Algeria: Selected Issues

Abstract

Algeria: Selected Issues

A Structural Model for Algeria1

A. Introduction

1. The sudden and sharp decline in oil prices presents important challenges for macroeconomic performance and financial stability in Algeria. Algeria’s economy is highly dependent on hydrocarbons for growth, budget revenues, and exports. The nonhydrocarbon sector, consisting largely of industry and services, is driven by public spending financed by hydrocarbons revenues, and therefore its performance is also highly correlated with oil prices. The collapse in oil prices has exacerbated an already unsustainable fiscal position and swung once-comfortable current account surpluses into deep deficit.

2. The oil price shock also poses challenges for macroeconomic policy formulation. Given the size and likely duration of the oil price shock, maintaining current high levels of fiscal spending would cause a considerable widening of the budget and current account deficits and a rapid depletion of fiscal savings and international reserves. The adjustment to the shock will require a reduction in domestic absorption that, for the time being, can mainly be achieved by tightening the fiscal stance. However, this will reduce growth given the high dependency of the economy on fiscal spending, although Algeria could shape fiscal adjustment in a gradual manner thanks to its sizeable fiscal and external buffers. To reduce the impact on growth, the fiscal adjustment should be supported by a well-designed monetary policy, which will soon be able to rely on short-term interest rates as a policy instrument once excess liquidity gives way to structural shortages, as well as an appropriate exchange rate policy.

3. This paper develops a structural macroeconomic model for Algeria that can help inform the discussion of the policy choices faced by the authorities. The model captures the core dynamics of Algeria’s macro-economy and provides an organizing framework for forecasting and policy analysis that can facilitate an assessment of the optimal policy responses to oil shocks and the implications for macroeconomic stability. A key question is what is the right mix of fiscal consolidation, exchange rate depreciation, and monetary policy adjustment that would deliver fiscal sustainability, minimize the cost of adjustment on growth, and keep inflation in check.

4. Results from the model suggest that authorities face a trade-off between, on the one hand, a somewhat larger output contraction and, on the other, a lower debt burden and greater price stability. Exchange rate depreciation cannot substitute for real adjustment, whose magnitude depends on the size of world oil price decline and, as such, will be significant. Maintaining a fixed exchange rate would shift the burden of adjustment entirely to fiscal policy, leading to a larger aggregate demand decline and public debt accumulation. By contrast, allowing the dinar to freely float would increase inflation and trigger a procyclical monetary policy response that would reinforce the impact of fiscal tightening on demand. A well-calibrated exchange rate policy could align the objectives of monetary and fiscal policies, resulting in lower GDP losses and price volatility while supporting fiscal sustainability.

5. The remainder of the paper is structured as follows. Section 2 provides an overview of Algeria’s economic structure, recent macroeconomic developments, and the policy framework. Section 3 presents the model. Section 4 discusses alternatives for monetary policy to support fiscal policy in the face of a persistent oil price shock, and assesses the potential costs of the different options. Section 5 concludes.

B. Stylized Facts and Recent Developments

Recent economic developments

6. The Algerian economy depends heavily on hydrocarbons. In 2015, hydrocarbons accounted for about 25 percent of GDP, 94 percent of export earnings, and 48 percent of budget revenues. The nonhydrocarbon sector consists mainly of industry and services fueled by oil-related government spending. It accounts for about 75 percent of the economy.

A01ufig1

Oil Share of Exports and Government Revenue, 2010-14

(Percent of GDP)

Citation: IMF Staff Country Reports 2016, 128; 10.5089/9781484358351.002.A001

Sources: Country authorities; and IMF staff estimates.

7. The oil price boom over the last several years enabled Algeria to build up substantial external and fiscal buffers. Rapidly rising oil prices led to strong growth of hydrocarbon exports, which in turn generated large current account surpluses and a surge in hydrocarbon revenues. International reserves rose to a peak of US$194 billion in 2013, equal to nearly three years of imports. Part of the hydrocarbon revenue windfall was used to pay down external debt; part was saved in the country’s oil savings fund, which reached 43 percent of GDP in 2009.

8. However, the sharp decline in hydrocarbon prices since 2014 is unmasking longstanding vulnerabilities. The fiscal position—already weakened by a ramp-up in spending in the wake of the Arab Spring—deteriorated significantly in 2015 as oil revenues plummeted. The fiscal deficit doubled to 16.4 percent of GDP. Hydrocarbon exports fell by nearly half, causing the current account deficit to widen sharply. Fiscal savings fell to 12.3 percent of GDP and could reach their statutory floor in 2016. Reserves, while still substantial (2½ years of imports), declined by US$35 billion. Public spending continued to increase, fueling imports and a loss of competitiveness, and placing fiscal policy on an increasingly unsustainable path (Figure 1).

Figure 1.
Figure 1.

Algeria: Macroeconomic Developments, 2010–16

Citation: IMF Staff Country Reports 2016, 128; 10.5089/9781484358351.002.A001

9. The policy response to the oil price shock initially focused on allowing nominal exchange rate depreciation. To moderate import demand and reduce pressure on reserves, the Bank of Algeria (BA) allowed the currency to depreciate by 25 percent against the US dollar and by 6.7 percent against the euro in 2015. The 2016 budget, however, calls for a sharp reduction in spending, and the authorities have initiated some reforms including higher taxes on energy products. To finance the increasing current account deficit, Algeria has been drawing down its foreign exchange reserves. The authorities have been reluctant to issue new external debt despite a low current level of public debt.

