Printer Friendly

From rhetorics to practice in monetary policy: a Romanian perspective.


In its monetary policy standing, the National Bank of Romania (NBR) shares the commitment to long-term price stability common to all central banks throughout the world. However, the Romanian transition to a market economy system has been marked by high and volatile inflation rates. Monetary policy choices have been rooted in an understanding of inflation as an excess aggregate demand phenomenon. The policy framework has changed throughout transition, from broad money targeting (1990-1996) to high-powered money targeting (1997-2005), and then to inflation targeting (August 2005). The NBR explained the latter policy move as a response to the increasing inadequacy of monetary targeting approaches (NBR, 2005).

The paper will use the 1997-2005 period, when the NBR formally committed itself to targeting the monetary base of the system, as the basis for exploring Romanian policy-making in practice and the challenges to the newly adopted inflation-targeting framework. It will situate the transition to the new policy regime in the context of the April 2005 further liberalisation in the capital account and the attending policy dilemmas. It will argue that the new policy regime carries on the tension between theorising and practice of monetary policy, with monetary authorities on the one hand failing to overcome the policy trade-offs of the monetary targeting years, and on the other hand unable to alter the money markets' structural features in order to consolidate the relevance of the policy rate for spending decisions as required by the new regime.

It will show how the inflation targeting framework faces the central bank of a small open economy with a paradoxical situation: interest rate decisions that in theory ought to be geared toward internal considerations, but in practice are greatly limited by, and depending upon external factors. Hoping to stimulate debate, it will finally outline some alternative policy options.

Although other macroeconomic developments will be presented to explain the circumstances surrounding policy-making, the main aim is to critically assess the assumptions behind NBR's monetary policy strategy, without systematic questioning of its effects on inflation rates or the appropriateness of its chosen instruments for creating a well-functioning financial system.


Modern central banking theory is based on a widespread consensus that monetary policy is the most efficient mechanism of ensuring a low and stable level of inflation (Woodford, 2003). Thus, the central bank chooses an operational target, an economic variable with a predictable impact on the inflation rate, which it can best control through its monetary policy instruments (1) (Bindseil, 2004). It is a quantity versus price choice, that is, which of the possible operational targets, be it a monetary aggregate (the monetary base (2) of the system) or a market interest rate, should be used in daily policy practice.

In developed financial systems, there is nowadays little controversy that the short-term interbank interest rate is the only operational target that can be controlled to any meaningful extent. Monetary targeting regimes have been abandoned, primarily due to the inability of central banks to reach targets set for money stock growth.

However, by no means is the ineffectiveness of monetary targeting regimes universally accepted. Monetary authorities in the developing world, particularly in transition economies, still rely on its effectiveness for controlling the monetary conditions in the economy and further, the movements in inflation rates. Thus, while the short-term interest rates versus monetary targeting regimes debate seems to have long been settled with reference to the developed financial systems, the large majority of formerly planned economies outside the European Union (EU), that is Albania, the 12 members of the Community of Independent States (3) and until recently Romania have been running programmes supported by the International Monetary Fund (IMF) in which monetary targets are standard, the performance criteria being specified either on base money or its sources targets with a view of achieving a certain rate of expansion of the money stock and thereby a certain target for the price level (Keller and Richardson, 2003).

Furthermore, the increasing adoption of inflation targeting regimes across developing countries, popular in policy circles as it sanctions a unique focus on inflation and purportedly isolates policy decisions from political considerations, does not render monetary targets obsolete. Within an inflation targeting framework, 'sound' monetary policy is grounded in a 'sound trinity': flexible exchange rates, an inflation target and a monetary policy rule (Taylor, 2000). The policy rule should be specified according to the institutional characteristics of the country in question: thus, the monetary authority could be using either a market rate or the reserve base as operational target (Clarida et al., 1999).

Targeting base money

Associated with an interpretation of inflation as a consequence of excess aggregate demand, monetary targeting theories developed within the monetarist tradition rely on the assumption that inflation is a demand phenomenon, with the demand pressures yielding from excessive money stock expansion (Friedman, 1956). A strategy of base money targeting is grounded in a predictable relationship between the money stock and inflation through the following monetary transmission mechanism:

* Banking system's liabilities determine the inflation rate /assuming a stable velocity of circulation/.

* The volume of liabilities of the banking system is determined by the reserve base of the system through the money multiplier in a predictable and stable manner.

* The central bank exogenously (ie independent of economic variables) determines the desired level of monetary base through its monetary policy instruments.

Thus, using the monetary base as operational target rests on the assumption that central bank-induced quantitative adjustments in the reserves of the banking sector will prompt, via the money multiplier mechanism, immediate adjustments in the liabilities side of the banking system to the level of required reserves.

How do these quantitative adjustments function? Changes in the monetary base are driven by either net foreign assets (NFA) or net domestic assets (4) (NDA) of the monetary authorities. Thus, any balance of payments or exchange rate policy induced changes in the NFA component of the monetary base must be sterilised by an offsetting movement in NDA in order to isolate domestic monetary control from exogenous factors. Furthermore, a policy intervention seeking to increase the money stock is assumed to run as follows:

* Changes in reserve requirements or open market operations (5) (OMOs)

* Create excess reserves, that is banks' current account holdings with the central bank over the required reserves

* The excess reserves automatically set off the multiplier mechanism, in this case an increased lending activity, and further

* The asset adjustment triggers deposits liabilities adjustment, bringing the reserves to the required level and the liabilities side of the banking sector, the money stock, to the level targeted by the policy intervention.

