## 1idnea2022002

## Source details

**Canonical URL:** [1idnea2022002](https://www.imf.org/-/media/files/publications/cr/2022/english/1idnea2022002.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/cr/2022/english/1idnea2022002.pdf.md)
- [Structured JSON version](/-/media/files/publications/cr/2022/english/1idnea2022002.pdf.json)

---

### Regression specification and methodology
- Estimated equation:
  - Y_{ij} = α + β1 ∙ EMP_{ij} + β2 ∙ LMC_{j} + β3 ∙ HC_{j} + β4 ∙ SP_{j} + γ_{i} + ε
    - Y: real GDP growth
    - α: constant term
    - EMP: sectoral employment as a share of the population (by province)
    - LMC: indicator for labor market conditions
    - HC: proxy for human capital
    - SP: variable capturing the severity of the pandemic
    - γ: fixed effects at the sectoral level
    - i: sectoral subscript; j: provincial subscript; ε: error term
- Data and sample construction:
  - Panel with quarterly GDP growth at sector and provincial level.
  - Periods analyzed:
    - Pandemic period: output change from 2019Q4 to 2021Q3.
    - Downturn: 2019Q4 to 2020Q2.
    - Recovery: 2020Q2 to 2021Q3.
  - Seasonal adjustment: X12 method applied to data for 2010 through 2019Q4 to obtain pre-pandemic seasonal factors.
- Variable grouping (one variable chosen from each group):
  - Group 1 (labor market proxy): labor force participation; unemployment rate.
  - Group 2 (human capital proxy): wage level; share of workers in informal sector; years of education; poverty rate.
  - Group 3 (pandemic severity proxy): residential mobility (Google’s Community Mobility Report); COVID-19 cases per capita; COVID-19 deaths per capita.
  - Each Group 3 indicator is calculated as the change over the reference period.

### Main empirical messages and findings
- Multifaceted shock with sectoral winners and losers:
  - Health and communication sectors grew during the peak of the recession and gathered momentum during recovery.
  - Transportation/storage and hotel/restaurant sectors lagged due to mobility restrictions.
  - Large heterogeneity across provinces only partially explained by sectoral composition; mobility dynamics and labor market conditions pre-pandemic also important.
- Importance of pre-recession buffers and financial resilience:
  - Pre-recession policy buffers and financial resilience are associated with reduced scarring.
  - Indonesia’s pre-pandemic fiscal buffers provided sizable fiscal policy space; financial resilience prevented a systemic banking crisis.
- Recessions as opportunities for structural reforms:
  - Well-structured structural reforms and targeted active policies (cash transfers, financial help to viable firms, retraining) can offset scarring and potentially improve medium-term potential growth.

### Sectoral and provincial impact (disaggregated results)
- General patterns:
  - Sectors dependent on presence/movement of people hit hardest; sectors compatible with social distancing less affected or benefited.
  - Primary sectors (agriculture, mining) generally resilient because they require limited physical interaction.
- Sector-level outcomes (as of 2021Q3):
  - Most goods sectors: back to or above pre-pandemic output levels.
  - Transportation/storage: 19 percent below pre-pandemic level.
  - Hotel/restaurant: 14 percent below pre-pandemic level.
  - Health & social work and information & communication sectors gained momentum.
- Provincial heterogeneity:
  - All 34 provinces except Papua saw output on a lower path than pre-pandemic trends.
  - Bali: 2021Q3 GDP down 16 percent from its pre-pandemic level.
  - Papua: stronger growth during the pandemic than suggested by sectoral composition.

### Regression-based determinants of growth during the pandemic
- Sectoral fixed effects:
  - Transportation & storage, hotels & restaurants, business services, and “other services” underperformed across provinces; underperformance concentrated in the early phase of the pandemic and flips sign during recovery.
  - Health & social work and information & communication sector outperformance attributable to the recovery period.
- Sectoral employment share (EMP):
  - Pre-pandemic sectoral employment as share of provincial population: negative association during the downturn; positive during the recovery.
- Labor market conditions (LMC):
  - Lower pre-pandemic labor force participation associated with stronger growth during recovery, suggesting absorption of idle labor.
- Pre-pandemic wages (HC proxy):
  - Higher pre-pandemic wages have a negative association with growth over the pandemic period (not statistically significant when split into downturn or recovery).
- Pandemic severity proxy (SP):
  - Residential mobility increase associated with reduced growth, especially during downturn (2020H1).

### Historical perspective on scarring and international evidence
- Indonesia recessions since 1970s:
  - 1981 recession: started last quarter of 1981, lasted two quarters; fall in level of output vis-à-vis pre-recession trend: about 13 percent.
  - 1987 recession: lasted three quarters; about ½ percent fall in real GDP; no scarring — GDP above pre-recession trend two years after the recession.
  - 1997 recession (Asian financial crisis): seven quarters (19997Q3 to 1998Q4); nominal GDP fall up to 20 percent; fall in level of output vis-à-vis pre-recession trend: about 40 percent.
- Methodology for identifying recessions and quantifying scarring:
  - Applied Harding and Pagan (2002, “BBQ”) to log level of quarterly real GDP (Aslan and others, 2019) with symmetric window parameter: two quarters; minimum phase length: two quarters; minimum cycle length: five quarters.
  - Scarring quantified via Blanchard, Cerutti, and Summers (2015) two-step procedure using real GDP per working-age person (16–64); pre-recession trend estimated excluding two years before peak; post-recovery output loss averaged over 3 to 7 years after trough.
- International sample findings (Aslam and others, 2019):
  - Share of recessions exhibiting scarring/hysteresis:
    - AEs: 82 percent.
    - EMs: 69 percent.
  - Average output loss relative to pre-recession trend:
    - AEs: 8.8 percent.
    - EMs: 6.8 percent.
  - Recessions with banking crises increase average post-recession output loss by about 4.8 percentage points.

### Channels for hysteresis/scarring
- Labor market channel:
  - Longer unemployment duration erodes skills; scarring concentrated in particular worker groups (e.g., seniors, young self-employed).
- Capital and investment channel:
  - Capital obsolescence and subdued investment, including intangible assets, depress productivity.
- Confidence and financial channel:
  - Banking crises raise funding costs, reduce credit availability, and lower net worth, reducing investment and consumption.

### Policy recommendations and priorities
- Preserve and strengthen financial sector resilience:
  - Financial sector resilience is key to avoiding large scarring effects; Indonesia’s pre-pandemic banking strength mitigated impact.
- Maintain fiscal space and use targeted active policies:
  - Fiscal space before recessions diminishes impact and supports active policies during recessions.
  - Targeted support to hard-hit sectors/provinces can provide adjustment time and facilitate reallocation.
  - Active policies should balance costs and benefits and be calibrated to evolving conditions.
- Implement structural reforms effectively and at suitable pace:
  - Successful, well-paced reforms essential for overcoming scarring.
  - Reforms to increase fiscal revenues (broad medium-term revenue strategy, streamlining business income tax, reducing special regimes and discretionary VAT exemptions) would finance high-social-return spending (education, health, infrastructure) and improve allocation of capital and labor.
- Policy sequencing considerations:
  - Job retention policies reduce scarring and mitigate unequal impacts.
  - Reallocation policies supporting job creation can ease labor market adjustment.
  - Pace of labor market reforms should consider state of recovery; some measures could be counterproductive mid-recession.

### Macroeconomic policy recalibration, BI actions, and exit considerations
- Pandemic-era suspension of statutory constraints:
  - Annual budget deficit ceiling of 3 percent of GDP and restriction on monetary budget financing by Bank Indonesia (BI) suspended since early 2020; scheduled reinstatement by 2023.
- Fiscal stance and monetary outlook:
  - Fiscal policy expected to turn moderately contractionary over 2022–2023 to meet 3 percent deficit target; BI financing ends by end-2022.
  - Monetary policy expected to remain broadly accommodative in short term; gradual exit as recovery gathers steam.
- BI crisis response measures:
  - Early interventions in FX spot and DNDF markets and purchases of IDR-denominated government bonds in the secondary market.
  - Policy rate reductions in February and March 2020 (25 bps each).
  - 200-bps reduction in the rupiah reserve requirement ratios for banks (May 2020).
  - BI purchased IDR 473 trillion (3.1 percent of GDP) from the primary market over April-December 2020, and IDR 166 trillion (1.1 percent of GDP) in secondary market purchases in Q1 2020.
  - Banks’ liquid assets rose from about 20 percent at end-April 2020 to about 33 percent at end-November 2021.
  - BI’s additional primary market purchases in 2021: IDR 358 trillion, or 2.1 percent of 2021 GDP.
- Considerations for a gradual exit:
  - Initial focus: absorb excess liquidity “in a measurable and very prudent manner” before raising the policy rate.
  - Preferred initial tools: nimble, easy-to-reverse OMOs (reverse repos); RRR increases possible but with trade-offs.
  - Policy rate increases may reduce market value of BI’s government bond holdings and raise sterilization costs; Indonesian banking sector regulatory capital to risk-weighted assets ratio as of end-September 2021 stands at about 25.2 percent.
  - BI’s primary market purchases cumulative: IDR 616 trillion (4 percent of 2020 GDP) as of end-November 2021; expected cumulative total under KB I–III: about IDR 1,055 trillion (6.8 percent of 2020 GDP) by end-2022.
  - Illustrative cost exercise (“Steep” path): total cumulative cost to BI over 2023−25: about IDR 164 trillion; BI’s interest income estimated at about IDR 122 trillion; net cumulative cost for BI: IDR 42 trillion, equivalent to about 19 percent of BI’s end-2020 capital and reserves (IDR 219.8 trillion); BI’s capital ratio of 8.6 percent at end-2020 would be reduced to about 7 percent under that scenario.

