Calibration of Fiscal Objectives for Malaysia1
1. The analysis in the paper illustrates the implications of the debt limit for the conduct of fiscal policy. It finds that, under unchanged policies, maintaining public debt below the policy objective over time would require a significant reduction in today’s debt-to-GDP ratio. Otherwise, the policy response to shocks may be more restricted than in the past. The debt reduction can be achieved with a low deficit sustained over the long term.
2. Fiscal deficit reduction is driving Malaysian debt down consistently. As shown in the Debt Sustainability Analysis (DSA) in Appendix IV of the companion Staff Report, the Federal Government Debt is expected to fall to less than 45 percent of GDP by 2022. The debt is expected to fall in the baseline and under a variety of shocks. Even the most severe shocks would lead to sustainable debt levels.
Malaysia: Gross Nominal Public Debt
3. Fiscal policy is currently anchored by debt and deficit targets. The Malaysian authorities are conducting fiscal policy with an objective of achieving a near balance over the medium term. This objective has been the driver of the continuous fiscal deficit reduction over the past several years. Together with this objective, the ceiling of 55 percent of GDP for federal government debt, has been a major anchor for fiscal policy.
Malaysia: Medium-Term Fiscal Outlook
Source: IMP staff estimates.
4. The framework that limits the overall level of public sector debt is based on several laws that limit several aggregates. There is a legal ceiling of 55 percent of GDP for government securities, that is, Malaysian Government Securities (MGS), which are regular securities, and Malaysian Government Investment Issue (MGII) and Malaysian Islamic Treasury Bills (MITB), which are securities based on Islamic legislation. In addition, external debt and Treasury Bills are subject to ceilings of 35 billion ringgit and 10 billion ringgits respectively. The authorities have conveyed their policy of maintaining overall federal government debt below 55 percent of GDP but have not established this objective in legislation.
5. The definition of a fiscal framework for the long term would be very beneficial. A framework that ensures fiscal sustainability and is well communicated to markets would foster investment and growth. By anchoring market expectations, it can increase demand for government securities, reducing interest rates. Lower interest rates can increase domestic investment by reducing the cost of financing. A well-defined fiscal framework can also increase private investment by informing market participants about the level of taxation in the long term. This paper aims at providing a quantitative analysis that can help the authorities in defining long-term policy targets. The analysis is illustrative and complements the advice provided in the companion Staff Report, which factors in the entire macroeconomic policy mix.
A. Debt Targets and Likelihood of Breaching the Debt Ceiling
6. The analysis follows a methodology developed by FAD. A detailed explanation is presented in “How to Calibrate Fiscal Rules: A Primer” by FAD. This stochastic analysis estimates a multivariate normal distribution of the variables involved in a debt accumulation equation. It complements the DSA, which is based on the baseline debt projections and deterministic shocks applied to the baseline. Annual data covering the 20-year period from 1997 to 2016 are used to estimate the distribution of the variables involved in the evolution of public debt. The multivariate normal distribution of a k-dimensional vector of macroeconomic variables can be written as:
with the k-dimensional mean vector
μ = (E[X1,E[X1, ... E[Xk)
and the k × k covariance matrix
Σ = (cov(Xi,Xj)), for all i = 1,2, ... ,k; j = 1,2, ... ,k
7. A set of macroeconomic variables are forecast over a 6-year projection horizon N times by drawing from the calibrated multivariate distribution of macroeconomic variables each year. The N sets of macroeconomic variable forecasts are used to generate N trajectories of the primary balance, using the estimated fiscal reaction function (FRF) and the level of debt in the preceding year. The FRF applicable for either an advanced or emerging market economy is based on the specification of Bohn (1998)2. The coefficients of the FRF are estimated econometrically to capture historical fiscal behavior. The estimation is carried out using separate panels for advanced and emerging market economies, so that the estimated FRF coefficients differ between income groups. The estimated specification is:
pbit = ai + β1pbit-1 + β2ygapitDit + β3ygapit(l – Dit) + pdit-1 + εit
where pbit is the primary balance (as a ratio of GDP) of country i in year t, Dit is debt (as a ratio of GDP), ygapit is the output gap, Dit is an indicator variable taking the value of one when the output gap is positive, ai is the country specific intercept term (fixed effect) and εit is a random error term, εit ~ N(0, Σ2). The FRF allows for an asymmetric response to the output gap, so that the primary balance may deteriorate more when the output gap is negative, than it improves when the output gap is positive (β2>β3). The output gap is projected over the forecast horizon using GDP growth forecasts obtained from simulations (based on the joint distribution of macroeconomic variables) combined with an HP filter to estimate potential output Each type of fiscal reaction function includes a fiscal shock realized each period. The distribution of fiscal shocks is calibrated based on the estimated residuals of the fiscal reaction function; these residuals correspond to the deviations between actual fiscal responses observed (i.e. actual levels of the primary balance) and the fiscal response predicted by the fiscal reaction function within the sample.
8. Annual changes in the primary balance implied by the fiscal reaction function are constrained based on historical experience, to ensure that projected primary balances are realistic. Fiscal shocks can also be added directly in the fiscal reaction function. The distribution of fiscal shocks is calibrated based on estimated deviations between observed fiscal responses (i.e. actual levels of the primary balance) and the fiscal response predicted by the fiscal reaction function within the sample. The N corresponding trajectories of debt (starting at the current debt level) are obtained by the system of simultaneous equations formed by the debt accumulation equation (government budget constraint) and the estimated fiscal reaction function. The debt accumulation equation is:
where dt is debt (as a ratio of GDP), rt is the average effective real interest rate on debt, gt is the real GDP growth rate, pbt is the primary balance (as a ratio of GDP) and SFAt is the stock-flow adjustment (as a ratio of GDP). SFAt is a constant stock flow adjustment for each period t, that could potentially account for the realization of contingent liabilities.
9. The debt target is chosen by adjusting the level of debt consistent with the chosen threshold. If the 95th debt percentile (or other chosen percentile) of the debt ratio distribution in any year over the projection horizon is not sufficiently close to the debt ceiling, the “starting level” of debt is adjusted by a small amount (up to 0.4 percent) and a new iteration is done based on this new “starting level” of debt. The process is repeated until the chosen percentile of the debt level falls into a small margin around the ceiling in any year over the medium-term projection horizon: [Debt95] ∈ [ceiling – 0.4; ceiling + 0.4]. The “starting level” of debt satisfying this criterion is called the debt target, i.e. the level of debt whose projection does not exceed the ceiling with 95 percent likelihood over the medium-term projection horizon. The safety margin is computed as the ceiling minus the debt target.
10. Fan charts can also be used to determine the probability of breaching the maximum debt limit, conditional on any starting level. For example, using the current debt level as the starting level, the fan chart can be used to determine the probability that debt exceeds the debt limit, over the projection horizon. This can be useful to gauge the extent of risk associated with a country’s current debt position. In the following charts, the likelihood is computed for a range of six years from an initial state set at the target level.
11. Staff analysis shows that, under unchanged policies, macroeconomic shocks can drive the overall level of debt above the ceiling. A calibration of the level of debt based on the distribution of macroeconomic and policy shocks observed in Malaysia over the last two decades shows that, at the current level of debt, there is a 25 percent of probability that the debt level could exceed the ceiling of 55 percent of GDP over a 6-year horizon. These probabilities reflect the past behavior of fiscal policy. In other words, the analysis shows that at current debt levels, unless the authorities decide to react differently to shocks than in the past, there is a 25 percent probability that the debt would breach the ceiling. However, in the future, and given the current level of debt and policy commitments, the authorities may choose to respond to shocks differently than in the past, which would help preserve debt below the ceiling.
Malaysia Debt Starting at Current Level
Source: IMF staff calculation
12. There are several policy options to reduce the likelihood of high debt. The authorities can build fiscal space in good times, particularly when output is above potential. Another option is to exercise greater restraint than in the past when reacting to negative shocks. Such a path may be possible, but it may impose some risks or undesirable consequences, as it would mean less support to aggregate demand when output is below potential and could imply less support to vulnerable groups. Risk mitigation strategies are preferred, including the implementation of policies that reduce the frequency of negative shocks and their effects. Such policies could include regulatory or structural measures that would aim at maintaining a sound financial sector, exchange rate flexibility, and building up fiscal and monetary policy buffers, among others.
13. A level of debt lower than projected for the medium term would be consistent with the current ceiling and past policy reactions. According to the calibration done by staff, under unchanged policies, a debt level of 34 percent of GDP would imply that overall debt remains under the ceiling of 55 percent of GDP with 95 percent probability. With such a debt level, fiscal policy can react to shocks in the same way as in the last two decades without imposing additional restrictions to policymaking.
Debt Target 34 Percent
14. Alternatively, a good level of safety can be achieved with a higher debt target. If the authorities build less fiscal space, and target a level of debt of 39 percent of GDP, the likelihood of debt remaining below the ceiling would be 90 percent.
15. Gradual fiscal consolidation would be an alternative and more prudent way to raise the probability of maintaining the debt-to-GDP ratio under the 55 percent ceiling.3 Debt has been falling at current deficit levels and is expected to fall, under staff’s baseline projections for the medium term, to 45 percent of GDP as the deficit is expected to fall to 1.5 percent of GDP. Such a level of debt is associated with a likelihood of 85 percent of remaining below the ceiling. Alternative consolidation paths could also be appropriate depending on the macroeconomic circumstances.
Debt Target 39 Percent Debt Target 45 Percent
B. Operational Targets: Calibration of the Fiscal Deficit
16. There are multiple ways to reach a chosen debt target. For each of the debt targets presented earlier, four alternative methods are followed to relate the debt target to the operational targets. First, the fiscal deficit is calibrated to bring the debt level to its target asymptotically over the long term. Second, the debt target is achieved in the long term after maintaining the deficit constant. Third, the deficit is adjusted gradually from the current level to the level that will be sustained for a longer term. Finally, the deficit is sustained at a level that allows for increased spending in the last part of the period of analysis.
17. The assumptions for the calibration are in line with the baseline and the most challenging of the debt targets presented. This choice helps anchor the analysis in the current state of the economy and illustrates the change in fiscal deficit required to achieve the most challenging scenario discussed above. The starting point is characterized by debt at 51 percent of GDP, a debt target at 34 percent of GDP, long term nominal interest rate at 5.5 percent, long-term nominal growth at 8 percent, an initial primary deficit at 0.8 percent of GDP, and an initial overall deficit at 2.8 percent of GDP.
|Current debt ratio (% GDP)||d0||51%|
|Target debt ratio (% GDP)||d*||34%|
|Long-term nominal interest rate (%)||i||5.5%|
|Long-term nominal growth rate (%)||γ||8.0%|
|Current fiscal balance (% GDP)||b0|
|Primary balance (pb)||−0.8%|
|Overall balance (ob)||−2.8%|
|Convergence period for debt (years)||N||15|
|Initial fiscal adjustment period (years)||T5|
|Period until spending increases in (years)||P||10|
|Incremental spending (% GDP)||δ||0.5%|
18. Sustaining the deficit at a level close to that projected in the medium term in the baseline can achieve a great level of security. A deficit of 2.5 percent of GDP sustained in the long term can take debt asymptotically to the target of 34 percent of GDP. Such a deficit, would bring debt to about 40 percent of GDP in fifteen years. A deficit of 1.9 percent of GDP sustained for fifteen years can bring debt to 34 percent of GDP. If the deficit adjusts gradually over five years from the current level to a level of 1.9 percent of GDP, the same debt level can be reached over the same period. Finally, adding 0.5 percent of GDP of extra spending per year between the tenth and the fifteenth year would require that the deficit be initially at 1 percent of GDP in order to remain within the debt limit.
Figure 1.Malaysia: Balance Calibration Consistent With 34 Percent of GDP Debt Target
Source: Author’s calculations.
19. Under lower growth and higher interest rates the fiscal effort is higher. Assuming 1 percent lower nominal growth and 1 percent higher interest rates would imply that a lower deficit is necessary to reach the debt target, although the difference is less than 0.5 percent of GDP.
20. A credible long-term policy commitment can substitute for fiscal adjustment. A medium-term fiscal framework and credible policy commitments with well calibrated anchors can be beneficial. Sustaining the current level of the deficit for the long term could achieve the same level of debt currently envisaged in the baseline at the end of the projection period, but over a longer period of time. An additional adjustment can create fiscal space for increased future spending.
Figure 2.Malaysia: Sensitivity Analysis
Source: Author’s calculations.
Figure 3.Malaysia: Balance Calibration Consistent With 45 Percent of GDP Debt Target
Source: Author’s calculations.
21. A gradual consolidation is beneficial. The present analysis serves as a validation of the consolidation plans of the authorities. Consolidation would increase the likelihood of maintaining the debt below the ceiling and help build useful fiscal space. Having limited fiscal space restricts fiscal policy response to negative shocks as the debt ceiling could become binding. It is important to note that consolidation efforts should be done when growth is strong. The Staff Report explains staff’s fiscal policy recommendation and the proposed path for the fiscal deficit under the baseline.
22. The analysis is indicative of the degree of fiscal policy flexibility in response to shocks. The likelihood of the debt remaining below the ceiling of 55 percent of GDP varies between 75 and 85 percent by targeting debt levels between 34 and 45 percent of GDP. The illustration takes the debt ceiling as given and serves as an illustration to help guide the authorities in defining long-term fiscal anchors and operational policy objectives.
23. Deficit levels within the range of the baseline projection are consistent with the calibrated debt targets. Different alternatives are possible to gradually achieve the deficit targets and to build space for future spending. A long-term commitment is required to achieve these targets. In order to appropriately manage aggregate demand, these targets should be understood as average deficit over the business cycle.