Practitioners from central banks and heterodox economists have rejected this exogenous explanation of money stock formation, arguing that in a fully developed financial system, the money stock can only be endogenous (Goodhart, 1989; Moore, 1988). The productive sector's demand for credit and the banking sector's liquidity preference, that is, the risk assessment of the credit worthiness of the borrowers and the level of perceived risk the banks are willing to assume, drive credit creation (Dow, 2004). The causality of the orthodox monetary theory is reversed: loans create deposits and thus expand the money stock. Furthermore, reserve positions endogenously adjust to the new required level.

The endogenous determination of the money stock goes hand in hand with an interpretation of inflation as mainly a cost-push phenomenon and a mark-up theory of price formation: prices are set to cover costs of production and profits (Davidson, 1977). Loans are provided by the banking system to businesses and households as a consequence of the latter's increased expenditure plans, emanating from increased investment or working capital needs on the account of higher costs of production due to higher wages and/ or higher prices for domestic/imported intermediary inputs. Nevertheless, the money stock adjustment to increases in money prices and nominal income to finance higher expenditure requirements depends on the banking system's willingness to accommodate the increase in demand for loans. This in turn is a function of the changes in the perception of risks, with banks preferring investments in financial securities to loans at different stages of the credit cycle (Dow, 2004). Thus, in a developed financial system, reserves cannot constrain the credit creation, but rather endogenously adjust to it. Given banks' liquidity preference, the demand for credit determines the volume of loans and deposits. Commercial banks subsequently adjust their reserve positions by either turning to the discount window of the central bank, forced by its role of lender of last resort to satisfy it, or by borrowing/lending reserves on wholesale markets.

Wholesale markets and the monetary transmission mechanism

Banks trade liquidity with each other on the wholesale interbank market. It is on this market where the challenges to a quantity-constrained policy strategy become apparent. In a reserve-driven system, the demand for liquidity on the wholesale market comes from banks that need to supplement their reserve positions, while supply comes from banks with excess liquidity. Technically, the central bank does not operate on this market, unless it chooses to do so by using a third party. Rather, the central bank conducts open market operations with counterparts from the banking sector (not all banks are counterparts), usually by means of repurchase agreements (repos) or reverse repos.

Because of volatile liquidity positions, a salient characteristic of the interbank market is its inherent instability (Bagehot, 1910). Firstly, both supply and demand side of the interbank market are subject to transitory shocks linked to the so-called autonomous liquidity factors (cash in circulation, the Treasury's deposits with the central bank, excess/deficient reserves). Secondly, the demand for reserves is highly interest inelastic (Bindseil, 2004). These largely unpredictable changes in liquidity positions have several consequences for the choice of a monetary policy strategy.

Firstly, it makes it virtually impossible to target a certain level of reserves. OMOs conducted within an operational framework of quantity-constrained reserves create excess/deficient reserves, for which the banks will quickly search the most profitable placements, invariably to be found on the wholesale market. Banks that find themselves with excess reserves that receive no or at best below-market interest rates will not engage into time-consuming hunts for credit worthy borrowers, but turn to the money market for alternative placements with higher yields (Goodhart, 1994). Similarly, banks that need to supplement their reserves to the required level will rather pay punitive rates on the money market than cancel loans.

Secondly, the interbank rate is bound to be volatile when, as in a monetary targeting regime, the central bank does not have its stability as policy objective. The magnitude of the volatility depends, however, on the period of calculation for compliance with reserve requirements, the maintenance period. The highest volatility is registered with a daily maintenance period, as excess reserves push overnight interbank rates downward to zero or to the level of the deposit facility, (6) while deficient reserves would have to be accommodated at a penal rate either by the market or the central bank. Averaging reserve requirements over a period larger than 1 day creates a more stable pattern of overnight rates only up to the end of the calculation period, when excess/deficient reserves would produce similar effects as in the daily calculation regime (Bindseil, 2004).

Furthermore, the inherent volatility of wholesale market requires the central bank to take an active role in smoothening rates where the inflation targeting regime uses an interest rate policy rule. This question will be further addressed in the analysis of inflation targeting policy practice in Romania.


The communist regime that had ruled Romania for 45 years fell in 1989. The transition to a market-driven system came with the typical adjustment programme (the so-called shock therapy) based on the Washington Consensus recommendations: stabilisation--liberalisation--privatisation.

Throughout transition, the main objective of macroeconomic policy was the curtailment of inflation, deemed to be essential for 'putting the country back on a path of sustainable economic growth' (GoR, 1994, p. 7). This was to be achieved by means of monetary anchors, 'which preserve the independence of the monetary policy and are suitable for an economy facing both money demand shocks and external competitiveness difficulties' (NBR, 1998, p. 257). This position is clearly consistent with a monetarist view of inflation as a monetary phenomenon caused by excess liquidity that necessarily and sufficiently needs restricting excessive growth in the relevant monetary aggregate.

The assessment of the shifts in macroeconomic policy stance in the transition period in Romania allows the identification of four distinct subperiods (Gabor, 2003). While the main focus has time and again been inflation stabilisation, policy practice differed markedly throughout the period. Despite a formal commitment to monetary targeting, there have been substantial changes in policy emphases, as at times the exchange rate was perceived to be the best anti-inflationary nominal anchor.

