity of the Bulgarian CB to BoP movements and output has not yet been empirically tested. Since its introduction it has been operating well in the presence of a balanced BoP and persistent economic growth (Avramov, 1999).
Figure 1 Trade balance (Exports – Imports)
12
Figure 2 FDIs
Source: BNB
3. Testing the stability of the Currency Board – hypothetical case of
applied shock to the Financial Account
In this section, an eventual reduction of the Financial Account to 0 (financial inflows into Bulgaria are reduced to 0) is assumed and the BoP adjustment mechanism and its liquidity are examined.
In the evolution of the Balance of Payments ( BoP) accounting for the functioning of the currency board, two scenarios of liquidity problems are possible – one in the short run and another in the long run. In each case, the economy will react in a different way.
The key difference between short- term and long- term states is based on the
macroeconomic variables’ response to an applied shock of abrupt and major reduction in the Financial Account to 0 (through a decrease in the FDI ). In particular, in the short-run neither the trade balance nor GDP have the time to react and adjust to the applied shock on FDI so that the monthly values of the balance of trade and of the Gross Domestic Product by
13
the time of the reduction can be modelled separately before the sudden reduction spreads throughout the system.
On the contrary, there is a sharp contrast in the behaviour of the economy in the long-run where for each contraction in the foreign exchange reserves( FX) the money supply equation provides an exact pace of its similar reduction, which also has an impact on GDP growth. Given that the monetary shock has time to affect GDP (the response lag GDP-money supply is assumed to be 3 months) the contraction causes a fall in demand and consumption of imports which automatically corrects the trade balance and reduces the liquid pressure on the currency board. Therefore, if the currency board is liquid enough to meet a short-run shock, the subsequent monetary contraction immediately reduces the liquid pressure on BNB and lowers it for the subsequent periods. If after the initial shock the BoP is still in a deficit, the currency board mechanism causes another contraction of the money supply and an additional reduction in GDP and imports, consequent correction of the balance until equilibrium is restored. As Hanke and Schuler (1994) state in their book, the currency board system has the feature of balancing the current account as long as the trade imbalance is appearing as balance of payments disequilibrium as well. Subsequently, any long term scenario of slow net outflow of significant amounts of currency reserves (i.e. continuous BoP disequilibrium) is unsustainable.
Therefore, the long-run scenario of an abrupt decline in FDI (or uncompensated difference between CA and FA) can pose a threat to the CB in case that BoP falls immediately below zero in its adjustment process. Instead, it should gain some momentum until GDP and the CA reach equilibrium. In practice, this comes close enough to the short-run scenario which is examined further on and in great deepness. In particular, the very mechanism of the CBA allows for rejection of forecasts for a slow and continuous outflow of foreign exchange reserves from BNB for long periods of time instead of a sudden abrupt decline which threatens the board.
Following is an examination of an eventual short-term scenario of a sudden stop in FDI’s and the whole FA to zero for a period of 3 months keeping the current values of the TB and the CA. This case is quite extreme because it suggests a fall of around 100-300 mln euros in monthly investments (these are the last reported monthly values of the balance of inflows).
14
As noted earlier, it is assumed that GDP cannot react quickly and significantly to restore the BoP deficit that will occur. This paper makes another assumption for the monthly value of the current account to have stayed at -1.1674 billion euros – the lowest observed until now in December 2007. Furthermore, this value is assumed to be transferred directly into the BoP (the FA has dropped to 0, ignoring the Capital Account). Up until now (07.2013) -1.6456 billion euros has been the lowest BoP deficit observed in November 2007. Transferring the lowest CA value for 3 months making up for a total of 3x (-1.1674) = -3.5022 billion euro deficit in the BoP which is approximately 42 percent of the Q1 2013 GDP = 8.2885 billion euros. The logic and reasoning behind those assumptions is the intention of evaluating a case which is so bad and improbable that if the Currency Board withstands it, it will most probably endure every other short-term scenario in reality.