### Alternative illustrative scenarios and policy implications
- Scenario design:
  - Four scenarios constructed to assess uncertainty around baseline and data-dependent policy actions; implemented using an estimated log-linearized model with endogenous FXI rule and Taylor-type interest rule.
  - All numbers reported are deviations from the baseline; shocks are unanticipated and specified quarterly.
- Scenario summaries (deviations from baseline):
  - Scenario 1 — More favorable external conditions:
    - Domestic demand shock: +0.5, +0.5
    - Foreign demand shock: +1, +1
    - Foreign inflation shock: -0.1, -0.1
    - Foreign interest rate shock: +0.05, +0.05
    - Policy implication: faster increase in policy rate than baseline.
  - Scenario 2 — Less favorable external conditions delaying the exit:
    - Financial spread shock: +0.25, +0.25
    - Foreign demand shock: +0.2, +0.2
    - Foreign inflation shock: +0.1, +0.1
    - Foreign interest rate shock: +0.05, +0.05
    - Policy implication: policy rate would fall below the baseline.
  - Scenario 3 — New COVID-19 surge combined with tightened EM financial conditions:
    - Domestic demand shock: -0.20
    - Exchange rate shock: +0.80
    - Financial spread shock: +1.0
    - Foreign demand shock: -1
    - Policy implication: BI intervenes in FX market, lowers policy rate below baseline.
  - Scenario 4 — Integrated policy package including fiscal support:
    - Domestic demand shock: -0.20
    - Exchange rate shock: +0.80
    - Financial spread shock: +1.0
    - Government expenditure shock: +1, +1, +1, +1
    - Foreign demand shock: -1
    - Policy implication: a fiscal package amounting to about 1 percent of GDP can significantly mitigate output losses; preserving monetary policy space if fiscal dominance perceived as low.
- General policy tradeoffs:
  - Need for policy flexibility and coordination.
  - Allowing exchange rate to act as shock absorber except when external financial conditions are disruptive.
  - Use of macroprudential tightening and targeted microprudential measures to address excessive credit growth and reduce premature policy rate hikes.
  - Exceptional circumstances may require fiscal stimulus alongside monetary policy.

### Money market, CBDC considerations, and repo market development
- Rupiah money market overview and COVID-19 impact:
  - Call money and FX swap dominate transactions; repo market still developing.
  - Call money contracted sharply during COVID-19 due to reduced incentives to trade; daily average interbank repo volume: IDR 4.3 trillion in 2021, up from IDR 0.5 trillion in 2020.
  - Large state-owned banks are main liquidity providers; foreign banks supply FX; small banks face volatile liquidity.
  - Banks’ excess reserves rose during pandemic; overnight IndONIA declined rapidly relative to other tenors.
- Blueprint for Money Market Development 2025:
  - Deliverables: market; market infrastructure; payment infrastructure; data digitalization; regulation, licensing and surveillance.
  - Repo market reform priorities: product, participants, pricing, infrastructure (3P+1I).
  - Planned CCP for repo expected in 2022 pending legal and operational alignment.
- CBDC implications:
  - rCBDC (retail) could be a cash-like digital extension to promote inclusion and payment diversity; design choices include access, anonymity, operational hours, and remuneration.
  - wCBDC (wholesale) can improve settlement efficiency, enable direct settlement in central bank money for participants not currently on RTGS, and broaden NBFI participation.
  - Key design trade-offs:
    - Collateral choice: reserves vs HQLA (government bonds) affects reserve impact and central bank balance sheet.
    - Remuneration: non-interest-bearing rCBDC reduces bank disintermediation risk; interest-bearing wCBDC could become a hard floor for money market rates if widely held.
    - Redeemability limits can mitigate deposit outflow risks (examples: caps in Nigeria’s eNaira and PBOC’s E-CNY).
  - Indonesia-specific statistics and considerations:
    - Aggregate liquidity surplus toward central bank: around IDR 650 trillion in September 2021, down from more than IDR 1,000 trillion in August 2021.
    - If banking system has excess liquidity, small conversions to rCBDC unlikely to deplete reserves.
    - rCBDC could be used to absorb excess liquidity and incentivize interbank transactions; wCBDC could increase NBFI inclusion and money market depth.
  - Policy recommendation: further research on CBDC design choices and macro impacts warranted, given ongoing payment digitalization (BI-Fast launched December 2021) and money market reforms.

### Climate commitments, carbon pricing, and green financing
- COP26 commitments and NDCs:
  - Forestry sector pledged to become net carbon sink by 2030.
  - Indonesia exploring plan to reach net-zero emissions in 2060 or sooner; updated NDC (July 2021) kept GHG targets unchanged from 2016:
    - Unconditional GHG reduction target: 29 percent compared to BAU.
    - Conditional reduction target: 41 percent compared to BAU.
  - Forestry targets: peat land restoration of 2 million hectares by 2030; rehabilitation of degraded land of 12 million hectares by 2030.
- Coal and energy mix:
  - Ban on new coal-fired power plants from 2022; phase-out by 2056 at latest (does not apply to already approved plants).
  - RUEN energy mix target by 2025: oil 25 percent; gas 22 percent; coal 30 percent; new and renewable energy 23 percent.
  - Electricity demand growth assumption in RUPTL 2021−2030: 4.4 percent annually.
  - Total power investment needs estimated at around US$9.14 billion per year over 2021−30; PLN expected to contribute US$5.14 billion per year.
- Carbon pricing design and limitations:
  - Carbon tax introduced April 2022 at IDR 30,000 per ton CO2e (about US$2 per ton CO2e); applied on limited basis to coal-fired power plants.
  - ETS planned to start by 2024; design under preparation (pilot included 32 coal-fired power plants).
  - Cap-and-tax design limits incentives for below-cap firms; coverage currently leaves industry and transportation largely outside.
  - Energy subsidies and fixed domestic energy pricing undermine effectiveness of carbon pricing.
  - Modeling: a carbon price of US$25 would reduce GHG emissions by 16 percent and generate revenues of 0.7 percent of GDP.
- Green financing needs and market status:
  - Estimated financing needs:
    - achieving NDC targets: 2.8 percent of GDP annually.
    - infrastructure needs: 6.0 percent of GDP annually.
  - Indonesia’s green bond market: US$5 billion, around 0.5 percent of GDP; market dominated by government bonds and largely denominated in FCY.
  - Recommendations: develop green taxonomy, align disclosure standards internationally, deepen local currency bond market, implement 2017 FSAP recommendations, and maintain sound macro policies to mobilize private green investment.
- Policy recommendations:
  - Promote green financing combined with effective carbon pricing.
  - Redesign carbon pricing: expand participants in ETS; increase carbon tax rates predictably; expand sectoral coverage.
  - Establish institutional frameworks to manage carbon pricing consistently and protect lower income groups.
  - Enhance transparency, deepen financial markets, and maintain macro stability to foster green investment.

*Italic: Indonesia — Chapter excerpt and Box 2 methodology, as contained in the provided IMF chapter PDF content.*

### 1. Regression Specification ____________________________________________________________ 9

### 1. Regression Specification

### A. Introduction
- The COVID-19 pandemic and containment measures severely affected the Indonesian economy.
- Key numeric highlights:
  - After many years of strong growth of around 5 percent per year, output contracted by 2.1 percent in 2020 and is estimated to have expanded by about 3.2 percent in 2021.
  - COVID-19 was first detected in Indonesia in March 2020.
  - Output contracted by nearly 8 percentage points during 2020Q1-Q2.
  - The recovery began with a rebound in 2020Q3, which saw output rise by 3 percent q/q s.a.
- Definitions and objective:
  - Scarring or hysteresis is defined as when the level of output does not recover back to the pre-recession trend.
  - Objective: analyze pandemic impact across sectors and provinces and gauge implications for scarring effects; derive policy implications to reduce expected scarring.