1990-1993: In line with the view that the high and variable inflation rates in Romania are due to excessive liquidity in the system, the NBR adopted a policy of targeting the broad money stock. This implied a tight monetary policy by means of setting contractionary targets for both broad money stock growth and domestic credit. A floating exchange rate regime was adopted (IMF, 1991).

However, the anti-inflationary policies were by no means successful. Both consumer and producer prices hovered around 150%-200% in this subperiod. GDP fell by an average annual of 6.45% (worse still, the industrial sector contracted by a staggering 10.5% on average) while current account deficits expanded. With inadequate access to international finance and foreign reserves depleted by current account financing requirements, policymakers were forced to resort to large devaluations as part of the highly contractionary stabilisation programmes agreed with the IMF.

1994-1996: Although M2 was kept as the official intermediate target, the policy stance changed to targeting the nominal exchange rate, recognised as an important anchor of stabilisation. The macroeconomic situation improved patently, with an average annual GDP growth rate of 5% and inflation rates slowing down to 50%. Nevertheless, by the end of this sub-period, current account imbalances started pressuring again the NBR's foreign reserves position.

1997-2005: Growing current-account imbalances financed by depleting NBR's foreign reserves and the election of a centre-right government laid the basis for signing a new IMF-led shock therapy in early 1997. Similarly to earlier stabilisation programmes, it led to a yet stronger depression: three consecutive years of negative GDP growth, spiralling prices and exchange rate devaluations, a serious collapse in industrial output, instability in the financial sector. Only in 2000 did the macroeconomic climate begin to improve.

Most importantly, the new stabilisation programme included a shift in the monetary policy stance to targeting the cash base of the system, a strategy move complemented however with a strongly managed float. 'Officially, the NBR targets money base growth, but continues to be seen by the market as having an implicit exchange rate target' (OECD, 2002, p. 76). Targeting in this case implies smoothing short-term fluctuations with a view of the exchange rate as a nominal anchor in the anti-inflationary policy.

2005 (August) onwards: NBR officially adopted inflation targeting, with a short-term interest rate policy rule. In line with the new policy framework, monetary authorities abandoned direct interventions on the foreign exchange (forex) market starting with the last quarter of 2005. However, the official discourse attaches the word 'flexible' to inflation targeting, by which the monetary authorities allow exchange rate considerations into the policy process (NBR, 2005). Capital account liberalisation was further extended, allowing non-residents access to domestic currency denominated deposit accounts opened with resident banks (April 2005).

Why is the exchange rate important? A heterodox explanation of the Romanian inflation

Throughout transition, the exchange rate has been a tacitly acknowledged yet highly relevant variable for controlling inflation. Its rather obscure status in policy discourse does not correspond to the essential place it has had in policy-making. Thus, the NBR has been a long-term confessed adherent of the orthodox transmission mechanism that requires control of liquidity to control aggregate demand and further inflation, a policy choice strongly supported by international financial institutions, whose seal of approval is essential for Romania's access to international financial markets for financing the current account deficit.

That in practice NBR's anti-inflation strategy heavily relied on the interventions on the forex market to control the movements in exchange rates points to a different explanation of the inflationary process. A substantial body of literature, including several IMF country reports, have highlighted the importance of the exchange rate pass-through into prices, with a larger and quicker effect in producer rather than consumer prices (IMF, 2003, 2004). This implies that depreciations in the exchange rate of the domestic currency quickly reflect into prices. The monetarist explanation would be that excess money growth (and hence domestic inflation) inevitably pushes itself into trade deficits. These would either be eliminated through currency depreciations in the context of a flexible exchange rate regime or sustained to the extent that foreign reserves are available under a fixed exchange rate. However, I will argue that the cause-effect monetarist relationship is reversed in the Romanian case as the current account deficits, of a structural rather than monetary nature, trigger exchange rates adjustments. Furthermore, the nature of the production structure pushes exchange rate movements into prices, indicating to a cost-push explanation of inflation and endogenous money stock movements.


Firstly, Figure 1, a plot of the changes in the nominal leu/US Dollar exchange rate against the producer price index, confirms the strong correlation between the movements in the two.

The existence and large magnitude of the pass-through from exchange rates to prices is explained by Romanian production patterns, inherited from central planning, which create persistent, structural current account disequilibria and push exchange rate depreciations into increased costs of production and thus overall inflationary pressures. Of note in this context is that no depreciation, policy or market induced, has ever restored the current account surpluses registered during the planned period (see Figure 2).

The external imbalances the post-communist economy has been experiencing derive from a combination of factors. The first to explore is export performance. After initially collapsing to less than half of the 1989 value due the disappearance of the socialist markets, exports earnings recovered extremely slowly, reaching the 1989 level only in 2000. Secondly, the intensive industrialisation process characteristic to planned systems left the industrial sector highly dependent on imports of intermediary goods, which constitute the largest share of the total imports. (7) Central planning allowed imports for unavailable raw materials and fuels, investment goods not manufactured at home and intermediary products unavailable domestically in demanded quantities (Winiecki, 1988). Thus, the specific import structure made all imports necessary, any reduction in volume strongly affecting industrial output. A devaluation-induced increase in the price of imports would have little success in shifting demand to lower priced, competitive products manufactured domestically, as the philosophy of central planning allowed only for supplementary and not competitive imports. This explains the pass-through effect and its larger impact on producer prices: with a production structure rigidly dependent on imports, any exchange rate devaluation reflects in the domestic price of intermediary inputs, quickly increasing costs of production and prices.


