Oil Price Cycles, Fiscal Dominance and Counter-cyclical Monetary Policy in Iran

Transcription

1 MPRA Munich Personal RePEc Archive Oil Price Cycles, Fiscal Dominance and Counter-cyclical Monetary Policy in Iran Ahmad Reza Jalali Naini and Mohammad Amin Naderian Institute for Manaement and Plannin Studies (IMPS), Allame Tabatabaei University 1 Auust 2017 Online at MPRA Paper No , posted 13 February :07 UTC

2 Oil Price Cycles, Fiscal Dominance and Countercyclical Monetary Policy in Iran Abstract: Ahmad Reza Jalali Naini 1 Mohammad Amin Naderian 2 September 2017 Impulse for business cycles in Iran are larely enerated from oil price (terms of trade) shocks and propaated throuh fiscal policies. The classic mission of monetary policy is to conduct countercyclical policy, however, this is not a universal norm. Pro-cyclical fiscal and monetary policies durin boom periods has been observed in a number of developin countries. Such policies tend to amplify the impact of positive oil price (terms of trade) shocks throuh areated demand expansion. The consequence has been strenthenin of domestic inflationary pressures and appreciation of the real exchane rate. This paper attempts to examine if monetary policy in Iran is countercyclical and what is the impact of fiscal policy in this reard. It will be arued that the stance of fiscal policy and how overnment expenditures are financed can have a sinificant effect on how monetary policy is conducted. Our empirical observations reardin the experience of the Iranian economy indicates that, in a fiscally dominated structure, fiscal and monetary policies are enerally expansionary, particularly durin economic booms. This entails subsequent very lare manaed depreciation of the exchane rate, hiher inflation rates, and an economic downturn. Under fiscal dominance monetary policy will be ineffective and both tarets and instruments of monetary policy makin will not be under the control of monetary authority. The policy packae of a structural balanced fiscal rule combined with smoothin of quasi-fiscal operations is the appropriate policy measure that enhances the ability of central bank to conduct more effective countercyclical monetary policies. Key words: Pro-cyclicality, Fiscal Dominance, Monetary policy, Ricardian. JEL: E52, E63. 1 Associate professor, Institute for Manaement and Plannin Studies (IMPS), a.jalali@imps.ac.ir 2 Ph.D candidate, Allame Tabatabaei University, naderian.amin@mail.com 1

3 1-Introduction: The classic mission of monetary policy is to provide price stability and to operate as an anti-cyclical stabilizin force over the business cycles by choosin an appropriate nominal anchor to conduct policy. What is the suitable nominal anchor for the economy? That depends on the structure of the economy and development of the financial markets. The older literature reconized the central bank s tareted rowth rate of money supply as the nominal anchor and monetary policy as an exoenous process. In their joint work Friedman and Schwartz (1963) suested that the cause of inflation is excessive rowth of the money supply, and deceleration of nominal- GNP rowth rate and the rate of inflation occurs subsequent to a prior money rowth deceleration. Moreover, occurrence of a recession (depression) and deflation can be traced back to a prior contractionary monetary policy in a period of credit restriction (credit crunch). Empirical observations made by the above cited study led to a more eneral question of rules versus discretion as a way to conduct monetary policy. Friedman s k percent rule stipulated a lon-run policy induced rowth rate of monetary base. If, projected (averae) GDP rowth is θ percent and the (averae) money supply is allowed to row at k per cent, inflation rate on averae will be k-θ per cent--if the income elasticity of demand for money is unity.3. Several criticisms have been leveled aainst the K-percent rule. Obtsfeld and Rooff (1983) arue that the desired amount of money a representative household keeps is subject to inflation expectations, in other words, the demand for money reacts to chanes in expectations hence there are different equilibrium inflation trajectories. This line of arument questions whether control of the money supply is sufficient to determine the equilibrium inflation path. Blinder (1988) arues aainst a fixed rule by comparin it to fixin a rudder in a stormy sea. McCallum (1987) and Meltzer (1987) arue that the monetary policy rule should consider and adjust for financial disorders (and innovations), which allows for some flexibility while maintain a precommitment to the announced policy rule. Mc Cullum rule (1987) is an example of a money quantity rule which is more flexible than the k percent rule but adheres to a policy stance. Mc Cullum s rule can be differentiated with k percent rule in that it 3 In a different settin, by assumin that the marinal social cost of creatin money is zero so its opportunity cost should be equal to zero, Friedman (1969) attempts to answer the question: what is the optimum rowth rate of the quantity of money? He arues that the optimum quantity of money should row at a rate such that the nominal interest rate is set to zero. 3 The monetary authority can reach this milestone by ensurin that the averae (expected) rate of inflation is equal to the neative of the rate of return on riskless assets such a short-term overnment bond (say µ per cent). In other words, enineerin a deflation rate of at µ per cent. In the above two contexts, it is presumed that the monetary authority can control the rate of inflation throuh its control over the money supply. 2