### B. Main empirical messages
- First: Multifaceted shock with sectoral winners and losers
  - Health and communication sectors grew during the peak of the recession and gathered momentum during recovery.
  - Transportation/storage and hotel/restaurant sectors lagged due to mobility restrictions.
  - Large heterogeneity across provinces only partially explained by sectoral composition; mobility dynamics and labor market conditions pre-pandemic also important.
- Second: Importance of pre-recession buffers and financial resilience
  - Historical evidence indicates pre-recession policy buffers and financial resilience are associated with reduced scarring.
  - Indonesia’s pre-pandemic fiscal buffers provided sizable fiscal policy space; financial resilience prevented a systemic banking crisis.
- Third: Recessions as opportunities for structural reforms
  - Well-structured and paced structural reforms, and targeted active policy actions (cash transfers, financial help to viable firms, retraining programs), can help offset scarring and potentially improve medium-term potential growth.

### C. The impact of the pandemic across sectors and provinces — disaggregated view
- General pattern:
  - Sectors dependent on presence/movement of people hit hardest; sectors compatible with social distancing less affected or even benefited.
  - Primary sectors (agriculture, mining) generally resilient because they require limited physical interaction.
- Sector-level outcomes (as of 2021Q3):
  - Most goods sectors: back to or above pre-pandemic output levels.
  - Transportation/storage: 19 percent below pre-pandemic level.
  - Hotel/restaurant: 14 percent below pre-pandemic level.
  - Health & social work and information & communication sectors gained momentum.
- Provincial heterogeneity:
  - All 34 provinces except Papua saw output on a lower path than pre-pandemic trends.
  - Bali: 2021Q3 GDP down 16 percent from its pre-pandemic level.
  - Papua: stronger growth during the pandemic than suggested by sectoral composition.

### D. Empirical analysis — methodology and main findings
- Data and approach:
  - Panel with quarterly GDP growth at sector and provincial level.
  - Explanatory variables: sectoral fixed effects, labor market conditions, mobility changes across provinces.
  - Periods analyzed:
    - Pandemic period: output change from 2019Q4 to 2021Q3.
    - Downturn: 2019Q4 to 2020Q2.
    - Recovery: 2020Q2 to 2021Q3.
- Main regression-based findings (summarized):
  - Sectoral fixed effects:
    - Transportation & storage, hotels & restaurants, business services, and “other services” underperformed across provinces; underperformance concentrated in the early phase of the pandemic and flips sign during recovery.
    - Health & social work and information & communication sector outperformance is attributable to the recovery period.
    - Interpretation: these “high beta” sectors are sensitive to social restrictions; recovery saw resource shifts toward health and ICT.
  - Sectoral employment share (EMP):
    - Pre-pandemic sectoral employment as share of provincial population: negative association during the downturn; positive during the recovery.
    - Interpretation: greater concentration associated with sharper downturn; effect partially reverses during recovery as reallocation pressures diminish.
  - Labor market conditions (LMC):
    - Pre-pandemic labor force participation: low participation associated with stronger growth during recovery, suggesting absorption of idle labor.
  - Pre-pandemic wages (proxy for human capital, HC):
    - Higher pre-pandemic wages have a negative association with growth over the pandemic period (not statistically significant when split into downturn or recovery).
    - Possible interpretation: reallocation of labor from high-wage areas to lower-wage activities or geographic reallocation.
  - Pandemic severity proxy (SP):
    - Residential mobility increase associated with reduced growth, especially during downturn (2020H1).
    - Provinces with greater increases in residential mobility tended to grow more slowly over the pandemic.
- Table reference:
  - Estimation details are summarized in Table 1: "Baseline Estimation Results for the Determinants of Real GDP Growth" (periods and variables as above).

### E. Regression specification (Box 1)
- Estimated equation:
  Y_{ij} = α + β1 ∙ EMP_{ij} + β2 ∙ LMC_{j} + β3 ∙ HC_{j} + β4 ∙ SP_{j} + γ_{i} + ε
  - Variables:
    - Y: real GDP growth
    - α: constant term
    - EMP: sectoral employment as a share of the population (by province)
    - LMC: indicator for labor market conditions
    - HC: proxy for human capital
    - SP: variable capturing the severity of the pandemic
    - γ: fixed effects at the sectoral level
    - i: sectoral subscript
    - j: provincial subscript
    - ε: error term
- Seasonal adjustment:
  - All GDP data are seasonally adjusted using pre-pandemic seasonal factors by applying the X12 method to data for 2010 through 2019Q4.
- Variable grouping (one variable chosen from each group; preferred specification variables shown in italics in original):
  - Group 1 (labor market proxy): labor force participation; unemployment rate.
  - Group 2 (human capital proxy): wage level; share of workers in informal sector; years of education; poverty rate.
  - Group 3 (pandemic severity proxy): residential mobility (Google’s Community Mobility Report); COVID-19 cases per capita; COVID-19 deaths per capita.
  - Each Group 3 indicator is calculated as the change over the reference period.

### F. A historical perspective on scarring — previous Indonesian recessions
- Indonesia experienced three recessions since the 1970s:
  - 1981 recession:
    - Started in last quarter of 1981 and lasted two quarters.
    - Linked to abrupt decline in oil prices.
    - Fall in level of output vis-à-vis pre-recession trend: about 13 percent.
  - 1987 recession:
    - Lasted three quarters (covering 1987Q2 and 1987Q4).
    - Relatively shallow: registering only about ½ percent fall in real GDP.
    - Linked to global stock market shock (Black Monday); there was no scarring impact — GDP above pre-recession trend two years after the recession.
  - 1997 recession (Asian financial crisis):
    - Covered seven quarters (from 19997Q3 to 1998Q4).
    - Nominal GDP fall of up to 20 percent along with a systemic banking crisis.
    - Fall in level of output vis-à-vis pre-recession trend: about 40 percent — the deepest recession since the 1970s.

_Italic: Prepared by Eugenio Cerutti, Robin Koepke, and Rani Setyodewanti (all APD)._

### 11.      While the recessions of 1981 and 1997

### 1idnea2022002 - 11.      While the recessions of 1981 and 1997

### Scarring in Past Recessions: Indonesia evidence and interpretation
- The recessions of 1981 and 1997 display hysteresis/scarring; the 1987 recession does not.
- The absence of scarring after 1987 is linked to its shallowness and short duration and to an observed increase in the GDP trend growth just after 1987.
- The post-1987 improvement may be linked to structural reforms that accelerated in 1987:
  - Broad deregulation of the economy process (Halim, 1988).
  - Start of a trade liberalization process (Fane and Condon, 1996).
  - These reforms facilitated a rapid export-led growth process.

### Pandemic-era scarring projections and recent studies for Indonesia
- IMF Article IV staff projections for real GDP growth do not forecast a loss in medium-term growth rates, but do project a loss in the output level of about 6 percent in the medium term relative to the level envisaged in the January 2020 WEO.
- Other studies highlighting potential scarring effects in Indonesia:
  - World Bank (2021): the crisis could damage Indonesia’s growth potential through lower investment, weak productivity growth, and loss in human capital.
  - Pritadrajati (2021): long-term scarring effects due to unemployment and informal self-employment; scarring effects were significant among senior workers who were unemployed, and among young workers who were self-employed.
- Footnote clarification: "In other words, Indonesia would need to growth at about 7 percent per year for at least 5 years to reach back to a pre-pandemic level trend that had imbedded about 5 percent growth per year."

### Methodology for identifying recessions and quantifying scarring (Box 2)
- Identifying a recession:
  - Applied the non-parametric methodology of Harding and Pagan (2002, “BBQ”) to the log level of quarterly real GDP as compiled in Aslan and others (2019).
  - BBQ settings used:
    - Symmetric window parameter: two quarters.
    - Minimum length of the phase for both expansions and contractions: two quarters.
    - Minimum complete length for a cycle (expansion plus contraction): five quarters.
- Quantifying scarring effects:
  - Followed Blanchard, Cerutti, and Summers (2015) two-step procedure using real GDP per working-age person (GDP over population aged 16 to 64).
  - Step 1: Estimate pre-recession trend recession-by-recession using simple exponential trends and exclude the two years before the peak from computation of the trend (to account for pre-recession booms).
  - Step 2: Compare the real GDP level following the subsequent recovery with the pre-recession trend; post-recession output loss measured as an average over a window of 3 to 7 years after the trough.