Both the inflation process and the persistent current account deficits have been produced and sustained by structural characteristics of the Romanian economy: inflation has been so closely linked to exchange rate movements because of the external orientation of the domestic production structure. This has substantial bearing on policy choices, facing monetary authorities with a policy trade-off: reducing inflation requires a strategy of allowing real appreciations against the currency of reference. (8) Nonetheless, this strategy affects external competitiveness and might further aggravate the structural nature of the trade deficit.


The close relationship between exchange rates and inflation rendered developments in the external position essential for macroeconomic policy. Reducing the inflation to the low one-digit level required for entering the Eurozone (9) has created a vicious circle for policy-makers: lower inflation can only be achieved by maintaining an overvalued exchange rate, which in turn worsens trade imbalances, not a hazard as long as either foreign reserves are plenty or foreign financing is available. Fundamentally, the NBR would target a level of overvaluation high enough to warrant progress in the antiinflationary policy and low enough to avoid larger than necessary trade imbalances, which could affect foreign investors' confidence and thus force devaluations and price increases. In times of foreign capital scarcity, it would use its foreign reserves to prop up the domestic currency, while in times of foreign capital abundance it would prevent large appreciations by buying foreign currency on the forex market.

Figure 3 presents the evolutions of the Romanian Leu/US $ nominal exchange rate against NBR's foreign reserves. It would appear that the year 2000 marks a change in the patterns of exchange rate volatility. Contractions in forex reserves positions go hand in hand with high fluctuations in the exchange rate, as exchange rate smoothening depends on the availability of foreign financing. And foreign financing was at best erratic during that period. Thus, in 1995 and 1996, with little inward flows, foreign reserves were run down to finance the current account deficit. Moreover, the April 1997 IMF-sponsored shock therapy dealt a serious contractionary blow to the economy. To make matters yet worse, in 1997 the NBR was facing not only 3 years of peaks in the foreign debt service (amounting to around 30% of export earnings) but also a difficult international climate: the Asian and Russian financial crises reduced the international markets appetite for investing in emerging economies. The situation was aggravated in 1999 by IMF's refusal to grant funds through a new stand-by agreement unless access to private international financial markets was secured independently. Romania narrowly avoided default on external debt servicing by contracting foreign loans at a high-risk premium.


Riding on a tide of large foreign inflows, foreign reserves picked an upward trend after mid-2000, briefly reversed in the first semester of 2003. In this time interval, the exchange rate movements are significantly smoother than in the 1997-2000 period.

A strategy of base money targeting with exchange rate constraints necessitates sterilisations of foreign market interventions. Forex interventions, aimed at anchoring the currency within a given band, change the international reserves component of the cash base. For instance, upward pressures on the exchange rate arising from balance of payment deficits will trigger a sale of foreign currency, reducing NFA and overall liquidity in the system. The counterbalancing sterilisation operation would require an increasing in liquidity by increasing NDA through open market operations. In the opposite case, increases in NFA arising from an oversupply on the forex market would be sterilised by reducing NDA.

Table 1 presents the evolution in the sources of the monetary base in Romania between 1996 and 2004. Up to 2000, the NFA contribution changes from negative in 1996, when it registered a net liability position amounting to 30% of the monetary base, to positive but fluctuating levels. It picks up at a significant pace after 2000, reaching a more than double coverage of the high-powered money from 2003 onwards. This suggests that the variations in liquidity positions were mainly stemming from NBR's massive interventions on the forex market to avoid the appreciations attending large capital inflows. Since the treasury was slowly forced to turn to domestic or external money markets for deficit financing, the credit to government's negative contribution since 2002 reflects the Treasury's net depositor position with the NBR. The corresponding sterilisation operations have been mainly operated by modifying the credit to banks source (NCB) that registered an increasing negative contribution (net liability position) throughout the period.

To answer the question of whether monetary authorities were indeed able to consistently control the monetary conditions in the economy, I will examine the monetary policy instruments used to change reserve positions. The essential requirement of an effective monetary targeting strategy is a consistently full sterilisation of forex interventions.

NBR's portfolio of open market paper (government securities and Treasury bills) has historically amounted to a very small proportion of its domestic assets, which by and large consisted of NBR loans to the banking sector. In order to reconcile the two policy objectives, base money targeting and exchange rate management, the NBR had to make use of two policy instruments: changes in reserve requirements and taking deposits from the commercial banks, at times complemented by much smaller in volume repurchase agreements. These instruments were used at different times, depending on the international reserves position (see Figure 4).

Up to 2000, with foreign reserves dwindling and little access to foreign capital to finance structural deficits in the current account, the NBR made intensive use of all mechanisms available to quickly reduce the current account gap by practically stifling economic activity. (10) Not only did it drastically reduce access to its refinancing facility (see last column in Table 1) but also operated large changes in reserve requirements, (11) increasing the required ratio from 7.5% to 30%. Choking the system of liquidity would reduce aggregate demand, imports and thus minimise the current account gap. Foreign inflows once restored, OMOs were by and large resumed to sterilisations of forex interventions by deposit-taking operations, complemented with a reduction in reserve requirements at more acceptable levels.