Table 2.2 Joint distribution of CA and FA for 2000-2013
2000-2013 2000-2013 Financial Account Current Account Number of
observations Equal share
> 0 (values of FA>0)
<= 0 (values of
FA<=0) Total > 0 <= 0 Total
> 0 (values
of CA >0) 10 24 34 0.062893 0.150943 0.213836
<= 0 (values
of CA<=0) 111 14 125 0.698113 0.08805 0.786164
Total 121 38 159 0.761006 0.238994 1
Table 2.2 shows the joint distribution of the CA and FA for the period 2000-2013 (Even if Bulgaria introduced the CB on July the 1st 1997, the sample starts with the 1st January 2000, to allow for variables to “stabilize” after this important shock). The total number of
observations is 159 and both their monthly values are grouped into two categories – bigger or smaller/equal to 0 in an attempt to estimate the probability of them jointly falling below 0. 14 out of 159 observations fall in this category which stands for a monthly probability of 0.08805. Assuming that those joint distributions are independent in the short run, quarterly
15
probability of both accounts being negative can be calculated as the product of three monthly probabilities, with the arrangement that it is an approximate value. The quarter probability thus becomes 0.000683. Even if the assumption of independent joint distribution has caused some error in the estimation of around 10 times, the probability is still going to stay low enough, implying that even with rough estimates, the event is merely impossible. Moreover, taking into account that this paper focuses on a 3-month BoP deficit of about 3.5 milliard euros, it is obvious that such a case is improbable.
Even though that trade balances and the FA vary quite significantly, the variation is independent. In particular, when the trade balance falls, the FA tends to rise because the imports of some investment products are marked both as an Import and an FDI (Figure 3 depicts the CA and FA simultaneous movements throughout the years). Since the country is actively attracting FDIs, the correlation between both accounts is significant. The empirical distribution gives grounds to assume that the covariance between FA-TB is essential and should be examined in more details. In particular, regression analysis is carried out to estimate by what amount the FA rises for each 1% increase in the trade balance deficit.
Figure 3 Trade balance (yellow) and Financial Account balance (dark red)
16
3.1 Empirical Analysis
Data
The data used in this paper is generated by the Bulgarian National Bank. The data includes the values of the current account, capital account and financial account. It is a time series data and covers the period from 01.2000 – 03.2013(the latest value published on the BNB website) amounting to a total of 159 observations.
Model, Model specification & variables
The following linear equation is going to be empirically tested:
Financial Account = Trade Balance +
The independent variable is the financial account in the first equation and the financial account plus capital account in the second one. The dependent variable is the trade balance. It is estimated by what amount the FA rises for each 1% increase in the trade balance deficit. Since it is a time series data and a linear regression, OLS method could be carried out only if the errors are homoscedastic and serially uncorrelated. Thus, first a test for
heteroscedasticity is performed. For that purpose, the Breusch-Pagan test is used in this paper. It tests whether the estimated variance of the residuals from a regression are dependent on the values of the independent variables. If an F-test confirms that the independent variables are jointly significant then the null hypothesis of homoscedasticity can be rejected. Subsequently, Breusch – Godfrey test for autocorrelation in the errors is executed. It makes use of the residuals from the model being considered in a regression analysis, and a test statistic is derived from these. The null hypothesis is that there is
no serial correlation of any order up to p. ‘’The test is more general than the Durbin–Watson statistic which is only valid for nonstochastic regressors and for testing the possibility of a first-order autoregressive model for the regression errors’’ (Wikipedia, 2013).
If the first test rejects the null hypothesis of homoscedasticity and the second shows presence of autocorrelation then the OLS could not be performed and there is a need for another test. The regression of Newey-West HAC estimator can be used in this situation to
17
allow for time series correlation of errors. A Newey–West HAC estimator is used when the standard assumptions of regression analysis do not apply. The estimator is used to try to overcome autocorrelation and heteroskedasticity in the error terms in the models often found in time series data as in the case of this paper. Prior to performing the last test, it is necessary to execute one last test, namely augmented Dickey – Fuller test for the presence of a unit root in an autoregressive model which would make the result meaningless. It removes all autocorrelation in the time series. ‘’The null hypothesis is that there is a unit
root. If the series is stationary, then there is a trend for returning to a constant mean. Therefore, large values will tend to be followed by smaller values (negative changes), and small values by larger values (positive changes). Consequently, the level of the series will be a significant predictor of next period's change, and will have a negative coefficient. If, on the other hand, the series is integrated, then positive changes and negative changes will occur with probabilities that do not depend on the current level of the series; in a random walk, where you are now does not affect which way you will go next.’’(Wikipedia, 2013).