4 is also a feed-back rule like the Taylor rule (1999).4 These feedback rules endoenise the monetary policy instrument--usually via a short-term interest rate-- with respect to the state variables in the economy--the tarets of monetary policy. The canonical New Keynesian model provides a suitable framework for optimal feedback rules. More specifically, this model yields a taretin rule in contrast to simple instrument rules as an optimal reaction function to the taret variables specified in the loss function of the central bank. 5 In this context the central bank sets the policy rate as a function of inflation and output ap in line with their anticyclical or stabilization policy. 6 The prevalent tradition in the New-Keynesian as well as the broader literature has been to model inflation and its trajectory over time by focusin on monetary policy and leave aside fiscal policy. The standard New-Keynesian inflation taretin models for monetary stabilization policy often do not contain the effect of overnment budet and the path of the public debt. In this setup, the consequence of monetary policy decisions on the fiscal side is omitted. This assumption miht be harmless for an economy that can sufficiently raise non-distortive revenues to finance expenditures but it may not be suitable for those developin countries that have had to cope with hih inflation mainly due to their protracted fiscal imbalances and wish to implement an inflation taretin framework. In the above mentioned dichotomized approach, monetary policy s mission is inflation stabilization and the task of fiscal policy is manaement of the stock of overnment debt. This implies a clear and transparent form of deleatin and assinin the conduct of policy to monetary and fiscal policy institutions. However, in the actual practice the posture and also limitations of fiscal policy, as has been observed in the post 2007 financialeconomic crisis, influences the stance of monetary policy. 4 Taylor rule is a well-known example of an instrument rule which is a formula for settin the policy rate as a iven function of observable variables. A simple instrument rule makes the instrument rate a simple function of a few observable variables which happen to be monetary policy tarets. 5 The familiar quadratic loss function is based on Benino and Woodford (2012) derivation of the loss function for the representative aent. For a closed economy, the usual loss function is 3 E t t 0 t * {( ) t 2 } where * t and t is the current and tareted inflation rates, respectively, and is the output ap. 6 Assumin that the structural equations describin the economy consists of a Phillips relationship: t t 1 t 1 t and the areate demand (IS) equation: t t 1 ( it 1 t 1 ) t, the instrument rule is obtained from the first order condition obtained from the Phillips equation as: * ( it t 1 t t ) 2 t 2 t

5 One distinct area where the ability of the central bank to control inflation throuh its monetary policy instrument has been subject to question and scrutiny is the influence of fiscal policy. The disposition of fiscal policy has sinificant implications for the conduct of monetary policy and vice versa. Given a path for fiscal policy, monetary policy via interest rates influences the intertemporal solvency of overnment debt. As a result, a chane in monetary policy induces chanes in fiscal policy. Moreover, fiscal policy chanes can alter the trade-off between inflation and output. Sarent and Wallace (1984) show that while the central bank can control the tareted rate of inflation, in certain circumstances the taret rate is dictated by fiscal policy. At the heart of the matter is the solvency condition for the overnment. Over the plannin horizon, the real value of overnment debt is equal to (and must be financed by) the discounted sum of seiniorae and taxes collected by the central bank and the fiscal authority, respectively. However, this consolidated budet constraint is not viewed as bindin by the monetary and fiscal authorities. Without a specific framework for cooperation it is not clear which department leads and which one follows in settin policy. If the central bank is able to lead and set the path for the tareted policy variable in its domain, (the trajectory for inflation and seiniorae), the central bank is able to choose and set its tareted inflation rate and determine the quantity of seiniorae throuh monetary policy. Given the transversality condition for the consolidated budet constraint, once the inflation taret rate is determined, the central bank dictates it to the fiscal authority, and the fiscal authority would then determines the discounted value of the primary surplus. Thus, determination of the rowth rate of the money supply by the monetary authority delivers both the nominal anchor and inflation control. In this settin, the monetarist explanation for price determination and control holds (Canzeroni, et al 2010). If fiscal policy has the upper hand and leads the policy process hence determines the path for fiscal policy (the trajectory of overnment revenues minus expenditures or more specifically the discounted sum of primary surpluses), the taret rate of inflation will be set by fiscal policy. In this case, the central bank only has control over the inflation taret handed to them by the fiscal authority. In a somewhat similar but distinct ways, the Fiscal Theory of Price Level (FTPL) arues that the conduct of fiscal policy is a constraint for determinin the optimal inflation path decided by the monetary authority and by itself does not deliver the nominal anchor for the economy (Leeper 1991, Woodford 1995, Kocherlakota and Phelan 1999, Chochran 2005). FTPL discusses the requisite forms of coordination (pairins) between the monetary and the fiscal policies to determine an equilibrium price level path. Not all 4

6 policy pairins can deliver stable prices. Other pairins may provide sunspot equilibria or explosive price paths (Canzeroni et al 2010). The type of fiscal policies pursued has crucial consequences for formulatin the optimal monetary policy. The problem of monetary stabilization would be more complex when different fiscal institutions and reimes are taken into account, and standard inflation taretin may no loner yield the optimal plan. None the less, for a closed-economy, "[O]ptimal monetary policy can be implemented throuh a commitment to use policy to uarantee fulfillment of a taret criterion, which specifies the acceptable level of an output-ap-adjusted price level iven the central bank s current projections of the economy s possible future evolution. A credible commitment to such a rule should serve to anchor inflation."7 However, in a number of developin countries the institutional setup is not suited to handle such complex situations. In some countries the underlyin institutions does not allow for clear demarcation between fiscal and monetary policies. In others, potential cooperative schemes are undermined by fiscal dominance. Even in the presence of such schemes, the existin monetary policy frameworks are not sufficiently complex as to handle different fiscal institutions and policy characters. In particular, for commodityexportin countries fiscal policy has a pro-cyclical character that, in many instances, can induce co-movements in monetary operations or policies. This combination tend to exacerbates expansionary areate demand forces durin booms and weakens them durin an economic downturn (McGettian et al 2013), and overwhelms the anti-cyclical nature of monetary policy. In this paper we take up the question why the Iranian central bank has not been successful in conductin systematic anti-cyclical and anti-inflation policy. The paper focuses on the role of fiscal dominance as it relates to the question posed in the above. We arue that over the period fiscal policy has been a major contributin factor that impined on monetary policy conduct. In particular, fiscal operations influenced the time path of monetary policy instruments like monetary base, policy rates, and the exchane rate as instances of fiscal dominance. We will arue that the nature of fiscal dominance has for the last four decades underone chanes, mainly due to the behavior of an exoenous variables, namely, oil revenues. State-dependency of overnment expenditures and its finance has presented 7 Benino and Woodford (2007), P. the authors warn that even if a constrained-optimal monetary policy reime can be formulated and the fiscal authorities know about it, that should not be an incentive for fiscal policy to be "profliate" and rely on the monetary authority to assist with modifications of its policy to "accommodate any deree of spendin". 5