### International evidence on recessions and recoveries
- Data and sample:
  - Rely on Aslam and others (2019) calculations for a sample of 23 advanced economies (AE) and 27 Emerging Markets (EM) over 1960Q4-2016Q4.
  - Sample includes 255 recession-and-recovery episodes.
- Key empirical findings:
  - Share of recessions exhibiting scarring/hysteresis:
    - AEs: 82 percent.
    - EMs: 69 percent.
  - Average output loss relative to pre-recession trend:
    - AEs: 8.8 percent.
    - EMs: 6.8 percent.
  - Distributional differences:
    - EMs show more varied outcomes with fatter tails; more outcomes in the left tail denote very weak recoveries, and also a larger right-tail mass consistent with stronger recoveries linked to structural reforms.
- Banking crises and scarring:
  - Recessions coinciding with banking crises are deeper and longer lasting.
  - If a recession is associated with a banking crisis, the average post-recession output loss increases by about 4.8 percentage points, with no discernable EM-specific difference in impact.
  - Mechanisms cited include higher funding costs, reduced credit availability, declines in firms’ and households’ net worth, and confidence effects that reduce investment and consumption.
  - Literature support: Cecchetti, Kohler, and Upper (2009); Caldara and others (2016); Jorda and others (2011, 2013); Reinhart and Rogoff (2009); Teulings and Zubanov (2014); Aghion and others (2010, 2012).

### Channels for hysteresis/scarring
- Labor market channel:
  - Longer unemployment duration erodes skills; job losses may be concentrated in higher-skill sectors.
  - Scarring disproportionate across worker groups (e.g., seniors unemployed; young self-employed).
- Capital and investment channel:
  - Part of the capital stock may become obsolete.
  - Subdued or uncertain growth prospects reduce investment, including intangible assets, depressing productivity growth.
  - Tight credit constraints after financial crises imply procyclical investment in intangible assets (Aghion and others, 2010, 2012).
- Confidence and financial channel:
  - Financial crises raise funding costs, reduce credit availability, and lower net worth, which together reduce investment and consumption.

### Policy recommendations (three priorities)
- Preserve and strengthen financial sector resilience:
  - Financial sector resilience is key to avoiding large scarring effects.
  - Banking crises have very important impacts on scarring in both AEs and EMs.
  - Indonesia’s pre-pandemic banking sector strength and efforts to strengthen supervision and regulation have mitigated the pandemic recession’s impact.
  - Going forward, preserving financial stability is key to avoiding deeper future costs.
- Maintain fiscal space and use targeted active policies:
  - Availability of fiscal space before recessions diminishes their impact and is key for supporting financial resilience and active policies during recessions.
  - Targeted policy support to hard-hit sectors and/or provinces could help adjustments by providing time and facilitating resource reallocation.
  - Active policies should balance costs and benefits and be calibrated to evolving conditions.
- Implement structural reforms effectively and at a suitable pace:
  - Successful and well-paced implementation of structural reforms is essential for overcoming scarring in the medium to long term.
  - Recessions can provide opportunities for structural reforms; Indonesia enacted an omnibus bill on job creation.
  - Implementation and complementarities of reforms are key; market and labor reforms could be complemented with measures to enhance governance and address weak access to finance through financial deepening.
  - Reforms to increase fiscal revenues (e.g., a broad medium-term revenue strategy, streamlining business income tax structure, reducing special regimes and discretionary exemptions in the VAT system) would provide medium-term finance for high-social-return spending (education, health, infrastructure) and improve allocation of capital and labor for increased productivity.
- Additional notes on policy sequencing:
  - Job retention policies are powerful at reducing scarring and mitigating unequal impacts (Chapter 3, April 2021 WEO).
  - Reallocation policies supporting job creation can ease labor market adjustment to COVID-19’s more permanent effects.
  - Pace of labor market reforms should consider state of recovery; some measures could be counterproductive in the middle of recessions (April 2016 WEO).
  - Reforms in governance, domestic and external finance, trade, and labor and product markets can deliver sizable medium-term output gains (October 2019 WEO).

*Source: Indonesia — Chapter excerpt and Box 2 methodology, as contained in the provided IMF chapter PDF content.*

### 1.      With the economic recovery gathering pace in Indonesia, there is a growing need to

### With the economic recovery gathering pace in Indonesia, there is a growing need to recalibrate current macroeconomic policy settings.

### Background: pandemic-era policy adjustments and reinstatement of framework
- Since early 2020, policymakers suspended two central pillars of Indonesia’s macroeconomic policy framework: the annual budget deficit ceiling of 3 percent of GDP and the restriction on monetary budget financing by Bank Indonesia (BI).
- Under current laws, these pillars will be reinstated by 2023.
- Fiscal policy stance is expected to turn moderately contractionary over 2022–2023 as the fiscal deficit is lowered to comply with the 3 percent deficit target, and BI financing will end by end-2022.
- Monetary policy is expected to stay accommodative in the short term, with a gradual exit needed as the recovery gathers steam.

### BI’s monetary framework and crisis response
- BI’s framework: “Flexible Inflation Targeting” with emphasis on exchange rate and financial stability, use of macroprudential and capital flow measures, ample foreign reserves, bilateral swap lines, and a domestic non-deliverable forward (DNDF) market (introduced in 2018).
- Early pandemic actions (Q1 2020):
  - Active interventions in FX spot and DNDF markets and purchases of IDR-denominated government bonds in the secondary market (“triple intervention”).
  - Consecutive policy rate reductions in February and March (25 bps each).
- From April 2020 onward, policy priority shifted to monetary easing to support domestic activity:
  - 200-bps reduction in the rupiah reserve requirement ratios for banks (May 2020).
  - Increases in BI repo and FX swap issuance.
  - Easing of macroprudential measures.
- Unconventional measures to finance fiscal response:
  - Additional government spending widened the budget deficit to 5.9 percent of GDP from 1.8 percent in the initial budget.
  - Perppu No. 1 of 2020 (March) allowed BI to purchase long-term government bonds in the primary market.
  - BI purchased IDR 473 trillion (3.1 percent of GDP) from the primary market over April-December 2020, and IDR 166 trillion (1.1 percent of GDP) in secondary market purchases in Q1 2020.
- 2020–2021 outcomes:
  - Banks’ liquid assets rose from about 20 percent at end-April 2020 to about 33 percent at end-November 2021.
  - BI’s additional primary market purchases in 2021: IDR 358 trillion, or 2.1 percent of 2021 GDP.
  - Policy rate cut by 25 bps in February 2021.
  - Easing of loan-to-value ratio on property and car loans to up to 100 percent until end-2022.

### Baseline scenario (staff)
- Recovery is expected to strengthen in Indonesia and globally over 2022−23, despite headwinds from U.S. monetary policy tightening.
- Expected effects of U.S. monetary policy turnaround in 2022:
  - Gradual and orderly rise in global long-term rates and depreciation of EM currencies, including the rupiah.
  - Financial conditions in Indonesia and other EMs would tighten somewhat, but adverse domestic demand impact expected to be largely offset by stronger exports.
- Planned return to the 3-percent budget deficit ceiling in 2023 will impose a modest drag on growth over 2022−2023, but this is expected to be more than offset by rebound in domestic and external demand amid continued vaccination progress and easing of the pandemic by end-2022.
- IDR-denominated government bond market outlook:
  - Domestic banks expected to unwind some government bond holdings in favor of more private sector credit.
  - Discontinuation of BI’s primary market government bond purchases by end-2022.
  - Long-term yields expected to rise gradually over 2022−23 along with other EM yields under stable market conditions.
- Monetary policy: expected to remain broadly accommodative in the short term given a projected output gap of about 3½ percent in 2022, well-anchored inflation expectations, Omicron uncertainty, and expected fiscal consolidation.