As restrictions of refinancing from the central bank were insufficient to ensure quick liquidity drainage, reserve requirements were deemed to be a better instrument for 'confirming the toughness of the monetary policy' (NBR, 1997, p. 12). By triggering immediate credit and monetary adjustments via the money multiplier, changes in reserve requirements are perceived by monetary targeting promoters as a powerful, albeit rudimentary, instrument for controlling high-powered money (Ritter et al., 1997). The mechanism is straightforward: lower/higher required reserves create/reduce additional reserves for banks, enabling the banking system to expand/reduce their loans, which in turn will determine deposits and the money stock while leaving the monetary base unchanged.

The second instrument, deposit-taking operations, is employed in situations of downward pressure on the exchange rate that induce the NBR to take a buying position on the forex market. The attending liquidity increases the excess reserves commercial bank hold with the monetary authority. To bring these back to required level, the NBR engages in liquidity mopping operations by taking fixed-term deposits with maturity ranging from 2 weeks to 3 months (NBR, 2001). Thus deposit-taking operations have replaced the discount window as an instrument for managing liquidity, particularly after 2001. Interestingly, NBR used reductions in its outstanding loans to the commercial banks as a contractionary instrument in the 1997-2000 period, and for sterilising forex interventions up to the end of 2002. Ever since, discount window lending disappeared from NBR's balance sheet.

The reserve accounting mechanism was amended three times during the regime of high-powered money targeting in an effort to improve monetary control. It changed from averaging over a month (January 1997-March 1999) to daily reserve requirements (April 1999-July 2002) to again averaging over a month (since August 2002). For each of the three methods of reserve maintenance, before and after the introduction of the overnight interbank markets (April 1999), actual reserves consistently diverge from the required ratios, pointing to the difficulties the central bank faced in its sterilisation operations. As expected, interbank rates fluctuated in line with banks' liquidity needs and the period of accounting.

January 1997-March 1999: Averaging over a month

Figure 5 is a plot of the required and actual reserve ratio from January 1997 to March 1999, when reserves were averaged over a month. Although NBR made extensive upward changes in reserve requirements to affect the money supply, the control over the monetary base is by no means successful under this particular regulatory arrangement: the actual reserve ratio consistently diverges from its mandatory values. With no overnight market to turn to and restrictions in refinancing from the discount window, reserves follow a volatile pattern, depending on the liquidity positions of the banking sector. It would seem that for this period the money stock and the reserves position were indeed endogenously set by the commercial banks, with little control from the NBR.


April 1999-July 2002: Daily reserve requirements

The unpredictable impact policy decisions had on the intermediate target (broad money) drove the NBR to review its reserve requirements mechanism. The search for more efficient instruments was driven by a rather delicate macroeconomic situation: a peak in foreign debt service looming in the middle of 1999 combined with very limited access to foreign capital were prompting many speculations of default. Still trustful in the virtues of base money targeting, NBR only believed it necessary to adjust the accounting methods for reserve requirements. Accordingly, the maintenance period was reduced in the first quarter of 1999 to 15 days. Moreover, the NBR imposed a daily floor and ceiling for reserves practically amounting to a daily reserve requirement, while simultaneously setting up the overnight interbank market.

The change did little to improve control. Indeed, the new overnight market provided commercial banks with an outlet for the excess or deficient reserves resulting from own financing needs and NBR's interventions. Under the quantity targeting strategy, this translated into significant volatility of overnight rates (see Figure 6).

August 2002 onwards: Averaging over a month again

In August 2002, the de facto daily reserves requirement was replaced with 1 month averaging, from the 23rd of the month to the 23rd of the next month. Comparing Figures 6 and 7, it appears that the length of the accounting period does have an impact on overnight rates, as averaging over a period does reduce volatility. Still, on the last working day before the end of the maintenance period, the interest rate either falls dramatically to the level of the standing deposit facility, indicating excess reserves, or it increases sharply as commercial banks attempt to restore deficient reserves (as in January and March 2003). Wholesale markets remain volatile in the absence of an interest rate smoothening policy (Bagehot, 1910).

The money multiplier's incompliance with the theoretical underpinnings of the monetary targeting strategy is consistent with a theory of endogenous money (Dow, 2004). As the textbook assumptions behind a high-powered money targeting strategy do not leave any room for excess/deficient reserves, OMOs must be translated into fully predictable movements in both multiplier and money supply. Nevertheless, it appears that the transmission mechanism does function through excess/deficient reserves, creating volatility in overnight rates. Commercial banks choose their liquidity position given NBR's sterilisation operations and the demand from the productive sector, confirming the endogenous money argument that 'bank earning asset decisions are largely independent of their reserve position' (Moore, 1984, p. 106).



What explains this pattern of reserves?