7 different policy trade-offs to the monetary authority and at times has shaped the time path of monetary policy instruments and tarets and in this process has undermined the ability of the central bank to conduct stabilization policy. In section (2) we discuss some fiscal characteristics of a commodity (oil) exportin country such as Iran. Section (3) provides a eneral accountin framework to show the connections between fiscal and monetary sector operations. In section (4) we provide evidence reardin the pro-cyclicality of fiscal and monetary policies. Section (5) discusses the basic reasons why fiscal and monetary policies tend to be pro-cyclical durin a commodity boom. Section (6) we discuss the consequences of not adoptin countercyclical policy. Section (7) discusses the role of monetary policy in the environment pictured in the previous sections. Section (8) presents the concludin remarks. 2-Composition of Fiscal Revenue and Expenditures Fiure (1) shows the composition of overnment revenue in Iran durin the period As in most commodity (oil) exportin economies, resource exports are a major part of overnment revenue and finance and portrays the same picture. As indicated by fiure (1), the share of oil revenues (in current rials) in total overnment revenues (in current rials) has been larer than other components in most of the observation years. The share of oil revenues in total overnment revenues peaked in 1994 with 73.4 percent. Subsequently it followed a downward trend to a trouh of 25.2 percent in 2009, followed by a partial increase durin the subsequent years reachin 33 percent in Note that the share is dependent on the volume and price of oil exports and the rate at which oil revenues are converted into rials. Zarei and Najafi (2014), apply concordance measure on the relationship between fiscal variables and business cycles in Iran and shows that overnment oil revenues lead real GDP positively within two quarters. Tax income on the averae comprises 42 percent of total revenues durin Zaeri and Najafi (2014) show that tax revenues are larely pro-cyclical in Iran and only in 31 percent of the boom and 25 percent of bust years were concurrent with countercyclical tax policies. The overall budet (OB) balance was always in deficit except four years 8 durin the period. The OB deficit in Iran first peaked in 2008 owin to an oil price plune. Followin the recent oil price collapse durin 2014, the level of deficits 8 In 1995, 1996, 1997, and 2001 overall balance were in surplus. 6

8 aain rose to a new peak in Note that the increase in the OB deficit since 2014 has been financed primarily throuh the accumulation of debt (fiure 1). It must be noted that the share of oil revenues in terms of the domestic currency is a function of oil prices and the rate at which they are converted into the domestic currency (rials). Moreover, durin the periods when oil revenue share declines sinificantly, accumulation of debt and sellin of overnment owned assets increases. 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% overnment oil revenue (%) Tax Income (%) Other Revenues (%) Overall Fiscal Deficit (1000 billion Rials, RHS) Fiure (1): The composition of overnment revenues in Iran. Source, Bank Markazi Iran, (BMI), Economic Time Series Database, and Economic Trends. Fiure (2) shows the upward trend in nominal overnment expenditures and the composition of overnment expenditures. Current expenditures on the averae involves 74.9 percent of total overnment expenditures. Current overnment spendin always rew durin the period. Total overnment expenditures experienced only two contractions in 2009 and 2012 as a result of sharp declines in development expenditures. On the averae, current expenditures rew faster than development expenditures, implyin rowin need for revenue to keep the overnment runnin. Current expenditures were enerally downward-sticky in this period due to contractual obliations, e.. wae and salaries payments to public employees and commitments for direct cash transfer payments. 7

9 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Current Expenditure (%) Development Expenditure (%) Total Government Expenditure (1000 billion Rials,RHS) Fiure (2): Trend of total overnment expenditures and expenditure shares. Source: BMI. 3-Interconnection between Fiscal, Monetary and Bankin System Accounts Financin of the overall fiscal deficit in an oil exportin economy such as Iran is executed mainly in three distinct ways: 1)-hiher oil revenues (in dollars) and/or hiher exchane rates; 2)- hiher taxes; 3)-hiher borrowins (from the domestic residents, forein borrowins, and borrowin from the central bank). With the help of the followin balance-sheet identities, we describe the relationship between sectoral financial accounts and the effect of chanes in each account on monetary base rowth (the policy instrument of the central bank). In the followins, chanes in the balance-sheet of the main players, the overnment, the central bank, and the commercial bankin system will be presented. I-The overnment sector Factors determinin chanes in overnment sector net worth: NW t = A,no NF,t + A,o NF,t (DP ComB,t + E t (A F,t L ComB,t ) + (A np,t B np,t L F,t ) (1) ) + (DP CB,t L CB,t ) + Chanes in overnment net worth from income statement is iven by: 8

10 NW t = T t G t + i t (DP CB,t i t (A np,t L CB,t B np,t ) + E t i t (A F,t B F,t ) + i t (DP ComB,t ) (2) L ComB,t ) + Operational fiscal balance: T t G t (3) Government net interest receipt: i t (DP CB,t i t (A np,t B np,t ) + E t i t (A F,t L F,t ) (4) L CB,t ) + i t (DP ComB,t L ComB,t ) + Substitutin for NW t in (1) from (2) and definition of overall fiscal balance= Operational fiscal balance+ Government net interest payment = (2)+(3), then the overall fiscal deficit OFD, is obtained: OFD t = A,no NF,t A,o NF,t (L ComB,t DP ComB,t ) + (B np,t + E t (L F,t A F,t ) + (L CB,t DP CB,t ) + A Priv,t ) (5) Financin of OFD can be done throuh the items on the riht hand side of (5). Specifically, sellin of non-financial non-oil assets (A,no NF,t ), sellin of non-financial oil assets (A,o NF,t ), currency devaluation or manaed depreciation (E t ), increasin net borrowin from abroad (L F,t A F,t ), increasin net borrowin from the central bank (L CB,t (L ComB,t (B np,t DP CB,t DP ComB,t A np,t ). II-The central Bank ), increasin net borrowin from the commercial banks ), increasin net borrowin from non-financial private sector Chanes in the net worth for Central Bank is due to increased net claim on forein sources or net forein assets=nfa, (A CB F,t L CB F,t ), currency devaluation or manaed depreciation, or net claims on the overnment sector (L CB,t DP CB,t ), net claims on commercial banks (L ComB CB,t DP ComB CB,t ), minus chanes in the monetary base (central bank liabilities) 9. The first sement on the riht hand side of (6) is equal to net interest income from central bank s net assets. NW CB t = E t (A CB F,t L CB F,t ) + (L CB,t MB t (6) DP CB,t ) + (L ComB CB,t DP ComB CB,t ) 9 For simplicity, it is assumed that the chane in non-financial asset of the central bank is null. 9

11 From the income perspective, chane in central bank net worth can be written as NW CB t = i t (L CB,t DP CB,t ) + i t (L ComB CB,t DP ComB CB,t ) + i t (A CB F,t L CB F,t ) (7) Viewed from sources and uses (6), (7) can be re-written as: MB t + i t (L CB,t DP CB,t E t (DP CB F,t L CB F,t ) + (L CB,t ) + i t (L ComB CB,t DP ComB CB,t ) + i t (DP CB F,t L CB F,t ) = DP CB,t ) + (L ComB CB,t DP CB,t ComB ) (8) For the sake of simplicity, we assume that net interest payment accrued to the central bank is neliible, i t (L CB,t DP CB,t ) + i t (L ComB CB,t DP ComB CB,t ) + i t (DP CB F,t CB ) = 0, therefore, from the sources side, monetary base chane can be written in L F,t terms of chanes in net forein and domestic assets: MB t = E t (A CB F,t L CB F,t ) + (L CB,t = NFA t + NDA t, III-Commercial Banks: 10 DP CB,t ) + (L ComB CB,t DP ComB CB,t ) NDA t = (L CB,t DP CB,t ) + (L ComB CB,t DP ComB CB,t ) (8.1) Chanes in the net worth of commercial banks is equal to the sum of : net forein deposits, net lendin to overnment, net claims of the central banks on the commercial banks, net loans to the private sector, and chanes in the required reserves placed with the central bank. NW ComB t = E t (DP ComB F,t L ComB F,t ) + (L ComB,t L ComB priv Priv CB,t ) + (L ComB,t DP ComB,t ) + RR t (9) From the income perspective, (9) can be written as DP ComB,t ) + (DP ComB CB,t NW ComB t = i t (L ComB,t DP ComB,t ) + i t (DP ComB CB,t L ComB priv CB,t ) + i t (L ComB,t Priv DP ComB,t ) + i t (DP ComB F,t L ComB F,t ) (10) For the sake of simplification, we set commercial banks net interest payment equal zero: i t (L ComB,t DP ComB,t ) + i t (DP ComB CB,t L ComB priv Priv CB,t ) + i t (L ComB,t DP ComB,t ) + i t (DP ComB F,t L ComB F,t ) = 0 (11)

12 And the followin identity can be derived for commercial banks: E t (DP ComB F,t L ComB F,t ) + (L ComB,t priv (L ComB,t Priv DP ComB,t ) = RR t (12) DP ComB,t ) + (DP ComB CB,t L ComB CB,t ) + Given identities (5), (8.1), and (12), it is possible to trace throuh the impact of a fiscal deficit on the monetary accounts. As indicated by identity (5), the overnment can resort to different means to finance its overall fiscal deficit. Sellin non-financial oil or non-oil assets, devaluation of exchane rate, borrowin from central bank, borrowin from commercial banks, borrowin from non-financial private sector, and external borrowin from forein countries or international institutions. These alternatives are trade-offs facin the fiscal authorities. For instance, if increased net borrowin from the central bank [ (L CB,t DP CB,t )] is the method of financin the over-all deficit, its impact would be reflected on the central bank balance sheet in (8.1) with consequences on the inflation rate. However, this can frustrate monetary policy tarets set by the monetary authority. Alternatively, the overnment can resort to exchane rate devaluation or borrowin from the commercial banks. Their effects will be reflected on the composition and expansion of monetary base throuh (8.1) and via composition of the assets of the commercial banks (12) with indirect ramifications on the exchane rate or on interest rates (crowdin-out effect), respectively. To the extent that the above financin methods impine on the central bank tarets and instruments, we have the presence of fiscal dominance. 11

13 100% 80% 60% 40% 20% 0% -20% -40% CB Net Forein Asset (%) CB Net Claims on Public Sector (%) CB Claims on Banks (%) Other Items (%) Fiure (3): Sources of chane in the Monetary Base. Source, BMI. Fiure (3) shows the trend of domestic and forein asset components of the monetary base. Prior to the Third Five-Year Development Plan, net claims on the overnment was the main component of the monetary base since overall fiscal deficit was simply monetized by the central bank--reflectin the influence of debt dominance on monetary expansion. To control monetary expansion in this period, the central bank imposed credit limits on the bankin system. The Third Plan law did not allow for automatic financin of fiscal deficit throuh monetary base expansion. The Third Plan allowed the overnment to finance fiscal deficits by sellin more forein exchane to the central bank hence rowth of its net forein assets, and/or devaluation of the domestic currency, particularly durin low oil-price periods, to et more rial per dollar of oil revenues. Hence, the form of fiscal dominance chaned to oil dominance, as net forein assets became the larest components of the central bank s asset-side of the balance-sheet 10. In the more recent years, the overnment 10 In an oil exportin economy, where oil receipts in forein currency are deposited in central bank, the reduction in central bank claim on overnment will be compensated with net forein asset rowth. Hence, monetary base do not chane. However, when overnment uses the oil-related receipts to finance domestic expenditures, there will be an increase in central bank claim on overnment and simultaneous expansion of monetary base (Da Costa et al, 2008). 12

14 has opted to limit depreciation of rial and to maintain the risin trend of nominal overnment expenditures throuh borrowin (fiure 1). While hiher tax revenues has enhanced overnment revenue stream, borrowin from the public has expanded rapidly since The eneral observation is that, fiscal operations either in the form of overnment borrowin from the bankin system (quasi-fiscal operations), debt monetization or forein-exchane monetization (accumulation of net forein assets) has been a major factor behind monetary rowth, underminin the central bank ability to control its main policy instrument. In the recent years, the central bank claim on the bankin system has increased, partly reflectin bankin system claims on the overnment. 3-Procyclicality of Fiscal and Monetary Policies: Some Evidence As discussed by McGettian (2013), enerally, advanced market economies run countercyclical policy and the majority of emerin economies' monetary stance is pro-cyclical. However, larely throuh adoption of inflation taretin, a lare number of emerin market economies have over the years moved on to countercyclical monetary policy. While the same eneral observation can be extend to fiscal policy, Gueruil et al (2016) arue that different fiscal rules have not been equally effective and have had different outcomes and that the desin of countercyclical fiscal rule is an important issue in this reard. Pro-cyclicality of policies is more manifest in oil (resource) exportin countries where income from the oil sector is sinificantly lare and tend to exert sinificant influence on the business cycle. The oil price (terms of trade) shocks are the main impulses of the business cycles and expenditure of oil windfalls throuh fiscal and monetary channels are the propaation mechanism. Note that, while the oil price (terms of trade) shocks are external and exoenous, conduct of fiscal and monetary policies (operations)--channels that propaates the oriinal shock into the economy are internal decisions. Moreover, the nature of the propaation mechanism is a function of the particular forms that the fiscal and monetary packae assumes. A type of propaation mechanism often observed in the oil (commodity) exportin countries results in pro-cyclical fiscal and monetary policies (Frankel 2011) and stron positive correlation between the terms of trade (as well as real oil price) and the real exchane rate durin expansionary cycles. 13

15 Pro-cyclicality of policy has been an issue of economic planners in Iran. There has been two different formats throuh which the overnment has tried to control procyclicality of fiscal policy: Oil Stabilization Fund and the National Development Fund of Iran (NDFI). However, their effectiveness has been limited. Due to its ineffectiveness, the Oil Stabilization Fund that was introduced with the Third Five- Year Development Act, was revoked and replaced with NDFI in the Fifth Development Plan in One way to account for and evaluate the deree of pro-cyclicality of policies (or lack thereof) is to examine their co-movement durin economic booms and recessions. To this end, the first step is to identify boom-bust cycles of economic activities for the Iranian economy with the HP filter and the business-cycle datin alorithm developed by Hardin and Paan (2002, 2006) 11. Based on this method, expansionary (boom) and contractionary (bust) phases of the economic activity are determined by cyclical turnin points, peaks and trouhs, in the time series data. An expansionary phase is defined as trouh-to-peak, while a contractionary phase is defined and measured from peak-to-trouh. However, the weakness of this approach is that it does not measure peaks and trouhs relative to a trend (potential) output. We utilized the HP method to identify the boom and bust cycle for quarterly real non-oil GDP and the ratio of nominal overnment expenditure to non-oil GDP, and the ratio of nominal money base to nominal non-oil GDP. 12 Then by specifyin a series of if statements we check for the cyclical relationship between fiscal and monetary policies over the economic cycles. Final results for eiht cateories are shown in Table (1). 11 For more details on the Hardin and Paan business cycle datin method see (Male, 2010). The R- packae called BCDatin developed by Einian (2013) is used for datin business cycles on quarterly nonoil real GDP rowth durin The results obtained by usin this method is hihly similar to the results reported in table (1). 12 Given a tax rate structure, tax revenues are larely a function of the level of economic activity (i.e. it is endoenous) hence we focus on the pro-cyclicality of the rowth rate of nominal overnment expenditures as the proxy for fiscal policy. 14

16 Table 1: Test of fiscal and monetary pro-cyclicality and co-movement: Economic Cycles quarters Fiscal Cateories quarters Monetary Cateories quarters Inflation Share (%) Currency Devaluation Share (%) ME FE 24 MC ME FC 22 MC ME FE 32 MC ME FC 21 MC FE/FC: Expansionary/Contractionary fiscal policy, ME/MC: Expansionary/Contractionary monetary policy, Source: Author Calculations. Positive Gap Neative Gap The results show that fiscal policy, represented by the ratio of nominal overnment expenditures to real non-oil GDP, is asymmetric durin the economic cycles; it is predominantly pro-cyclical durin the boom phase (52 percent) and predominantly countercyclical (expansionary in a downturn) durin a recessionary period (60 percent). Application of the same method for auin the stance of monetary policy is more difficult because monetary policy instrument is dependent on the existin exchane rate reime and there could be a mix of policy variables. Clean float has never been practiced as a policy, however, manaed float has been the eneral policy approach except when the central bank could not limit currency movements. Since rial is not a perfect substitute for major currencies, policy rates can be a proxy for monetary policy for periods where the exchane rate is fairly stable, partly due to the influence of fiscal operations, and when the rates are allowed to fluctuate more widely. Note that, since interest rate movements within the formal bankin system is constrained, policy rates are at best a ross proxy. 13 The central bank also observes rowth rate of the base money and for many years, it had tarets on the rowth of monetary 13 We do not have sufficiently lon time series on the interbank overniht rates to use it as proxy for policy. 15

17 areates. The behavior of the variable representin monetary policy is similar to fiscal policy. It is procyclical durin boom periods and countercyclical over recessions. When fiscal policy is expansionary durin the boom periods, monetary base is also expansionary 62.5 per cent of the time (15 out of 24 quarters), reflectin a fairly stron co-movement durin boom cycles. This pattern can also be observed durin the recessionary periods. When fiscal policy is expansionary, monetary policy comoves in more than 75 percent of the observation points (24 out of 32 quarter). This behavior can not only be understood as simultaneous shift of both the IS and LM curves that tend to amplify the impact of external shocks, but it can also be an evidence for the existence of fiscal dominance 14. We also tried to find out in which cateories the share of inflation and currency devaluation in their total variation is hiher. The results are shown in the last two columns of the table (1). Nearly 43 percent of the total chane in the CPI index over the periods is associated with periods durin which we observe expansionary monetary and fiscal policies. The reater part of variation in the CPI index occurred in those periods when the economy was in recession (due to a neative areate supply shock) but fiscal and monetary policies remained expansionary (22.4 percent). Interestinly, durin the same observation cateory, currency devaluation is most frequent and accounts for 83 percent of currency devaluation durin the period. Likewise, the bulk of currency devaluations took place durin the 2nd quarter of 1992 and the 3rd quarter of Currency devaluation in these bust phases were a source of fiscal revenue to fill the ap in the budet. This can be perceived as an indication for utilizin currency devaluation to finance budet deficit. 15 The empirical observations support that policy makers follow expansionary fiscal and monetary policies irrespective of boom-bust cycles in a fiscally dominated environment. The averae quarterly rowth rate of nominal fiscal expenditures durin boom periods is 7.54 percent and slihtly larer than the averae rowth rate of Concurrence of countercyclical monetary policy with expansionary fiscal policy in boom periods is the least frequent observation (5 out of 22 quarters). 15 Currency devaluation in connection with the implementation of exchane rate unification policy occurred durin the boom period of the 2nd quarter Expansionary fiscal policy coincided in this period with contractionary monetary policy. 16

18 percent durin recessionary periods. Note that as discussed in sections (2) and (3), the method of financin durin busts is different from the boom periods and the overnment resorts to manaed currency depreciation, quasi-fiscal operations, and borrowins to finance its expenditures. Inflationary pressures and manaed currency depreciations are the by-products of this macroeconomic environment. Our observations also indicate that accumulated fiscal expansion, misalinment of the real exchane rate and inflationary pressure resolve themselves in a jump of the rate of inflation and lare policy manaed depreciations when the neative stron terms of trade shock (and international payment disruptions) hit the domestic economy. Bankin profit rates have been another policy instrument of central bank of Iran. Fiure (4) shows the trend of inflation-adjusted policy (loan and deposit) rates in Iran durin the period. Aside from the period after 2015, real policy rates for both deposits and bank facilities have been neative most of the time, indicatin that policy rates did not follow an anti-cycle pattern and did not react to hiher inflation rates. The standard deviation for the rate of inflation is three times that for the policy rates, indicatin limited reaction of policy rates to fluctuations of the rate of inflation. Durin above HP trend period annual non-oil-gdp, the averae policy-facility rate is per cent while durin below trend (recession) periods it is On the face of it, this indicates stron pro-cyclical behavior of the policy loan rates. However, the important point to note is that, inflation rates on the averae are sinificantly hiher durin recessionary periods (22.5 percent) compared to boom periods (18.05). Since nominal policy rates chanes are limited and do not react sufficiently to movements in the inflation rate see fiure 4--inflation rate variations dominate real policy rates. Additionally, Jalali-Naini and Hematy (2013) and Hematy-Jalali-Naini (2015) estimated several version of the Taylor and McCullum rules 16 to test for reaction of the central bank to output ap, inflation ap, and exchane rate misalinment. The findins showed no systematic reaction of monetary authority to the above macroeconomic variables. Overall, the empirical observations cited here indicate that monetary policy is predominantly pro-cyclical durin boom periods but that cannot be clearly extended to recessionary periods. It is weakly counter-cyclical durin recessionary periods. It should be emphasized that as far as policy rates are 16 Estimation methods included OLS, GMM, and time varyin parameter (TVP) methods. 17

19 concerned, lower real rates durin recessions is not a policy outcome but the result of inflationary conditions which tend to be stroner durin economic slacks Real policy deposit rate (%) Real policy bank facility rate (%) Nominal policy deposit rate (%) Medium term proxy real rate Fiure 4. Nominal and Real Policy rates. Source BMI, Annual Reports and Trends. 4-Factors that tend to Generate Pro-cyclical Fiscal and Monetary Policies The main factors that contribute to creatin pro-cyclicality in boom periods are two broad cateories: political economy motivations and structural factors. a) Political-economy of pro-cyclical budetary operations and the "veracity effect". Availability of more resources induces pressure by the public at lare for hiher welfare spendin and subsidies. Moreover, lobbyin by influential roups to be the beneficiary of more public projects ranted by the overnment and the political benefits of larer spendin by the overnment are the main drivers for oil-induced fiscal expansions. Measures to control this correlation throuh creation of Stabilization Funds (SF) or soverein wealth funds (SWF), have been introduced by overnments of resource exportin countries includin Iran to counter above tendencies. However, the experience has been mixed. b) Asymmetry of access to lobal financial markets. Developin economies (and a small number of developed economies) do not have the capacity to issue debt in their 18

20 own currencies the lobal capital markets (Eichenreen et al 2007). Durin resource booms, resource (oil) exportin countries have stron balance of payment positions and have a better access to world capital markets. Durin these periods, country-risk evaluation and the real exchane rate tend to be on the low side, hence borrowin by firms to finance imports of intermediate and capital oods is less costly. In an oilexportin country like Iran, the impact oil-financed expansion in fiscal expenditures often filters throuh the central bank balance sheet. 17 In particular sellin of forein currency receipts from oil exports to the central bank expands its net forein assets which can further expand the lendin capacity of the bankin system and hence the money supply. In a sense, hiher overnment expenditures that shifts IS to the riht also shifts LM to the riht, makin it a potent combination. 5-Consequences of not adoptin counter-cyclical policies What are the complications associated with pro-cyclical fiscal and monetary policies? Pro-cyclical fiscal policy combined with accommodatin monetary operations tend to strenthen an initial commodity boom on the up-cycle. Expansion of areate demand, hiher inflation rate for non-traded oods combined with relative stability of the nominal exchane rate due to reater availability of forein currency durin oil booms lowers the level of real exchane rate. Fiure (5) shows a positive relationship between the terms of trade (real oil price) and the real exchane rate durin boom periods 18. Persistent decline in the real exchane rate propels the economy onto an unsustainable expansion path particularly, if the boom is not accompanied by risin productivity and/or is tainted with rent-seekin behavior and corruption. In the absence of internal dynamism in the economy e.. a productivity pickup or benefits from economies of scale the oil windfall upcycles tend to fizzle out with maturin of the commodity boom, and increased severity of the Dutch disease, and also as a consequence of deleterious effects of budetary rent seekin. 17 In fact a sinificant contributor to chanes in the monetary base issues from such fiscal operations and the consequent variations in the monetary base. For more details see Jalali-Naini et al (2015). 18 Followin our result from Table (1), monetary and fiscal policies are disproportionately pro-cyclical in expansionary phases resultin in stron positive correlation between real exchane rate and terms of trade. However, since fiscal and monetary policies are frequently counter-cyclical durin contractionary periods, the above-mentioned positive correlation weakens very sinificantly. 19

21 Real rial/market Exchane rate (normalized) (Import Price/Export Price) Real Market ER ToT Q2 1989Q2 1990Q2 1991Q2 1992Q2 1993Q2 1994Q2 1995Q2 1996Q2 1997Q2 1998Q2 1999Q2 2000Q2 2001Q2 2002Q2 2003Q2 2004Q2 2005Q2 2006Q2 2007Q2 2008Q2 2009Q2 2010Q2 2011Q2 2012Q2 2013Q2 2014Q2 0 Fiure (5): Terms of Trade and Real Exchane Rate Trends. There is an asymmetry in the propaation mechanism durin low-oil-price periods. Three different channels propaate the effects of an adverse terms of trade shock. The impact of a commodity-bust is channeled throuh a decline of the real areate demand. Durin an oil price bust, the fiscal authority besets with a lare revenue fall and iven the domestic tax/financin constrains, fiscal authorities cannot effectively expand or maintain the level of real overnment expenditures in the downturn phase of the economic cycle to offset areate demand fall. Attempts to offset declines in nominal overnment revenues with hiher exchane rates and borrowins often does not translate into hiher real overnment expenditures due to hiher inflation rates reflectin shrinkae of the areate supply. The second (exchane rate) channel has two distinct effects; an expansionary and a contractionary effect. An increase in the nominal exchane rate, due to deterioration of the currency reserve (balance of payments) can throuh Marshall-Lerner effect boost net exports and ameliorate the decline in areate demand due to decline of real overnment expenditures. However, this channel has a downside effect too. Since the balance of payment constraints are tihter, and access by domestic firms to forein capital is more limited. Domestic banks are also less willin to expand lendin under such conditions even if monetary policy becomes more 20

22 accommodative. If adjustments in the exchane rate are lare and are followed by speculative attacks on the domestic currency, risk-premiums in the currency markets et larer and the economy is exposed to "the balance sheet effect" which under certain conditions overwhelm the Marshall-Lerner effect and enerate a case of "contractionary devaluation". This situation can further deteriorate if the access to lobal capital markets are reduced due to hiher currency risks. 19 Financin the deficit throuh devaluation, when oil revenues are below trend, thus can be potentially hazardous. This latter, depends on the manitude of price chane in the currency market and the ratio of the value of forein liabilities to the sum of the value of forein and domestic assets in the household and corporate sector, and the size of forein liabilities in the bankin system's balance sheet. Worldwide experience reveals that pace of economic recovery after a financial crisis have been more timid than recovery from other recessions (Reinhart and Rooff 2010). The slow pace of economic recovery in Iran after the recession of 2012, also is partly due to the presence of balance-sheet effects and constriction in credit flows and international payments. Once the speculative attacks beins to strenthen and access to lobal markets becomes limited, the central bank has limited ability to maneuver in the currency market and to limit price movements in this market to control the exchane riskpremium. If the central bank has hue currency reserves and is willin to spend it to defend the exchane rate or limit its depreciation and is willin to raise domestic interest rates, the extent of depreciation can be reduced. However, raisin interest rates in a recessionary environment implies withdrawin accommodation for demand and output in exchane for a lower nominal exchane rate. Economic expansions stimulated by positive oil price (terms of trade) movements and pro-cyclical macroeconomic policies have short-to-medium life span if not accompanied by positive productivity shocks. They tend to fizzle out due to the emerence of the Dutch Disease and expiration of the oil boom. Occurrence of oil price (terms of trade) shocks and imposition of an economic/international payment sanction (that in certain respects works similar to a sudden stop) in Iran, as shown in fiure 5, have resulted in simultaneous lare depreciation of the domestic currency, 19 Basically, those countries that are disposed to the "oriinal sin" are prone to reduced access to lobal credit markets durin lean times. 21

23 lare jumps in the rate of inflation, and contraction of economic activity as corroborated by the results in Table (1). The evidence indicates that pro-cyclical policies durin boom tend to result in economic volatility and instability durin bust periods GDP Growth Rate (%) Currency Depreciation rate (%) CPI Inflation rate (%) Fiure (6). Macroeconomic variable Performance durin Economic Cycles. 5-The role of Monetary Policy: Some Policy Perspectives Greater independence has been iven to central banks and their role in stabilization policy has been enhanced over the last quarter of century. Popularity of flexible inflation taretin and shift from quantity to price (interest) based instruments and the compatibility of this policy framework with New Keynesian models led into formulation of a handful of widely held policy rules that describe under what conditions the monetary authority can achieve simultaneous internal and external stabilization. Assumin that fiscal policy is in balance and the stock of public debt stable, the canonical form arues that in small open economies with price stickiness in the oods market, under the presence of purchasin power parity (producer country pricin=pcp), and existence of deep and internationally interated financial markets, optimal monetary policy stabilization is done via interest rates. Correct policy interest rates and floatin exchane rates provide internal and external 22

24 stabilization (Corsetti et al 2011). When the domestic financial markets are not interated, Local Currency Pricin (LCP) is the norm for price settin by the firms, and home consumption bias exists, the canonical form does not yield the optimal policy. Under these conditions, the monetary authority should also consider reactin to exchane rate misalinments (Enel 2011, 2014). As arued by McGettian et al (2013), only those emerin market economies with deep financial markets and flexible exchane rates have been able to run countercyclical monetary policies. Monetary Policy in a developin economy exposed to oil price (terms of trade) shocks is somewhat more complicated. In an economy with incompletely interated financial markets, a balanced fiscal policy, and beset by a fall in oil prices (neative terms of trade shock), achievin optimal internal and external stabilization by settin the policy rate or the money supply is not possible. In this case, the above dual stabilization also requires adjustin the exchane rate, that is, simultaneous stabilization of two distinct tarets requires two instruments (Jalali-Naini and Naderian 2016, Ostry et al 2012). In the followin sections, we discuss the appropriate monetary policy under different external conditions Positive Oil Price Shocks. What is the appropriate policy reaction in economy with incompletely interated markets (incomplete risk sharin) and LCP pricin and also subject to fiscal dominance, such as Iran? Startin from an initial equilibrium position, we discuss the question of appropriate policy responses under two different set of exoenous shocks; a positive and a neative-oil price shocks. In the case of the former, the appropriate policy response is a counter-cyclical policy that limits expansion of areate demand, checks inflation rate on domestically produced oods, and does not allow for sinificant depreciation of the real exchane rate. A positive oil price (terms of trade) shock increases the volume of oil revenues in forein currency units which the overnment sells them to the central bank to obtain local currency to finance its expenditures. To limit expansion of the monetary base, the central bank increases the amount of forein exchane it sells to the public, thus keepin a lid on the nominal exchane rate. At the same, pro-cyclical fiscal and credit policy expands areate demand, pushin the domestic inflation rates. As explained by (Jalali- Naini and Naderian, 2017), to manae the real exchane rate, the central bank needs 23