### Considerations for exit under baseline
- BI intends a gradual exit involving adjustments in policy mix and stance to achieve price, currency, and financial stability.
- Initial exit focus: absorb abundant excess liquidity “in a measurable and very prudent manner” before raising the policy rate in response to “early signals of rising inflation.”
- Discontinuation of BI’s monetary budget financing by end-2022 (under Law 2 of 2020) will aid liquidity absorption and signal exit from unconventional tools.
- Sequencing and tool choice:
  - Preference for nimble, easy-to-reverse tools initially: open market operations (OMOs) such as reverse repos.
  - Reserve requirement ratios (RRRs) could be raised to absorb excess liquidity; advantages: no additional sterilization costs under current setup and better suited for structural liquidity absorption.
  - Trade-offs: raising RRRs risks discouraging intermediation (acts as a tax on lending) and is harder to change frequently, making it more suitable at later stages.
- Effects of policy rate increases:
  - Potential adverse effects: reduced market value of banks’ fixed-income assets (notably government bonds) and possible debt service problems in loan portfolios.
  - Offsetting effects: higher net interest margin that can improve banks’ profitability.
  - Indonesian banking sector well-capitalized: regulatory capital to risk-weighted assets ratio as of end-September 2021 stands at about 25.2 percent.
  - Overall financial stability impact likely limited if adjustments are measured and gradual.
- BI’s balance sheet considerations:
  - Policy rate hikes would reduce market value of BI’s government bond holdings and raise costs for liquidity absorption via OMOs.
  - BI transfers part of its interest income to the Ministry of Finance under burden sharing agreements, imposing additional financial burden on BI.
  - Appendix I (in source) examines illustrative interest rate paths and finds costs mostly manageable.
- Exchange rate and macroprudential policies:
  - Allowing ample exchange rate flexibility can facilitate keeping the policy rate supportive of recovery while preserving price stability.
  - Tightening macroprudential measures (and targeted microprudential measures) can address excessive credit growth and reduce need for premature policy rate hikes.
- Unwinding BI’s government bond holdings:
  - Preferred approach: long-term, organic unwinding by not rolling over maturing bonds.
  - International experience favors passive runoff over outright asset sales due to price impact and limited effectiveness of sales relative to policy rate adjustments.
  - If needed, BI could partly roll over maturing bonds through secondary market purchases.

### Policy response under alternative scenarios and model simulations
- Indonesia relatively well positioned to weather external shocks due to:
  - Ample FX reserves.
  - Current account estimated at close to balance in 2021.
  - Relatively low public debt ratio.
  - Improved capacity to deal with COVID-19 infections.
  - Long-term local-currency government bond yield maintains a comfortable margin over the global benchmark.
- Risks requiring policy readjustment:
  - Faster-than-expected U.S. monetary tightening could have differing global impacts depending on drivers:
    - If driven by stronger U.S. growth: positive trade spillovers could offset tighter global financial conditions.
    - If driven by U.S. inflation surprises from global supply-side disruptions: net impact on emerging markets could be negative.
  - Slower-than-expected U.S. tightening could occur if another global infection wave emerges, requiring deviations from the baseline exit path.

*Prepared by Minsuk Kim (APD) and Hou Wang (MCM), with inputs from Agnes Isnawangsih (APD), Rani Setyodewanti, and Wahyu Ari Wibowo (RRO Jakarta).*

### 20.      Four illustrative scenarios are constructed to emphasize the uncertainty of shocks

### Four illustrative scenarios are constructed to emphasize the uncertainty of shocks around the baseline scenario and to highlight the data-dependent nature of future policy actions

### Purpose and method
- Scenarios are constructed to emphasize uncertainty around the baseline and to highlight the data-dependent nature of future policy actions.
- Scenarios are based on different assumptions of U.S. monetary policy spillovers and the pandemic (Appendix II, Table 2).
- All numbers are deviations from the baseline.
- Policy instruments in the simulations are much narrower and more simplified than in the real world; the exercise is intended to shed light on policy tradeoffs to external shocks.
- The scenarios are implemented using an estimated log-linearized formulation of Adrian and others (2021) with an endogenous FXI rule and a Taylor-rule-type interest rate reaction function (Appendix II).
- The model is estimated by Bayesian likelihood methods using Indonesia data during 2003Q4-2020Q3, conditional on a pre-estimated foreign economy model using the U.S. as proxy.

### Scenario summaries (deviations from baseline)
- Scenario 1 — More favorable external conditions
  - U.S. interest rates increase as a reflection of a stronger-than-expected U.S. economy.
  - Positive spillovers to Indonesia through higher exports offset contractionary effects from higher interest rates.
  - Supply-driven inflationary tensions ease, causing U.S. inflation to fall.
  - Indonesia: output gap narrows as domestic demand and net exports strengthen.
  - Inflation rises but only slightly, reflecting a rather flat Phillips curve.
  - Monetary policy implication: faster increase in the policy rate than in the baseline.
  - Impact on the exchange rate and long-term interest rates: small.

- Scenario 2 — Less favorable external conditions delaying the exit
  - Continued supply-side disruptions and higher commodity prices weigh on the global economy.
  - Inflation in major economies including the U.S. surprises on the upside; the Fed tightens monetary policy faster than expected.
  - Higher U.S. interest rates and tighter financial conditions trigger capital outflows, weakening of EM currencies, and higher long-term interest rates.
  - Tighter financial conditions without positive spillovers imply weaker-than-expected output; policy rate would fall below the baseline.

- Scenario 3 — New COVID-19 surge combined with tightened EM financial conditions
  - Assumes another round of mobility restrictions and reduced confidence; output falls below the baseline in Indonesia and the rest of the world.
  - U.S. monetary policy: looser than expected.
  - EM financial conditions tighten more than in Scenario 2, with larger exchange rate depreciation and higher long-term interest rates due to higher risk premia.
  - BI responds by intervening in the FX market, limiting depreciation at the cost of some decline in BI’s foreign exchange reserves.
  - BI lowers the policy rate below the baseline to keep inflation within BI’s target band at 2−3 year horizons.

- Scenario 4 — Integrated policy package including fiscal support
  - Illustrates benefits of combining fiscal policy with monetary policy under adverse shocks.
  - In Indonesian context: delayed return to the budget deficit ceiling of 3 percent to provide additional support.
  - Assumes a fiscal package amounting to about 1 percent of GDP, financed by an increase in the fiscal deficit and issuance of new government bonds at market price.
  - The package significantly mitigates the impact on output from the pandemic.
  - If market participants consider the risk of fiscal dominance is low, the fiscal package may only imply small increases in the long-term interest rate.
  - Use of fiscal space helps preserve monetary policy space when external stability concerns are heightened.

### Policy tradeoffs and general implications
- Need for policy flexibility and coordination under the current uncertain external environment.
- Allowing the exchange rate to act as a shock absorber is important under most circumstances, except when external financial conditions become disruptive and warrant FX interventions.
- Policy rate adjustments needed to preserve price stability:
  - Could require a faster hike path than the baseline (Scenario 1).
  - Could require a slower and more gradual path or even lowering the policy rate when output gap widens (Scenarios 2−4), with 2- to 3-year-ahead inflation possibly falling below the baseline (e.g., Scenarios 2 and 4).
  - If inflation surprises on the upside despite a wider output gap (possibly due to higher exchange rate passthrough), BI must respond with a higher policy rate path to anchor inflation and expectations.
- Exceptional circumstances (e.g., new COVID-19 variant) may make monetary policy alone insufficient to sustain recovery; fiscal stimulus may be useful to provide additional demand support.

### Key model features and calibration
- Interest rate rule: i_t = ρ i_{t−1} + (1−ρ)[(1+γ_π) π̅_{c,t+4|t} + γ_y y_t] + ε^i_t.
- FX intervention rule: ΔR_t = ρ_{ΔR} ΔR_{t−1} − (1−ρ_{ΔR}) ΔS_t γ_{ΔS}/(1−γ_{ΔS}) + ε^{FXI}_t.
- FXI proxy: change in central bank foreign exchange reserves (ΔR_t); if central banks use FX swaps that do not affect reserves, those interventions are not captured.
- Exchange rate pass-through estimated to be modest, reflecting BI’s monetary policy credibility.

### Scenario assumptions (Table 2: size of each unanticipated quarterly shock)
- Scenario 1
  - Domestic demand shock: +0.5, +0.5
  - Foreign demand shock: +1, +1
  - Foreign inflation shock: -0.1, -0.1
  - Foreign interest rate shock: +0.05, +0.05
- Scenario 2
  - Financial spread shock: +0.25, +0.25
  - Foreign demand shock: +0.2, +0.2
  - Foreign inflation shock: +0.1, +0.1
  - Foreign interest rate shock: +0.05, +0.05
- Scenario 3
  - Domestic demand shock: -0.20
  - Exchange rate shock: +0.80
  - Financial spread shock: +1.0
  - Foreign demand shock: -1
- Scenario 4
  - Domestic demand shock: -0.20
  - Exchange rate shock: +0.80
  - Financial spread shock: +1.0
  - Government expenditure shock: +1, +1, +1, +1
  - Foreign demand shock: -1
- Note: Each number represents the size of the shock in a given quarter. All shocks are unanticipated shocks.