Consequently, after performing the Newey- West HAC estimator and observing the results, an additional F-test is executed so as to test whether the financial account and the trade deficit exhibit significant relationship and move 1-1 or they differ. Expectations are that each 1 percent increase in the trade deficit brings about the same increase in the financial account and thus ensuring the BoP is in continuous equilibrium as predicted by the
automatic adjustment mechanism.
3.2 Results, evaluation and limitations
The results are listed in the Appendix.
Taking a look at the first regression estimation which takes into account the financial account and trade balance, the Breusch-Pagan test for heteroscedasticity rejects the null hypothesis for homoscedastic errors at 1 percent significance level with = 11.21. Following is the Breusch-Godfrey test for autocorrelation which also rejects the null at 5 percent and 10 percent significance level for no serial autocorrelation up to 4th 5th and 6th lags. The last test for a unit root, the augmented Dickey-Fuller test rejects the hypothesis for the presence of a unit root at Z= - 5.747 < 1% critical value of -3.491. Thus, next follows the Newey-West HAC estimator test, using 4 as maximum lags from the following equation: # =
18
0.75 * T^ (1/3) = 0.75*159^1/3 = 4.0631 (Simons, 2013). The coefficient of the trade balance is -1.060361 which shows that for each 1 euro increase in the trade deficit, the financial account increases with 1.060361 euros. In practice, the constant in the regression is not statistically significant and the regression coefficient is very close to 1. To confirm this, an F-test is performed to see whether the coefficient is significantly different than -1. The result is a p-value of 0.5085 which cannot reject the null hypothesis for equality.
The result is interesting for several reasons. First, it provides quantification of the empirical covariance between the trade balance and financial account which precisely coincides with the intuition of analysts claiming that the two accounts are interdependent. Secondly, the value of the resulting coefficient shows that the trade balance tends to manifest the average balance of the financial account with surprising accuracy. If the coefficient is significantly different from 1, it would mean a long-term trend of balance of payments disequilibrium. In the current case of the coefficients observed being close to 1, however, it confirms the numerical aspect of the intuition in mind that the investments are marked simultaneously as an import as well. Finally, although the values obtained are purely empirical, they give an idea why in the previous section such low probabilities of a crisis were observed (when calculating whether both accounts were falling below 0 at the same time causing BoP disequilibrium). This is due to the large covariance of the two accounts, which vary together in the exact opposite directions.
The above considered observations provide a reason to assume that from the resulting coefficient of -1 it is generally possible that the observed deficit in the trade balance largely reflects the impact of the foreign investments. If such hypothesis is correct, it would mean that the import of goods, driven by domestic demand (and by foreigners investing in Bulgaria), it is possible to be balanced or close to balanced and what is noticeable in the current and financial accounts is nothing more than an accounting effect without consequences for the economy.
The trade balance and financial account exhibit a very high degree of correlation and greatly reduce the likelihood of long lasting changes in the balance of payments.
Now, the effects of a short-term three-month crisis on the liquidity of the currency board are examined in more details. As noted already, the likelihood of such a crisis is negligible but it is illustrative to consider the possible effects on the CB liquidity.
19
As of 31.07.2013, the latest date for which data is available, the available cash and foreign currency denominated deposits at the BNB amounts to 3.4351 billion. Even if the current account deficit is to remain at -1.1674 billion for 3 months (which, as we have seen, is very unlikely), a simple calculation shows that the cash position of the Bulgarian National Bank is sufficient to finance the deficit for 2.94 months, without turning to any restructuring of the portfolio of securities or the use of monetary gold. In practice, it can be expected securities with 3-months maturity to mature which will induce the accumulation of additional
currency. In the BNB assets still remain unused monetary gold reserves of about 1.26 billion, part of which, if necessary, can be converted into reserve currency so that through its use the deficit financing can be extended up to 4 months without restructuring of the portfolio of securities. Now, for 4 months the BNB has enough time to restructure the rest of its investments into highly liquid ones even if it has to realize some losses at the expense of the excess B deposits occurring during the exchange.