The structural nature of the current account deficit and the strategy of exchange rate targeting induce a high dependency on foreign inflows. During the years of formal monetary targeting, policy practice for bringing down inflation was founded in a rather unorthodox use of two interdependent instruments: the exchange rate and sterilisations. However, sterilisations were not undertaken in the strict sense of monetary targeting to isolate reserve positions from forex interventions (since sterilisations failed to eliminate the massive excess liquidity in the system) but as the complementary instrument of the managed float: a vehicle for attracting foreign capital by offering high returns. This further placed the NBR in a rather uncomfortable relationship with the commercial banking sector, a vicious circle where foreign capital was reaping profits not only from expectations of exchange rate on an appreciating trend but also essentially from the risk-free high returns NBR offered on its deposit-taking operations. As commercial banks depended on the forex position for liquidity, any loss of foreign confidence, such as happened in the second quarter of 2003, would prompt the NBR to impose tight liquidity policies, driving commercial banks to the money markets for satisfying the reserve requirements and triggering upward adjustments across the structure of interest rates.


Under these circumstances, the monetary authorities' decision to adopt a new policy strategy raises a series of challenges. It would firstly require an abandonment of the exchange rate as a policy objective, a policy stance greatly complicated by the systematic influence of exchange rate movements on costs and prices. Indeed, while under the new policy regime the NBR has formally renounced any direct forex interventions, the extent to which policy rate decisions can be freed from exchange rate considerations is not entirely straightforward.

Secondly, it would call for a new approach to policy rate manipulation and sterilisations. The new policy framework, built on an explicit commitment to a single 'credible' inflation target, is rooted in the use of a short-term interest rate (NBR, 2005), usually the overnight interbank rate, as the operational target of the new policy rule. This represents the first of a series of assumptions behind the inflation targeting model (Clarida et al., 1999):

* Inflation is constructed as broadly an aggregate demand phenomenon, with cost-push factors entering the model as exogenous shocks.

* Monetary authorities, by controlling short-term rates, can influence long-term rates relevant to aggregate demand control.

* There is predictable transmission mechanism from interest rates to aggregate demand, reflected in a policy rule.


For the purpose of this paper, the analysis will focus on exploring short-term interbank rates control, on which the entire transmission mechanism depends.

The new policy framework has to be understood in the context of capital account liberalisation. Indeed, EU accession negotiations required Romania to open access to domestic currency denominated bank accounts for nonresidents by early 2004, a move monetary authorities feared would reverberate in the speculative transactions segment. Thus, a postponement to April 2005 was negotiated, a breathing space the NBR demanded in order to adopt containing measures. On the one hand, it undertook a series of aggressive policy rate cuts to size down the interest rate differential (see Figure 8). On the other hand, fearing the potential inflationary consequences of stimulating aggregate demand, the NBR decided to simultaneously allow inflation-curbing nominal appreciations in the domestic currency (12) by gradually withdrawing from the forex market. (13) Consistent with the new policy framework, it would completely abandon its forex interventions by the fourth quarter of 2005. Massive appreciations in the domestic currency accompanied the withdrawal.

Under these circumstances, the 2005 policy shift has been accompanied by sustained attempts to consolidate the overnight rate as an operational target. As shown previously, interest rate policy practice during the years of monetary targeting consisted of manipulating the interest rate on sterilisation operations. Coupled with the forex interventions to moderate real appreciations (and limit the further deterioration of the current account position), it created excess liquidity on the interbank market, limiting the policy rate relevance for liquidity pricing decisions, and hence for aggregate demand control. With the 2005 capital account liberalisation, considerations of speculative inflows would become more prominent in setting the policy rate.

In order to strengthen policy fate's signalling role for the overall interest rate structure, the monetary authority essentially needed to break the vicious cycle of speculative capital inflows, excessive liquidity, sterilisations and the attending volatility on the overnight market. Indeed, immediately after the August 2005 regime switch, the NBR decided to significantly downsize sterilisation operations, and simultaneously reduce both the intervention rate (ie the sterilisation rate) and the rates on standing facilities (see Figure 8). This caused a massive plunge in overnight rates, as commercial banks were forced to take all excess liquidity to the deposit facility. However, the strategy was short-lived: short-term capital dried up quickly, the domestic currency started depreciating first in nominal and then in real terms. This threatened the credibility of the inflation objective and forced the NBR back to its old habits: by November 2005 business as usual on the interbank market, with NBR increasingly sterilising liquidity through deposit-taking operations and speculative flows restored. Overnight rates resumed the familiar pattern of the previous policy regime, fluctuating and most often dropping to the level of the deposit facility towards the end of the maintenance period.

This has further implications for policy-making. While the NBR is formally professing to use an interest rate-based policy rule, policy practice seems little changed from the days of monetary targeting. There is indeed one important difference, namely the withdrawal of the monetary authority from the forex market as required by the inflation targeting framework. (14) Nevertheless, the exchange rate still enters policy rate decisions, recasting the policy rate dilemma in terms of a trade-off between internal {fuelling aggregate demand) and external {moderating real appreciations) considerations.

Furthermore, almost 2 years after the adoption of the new policy regime, the NBR has not managed to eliminate either excess liquidity or the attending overnight volatility. Thus, its policy rate can hardly be construed as relevant for spending decisions. A move from rhetoric to actual practice of inflation targeting has to be rooted in forging a closer relationship between the policy rate and short-term money market rates, in other words eliminating the excess liquidity on overnight markets. This clearly requires a reconsideration of the strategy towards speculative inflows and sterilisations, a problem with two apparent solutions: either lowering interest rates dramatically to slim down speculative flows or designing a new strategy to attract foreign inflows by other means than the policy rate.