### Relevant fiscal/financial magnitudes from Appendix I (monetary budget financing)
- BI’s primary market purchases of IDR-denominated government bonds: cumulative IDR 616 trillion (4 percent of 2020 GDP) as of end-November 2021.
- Expected cumulative total purchase amount under KB I–III: about IDR 1,055 trillion (6.8 percent of 2020 GDP) by end-2022.
- Of this amount, about 79 percent (IDR 837 trillion) feature a coupon rate equal to the variable 3-month reverse repo rate; remaining amount IDR 218 trillion purchased under the Market Mechanism.
- Immediate market reactions to KB announcements:
  - KB II and III: IDR/USD exchange rate depreciated by 28 and 14 basis points (bps) on a daily basis, respectively.
  - KB I: 128 bps depreciation on announcement day.
  - 10-year government bond yield declined by about 10−13 bps following announcements.
- Illustrative cost exercise for KB II and III under 2023−25 policy paths:
  - Most hawkish scenario (“Steep”): policy rate rises from current 3.5 percent to 6 percent by end-2025 while sterilizing 70 percent of liquidity injection.
  - Total cumulative cost to BI over 2023−25: about IDR 164 trillion.
  - Of that, IDR 79 trillion due to revenue transfers to the MOF.
  - BI’s interest income from these government bonds estimated at about IDR 122 trillion.
  - Net cumulative cost for BI: IDR 42 trillion, equivalent to about 19 percent of BI’s end-2020 capital and reserves (IDR 219.8 trillion).
  - With this cost, BI’s capital ratio of 8.6 percent at end-2020 would be reduced to about 7 percent.

*Source: IMF staff report chapter text and appendices (scenarios, model description, and illustrative fiscal/monetary financing figures).*

### Chapter 3 in Global Financial Stability Report, April 2013: Old Risks, New Challenges

### Chapter 3 — The Rupiah Money Market in Indonesia: Recent Evolution and Implications of Introducing a Central Bank Digital Currency

### A. Money Market in Indonesia: An Overview
- The money markets consist of the rupiah market, the foreign currency market, and their derivative markets; they are where short-term liquidity is traded.
- The size of the rupiah money market has grown rapidly since 2016; daily average volume and outstanding transactions have steadily increased since 2016.
- During the COVID-19 pandemic the market experienced a sharp contraction in volume, mostly driven by a decline in interbank call money market transactions as banks’ incentives to trade liquidity decreased.
- Market composition and instruments:
  - Transactions are dominated by interbank transactions; call money and FX swap are the mostly used instruments.
  - Call money: easy to conduct, largest number of bank participants, unsecured, predominantly overnight or up to 1 week tenors.
  - FX swap: most liquid instrument in almost all tenors from overnight to 12 months; used for liquidity management and to hedge FX risk.
  - Repo market: still developing; users fewer than call money and mostly banks and some licensed brokerage companies. Non-bank financial institutions (insurance and pension funds) are not using repo due to regulation and tax reasons.
  - Most repo transactions use government bonds as underlying securities; other acceptable instruments include BI certificate, BI deposit certificate, sukuk BI, and sovereign bonds in foreign currencies.
  - Despite lower credit risk from collateral transfer, repo instruments in Indonesia are priced at a similar, if not higher, rate than unsecured interbank borrowing for the same maturity (Aditya, 2021).
- Counterparty and size dynamics:
  - Large state-owned banks are the main liquidity providers due to better access to deposit funding and network infrastructure.
  - Foreign banks have stable foreign exchange supply but lack rupiah liquidity and frequently access the money market.
  - Small banks face more volatile liquidity and borrow from large banks and among themselves; large and medium-sized banks tend to trade among themselves rather than lending to small banks.
- Banks’ excess reserves and liquidity management:
  - Banks’ excess reserves are relatively insensitive to their opportunity cost, likely reflecting precautionary savings motive and less efficient liquidity management through the money market.
  - A simple VAR for January 2017–December 2019 (weekly frequency, detrended with a time trend) shows that a positive shock to excess reserves leads to a temporary reduction in the spread between IndONIA and the deposit facility rate, while a positive shock to the spread has no meaningful impact on excess reserves.
  - Distribution by bank size: BUKU 4 banks hold the most excess liquidity in the system in aggregate; normalized by size, BUKU 4 banks’ excess liquidity relative to total assets is about 8.5 percent, somewhat lower than other banks on average. BUKU 2 group displays high heterogeneity and dispersion in excess liquidity positions.

### B. Recent Evolution: Impact of the COVID-19 Pandemic
- COVID-19 effects:
  - Contraction in the size of the rupiah market, mainly from decline in call money transactions due to lower incentives to participate.
  - Ample liquidity from weak loan demand and declining loan disbursement reduced the need for some banks to borrow.
  - Flight to safety increased perceived credit risk and reduced interbank lending; FX swap transactions saw a similar but smaller contraction.
- Policy support and liquidity injections:
  - BI injected large amounts of liquidity via a 200 bps reduction in the statutory reserve requirement in April 2020 and by buying government bonds in primary and secondary markets.
  - The government placed funds in selected banks as part of fiscal response; with suppressed credit growth, only part of funds was channeled to the real economy, resulting in large excess liquidity.
  - Reflection of excess: rapid decline of the overnight JIBOR (IndONIA), much more than decline in other JIBOR tenors.
- BI’s open market operations:
  - BI significantly increased open market operations to absorb excess liquidity.
  - Banks’ placement at BI (demand deposit, fine tune operations, and deposit facility) increased markedly between end-2019 and August 2021.
  - Larger banks (BUKU 4) were more engaged in BI’s open market operations than smaller banks, reflecting more favorable liquidity positions and greater repo expertise.

### C. Towards a Modern Money Market
- Importance of development:
  - A well-developed money market provides alternative funding, distributes short-term liquidity efficiently, facilitates price discovery, mobilizes new savings, and supports other financial market segments.
  - Money markets can also amplify shocks due to short-term instruments and linkages with the broader financial system (example: U.S. money market stress in March 2020).
- Monetary policy transmission:
  - Money market development is key to effective monetary policy; money markets are the first link in transmission and allow the central bank to keep short-term interest rates at or near its operational target.
- Reforms and operational framework:
  - BI reverted to a mid-corridor system in 2017 with the 7-day reverse repo rate as the operational target.
  - Operational outcome: alignment of the overnight interbank rate with the operational target—the midpoint of the 100 basis-point corridor—except in the post-COVID-19 period.
  - Reserve averaging mechanism in place since 2017 intended to increase banks’ incentives to trade and align with international best practices.
- Payment systems and spillovers:
  - Advanced digital payment methods can lower transaction costs, reduce risk, speed processing, and free up liquidity.
  - BI launched BI-Fast as a real-time, 24/7 retail payment infrastructure in December 2021.
  - Plan to upgrade RTGS by 2025, including multi-currency feature and strengthened risk management.
- Blueprint for Money Market Development 2025:
  - Aims: establish reliable infrastructure, offer a variety of instruments, foster high market integrity, and create a well-informed investor base.
  - Progress on five deliverables: market; market infrastructure; payment infrastructure; data digitalization; regulation, licensing and surveillance.
  - Market infrastructure: 2021 implementation of Multimatching ETP for FX spot; conceptual design of BI-SSSS (Gen III) and trade repository.
- Repo market development (3P+1I: product, participants, pricing, and infrastructure):
  - Daily average interbank repo volume: IDR 4.3 trillion in 2021, up from IDR 0.5 trillion in 2020.
  - Market education: BI uses continuous market education and moral suasion; during the pandemic BI encouraged repo use and big banks provided knowledge transfer to small banks on repo accounting, settlement, and infrastructure.
  - Regulation: intention to strengthen framework (close-out netting regulation, financial sector omnibus bill, repo taxation regulation) to broaden participants, especially NBFI and large corporations. New clauses in harmonization tax law aim to equalize tax treatment between short-term and long-term instruments and between banks’ repo and nonbanks’ repo; further regulations planned for technical tax implementation.
  - Infrastructure: focus on standardizing repo trading via Electronic Trading Platform (ETP) and settlement via a central counterparty (CCP). Post-trade improvements target transparency and straight-through processing. CCP for repo expected in 2022, pending business model and IT design, alignment across authorities, and legal challenges on close-out netting.

### D. Implications of Introducing CBDC in Money Market
- CBDC overview and design choices:
  - CBDC defined as a central bank’s liability denominated in the local currency, usable as medium of exchange and store of value.
  - Potential benefits: promote diversity in payment options, increase financial inclusion, facilitate cross-border payments.
  - Design choices include access (widely vs restricted; retail CBDC vs wholesale CBDC), degree of anonymity, operational availability (hours/days), and interest-bearing characteristics (yes or no) (CPMI, 2018).
  - As of the source, two countries formally introduced CBDC fully open to the public (Bahamas and Nigeria); other countries/regions (China; Eastern Caribbean Currency Union) are running pilots open to a smaller set of the population.
- Indonesia’s payment landscape and cash use:
  - Cash as a share of nominal GDP fell very slowly from the early 2000s until the global financial crisis, after which it rose steadily; cash is losing importance relative to alternatives such as cards and e-money.
  - Aggregate value of e-money transactions reached around IDR 205 trillion as of 2020 and has risen exponentially since 2010 (Statista, 2021).
  - Payment method shares in e-commerce payments: digital/mobile wallet nearly 30 percent; bank transfers 23 percent; cash on delivery 15 percent; credit card 14 percent; debit card 10 percent.
  - Consumer concerns about digital payments (Statista, 2021): security 59 percent; more comfortable with cash 49 percent; concern over scams 36 percent.
- Potential role for a retail CBDC (rCBDC) in Indonesia:
  - Indonesia has potential for a secured, cash-like rCBDC to play an important role in the retail payment landscape.
  - A rCBDC could explore addressing remaining concerns (security, comfort with cash, scams) and expand digital payment options, potentially supporting financial inclusion and payment-system resilience.