10. High dependence on imports is expected to result in some pass through of exchange rate depreciation to inflation. Inflation rose to double-digit levels (on a year-on-year basis) in 2012 following public sector wage increases. While it retreated in 2014 to average 2.9 percent, average year-on-year inflation exceeded the 4 percent target of the BA in 2015. It was partly driven by higher import price inflation, suggesting some degree of exchange rate pass-through as the dinar depreciated significantly against major currencies in 2015 (Figure 2). In the absence of accurate estimate of the exchange pass-through for Algeria, we assume in the rest of the paper that the pass-through is less than one.

Figure 2.
Figure 2.

Algeria: Inflation Developments, 2014–15

Citation: IMF Staff Country Reports 2016, 128; 10.5089/9781484358351.002.A001

Monetary policy framework

11. The ultimate objective of the BA’s monetary policy is to achieve price stability. The BA has since 2010 explicitly targeted price stability, in addition to external stability of the currency. There is an explicit annual inflation target of 4 percent. The monetary policy framework has adjusted over time to reflect the developments of the economic environment caused by oil price fluctuations. Since 2003, base money has been the main intermediate instrument of monetary policy, and liquidity management tools have dominated the monetary policy toolkit. Exchange rate policy targets the equilibrium value of the real effective exchange rate, but has occasionally been used to contain price pressures, a policy that is easily implemented due to the price-maker status of the BA on the forex market.

12. The monetary framework relies on transmission channels of uneven effectiveness. A recent IMF study on the effectiveness of monetary policy in Algeria finds that expectations might have played a powerful role in the price-setting mechanism during periods of high inflation.2 The interest rate channel is currently muted and unresponsive to changes in monetary conditions, due to excess liquidity and the insufficient development of the financial system, blurring the signaling of the policy stance. However, prior to 2004, the policy rate was found to significantly affect bank lending and saving rates. Going forward, excess liquidity conditions are expected to switch to structural refinancing needs, offering the BA an opportunity to re-establish control over domestic liquidity conditions and short-term interest rates.

13. Inflation dynamics in Algeria are sensitive to administrative controls and exchange rate changes. Food items, a third of which are subject to price controls, account for over 43 percent of the CPI basket. Prices of non-administered food items tend to be volatile either because of domestic (e.g., weather) or external factors (e.g., international prices). Altogether, consumption goods and services with administered prices account for 26 percent of the CPI basket. Goods and services with high import content also represent a large share of the CPI basket (26 percent).

Fiscal policy framework

14. Algeria’s de jure fiscal framework is based on a saving rule based on the current oil price. Algeria created an oil stabilization fund (Fonds de Regulation des Recettes, or FRR) in 2000 to insulate the Algerian economy from volatility in hydrocarbon prices. There is a saving rule that stipulates that oil revenue is saved into the FRR above the oil price threshold of US$37 per barrel.

15. However, in practice, Algeria’s fiscal stance has been heavily influenced by hydrocarbon prices. Public spending is disconnected from the saving rule since the FRR can be freely drawn upon for budget support. The nonhydrocarbon primary deficit (NHPD) and spending have been highly correlated with oil prices for the past 15 years, widening during good times and contracting in bad times.

16. Subsidies not only carry large fiscal cost but also weigh on external sustainability. Subsidies cost an estimated 13.6 percent of GDP in 2015, with energy subsidies accounting for over half this amount. In addition to their fiscal cost, subsidies have increased domestic energy consumption, squeezing exports and encouraging large-scale smuggling to neighboring-countries.

C. The Model

17. The analysis is based on a structural model. The key equations of the model derive from micro-foundations.3 The model’s equations jointly determine the dynamics of inflation, output, short-term interest rates and the real exchange rate.4 The system of equations is subject to various shocks, the variance of which can help derive measures of uncertainty in the baseline forecast.

18. The model exhibits correct accounting of both stock and flow variables. Namely, it determines the country’s net exports and net foreign assets along with government debt and deficits. Without a stock-flow consistency, the model would not be able to determine the effects of the terms-of-trade shocks on variables of interest such as debt. This feature contrasts the model to Berg et al. (2006), commonly used as a workhorse model working only with flows.

19. The model employs a new-Keynesian methodology with rational expectations, taking into account the stylized facts of Algeria’s economy. The model consists of four building blocks:

  • A real block defines both domestic demand (i.e., private and public absorption minus imports) and oil exports. It also includes a fiscal rule that influences the level and composition of public expenditures. The hydrocarbon sector plays a key role in shaping budget revenues and GDP.

  • A second block links the real economy with price dynamics, which is represented by a Phillips curve.

  • A third block comprises a financial link with the rest of the world through arbitrage conditions, whereby the country risk premium is assumed to react to changes in the net foreign asset position of the country.

  • The last block consists of monetary policy with a price stability objective. The model is closed by a “hybrid” policy reaction of the central bank. The policy rule, which is defined typically for countries with a closed capital account, assumes that the central bank can use exchange and interest rates as policy instruments to influence monetary conditions.

20. The rest of the model consists of definitions and accounting identities. Expectations play a crucial role in the model as future dynamics of macroeconomic variables matter for their current levels. Expectations are rational, i.e., the model, assumes that economic agents know the dynamics of the economy and the behavior of key macroeconomic variables (model consistent expectations). The structural characteristics of the model and the nature of expectations make the model immune to the Lucas criticism.