However, since there is no public information on the maturity structure of the reserve, it is difficult to make a quantitative estimate of the volume of transactions that the BNB should execute in this situation. Nevertheless, the available information allows for some qualitative conclusions. This is mainly due to the fact that the securities of foreign exchange reserves are of high quality and can, therefore, be exchanged for more liquid assets in reasonable time. BNB Law imposes certain restrictions on the quality of the securities in foreign currency. In particular, it requires the securities to be rated at the highest rating from two internationally recognized credit rating agencies (e.g. the rank of "Moody's"), making their eventual replacement when needed the most non-problematic, although it requires some physical time. In the current outlined framework of 3-4 month, however, it can be expected that this time will be enough for the BNB to restructure the bulk of its reserves in a form that provides the necessary liquidity, even if it suffers some short-term losses from unrealized interest. As mentioned above, these losses can be covered by the existing B foreign currency deposit of the Banking department in the Issue Department which in 2013 is equivalent to an average of 2.61 billion euros. Subsequently, for a period longer than 4 months it can be stated with high probability that the liquidity restrictions are almost all eliminated. Therefore, although the scenario examined here is so bad that it is actually merely impossible, it is nevertheless not enough to break the stability of the CB. From these
20
simple calculations, it is clear that the survival of the currency board in Bulgaria in the short term is guaranteed in virtually every scenario that can arise in reality. Quite separate issue is the risk of potential macroeconomic destabilization, which would cause contraction in foreign exchange reserves and the monetary base.
As already outlined, under a currency board regime, the monetary base(C + R) in the country is follows almost precisely the movement of foreign exchange reserves. This means that any balance of payments crisis in which the board retains its existence, the Bulgarian lev mass will undergo an unprecedented contraction of very heavy monetary shock. While in the short term GDP is assumed to be fixed, in the long run monetary policy gives an effect that will undoubtedly lead to a sharp decline in GDP and contraction of the entire economic activity.
Several factors hamper the modeling of the situation.
Above all, the current fluctuation models (e.g. RBC with all its variants) use a log - linear methods, which are valid only in a certain vicinity of the steady state model. In a harsh monetary shock of the same magnitude as the one ‘played’ here, leaving the steady state is guaranteed, but there is no assurance that the system will return to the same equilibrium from which it escalated. It is therefore possible to monitor some effects occurring
completely unexpected from the model, which make the modeling of GDP impossible. However, based on the available information a non-quantitative assessment could be made which is limited to:
1. Such a scenario, with or without a decline in GDP, is anyway highly unlikely because of the relationship between trade balance and financial account, as discussed in the first part. 2. Even a significant drop in GDP in itself poses no risk to the operation of the currency board, as long as the BNB has enough time to restructure its foreign reserve. In the short-term scenario discussed above it was found that the liquidity of the BNB is sufficient to ensure that time. Therefore, it seems highly unlikely that the Bulgarian National Bank has the required short-term liquidity but will not be able provide long-term liquidity.
3. Under these circumstances, long-term risks imposed on the economy are emerging through income and output rather than through the stability of the board, providing that possible political pressure to abandon the CB are not taken into account.
21
To conclude, the stability of the currency board is not seriously threatened and the main risk to the economy comes rather from the effects that such an unlikely crisis would have on the real economy.
4. Conclusion
In the literature review, there was mostly a description and history of the Bulgarian currency board operation. This was used as a framework for the empirical analysis in the third part. The question in this paper is how stable is the Bulgarian currency board and is it liquid enough to endure a short-term 3-month crisis. From the theoretical framework the main advantage and disadvantage outlined appeared to be the balance of payments automatic adjustment mechanism. The variables used are the main components of the balance of payments equation, namely the current account, financial account and capital account. The outcomes in the third part were expected before performing the regressions and turned out significant in the regressions. The trade balance account and the financial account tend to move with the same precision in the exact opposite direction which confirms the Balance of Payments continuous equilibrium and its adjustment mechanism ensuring long-term stability.
In the third part, the hypothetically examined scenario of a crisis show that the CB could meet short-term and long-term liquidity problems and would still continue trying to ensure economic growth and attract investments. However, it should be noted that this is a situation where other factors are kept constant and several assumptions are being made which can not be guaranteed in the present situation. The board provides credibility but lacks flexibility which is necessary in dynamic periods like this.
A restriction in the model used and analysis made is that many historical facts, events and processes associated with the development of the currency boards around the world are left out.
22
Appendix
fa_k 159 237.112 355.2591 -683.7 1549.9 tb 159 -310.2201 202.5223 -952.5 39.6 fa 159 213.4987 386.5754 -725.7 2040.8 Variable Obs Mean Std. Dev. Min Maxfa = Financial Account
tb = Trade Balance = Goods account part of Current Account = X –M
fa_ k = Financial Account + Capital Account
Appendix 1 Financial account – Trade account regression tests
Prob > chi2 = 0.0008 chi2(1) = 11.21 Variables: tb
Ho: Constant variance
Breusch-Pagan / Cook-Weisberg test for heteroskedasticity
H0: no serial correlation 6 11.191 6 0.0827 5 9.713 5 0.0838 4 9.619 4 0.0474 lags(p) chi2 df Prob > chi2 Breusch-Godfrey LM test for autocorrelation
MacKinnon approximate p-value for Z(t) = 0.0000
Z(t) -5.747 -3.491 -2.886 -2.576 Statistic Value Value Value Test 1% Critical 5% Critical 10% Critical Interpolated Dickey-Fuller Dickey-Fuller test for unit root Number of obs = 158
24
Bibliography
Wikipedia. (2013, April 20). Retrieved July 31, 2013, from Wikipedia:
en.wikipedia.org/wiki/Breusch-Godfrey_test
Wikipedia. (2013, July 29). Retrieved July 31, 2013, from Wikipedia:
en.wikipedia.org/wiki/Dickey-Fuller_test
Avramov, R. (1999). The Role of a Currency Board in Financial Crises: The Case of Bulgaria. Sofia: BNB Discussion Papers, No.6.
Balino, J. T., & Enoch, C. (1997). Currency Board Arrangements: Issues and. Washington DC: IMF Occasional Paper no. 151.
Barlemann, M., & Nenovsky, M. (2004). Lending of first versus lending of last resort: the Bulgarian financial crisis of 1996/1997. Comparative Economic Studies 46, 245-271.
Desquilbet, J.-B., & and Nenovsky, N. (2004). “Credibility and Adjustment: Gold Standards Versus
Currency Boards''. Sofia: BNB Discussion Papers, No.39.
Dobrev, D. (1999). The Currency Board in Bulgaria: Design, Peculiarities, and Management of Foreign Exchange Cover. BNB Discussion Papers.
Dornbusch, R., & Giavazzi, F. (1999). Hard Currency and Sound Credit: A Financial Agenda for Central Europe. European Investment Bank Papers, 25-32.
Gedeon, S. J. (2009). Money supply endogeneity under a currency board regime: the case of Bosnia and Herzegovina. Journal of Post Keynesian Economics, 32, 98-114.
Hanke, S., & Schuler, K. (1994). Currency Boards for Developing Countries. San Francisco: ICS Press. Hanke, S., & Sekerke, M. (2003). How Bulgaria Is Destroying Its ‘Currency Board'. Central Banking
Journal, 14(1), 81-84.
IMF (2013), ‘’ World Economic Outlook’’, April.
Miller, J. B. (1999). The Currency Board in Bulgaria: The First Two Years. Sofia: BNB Discussion Papers. Minea, A., & Rault, C. (2011). External monetary shocks and monetary integration: Evidence from
the Bulgarian currency board. Economic Modelling 28, 2271-2281.
Mundell, R. A. (1997). Currency Areas, Common Currencies, and EMU. The American Economic
Review, 214-216.
25
Simons, K. (2013, January 25). Retrieved July 31, 2013, from Rensselaer: homepages.rpi.edu/simonk/pdf/UsefulStataCommands.pdf
Williamson, J. (1995). What Role for Currency Boards? Washington DC: Institute for International Economics.