*Prepared by Tao Sun and Hou Wang (both MCM), with inputs from Darryl King (MCM), John Kiff, Rani Setyodewanti and Wahyu Ari Wibowo (both RRO in Jakarta). Kaili Chen (MCM) provided research assistance.*

### 19.      On the other hand, wholesale CBDC (wCBDC), whose access is more restricted and

### 1idnea2022002 - 19.      On the other hand, wholesale CBDC (wCBDC), whose access is more restricted and

### wCBDC and settlement efficiency
- wCBDC, with access mainly restricted to financial institutions, may improve financial market settlement efficiency.
- Central bank reserves can be interpreted as a wholesale form of CBDC used exclusively for interbank payments (Mancini-Griffoli and others, 2018).
- wCBDC comparable to traditional central bank reserves in interbank systems could potentially reduce cost and improve liquidity management (CPMI, 2017).
- Settling outright or repo transactions using wCBDC on both securities and cash legs via an integrated or single ledger could reduce speed and complexity versus Indonesia’s current system, which involves two FMIs (RTGS and securities settlement system).
- Caveat: If transaction speed in existing FMIs further increases, the marginal gains from wCBDC are uncertain.

### Key design choices: collateral, remuneration, redeemability
- Collateral
  - Two theoretical options: issue CBDC against reserves or against high-quality liquid assets (HQLA) such as government bonds.
  - Issuing CBDC against reserves could lead to more volatile market interest rates and monetary conditions, mitigated by more frequent fine-tuning OMOs.
  - Example mechanics: conversion of reserves into CBDC raises loan-to-reserve ratios; banks may borrow in the money market, pushing money market rates higher and increasing call market transaction volume; extreme cases could lead to bank disintermediation.
  - Issuing CBDC against HQLAs could imply less impact on reserves but expands the central bank’s balance sheet and requires portfolio composition decisions consistent with the central bank’s risk appetite.
  - Increased demand for government bonds as HQLA could push up bond prices and lower yields.
- Remuneration
  - CBDC can be non-interest bearing or interest bearing.
  - Central banks may prefer non-interest-bearing rCBDC to avoid pressure on commercial bank deposits and risk of bank disintermediation.
  - wCBDC can, in principle, be interest-bearing and attractive as a liquid, risk-free asset facilitating final settlement.
  - If institutional investors can hold wCBDC without limits, the wCBDC remuneration rate could become the hard floor for money market rates.
  - Risk: setting wCBDC interest higher than reserves could reduce reserves, force banks to borrow in the money market, and shift funds out of other short-term instruments into wCBDC, pushing those yields higher.
- Redeemability
  - Redeemability = ability to exchange CBDC for cash, deposits, reserves.
  - Limits/caps on redeemability (e.g., caps on reserve-to-wCBDC exchange per period; selecting well-capitalized banks as rCBDC distributors) can mitigate bank disintermediation risks, especially as transitional measures.
  - Caps on conversion from bank deposits to rCBDC (daily transaction and/or balance limits) can mitigate deposit outflows; examples: Central Bank of Nigeria’s eNaira and PBOC’s E-CNY have such caps.

### Retail vs wholesale CBDC (disintegrated approach)
- Running rCBDC and wCBDC as disintegrated (non-convertible) systems can allow tailored design features for different policy objectives.
- Example design choices:
  - rCBDC: digital extension of cash; commercial banks convert reserves into non-interest-bearing rCBDC to meet user demand.
  - wCBDC: instrument for money market transactions; obtained from the central bank using HQLA such as government bonds as collateral via outright purchases or repos; interest-bearing with variable rate; can be initially set to zero and later non-zero.

### Indonesia-specific liquidity and money market context (key statistics)
- Aggregate liquidity surplus of the banking system toward the central bank (defined as net foreign assets minus reserve requirement and other autonomous factors such as currency outside banks and government claims) amounted to around IDR 650 trillion in September 2021, down from more than IDR 1,000 trillion in August 2021.
- Since the early 2000s, the central bank has been the net borrower in the interbank money market due to banking system excess liquidity.
- If the banking system already has excess liquidity, the risk of bank disintermediation from rCBDC introduction is low; small conversions of reserves into rCBDC are unlikely to deplete reserves or necessitate central bank liquidity injections.
- rCBDC could serve as a tool for the central bank to absorb excess liquidity and incentivize interbank market transactions.

### Channels through which wCBDC could affect the money market
- wCBDC enables participants who do not currently settle in central bank money to settle directly in central bank money rather than bank deposits, reducing counterparty risk; wCBDC is a direct claim on the central bank and carries no credit risk.
- For NBFIs, wCBDC provides a new instrument to facilitate money market participation.
- Conversion of CBDC with reserves creates an additional channel to reduce banking system excess liquidity, potentially increasing money market transaction volume.
- Providing CBDC using HQLA as collateral via permanent central bank balance sheet expansion (government bond purchases) could lower government bond yields; limited HQLA supply could shift demand toward riskier assets and promote repo market development.
- If wCBDC interest is set higher than excess reserve returns, banks could migrate excess reserves into wCBDC; wCBDC interest could become a hard floor for money market rates and affect deposit rates, making wCBDC a potential monetary policy tool.

### Imperfect substitutability, monetary operations, and NBFI inclusion
- Assumption: imperfect substitutability between CBDC and reserves underpins the analysis.
- Disintegrated rCBDC is part of M0 like cash; rCBDC demand depends on adoption and could be hard to predict.
- Some rCBDC will be kept by users and payment service providers, leading to consistent reserve reduction and incentives for banks to borrow in the money market; large banks may be able to lower reserves permanently due to networks and excess liquidity.
- wCBDC, being quantity- and/or price-controllable by the central bank, effectively becomes a new monetary policy instrument and could be cost-effective to absorb excess liquidity compared with OMOs, particularly if initially non-remunerated.
- Enhanced NBFI participation via wCBDC access using government bonds as collateral could increase financial inclusion and money market depth, offering NBFIs (that hold bonds to maturity) better liquidity management.

### Banking sector impacts and policy considerations
- Banks may face deposit outflows as money migrates into CBDC; they will face greater competition and may need to raise interest rates to attract deposits.
- Indonesia banks’ high NIMs suggest capacity to raise deposit rates exists.
- Disruptiveness depends on scale and speed of CBDC take-up, substitution pace, and offsets from third-party and non-bank service providers; smaller banks are likely more challenged.
- Bank consolidation may improve competitiveness and resilience.
- Design of CBDC should be tailored to country characteristics and policy objectives:
  - Expert opinion study suggests a cash-like general-purpose rCBDC best suits Indonesia to enhance financial inclusion and reduce shadow banking (Zams and others, 2020).
  - Kang (2017) argued that an interest-bearing CBDC may reduce structurally high NIMs and enhance banking competitiveness.
  - Further research on detailed design choices and macro impact analysis is warranted, considering Indonesia’s money market development, banking sector liquidity, availability of HQLAs, and associated risks.

### Infrastructure, reforms, and broader policy context
- Issuing CBDC requires and promotes digital infrastructure upgrades and leverages existing infrastructure, potentially increasing operational efficiency.
- Ongoing money market reforms and payment digitalization (e.g., BI-Fast) will raise public digital awareness and prepare the banking sector for potential CBDC introduction.
- Money market development measures—enhancing competition and payment digitalization—complement CBDC considerations.

### Conclusions and forward-looking notes
- 2021 saw a large decline in interbank call money transactions due to ample bank liquidity, while repo transaction volume increased aided by BI’s efforts and coordination.
- Virtual peer learning between big and small banks helped increase money market volumes.
- Indonesia continues to develop its money market and payment systems reform agenda under the Blueprint for Money Market Development 2025; central counterparty infrastructure improvements expected in 2022.
- CBDC introduction could benefit money market growth mainly by broadening instruments and increasing NBFI financial inclusion.
- If CBDC is considered feasible, further research on design choices and macro impacts is recommended.