21. The standard notation in the literature is adopted. For any given variable X, we denote a natural logarithm of this variable by a small letter x. The model is specified for quarterly frequencies, a delta (Δ) in front of the variable indicates quarter over quarter annualized seasonally adjusted changes, except for in inflation rates which are denoted by π, an asterisk (*) denotes a foreign variable, and finally (ss) denotes steady state variables.

Real block

22. The real block determines the dynamics of domestic private and public absorptions, and foreign trade.5 The economy consists of households who own all firms in the nonhydrocarbon sector. The government owns the hydrocarbon sector and collects its export revenues as well as other tax revenues, purchases goods and services, and makes transfers to households. The external sector is affected by fluctuations in oil prices and changes in consumption behavior of households and the government.

Households

23. Households are forward-looking (i.e., Ricardian). They smooth their consumption intertemporally and maximize their utility, subject to the following budget constraint:

PtAAt+Bt+StBt*=WtNt+exp(it1400)Bt1+exp(it1400)Bt1*+Tt+Πt(1)

Following equation (1), households’ revenues consist of wage income (WN)—where W is the nominal wage and N are labor hours—government transfers net of taxes (T), and profits (Π). Households use their income to finance final good consumption (PC A), where PA represents the deflator and A the domestic absorption, and invest in government bonds B (domestic) and B*(foreign).

24. Equation (2) describes the dynamics of private absorption (A). It represents the first-order optimality condition for maximizing their utility function with respect to (1), so-called Euler equation.6 Equation (2) implies that the marginal rate of substitution between consumption today and future consumption equals the real interest rate.7 To smooth absorption dynamics, we introduce a lag term reflecting habit preferences (i.e., households prefer to preserve past level of their consumption) that permits shocks to have persistent effects.8

(Atβ1At1)σ=(At+1β1At)σ1Rtexp(εtA)(2)

25. At the core of the private absorption dynamics is the negative relationship between the real interest rate and desired spending. When real interest rates are expected to be high, households would rather save than consume. At the same time, they are willing to spend more when future prospects are promising, regardless of the level of interest rates.

Government

26. The fiscal bloc determines the role of the fiscal policy in the economy. We assume that the government owns the hydrocarbon sector and all of its related export revenues (PX X).

Accordingly, any increase in world oil prices raise oil sector revenues and results in higher transfers to households, which in turn lift up their budget constraint and lead to a higher consumption. Budget revenues are as follows:

BRt=PtXXt(5)

Budget revenues are distributed through government spending of final goods (PtcGt), and transfers (Tt) to households.9

BEt=PtCGt+Tt(6)

Primary budget deficit (BD) is the difference between budget expenditures and budget revenues. Any deficit is financed by domestic government bonds issued at interest rate i, increasing government debt (Dt).

Dt=Dt1+BDt+it1Dt1(7)

27. The government can change the level and composition of its expenditures based on a fiscal rule.10 We assume that the government modifies the amount of its consumption of goods and services to keep the debt close to a sustainable level (Bss) as follows:

PtCGt=(Pt1Gt1)α1(PSSGSS)1α1(DtDSS)θ+εtg(8)

28. The rule has two main functions. The first is to respond to the business cycle. The rule allows the fiscal authority to adjust expenditures, and hence the overall fiscal balance, to changes in revenues induced by fluctuations in hydrocarbon prices. The second main function of the rule is to stabilize the government debt-to-GDP ratio to its long-run target that ensures intergenerational equity. This target pins down the long-run net asset position of the general government, and ensures dynamic stability.

29. The parameter θ determines the responsiveness of the fiscal rule to changes in the hydrocarbon revenues. A calibration of θ greater than zero implies that fiscal spending is maintained as long as the government’s debt is lower than its sustainable level and it is reduced when the government’s debt exceeds its sustainable level. In the latter case, the higher the level of θ, the larger and the faster would be the adjustment.

30. Besides ensuring debt sustainability, the government also smooths real economic activity (private absorption) through transfers. Transfers are kept close to a target level, defined in terms of share on private absorption. However, the government smooths transfers adjusting them only gradually to mitigate effects of income shocks on private consumption.11

Tt=(Tt1)α2(TSS)1α2(PtAAtPt1AAt1)α3+εtT(9)
External Sector

31. The trade block consists of two equations that specify the determinants of export (X) and import (M) volumes. The specification of these equations takes account of the structure of Algeria’s foreign trade, in which hydrocarbons account for the bulk of exports, and imports account for a large and stable share of domestic spending. Given the preponderance of hydrocarbons in Algeria’s exports, real exports are inelastic relative to standard determinants such as real effective exchange rate or foreign real income. We model the growth of real hydrocarbon exports rather than levels (Equation 3). It is hard to assess the long-term level of hydrocarbon production in Algeria because the amount of extracted hydrocarbons depends on investments in the sector, available technologies of extraction and the maturity of fields. We model real export growth by an autoregressive process that reflects the persistent effects of shocks to hydrocarbon production and domestic consumption, while allowing for changes in the long term growth rate of hydrocarbons (ΔX¯).12 The volume of exports is multiplied by the world price of oil and provides the main bulk of country revenues.

ΔXt=ρΔXt1+(1ρ)ΔX¯+εtX(10)

Import volumes are defined in Equation (4). As explained above, to match the data, we model real imports as a stable share of domestic production. This assumes that both private and public final good producers use imported goods and services in combination with domestic intermediate production to get final consumption goods and services.13

Mt=ωMYtexp(εtM)(11)

where Mt are real imports, ωM is the share of real imports in final good production (Yt). Multiplying import volumes by import prices yields nominal imports. Import prices (PtM) are defined as US CPI adjusted by the nominal exchange rate of dinar to the U.S. dollar. In order to capture observed data, there is import price Phillips curve that ensures smooth and gradual change of import prices in response to foreign prices and exchange rate shifts.

The external balance is determined using net exports and net foreign asset position, as described in Equation (8). Net exports represent the difference between hydrocarbon exports and imports. The net foreign asset position reflects the accumulation of all past net export deficits, meaning foreign debt and borrowing, adjusted by the foreign interest rate.

Bt=(1+itUS)Bt1+PtXXtPtMMt(12)

Finally, we sum up the different components of GDP to get the aggregate nominal and real GDP as follows:

YtN=PtAAt+PtAGt+PtXXtPtMMt(13)
Yt=At+Gt+XtMt(14)

where PtA,PtX, and PtM are the deflators of absorption, exports and imports respectively

Philips curve

32. Inflation dynamics are represented by a reduced-form (Phillips curve) with import prices capturing the impact of external influences. Headline inflation (π) depends on expected and lagged inflation and the current value of real marginal cost inflation. Real marginal costs are given by the production structure of final consumption goods in the model. Specifically, the Phillips curve is of the form:

πt=λ1E(πt+1)+(1λ1)πt1+λ2rmct+εtπ(15)
rmct=[ωwt+(1ω)(cpitUS+St)]cpit(16)

where ω represents the share of domestically produced goods in the CPI basket, wt is the nominal domestic wage, and st is the nominal exchange rate of the dinar against the US dollar.14

33. The specification of the Phillips curve is derived under the assumption of staggered price setting.15 We employ a Calvo price setting with full backward indexation. Forward-looking expectations are fully model-consistent (i.e., rational expectations). Parameters in the Phillips curve are set in line with micro foundations and the data. The data for Algeria shows a high degree of inflation inertia, indicating that economic agents typically view past inflation as a good predictor of future inflation dynamics. Therefore only large changes in real marginal costs would impact current inflation.

Monetary and exchange rate policy

34. The model captures the usage of the exchange rate as a monetary policy instrument and also allows for assessing different policy options. The choices of exchange rate and monetary policy regime are not independent. If capital is free to flow, a country that pegs its exchange rate cannot afford an independent monetary policy and vice versa. However, in our model, the central bank can use both short-term interest rates and the exchange rate as policy instruments, as Algeria’s capital account is closed.

35. The exchange rate and interest rates are linked using a modified version of the uncovered interest-rate parity condition. We capture the relationship with the rest of the world using an alternative version of the uncovered interest rate parity condition (Equation 10). This condition states that, in the case of perfect international capital mobility, the yield of investment into domestic economy has to be equal to foreign (U.S. dollar) yield adjusted by the country risk premium and the expected change in exchange rate valuation.

st=st+1e+(it+it*+premt)/4+εtUIP(17)

where st is the natural logarithm of the nominal exchange rate, st+1e is the expected exchange rate, it is domestic money market rate, it* is the foreign interest rate, and premt is a country risk premium.

To capture a managed exchange rate regime, we modify the above equation as follows:

st=e1stT+(1e1)(st+1e+it+it*+premt4)+ɛtUIP(18)
stT=e2st1T+(1e2)(st1T+e3Δpoilt)(19)

By setting the parameter e1, we can capture different exchange rate arrangements. While e1 = 1 corresponds to a pegged (or fully managed) exchange rate regime, e1 = 0 matches a fully floating exchange rate regime. Equation (19) determines the level of exchange rate target that is set by the central bank (stT). Among others, the nominal exchange rate depends on shocks to oil prices suggesting that the central bank can allow for some exchange depreciation (which is captured by the parameter e2) in response to negative terms of trade shocks.16

Furthermore, in order to smooth exchange rate dynamics under the floating regime, we define exchange rate expectations as follows:

St+1e=μ1St+1+(1μ1)((St1+2(Z¯t+πtπt*))(20)

The rational-expectation term is represented by St+1 and is given a weight μ1. When μ1 = 1, expectations are purely rational and the model will have Dornbush (1976) overshooting dynamics.

36. The specification reflects BAs monetary policy objective of price stability and reliance on exchange rate rather than interest rates. Therefore, we use the following form of monetary policy rule in the model:

it=δ1(st+1st+it*4+premt4)+(1δ1)(δ2it1+(1δ2)(itneutral+e3πt+1DEV)+ɛti(21)

Where πt+1DEV represents the expected deviation of inflation from the target level. Setting δ1 = 1 implies that monetary policy targets the exchange rate, and therefore is not independent in terms of setting an inflation target and steering domestic interest rates under capital mobility. In this case, domestic money market interest rate is implied by the UIP and the exchange rate target. In the contrary if δ1 = 0, we have an independent monetary policy that sets money market rates based on a forward-looking rule of Taylor type. In this case, the key monetary policy instrument is based on some short-term nominal interest rate that the central bank sets it in order to achieve a target level for inflation (π*).

Calibration

37. The model is calibrated to reflect properties of the Algerian economy (Table 1). The production function is intensive in labor and non-energy imports. The export industry is the most important income-generating sector. To capture the high degree of private consumption persistence (Equation 2) and the high inflation inertia (Equation 15) observed in the data, we set β1 to 0.8, which is at the edge of the range of parameter values used in the literature (i.e., between 0.5 and 0.8), and λ2 to 0.1. To reflect the low flexibility of fiscal expenditures, we set α1 and θ (Equation 8) to 0.5 and 0.25, respectively.17 Calibrating the Taylor rule parameters (Equation 21) is challenging, since the interest rate transmission mechanism of monetary policy in Algeria was muted during the oil price upturn. For the sake of illustrating the propagation mechanisms of the shock and policy tradeoffs, we assume that e3 equals 1.2, reflecting that price stability is the primary objective of the central bank. Parameters δ1, e1, and e2 characterize the monetary policy regime, and we will use different sets of values to assess implications of different macroeconomic policy options in response to the oil price shock.

Table 1.

Calibration of the Model

article image

D. Analysis of Policy Mix and Implications for Macroeconomic Stability

Illustration of the model: the case of permanent terms-of-trade shock

38. For illustrative purposes, this section starts with the analysis of the impact of a permanent oil price shock scenario. Oil prices fall by about 60 percent in 2015, and remain weak at US$47/barrel in the medium term. The oil price shock is considered as unexpected and permanent, implying that rational economic agents expect at any point in time that future oil prices stay low at the current level of US$47/barrel.

39. A permanent decline in oil prices has long-lasting real effects on the economy. The size of the world oil price decline determines along with the structure of national account, the magnitude of domestic absorption adjustment. The terms of trade worsen permanently, forcing real downward adjustments in consumption and nonhydrocarbon production. Monetary and fiscal policies can help smooth the adjustment but they cannot offset it. Furthermore, regardless of the exchange rate regime, a negative terms-of-trade shock creates depreciation pressures ceteris paribus.

40. The dynamics of inflation and the exchange rate depends on the monetary policy response, which varies with the exchange rate regime in place. A number of policy response scenarios are simulated to analyze the dynamics of relevant macroeconomic variables, in particular real growth, inflation and debt accumulation. We focus on monetary policy reaction under (i) a fully flexible exchange rate regime (i.e., under inflation targeting), (ii) a peg (i.e., the exchange rate is fixed at its pre-shock level) and (iii) a managed float that allows for a partial exchange rate depreciation. We also take into account the change in the fiscal policy stance that aims at smoothing real adjustment while ensuring debt sustainability. The results of these simulations are plotted in Figure 3.

Figure 3.
Figure 3.

Impulse Response Functions

(Permanent oil price shock)

Citation: IMF Staff Country Reports 2016, 128; 10.5089/9781484358351.002.A001

Source: IMF staff calculations.

41. When the central bank allows the exchange rate to depreciate, the transmission of the shock operates through two channels. On the supply side, exchange rate depreciation increases the marginal cost of production (Equation 16) and lowers demand for labor and imports, causing adjustments in the labor market (lower employment and real wages). On the demand side, consumers face higher inflation from rising prices of imports and a perceived reduction in income, due to lower wages and potentially lower fiscal transfers (Equation 2). At the same time, the depreciation increases oil export revenues in domestic currency, thus partially offsetting the adverse effects of the shock on fiscal revenues. This opens some space for countercyclical transfers that contribute to smoothing the consumption deleverage. However, as the government faces debt constraints, it adjusts its expenditure on domestic goods and services sharply to ensure that debt never exceeds the desirable (i.e., sustainable) level (Equation 8). Nonhydrocarbon production falls, both as a result of declining demand and supply.

42. A fully flexible exchange rate regime could lead to a procyclical monetary policy response to inflationary pressures, increasing the potential for output losses.18 The large and persistent term-of-trade shock worsens the country’s external position and causes the dinar to depreciate by 40 percent by 2021 (Subplot 5), bringing the real effective exchange rate close to its equilibrium value. Large inflationary pressures emerge, with inflation peaking at 5 percentage points above the central bank target. Simultaneously, exchange rate depreciation inflates government revenues in domestic currency, leading to lower debt accumulation than under less flexible exchange rate regimes and reducing the size of the subsequent fiscal adjustment. In essence, greater exchange rate flexibility supports economic activity and smooths real adjustment. However, the strong monetary policy response increases interest rates sharply and reinforces the real effects of fiscal policy. Of course, these effects would be smaller assuming a more sluggish monetary response caused by a weaker monetary policy transmission mechanism (i.e., with lower e3), or due to higher tolerance for higher inflation, with the central bank either not responding to the full impact of the shock on inflation or choosing to increase its inflation target.19

43. Under a pegged exchange rate regime, there is a relatively large debt accumulation that increases debt sustainability concerns.20 Smoothing private absorption through government transfers on the back of a large drop in hydrocarbon revenues (proportional to the fall in oil export revenues, Subplot 7), increases debt and creates debt sustainability issues (Subplot 9). As a result, the country’s risk premium rises, pushing real interest rates higher and leading to higher debt accumulation than under the floating exchange rate regime (Subplot 6). When public debt becomes unsustainable (i.e., higher than the debt level at the steady state), the government starts tightening its fiscal stance by cutting both expenditures on goods and services and transfers, causing a large contraction in demand through the channels described above. As a result, general prices drop sharply (subplot 3), particularly given the absence of inflationary pressures stemming from exchange rate depreciation and because of the contraction of the real economy.

44. Under the intermediate exchange rate regime, monetary policy keeps the magnitude of depreciation at a level that mitigates GDP losses and inflationary pressures. Assuming a partial exchange rate adjustment, the nominal effective exchange rate depreciates moderately compared to the IT regime (Subplot 5). The nominal depreciation makes imports more expensive relative to domestic production, and results in some substitution between imported and domestically produced goods. Therefore, real imports decline and contribute to reducing the trade deficit. At the same time, the fiscal policy response is stronger than under the IT regime due to the lower impact of exchange rate depreciation on fiscal revenues, resulting in lower debt accumulation. Compared to the peg regime, aggregate demand contraction is larger as the real effects of fiscal adjustment are amplified by the contractionary effects of monetary policy. The latter increases interest rates in response to higher inflationary pressures stemming from higher import prices.21 As a result, macroeconomic volatility is lower than the IT regime, and public debt is more sustainable than in the peg regime.

A scenario consistent with WEO assumption

45. A second set of simulations analyzes the impact of an oil price shock consistent with the January WEO update assumption (Figure 4). Oil prices reach their trough of US$36 in 2016, and gradually stabilize around US$47 by 2021 (Subplot 1). Compared to the permanent oil price shock scenario discussed above, the WEO scenario is more severe as it entails an additional temporary negative shock in 2016, increasing the size of needed adjustment and inducing more volatility in the economy.

Figure 4.
Figure 4.

Impulse Response Functions

(January WEO update assumption)

Citation: IMF Staff Country Reports 2016, 128; 10.5089/9781484358351.002.A001

Source: IMF staff calculations.

46. As in the permanent shock scenario, the policy response faces a difficult tradeoff between price, output, and debt sustainability objectives. The output fall (Subplot 2) and rise in inflation rate (Subplot 3) are larger, while debt accumulation (Subplot 4) is lower, under a pure float compared with a peg regime. Also, partial exchange rate flexibility achieves lower volatility in terms of deviation of inflation from the central bank target, while generating an output decline and debt-to-GDP ratio between the two extremes.

47. The authorities may aim at trading a somewhat larger output contraction for lower debt burden and greater price stability. Simulations comparing different exchange rate regime options suggest that:

  • A fixed exchange rate shifts the burden of adjustment to the fiscal policy, leading to larger public debt accumulation and aggregate demand losses.

  • Allowing the dinar to freely float leads to inflationary pressures, triggering a procyclical monetary policy response that reinforces the impact of fiscal tightening on demand, not only amplifying output contraction but also feeding back to higher inflation.

  • A well-managed exchange rate float could better align the objectives of monetary and fiscal policies, resulting in lower output losses and inflationary pressures while ensuring fiscal sustainability.

E. Conclusions

48. This paper develops a small structural model for Algeria. The model can be helpful in preparing baseline forecasts and analyzing policy options and trade-offs in response to a variety of shocks. By design, it is a relatively simple and tractable representation of the Algerian economy, allowing users to evaluate different policy responses to shocks. Furthermore, the model can be used as a forecasting tool, which provides complete national account forecasts including policy reactions. The model can serve to frame the discussion about the baseline forecast, by evaluating risks to the forecast and external shocks. In contrast to existing gap models a la Berg et al (2006), the model exhibits a stock-flow accounting, which allows us to analyze adverse effects of oil price shocks.

49. Algeria needs to adjust to the low oil price realities. Without sustained policy adjustment, a protracted period of low hydrocarbon prices could set the country on an unsustainable trajectory of macroeconomic imbalances. The heavy reliance on hydrocarbons poses structural risks to Algeria’s economy, and both monetary and fiscal policies face important challenges to respond to the ongoing oil price shock and preserve a positive outlook for growth, inflation and international reserves.

50. Well-managed exchange rate flexibility can help the economy better adjust to the low oil price environment. Given the shallow and imperfect exchange market, the central bank should “lead” the market towards a path of exchange rate adjustment consistent with a medium-term internal and external equilibrium. However, exchange rate adjustment cannot substitute for fiscal and monetary policy adjustment. In addition, these need to be complemented by a structural reform strategy that aims at diversifying exports and replacing imports with domestic production to help preserve growth

Appendix I. Algeria: Microeconomic Foundations of the Model

Households

There is a continuum of infinitely-lived households who consume final goods, supply labor, receive government transfers, and save in banks deposits. The j-th household maximizes the expected discounted utility choosing consumption Ct, labor Nt and the amount of domestic and foreign bonds, Bt respective Bt*:

E0Σt=0βt[ln(Ct(j)χC¯t1(j)1χ)Nt(j)]ct(j),Nt(j),Bt(j),Bt*max

Subject to a sequence of budget constraints at any point of time t:

PtCt(j)+Bt(j)+StBt*(j)=WtNt(j)+(1+it)Bt1(j)+(1+it*)StBt1*+Γt

The expected stream of utility is weighted by a deterministic discount factor β. A household sets his intertemporal path of consumption taking into account the past aggregate consumption C¯ —an assumption of the external habit formation with a smoothing parameter χ. The introduction of the habit formation setup is motivated by empirical evidence that monetary policy shocks trigger a hump-shaped response of consumption. The price index of households’ consumption bundle is denoted as Pt.

Households finance their spending using their labor income and revenues from government bonds. Furthermore, they receive Γt as lump-sum net payments from government—lump-sum taxes adjusted by transfers. There is government’s ownership of the whole economy, including oil industry in particular. Also, households receive Wt for each unit of their labor Nt. Nominal wages are fully flexible implying pro-cyclical real wages. 1 If needed, we make wages sticky later based on empirical observations.

Households save using government bonds. Alternatively, we can label government bonds as commercial banks deposits (we just need to check whether households/agents are allowed to hold foreign currency deposits or/and bonds). In addition to domestic bonds, households hold foreign currency bonds, Bt* Both domestic and foreign currency bonds bear an interest rate, it and it*, and they are for the purpose of derivation considered as risk-free. St denotes the nominal exchange rate in the standard notation (units of domestic currency which pay for one unit of foreign currency).

Maximizing the household problem with respect to the budget constraint a following set of optimality conditions can be obtained. The first condition states that a representative household supplies labor until the marginal rate of substitution between consumption and leisure equals to the real consumption wage:

WtPt=Ct(j)χC¯t11χ

The flow of household’s consumption is determined by a modified Euler equation which balances marginal utility from consumption today and tomorrow (derivative with respect to domestic bonds):

Ct+1χCt=PtPt+1(CtχCt1)β(1+it)

The non-arbitrage condition implies that households do not distinguish between domestic and foreign bonds as both bear the same interest rate revenues in domestic currency. In fact, taking derivative of the Lagrangian of the problem, we get: (1+it)St=(1+it*)St+1, which represents the common UIP condition.

Besides the above-mentioned equations, optimality conditions consist of the budget constraint and the transversality condition. As all households are identical, we can remove the index j from all equations above.

Domestic intermediate production

There is a continuum of intermediate goods producers, out of the oil sector, who are endowed with Cobb-Douglas production function. These producers use just households’ labor Nt to produce intermediate goods. Firms are fully competitive and there is no physical capital so far to keep the model simple.

Firms maximize their profits choosing labor demand Nt as follows:

E0Σt=0Ω0,t[PtDyYtDWtNt]Ntmax

while they face the following production constraint

YtD=(ANt)1γ

In the optimization problem above Pty denotes intermediate prices, YtD states for domestic intermediate product. As firms are owned by households the stochastic discount factor Ωt|t+1=Λt+1Λt is used to weigh profits where Λt denotes the Lagrange multiplier from households optimization problem. Finally, A denotes labor-augmenting technology.

The first order conditions of the firms’ optimization problem constitute demand function for the productive factor:

(1γ)YtDNt=WtPtDy

Final goods producers and aggregators

There is infinity of intermediate goods producers who use domestic intermediate production along with imports to produce final consumption goods that are consumed domestically either by households or government. These producers maximize profit and they are endowed with Leontief production function. Firms maximize their profits choosing the amount of labor,

E0Σt=0Ω0,t[PtyYtPtDyYtDPt*StMt]Ntmax

They face a production constraint

Yt=min([YtD,Mt])

The optimization problem implies that production factors are used in constant shares in production. Furthermore, these shares are not price elastic. Finally, the price of intermediate good, PtY, equals the linear combination of domestic intermediate production prices and import prices. Final good producers are monopolistically competitive firms experiencing a rare opportunity to set prices. The staggered price setting is introduced following Calvo’s setup.

Final goods producers transform intermediate production to final goods exploiting linear production function:

Y¯t(f)=Yt(f)

The f-th producer of final goods transforms Yt(f) of the intermediate product into Y¯t(f) units of final goods. The f-th producer maximizes profit choosing price Pt(f) and Yt(f) facing downward sloping demand for his output and a rare probability 1 – ξp to optimize prices. The optimization problem of the final goods producers can be written as:

E0Σt=0ξpt[Pt(f)Y¯t(f)PtyYt(f)]Pt(f),Yt(f)max

Subject to

Y¯t(f)=(Pt(f)Pt)1+vpvpY¯t

Final goods producers who cannot re-optimize use a full backward indexation on overall price inflation. The linear version of first order conditions for this optimization problem consists of the standard Phillips curve which is partially forward looking. Optimal flexible prices are set as a markup MRK over production costs which equal to Pty.

Δpt=1βΔpt1+β1+βΔpt1+(1ξp)(1ξpβ)ξ(1+β)log(PtYMRKPt)

Distributors

Outputs of final goods producers are collected by fully competitive distributors. These distributors aggregate different types of final goods and sell it as a homogenous product. As final goods are not exported, the final product is divided into household consumption and government spending Gt.

Yt=Ct+Gt

Distributors split the total amount of demanded consumption Y¯t among final goods producers given their prices Pt(f). They seek to minimize total costs:

01Pt(f)Y¯t(f)dfYt¯(f)min

while servicing the total demand

Y¯t=(01Y¯t(f)11+vpdf)1+vp

Where vp determines the price elasticity of substitution among types of goods.

Oil sector—Exporters

There is no optimization in the oil sector. There is only one firm assumed in the model. The firm is owned by the government and it just pumps oil. Real production of the oil firm follows a random walk process.

ΔYtOIL=ρΔYt1OIL+ɛtOIL

As a result, we might think about it as exporters are endowed with a good which price is determined in the international market. The price is not affected by the domestic business cycle. The firm sales all production priced by the world oil price level.

Revenuesoil=Pt*oilStYtOIL

Net foreign assets

Substituting profits and net government transfers into households’ budget constraint, the equation for net foreign asset accumulation can be derived as follows:

Bt=(1+premt1)(1+it1*)Bt1StSt1+PtoilXtoilPtmMt

The net foreign asset position Bt(a positive value denotes borrowing from abroad) is function of the current account where Ptoil denotes export prices of Xtoil, and Mt states for real imports. The net foreign asset position changes along with a foreign nominal interest rate it* adjusted by premium premt and exchange rate appreciation.

References

  • Berg, Andrew, Philip Karam, and Douglas Laxton, 2006, “Practical Model-Based Monetary Policy Analysis—A How to Guide,IMF Working Paper 06/81 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Calvo, Guillermo A., 1983,” Staggered Prices in a Utility-maximizing Framework, Journal of Monetary Economics, Vol. 12(3), 383398.

  • Clarida, Richard, Jordi Gali J, and Mark Gertler, 1999, “The Science of Monetary Policy: A New Keynesian Perspective,Journal of Economic Literature, Vol. 37(4), 16611707.

    • Search Google Scholar
    • Export Citation
  • Gali, Jordi, and Tommasso Monacelli, 2005, “Monetary Policy and Exchange Rate Volatility in a Small Open Economy Model,Review of Economic Studies, Vol. 72, 70734.

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund, 2011, “Target What You Can Hit: Commodity Price Swings and Monetary Policy,” in World Economic Outlook (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Kumhof and Laxton, 2008, “Chile’s Structural Fiscal Surplus Rule: A Model-Based Evaluation,IMF Working Paper 09/88 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Ostry, Jonathan D., Atish R. Gosh, and Marcos Chamon, 2012, “Two Targets, Two Instruments: Monetary and Exchange Rate Policies in Emerging Market Economies,IMF Staff Discussion Note 12/01 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
1

Prepared by Moez Souissi. The model was built in collaboration with Jan Vlcek (Czech National Bank).

2

See 2014 Selected Issues Paper: “Enhancing the Effectiveness of Monetary Policy in Algeria.”

3

A detailed description of micro-foundations of the model and its derivation is provided in Appendix I.

4

Key equations of the model are based on explicit assumptions about the behavior of the main economic actors in the economy (households, firms, and the government). These agents interact in market that clear every period, which leads to the general equilibrium feature of the model.

5

To keep the model tractable, we do not model investment. It would require introducing an additional state variable for capital, and its accumulation. Hence, we simply aggregate investment together with private consumption.

6

Equation (2) describes the dynamics of private absorption and not its level, which is determined by the country’s revenues, particularly from the hydrocarbon sector.

7

σ is a constant intertemporal elasticity of substitution (IES) of consumption.

8

We calibrate β1 to 0.8 which is at the edge of the range of parameter values used in the literature (i.e., between 0.5 and 0.8). This value indicates the high degree of private consumption persistence, consistent with the data for Algeria.

9

The share of government transfers accruing to each household type is determined by its share of total labor hours, and is kept constant.

10

In the absence of any explicit fiscal rule for Algeria, fiscal policy becomes neutral, in that it fully accommodates the oil price shock by adjusting expenditures accordingly.

11

As such, transfers in the model can be interpreted as subsidies net of taxes.

12

Equation (3) implies that that any shock to oil production, ɛtX, will increase or decrease oil exports persistently without returning to the original level, as would be the case of a level equation. The long-term growth rate of hydrocarbon exports is determined by new investments in the sector and fiscal measures that can reduce domestic consumption, such as subsidy reforms.

13

For domestic final good production, we assume a Leontief production function that combines both imported and domestic intermediate goods and services. As a result, the shares of domestic and imported goods in total production are invariant, and therefore the demand for these two types of goods does not depend on their relative prices. For the production of domestic intermediate goods, we assume a Cobb-Douglas production function with labor being the only production factor. The final good production is the sum of private and public consumption.

14

Equation (9) can be written as follows: rmct=ωWtr+(1ω)zt, where wtr is the real wage and zz is the real exchange rate.

15

This specification can be derived from micro-foundations. The real marginal costs represent the natural logarithm of marginal cost in deviation from the price index that maximizes the profit of the representative firm in that sector. We assume that the production of final consumption goods requires both domestic and imported inputs.

16

Setting e2 equal to 0, we capture the case of fully pegged exchange rate to a particular target level st1T. Calibrating parameter e2 different from zero, we assume that the central bank adjusts the level of the exchange rate target in line with the world oil price developments. Such a policy resembles the case of real exchange rate smoothing to avoid external imbalances.

17

We undertook a number of sensitivity analyses assuming different values of α1 and θ1. The results presented below remain qualitatively the same.

18

To simulate a fully flexible exchange rate regime, we set δ1 equal to 0.

19

Implicitly, the central bank pays attention to real economic developments. Such a course may be desirable for restoring employment as wages are relatively inflexible. A higher inflation would help achieve the necessary downward adjustment in real wages faster.

20

To simulate a peg regime, we set δ1 and e2 equal to 1 and 0.25, respectively.

21

Note that interest rates partly increase due to higher country risk premium reflecting the increase in public debt (Equation 21).

1

This feature can be easily changed assuming differentiated types of labor supplied by households. In such a case households exhibits a monopolistic power setting their wages.

Algeria: Selected Issues
Author: International Monetary Fund. Middle East and Central Asia Dept.