There are indications that the NBR is considering the first strategy, as indeed it has operated two massive cuts in its policy rates in early 2007 (75 and 50 percentage points, respectively). The argumentation behind suggests that money market rates have somehow entered the policy rule, as the cuts have been motivated by the growing differential between the policy and the market While this unorthodox approach stands proof that the monetary authorities are aware of the need to bring market rates in line with policy rates, it does not remove one essential predicament: the theoretical framework behind inflation targeting is grounded in an active control of market rates rather than adjustments of the policy instrument to money market movements exogenously determined by commercial banks' forex positions.

A possibly longer-term solution to attracting foreign capital inflows while establishing the policy rate as an instrument for controlling aggregate demand would require segmenting money markets so that foreign inflows become gradually less sensitive to movements in short-term rates. This could be achieved by issuing long-term foreign currency denominated bonds to build up reserves (also addressing the high sterilisation costs), while simultaneously gradually reducing excess liquidity in the market through frontrunner operations: using one particular commercial bank for sterilisation operations at rates consistent with a strategy aimed at ensuring that short-term rates track the policy rate. Admittedly segmenting markets would be controversial for two reasons. It would firstly require coordination with public debt issuing. It would be further complicated by the Maastricht criteria for adopting the euro. (16) However, the vulnerability to foreign market sentiment constraining the NBR can only be tackled through innovative policy-making. Opening up policy debates with other interested parties (the fiscal authority, exporters, banking sector, etc) on a certain policy option, while not necessarily providing grounds for concrete action could nevertheless contribute to the expansion of the imaginary of policy options.

Ultimately, given the structural factors that shape the close relationship between the inflation and the exchange rate, restoring meaning to monetary control requires a reconsideration of the understanding of inflationary processes and the monetary theories underlying the policy framework.


A fundamental weakness undermined the monetary policy strategy employed by Romanian monetary authorities between 1997 and 2005. Despite various adjustments in the policy framework, base money targeting failed to provide monetary authorities with a predictable control of either the intermediary target or prices. The monetary transmission under this policy regime could be best described as an endogenous process, with the banking sector playing an active role in the endogenous determination of the money stock. Furthermore, the large pass-through from exchange rate to prices positioned the exchange rate policy at the centre of the anti-inflationary strategy. That excess liquidity on interbank markets co-existed with a rapid disinflation process constitutes a clear indication that exchange rate movements, and not liquidity control, have been instrumental in reducing inflation rates.

Aside from its modernity appeal, the shift from monetary targeting to inflation targeting has not fundamentally changed policy-making. In a policy environment where the trade-off between inflation and the external position retains its relevance, the new policy regime has so far failed to strengthen the link between policy and money market rates. Repeated attempts at modifying the structural conditions in the money market have been constrained by and finally failed because of the dependency on foreign capital inflows and the inflationary consequences of depreciations. Partial sterilisations of liquidity through deposit-taking operations remain the main vehicle for policy interventions.

While the policy rate is far from providing the type of signalling role envisaged in the inflation targeting transmission mechanism, placing the new policy regime in the context of the April 2005 phase in the capital account liberalisation sheds new light on interest and exchange rate decisions. Initially monetary authorities' withdrawal from the forex market was designed to offset the inflationary implications of reducing the interest rate differential through real appreciations. The current policy regime has restricted the array of policy options available for moderating the impact of massive speculative inflows to cuts in the policy rate. This policy stance, constructed as expansionary in the inflation targeting conceptualisation of inflationary pressures (and despite the little predictability of the transmission mechanism) could potentially undermine the consolidation of the new policy regime by confusing market expectations and affecting policy credibility.

The National Bank of Romania is confronted with a policy cul-de-sac. To escape it, it should aim in a first instance at exploring innovative policy strategies and ultimately at designing an integrated approach to economic policy that takes into account the structural nature of the current account deficit and the desirability of imposing low inflation rates.

Finally, with an economy highly vulnerable to 'sentiment' on international financial markets, the real challenge for Romanian monetary authorities is how to conduct monetary policy in times of forex shortages. These are correlated with tight liquidity policies in developed countries, which usually occur in times of rising prices, particularly in commodities, which form the bulk of imports for Romania. Tightening liquidity in such circumstances, as prescribed by an excess aggregate demand understanding of inflation, will have severely contractionary consequences. This raises concerns when it comes to surrendering policy-making to the European central bank, as its 'hard nosed' reputation and its general understanding of inflationary processes seem to leave little room for a proper consideration of inflationary pressures for any particular Eurozone member.


Bagehot, W. 1910: Lombard Street, a description of the money market. Henry S King & Co: London.

Bindseil, U. 2004: The operational target of monetary policy and the rise and fall of reserve position Doctrine. Working paper no 372, European Central Bank.

Clarida, R, Jordi, G and Gertler, M. (1999): The science of monetary policy: A new Keynesian perspective. Journal of Economic Literature, American Economic Association 37(4): 1661-1707.

Davidson, P. 1977: Post-Keynesian monetary theory and inflation. In: Weintranb, S. (ed). Modern Economic Thought. University of Pennsylvania Press: Philadelphia.

Dow, SC. 2004: Endogenous money: Structuralists. In: Arestis, P and Sawyer, M (eds). Handbook of Alternative Monetary Economics. Edward Elgar Publishing: Cheltenham.