*Source: IMF chapter section on CBDC and money market development for Indonesia.*

### 2.      Indonesia emphasized its strong commitment to tackling climate change at COP26.

### 2.      Indonesia emphasized its strong commitment to tackling climate change at COP26.

### COP26 commitments and announced policies
- Pledged that the forestry sector will become a net carbon sink by 2030—absorbing more carbon dioxide from the atmosphere than it releases.
- Energy sector intentions:
  - develop an electric car ecosystem;
  - build the largest solar power plant in Southeast Asia;
  - promote new renewable energy; and
  - foster clean energy-based industries.
- Plans to mobilize climate finance and innovative financing such as green bonds.
- Stressed that carbon markets and carbon pricing must be part of efforts, and called for a carbon economy ecosystem that is transparent, has integrity, and is inclusive and fair.
- Signed the Glasgow Climate Pact (including acceleration of efforts towards the phasedown of unabated coal power and the phase-out of inefficient fossil fuel subsidies).
- Signed the Glasgow Leader’s Declaration on Forest and Land Use, committing to work collectively to halt and reverse forest loss and land degradation by 2030.

### Updated NDCs and sectoral projections
- Indonesia announced it will explore a plan to reach net-zero emissions in 2060 or sooner but left unchanged its NDC targets.
- Updated NDC submitted in July 2021 kept GHG emissions targets the same as the 2016 submission, implying further increases in greenhouse gas emissions from current levels:
  - unconditional target for GHG reduction: 29 percent compared to a Business as Usual (BAU) scenario;
  - conditional reduction target: 41 percent compared to BAU.
- Projected BAU and emission reductions from each sector category are unchanged.
- More than half of the reduction in GHG emissions will be contributed by forestry and other land uses, expected to see a significant reduction by 2030 compared to 2010.
- Emissions from energy use in 2030: lower relative to BAU but substantially higher compared to 2010 levels.
- Forestry-specific targets in the updated NDC:
  - peat land restoration of 2 million hectares by 2030;
  - rehabilitation of degraded land of 12 million hectares by 2030.
- Energy-sector action in the NDC: development of green refineries to produce green fuels from bio-resources and mix them with existing fuels to increase biofuel content and reduce fossil fuel consumption.

### Coal, energy mix, and infrastructure
- Authorities banned new development of coal-fired power plants from 2022 and intend to phase out such plants by 2056 at the latest.
  - Note: moratorium does not apply to plants already approved.
- Roadmap being devised to shift from coal-generated electricity to renewable energy, complemented by energy-efficiency measures, increased use of biofuels, and development of an electric vehicle industry.
- Existing contracts with coal-fired power plant operators could be a near- to medium-term roadblock.
- National Energy Plan (RUEN) energy mix target by 2025 remains unchanged:
  - oil: 25 percent;
  - gas: 22 percent;
  - coal: 30 percent;
  - new and renewable energy: 23 percent.
- RUEN mapped into new Electricity Supply Business Plan (RUPTL) for 2021−2030 devised by PLN, based on electricity demand growth assumption of 4.4 percent annually (compared to 6.5 percent in previous RUPTL 2019−2028). Additional power capacity is around 40 percent lower; 26.3 GW of the total 40.6 GW planned additional capacity is allocated to Independent Power Producers (IPPs).
- Total power investment needs estimated at around US$9.14 billion per year over 2021−30; PLN expected to contribute US$5.14 billion per year.

### Carbon pricing: current design and implications
- General purpose: reduce GHG emissions by making them costly; main mechanisms are carbon tax and Emission Trading Systems (ETSs).
- Indonesia’s carbon pricing timeline and parameters:
  - carbon tax introduced on April 2022 with a rate of IDR 30,000 per ton CO2e, imposed on GHG emissions exceeding a certain threshold (a cap and tax system).
  - About US$2 per ton of CO2e—one of the lowest tax rates among countries with carbon taxes currently in place.
  - There are no plans for further increases.
  - Authorities will start an ETS by 2024; design still under preparation. Pilot completed with participation of 32 coal-fired power plants with 14 buyers and 18 sellers.
- Design and coverage limitations:
  - Cap-and-tax design does not provide incentives for below-the-cap companies to reduce emissions further, unlike ETS.
  - Current carbon tax applies only on a limited basis to coal-fired power plants, leaving industry and transportation sectors—which account for about 29 and 26 percent of the GHG emissions by the energy sector, respectively—largely outside coverage.
- Effectiveness constraints without complementary reforms:
  - Energy subsidies apply to fuels and electricity and domestic market prices are effectively fixed with very infrequent adjustments to world market prices; this undermines carbon pricing effectiveness because costs may not be passed to end-users.
  - Electric company is a state-owned monopoly and rigid price mechanisms could lead to fiscal risks if carbon costs cannot be recovered.
- Policy-reform context:
  - Government proposed energy subsidy reform plan to parliament in 2020 to improve targeting and transform some energy subsidies into direct social assistance; implementation expected this year but depends on pace of economic recovery from the pandemic. No plans to reform the energy pricing mechanism.
  - Gradual transition from fixed to flexible pricing recommended, with clear communication to minimize distributional effects and mitigate inflation concerns.
- Modeling and revenue implications:
  - A carbon price of US$25 would reduce GHG emissions by 16 percent and generate revenues of 0.7 percent of GDP.
- Recommendations for carbon pricing design:
  - Consider redesign of the carbon tax to include further rate hikes and expansion of applicable sectors.
  - Ensure upcoming ETS incorporates major emitter sectors and involves as many sectors as possible.
  - ETS should include a market adjustment mechanism for energy prices.
  - Establish institutional frameworks to manage all carbon pricing mechanisms consistently and effectively.
  - Ensure burden of carbon pricing does not fall disproportionately on lower income groups.

### Green financing: needs, markets, and market development
- Estimated financing needs:
  - achieving NDC targets: 2.8 percent of GDP annually;
  - infrastructure needs: 6.0 percent of GDP annually.
- Mobilizing private investment is essential given the scale of needs.
- OJK sustainable financing roadmap priorities:
  - developing a green taxonomy;
  - implementing ESG aspects into risk management through reporting and KPIs;
  - sharing success stories of innovative green scheme development;
  - developing innovative schemes of sustainable project financing; and
  - building understanding of importance of activities that consider ESG aspects.
- Status of green bond/sukuk markets:
  - Indonesia has issued several sovereign green sukuks internationally; domestic green financing still underdeveloped.
  - Indonesia’s green bond market is second largest among Southeast Asia peers in nominal terms (behind Singapore) but small relative to GDP: US$5 billion, accounting for around 0.5 percent of GDP.
  - Market dominated by government bonds and almost all bonds are denominated in foreign currency (FCY).
  - Indonesia’s local currency bond market is the smallest among peers in terms of GDP ratio, implying limited domestic market for green bonds.
- Secondary market observations:
  - Government green sukuk issued in 2018 appears to carry a greenium since 2020; other government green sukuks do not show similar premium, potentially due to lack of liquidity and long maturity.
  - Government green sukuk could be used as a benchmark for design and pricing of corporate green market.
- Market development and transparency needs:
  - Transparency and reduced information asymmetry are crucial.
  - Disclosure standards of listed companies should be internationally aligned to ensure comparability and consistency.
  - Indonesia Green Taxonomy is essential for comparing information about green financial instruments for investment decisions.
- Broader market and macro conditions:
  - Need to overcome a shallow financial market and maintain sound macroeconomic fundamentals to mobilize private green investment.
  - Implement 2017 FSAP recommendations to enhance bond yield curve by consolidating debt issuance and improving secondary markets.
  - Maintaining strong macroeconomic policies is crucial to attract foreign direct investment and foreign financing and enable private sector to obtain favorable financing terms for green investment.

### Policy recommendations and roadmap
- Promote green financing combined with effective carbon pricing to reconcile GHG emission reductions with development needs.
- Develop a comprehensive mitigation strategy and roadmap by 2030 including:
  - energy subsidy and pricing reforms;
  - a carbon pricing revenue recycling plan.
- Redesign carbon pricing mechanisms:
  - expand number of participants in ETS;
  - increase carbon tax rates in a clear and predictable manner;
  - expand sectoral coverage.
- Establish institutional frameworks to manage carbon pricing mechanisms consistently and effectively.
- Protect lower income groups from disproportionate burdens of carbon pricing.
- Enhance transparency (including implementation of green taxonomy), deepen financial markets, and maintain sound macroeconomic policies to foster a green financing market and meet climate financing needs.

*Prepared by Koki Harada (APD) as presented in the IMF chapter.*

---


_Source: https://www.imf.org/-/media/files/publications/cr/2022/english/1idnea2022002.pdf_