Friedman, M. 1956: The quantity theory of money: A restatement. In: Friedman, M (ed). Studies in the Quantity Theory of Money. University of Chicago Press: Chicago.

Gabor, D. 2003: Inflation in postcommunist Romania. Unpublished Master's Thesis, Institute of Social Studies, the Hague.

Goodhart, CAE. 1989: The conduct of monetary policy. The Economic Journal 99: 199-346.

Goodhart, CAE. 1994: What should central banks do? What should be their macroeconomic objectives and operations? The Economic Journal 104: 1424-1436.

Government of Romania. 1994: Technical memorandum of understanding. Bucharest, Romania.

International Monetary Fund. Selected issues. IMF country reports 1991-2004, Romania, Washington International Monetary Funds.

Keller, P and Richardson, T. 2003: Nominal anchors in the CIS. IMF Working paper 03179.

Moore, B. 1984: Contemporaneous reserve accounting: Can reserves be quantity constrained? Journal of Post Keynesian Economics 7(1): 103-114.

Moore, B. 1988: Horisontalists and verticalists: the macroeconomics of credit money. Cambridge University Press: Cambridge.

National Bank of Romania. Annual reports 1990-2003. Bucharest, Romania.

National Bank of Romania. Inflation targeting quarterly reports 2005-2007. Bucharest, Romania.

Organization for Economic Cooperation and Development. 2002: Economic survey Romania economic assessment. OECD, Publishing: Paris.

Ritter, L, Silber, WL and Udell, GF. 1997: Principles of money, banking and financial markets, 9th Edition, Addison Wesley: Reading, MA.

Taylor, J. 2000: Using monetary policy rules in developing economies Bank of Mexico 75th Anniversary Conference.

Winiecki, J. 1988: The distorted world of Soviet-type economies. Routledge: London.

Woodford, M. 2003: Interest and prices: foundations of a theory of monetary policy. Princeton University Press: Princeton, NJ.

(1) Open Market Operations, standing facilities (the discount window and the deposit facility) and Reserve Requirements.

(2) Monetary base = reserves of the banking sector with the central bank plus cash in circulation.

(3) Armenia, Azerbaijan, Belarus, Georgia, Kazakhstan, Kyrgyz Rep, Moldova, Russia, Tajikistan, Turkmenistan, Ukraine, Uzbekistan.

(4) Net claims on the government and net claims on deposit money banks.

(5) With government paper or forex interventions.

(6) Facility offered by the central bank for depositing excess reserves at the end of the maintenance period.

(7) In 1990 approximately 80% of the total imports were represented by intermediary imports, ratio reduced to around 67% in 2001.

(8) Up to 2002, the US Dollar, then changed to the euro in view of the future accession to the EU.

(9) Having become a EU member state in January 2007, Romania must fulfil the so-called Maastricht macroeconomic criteria for the adoption of the euro, one of which is a level of inflation closely following the most performing countries of the European Union.

(10) Industrial production fell with 7% in 1997 and 14% in 1998.

(11) In order to reduce the costs incurred by commercial banks, interest rates below the market rates have been paid on the reserves held with the central bank.

(12) Complemented with administrative measures aimed at containing growth in foreign currency denominated credit.

(13) 'In an attempt at preventing an inflation flare-up following the easing of the interest rate policy, the National Bank of Romania left the domestic currency to appreciate significantly in nominal terms as a result of the increasing capital inflows by limiting gradually its intervention in the foreign exchange market' (NBR Annual Report, 2005, p. 12).

(14) Since exchange rate movements do feed into prices, this policy stance would come under serious strain from sustained demand-side pressures on the forex market.

(15) Developments in foreign exchange and money markets over the recent months have highlighted liquidity and interest rate levels triggering a faster appreciation of the domestic currency, which contributes significantly to a slowdown in aggregate price growth. Money market rates have constantly been below the monetary policy rate so far, which calls for the adoption of measures to enhance the signalling role of the latter' (NBR press release February 9, 2007).

(16) The Maastrieht criterion for long-term interest rates specifies that the long-term interest rate should not surpass by more than 2% the average rate on 10-year government bonds in the three countries with the lowest rates of inflation of the Eurozone.


Department of Economics, University of Stirling, Stirling, Scotland FK9 4LA, UK.

Table 1: Sources of expansion in high-powered money, 1996-2004

 HPM (bill. NFA (%) NCGVT (%) NCB (%) ONA (%)

1996 7,877 -29 3 98 28
1997 10,587 108 25 -69 37
1998 19,090 40 48 -16 28
1999 35,982 82 52 -39 5
2000 51,485 128 26 -44 -9
2001 67,791 204 0 -79 -25
2002 80,190 289 -9 -136 -44
2003 98,415 295 -12 -131 -52
2004 37,047 331 -37 -186 -18

 Refinancing from
 NBR (outstanding
 bill. lei)

1996 8,822
1997 3,367
1998 3,618
1999 2,433
2000 2,296
2001 1,148
2002 0
2003 0
2004 0

Source: Authors computation from NBR data.
COPYRIGHT 2008 Association for Comparative Economic Studies
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2008 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Symposium Paper
Author:Gabor, Daniela
Publication:Comparative Economic Studies
Article Type:Report
Geographic Code:4EXRO
Date:Sep 1, 2008
Previous Article:Credibility of inflation targets in Poland.
Next Article:13th Dubrovnik Economic Conference symposium.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters