## INDONESIA’S GOVERNMENT BOND YIELDS AND NONRESIDENT PARTICIPATION IN GOVERNMENT BOND MARKETS ## Source details **Canonical URL:** [INDONESIA’S GOVERNMENT BOND YIELDS AND NONRESIDENT PARTICIPATION IN GOVERNMENT BOND MARKETS](https://www.imf.org/-/media/files/publications/cr/2021/english/1idnea2021002.pdf) ## Other formats - [Markdown version](/-/media/files/publications/cr/2021/english/1idnea2021002.pdf.md) - [Structured JSON version](/-/media/files/publications/cr/2021/english/1idnea2021002.pdf.json) --- ### Executive summary and objective - Analyzes whether government bond yields in Emerging Markets (EMs) vary with (i) the share of nonresident participation in government bond markets and (ii) exchange rate volatility. - Documents recent evolution of Indonesian government bond yields, exchange rate volatility, and nonresident participation; compares Indonesia with other large EMs; provides an econometric analysis controlling for domestic factors and time and country specific effects. - Differentiates by controlling for domestic factors and time and country specific effects. ### Evolution of government bond yields and nonresident participation - Since the early 2000s Indonesian government bond yields declined; since the beginning of the 2010s Indonesian U.S. dollar bond yields have declined more than yields on local currency government bonds. - After the COVID-19 shock and spike in global risk aversion (VIX), both LC and FX bond yields increased sharply before declining as external pressures eased and liquidity was restored. - Decline in Indonesian U.S. dollar government bond yields exceeded that in local currency yields, increasing the LC–FX yield differential. - Nonresident participation trends: - Share of nonresident participation in Indonesia’s LC bond markets increased more than in other EMs since the 2010s. - In 2019, Indonesia was among EMs with the highest participation of nonresidents in LC bond markets. - First semester of 2020: - Indonesia: nonresident holdings in total outstanding debt securities declined from 50 percent to 45 percent. - EM average declined from 36 percent to 34 percent. - Cross-country composition of 2020 decline: - Indonesia, Egypt, Malaysia: most decline from outflows in LC bond markets. - Poland, Turkey: outflows evenly distributed between FX and LC securities. - Mexico: sharpest decline in LC participation but largely compensated by large issuances of FX debt. - In Indonesia, domestic banks substantially increased holdings of LC government bonds, more than compensating for reduced nonresident holdings. ### Nonresident participation during global stress episodes - Nonresident participation in EM LC bond markets typically declines during global turmoil, contributing to disorderly LC bond market conditions and higher yields. - EM selloffs in the Global Financial Crisis (GFC) and 2018 were particularly acute for EM government bonds. - ASEAN countries experienced declines in nonresident participation across several stress episodes, including the 2015 China devaluation episode. - For Indonesia, the decline in nonresident participation in LC bond markets during the COVID-19 crisis exceeded that of prior episodes. ### Key empirical findings on yields (quarterly data, 2004:Q1−2020:Q2; 18 large EMs) - Model features: - Dependent variables: government bond yields in local currency and in U.S. dollars. - Key regressors: share of nonresident holdings of total government bond securities (t-1), volatility of lagged daily changes in the exchange rate (t-1), monetary policy rate (t-1), real GDP growth (t-1), inflation (t-1), equity returns (t-1), VIX (t-1), US 10-year yield (t-1), interactions, time and country fixed effects. - Bond yields, monetary policy rate, and global factors enter in logs. - Local currency (LC) bond yields (selected coefficients and significance): - Nonresident in percent of total bonds (t-1): coefficient -0.009 *** - Nonresident in percent of total bonds (t-1) x VIX (t-1): coefficient 0.004 *** - FX daily volatility (t-1): coefficient 0.063 ** - Monetary policy rate (t-1): coefficient 0.334 *** - LC yield interpretations and quantitative impacts: - Higher presence of nonresidents is on average associated with lower LC yields. - Positive interaction with VIX implies that when VIX increases, a higher share of nonresident participation is associated with higher LC yields. - Quantitative impacts reported: - At the average sample level of the VIX, a one-standard deviation increase in the share of nonresident participation is associated with an increase of 1.0 percent in the average level of LC bond yield. - An increase of one-standard deviation in exchange rate volatility is associated with an increase of 1.03 percentage points in LC bond yields. - LC yields are positively associated with monetary policy rates. - U.S. dollar (FX) government bond yields (selected coefficients and significance): - FX daily volatility (t-1): coefficient 0.141 *** - VIX (t-1): coefficient 0.321 ** - US 10-year yield (t-1): coefficient 0.656 *** - Real GDP growth (t-1): coefficient -1.279 *** - FX yield interpretations: - FX bond yields respond to global factors (VIX, US 10-year yield) and decline with stronger economic activity. - FX bond yields are positively associated with exchange rate volatility. - Yield differential (LC minus FX, "currency spread"): - Positively correlated with monetary policy rate (example coefficient 0.518 ***), inflation (example coefficient 6.498 ***), and FX daily volatility (example coefficient 0.259 **). - Interpretation: higher policy rates, higher inflation, and higher exchange rate volatility widen the currency spread. - Country risk premium (EMBI spread): - Positively correlated with global factors rather than currency-risk variables. - Significant positive associations with VIX (example coefficient 1.008 ***) and US 10-year yield (example coefficient 1.376 ***). ### Policy implications and recommendations (bonds and nonresident participation) - Benefits of nonresident participation: increased access to financing, enhanced efficiency, financial sector competitiveness, facilitation of productive investment and consumption smoothing. - Risks: greater exposure to downside risks and potential for capital outflows amplifying domestic stress during crises. - Suggested measures to reduce vulnerabilities in domestic government bond markets: - Strengthen domestic market infrastructure, including hedging markets and liquidity. - Broaden inclusion in global bond indices. - Maintain a stable macroeconomic environment. - Keep inflation low to contribute to a lower currency risk premium. - Indonesia-specific recommendations: - Increase revenue mobilization. - Deepen domestic bank and nonbank financial markets (increase saving rates, further develop pension funds’ investments in government bonds). - Attract more FDI to offer alternative financing sources and reduce reliance on volatile nonresident portfolio flows. --- ### Substitution toward hard-currency issuance during COVID-19 - Hard-currency (U.S. dollar) bond issuance benefited from ample global liquidity after monetary easing by the U.S. Fed and other major central banks. - Indonesia experienced unprecedented hard-currency bond issuance during COVID-19 compared to other stress periods and outperformed other EMs. - Net substitution in 2020 (Indonesia): - Nonresident outflows from LC bonds: US$9 billion - Issuance of Indonesian government bonds denominated in U.S. dollar: US$10.9 billion ### Exchange rate volatility and investor behavior - Sharp exchange rate depreciations lower take-home returns for nonresident investors, affecting demand for LC EM bonds. - Nonresident investors optimize returns measured in foreign currency; exchange rate changes affect their entry/exit decisions and can induce synchronized flows reinforcing LC yield movements. - Since the beginning of COVID-19, the LC–FX differential increased in Indonesia, Brazil, and Russia, but declined in the Philippines and South Africa. - Historical large exchange rate volatility episodes for Indonesia: GFC (2008−2009), Taper Tantrum (2013), China devaluation (2015), COVID-19 crisis (2020). - Over 2004−2020, exchange rate volatility in Indonesia was close to the median across EMs but relatively higher than other ASEAN peers. --- ### FX markets, FX swaps, DNDFs, and BI intervention strategy ### Market structure and instruments - Average daily IDR turnover increased from US$4 billion to US$27 billion over the past decade (BIS, 2019). - Main instruments: FX swaps and outright spot deals; FX swaps trading volume exceeds outright forwards substantially. - Offshore trading has grown strongly and now accounts for the bulk of IDR turnover; offshore trading has grown over six times. - BI’s FX instruments and regime: - Flexible exchange rate with BI mitigating excessive volatility and sustaining market liquidity. - Instruments: direct spot interventions, FX swaps (simultaneous spot and forward), and DNDFs (non-deliverable forwards settled in rupiahs). - Size and usage: - Outstanding stock of FX Swaps (DNDF) was US$1.5 (3.8) billion at end 2020:Q3, with total issuance of US$123 (58) billion. - Average daily issuance: FX swaps US$290 million; DNDF US$98 million. - These represent about 17 percent of the onshore FX swap and outright forward market turnover respectively. ### Empirical evidence on liquidity and forward effects (daily data; 719 intervention days; total amounts ≈ US$150 billion) - Two price-based FX liquidity indicators: - Funding liquidity = (forward − spot) normalized by spot. - Market liquidity = (ask − bid) normalized by mid-price. - Stylized facts: - FX liquidity pressures spike during global stress episodes; COVID-19 raised both market and funding illiquidity relative to prior EM selloff periods. - Introduction/availability of FX swaps associated with higher FX liquidity (market and funding). - Introduction/availability of DNDF associated with lower FX liquidity (market and funding) in naive comparisons, but usage specifics matter. - Mitigation of VIX-induced funding liquidity stress: - An intervention of US$250 million associated with reducing VIX induced funding liquidity stress by approximately 18 percent (FX swaps) and 27 percent (DNDF). - Forward volatility (GARCH results): - An intervention of US$50 million associated with reducing VIX induced forward volatility by approximately 65 percent (FX swaps) and 21 percent (DNDF). - Uninteracted effects on level/value of forward rate are positive but imprecise; interaction with VIX shows interventions dampen forward volatility during stress. - DNDF market: - Launched November 2018; dominated by onshore banks; liquidity limited; DNDF typically below offshore NDF rates; DNDF interventions used frequently (roughly every other day). ### CIP deviations, arbitrage and capital flow channels - CIP deviations emerged during GFC and reappear during acute stress; they widened during COVID-19 reflecting funding liquidity pressures. - At the onset of the GFC in Indonesia, synthetic dollar rates turned negative; JIBOR rates were almost six times LIBOR. - During COVID-19, CIP deviations turned negative such that synthetic-dollar funding cheaper than direct-dollar, affecting hedging costs and creating arbitrage possibilities. - Pricing and hedging frictions can affect capital flows and incentives for arbitrage-driven outflows/inflows. ### Policy considerations on interventions - Forward interventions can reduce FX funding stress and forward volatility during global episodes but may have asymmetric effects vs spot interventions and affect arbitrage via CIP deviations. - Intervention design should account for: - Asymmetric impacts due to convertibility/counterparty risk. - Potential to narrow positive CIP deviations or widen negative CIP deviations, affecting arbitrage incentives. - Small size of onshore market (20 percent of trading market) relative to offshore markets. - Trade-offs between providing FX liquidity, influencing hedging costs, and possibly encouraging arbitrage or affecting capital flows. - Complementary crisis measures used internationally include: FX swaps and dollar repos, reserve requirement adjustments, expanding counterparties, easing macroprudential FX limits, and sales from sovereign wealth proceeds. --- ### Inflation dynamics during COVID-19 and outlook ### Phases and drivers - Pre-COVID-19: inflation on a downward trend; Bank Indonesia preannounced lowering target by ½ percentage point at end-2017 and end-2019. - Lockdown phase: aggregate demand plummeted, exerting downward pressure on prices; administered prices little changed; temporary passenger-travel administered price rise noted. - Early recovery (around June/July 2020): Q3 real GDP grew at 12.4 percent (q/q s.a.a.r); month-on-month consumer price changes continued to fall into negative territory. - Transition to new normal: demand expected to recover and output gap to narrow; inflation expected to pick up gradually depending on recovery pace, food supply, administered prices. ### Measurement complications and disaggregate findings - Consumption baskets changed substantially; fixed-basket measures may understate “true” price increases for items in higher demand. - CPI disaggregation: - CPI basket: 11 components, 43 subcomponents. - PPI basket: 9 components, 35 subcomponents. - Volatile food: 16 percent of CPI basket; administered prices: 18 percent. - Key disaggregate outcomes: - Most CPI components contributed to disinflation; food prices largest share of decline. - Contact-intensive sectors (education, restaurants, other services) saw declines of 2 to 5 percentage points y/y. - Inflation densities shifted left and narrowed in 2020:Q2, widened in 2020:Q3 indicating greater sectoral dispersion. ### Regional, producer price, and property price notes - Regional CPI: disinflation broadly across cities; larger cities more elevated inflation than smaller cities. - PPI: - PPI inflation fell from 4.2 percent y/y in 2018:Q3 to 0.7 percent in 2020:Q1 and −0.7 percent in 2020:Q2; recovered to 0.3 percent in 2020:Q3. - Mining deflation: 10.8 percent y/y as of September 2020; coal prices dropped 51 percent from peak in September 2018. - Property prices: - Price increases averaged around 2 percent for much of 2019, dipping to 1.2 percent in 2020:Q3. - Regional large contractions in 3Q in Batam, Balikpapan, Banjarmasin with examples: Batam residential price inflation fell 3.8 percent in March and 1.6 percent in September. ### Outlook and risks - Staff projections for 2021: - Headline inflation projected to climb from 1.7 percent in December 2020 to 3 percent next year. - Core inflation forecasted to climb from 1.6 percent in December 2020 to 2.6 percent at end-2021. - Bank Indonesia’s forecast envisions inflation returning to the 3.0%±1% target range in 2021. - Risks: - Downside: pandemic weighs on activity longer, disinflationary forces entrenched, sustained undershooting of BI target. - Upside: stronger rebound could lift inflation more quickly than expected. --- ### Banking sector lending dynamics and corporate sector vulnerabilities ### Bank lending trends and drivers - Loan growth fell from 8 percent (y/y) at end-2019 to below 2 percent (y/y) in July 2020. - Private and foreign banks’ lending contracted in June and July 2020; public banks (including SOEs) and regional banks maintained positive but declining loan growth. - Deposit growth rose from around 6 percent (y/y) at end-2019 to about 9 percent by July 2020; SOE banks recorded deposit growth of 13 percent (y/y) by July 2020. - Monetary policy transmission to lending has been weak and asynchronous at times, especially since 2018. ### Bank-level empirical findings (monthly bank data, Jan 2012−Jul 2020; 86 commercial banks) - Estimation equation: Δloan_it = α + γ X_it−1 + δ Stress_it + φ b_i + θ_t + ε_it. - Liquidity matters: - More liquid banks have higher loan growth; liquidity provides cushion during stress episodes. - Deposit growth is positively associated with loan growth. - Public vs nonpublic banks: - Stress episodes associated with lower loan growth from nonpublic banks but not materially from public banks. - Non-linearity: - Marginal positive effect of liquidity on loan growth significantly higher when liquidity ratio below 34 percent of total assets. - Distributional concern: - 10th percentile liquidity ratio declined to 15 percent in 2020. ### Policy recommendations for banking - Strengthen liquidity of solvent banks with sound fundamentals, especially those with low liquidity. - Monitor and provide funding to banks to contain funding and deposit concerns and facilitate lower lending rates. - Complementary government measures: - SME loan guarantees, SME temporary interest payment subsidies, state fund deposits with low interest conditional on lending. - If credit growth does not recover in early 2021, consider funding-for-lending programs with careful design (incentives, borrowing caps). - Given BI’s legal limits on direct lending to banks, expanding fiscal resources to eligible banks with safeguards could support credit supply. ### Nonfinancial corporate (NFC) vulnerabilities (firm-level S&P Capital IQ sample; 459 firms, end-2019) - Sample coverage: - Accounts for about 62 percent of total NFC debt and 22 percent of GDP in revenue terms. - 439 publicly listed, 20 private firms; 23 SOEs. - Pre-pandemic vulnerabilities: - About half of firms used more than half of operating income to cover interest payments. - Cash holdings covered less than 20 percent of outstanding current liabilities for these firms. - Median ICR ~2 at end-2019 (one of the lowest in region). - Median ROA ~5.1 percent as of end-2019. - Median debt-to-assets ratio 27 percent as of end-2019. - FX debt share ~39 percent of outstanding debt as of end-2019 (regional peers ~16 percent). - Debt-at-risk and projected COVID-19 impact: - Debt-at-risk (ICR <2) about 56 percent as of end-2019 (higher than peers). - Under October 2020 IMF WEO shocks: - Share of sample firms with ICRs below one could reach between 47−50 percent by end-2020. - About 45 percent of sample firms could face cash shortages at end-2020 without liquidity support. - Q2 2020 sub-sample actuals: share of firms at risk (ICR <1) ~49 percent; median ICR ~0.8; cash ratio fell to 19 percent from 20 percent end-2019. - Share of NFC debt held by firms with ICR <1 expected to rise to 32 percent by end-2020; share with ICR <2 expected to rise to 88 percent by end-2020 (2½ and 1½ times 2019 levels respectively). - Policy implications for corporates: - Sustain or scale targeted assistance until recovery: interest rate subsidies, loan moratoria for MSMEs, credit guarantees, corporate income tax reductions, liquidity support, regulatory easing for loan restructuring. - Strengthen insolvency framework, social safety nets, and active labor market policies. --- ### Corporate distress estimation approaches and cash-flow calibration ### Two approaches to forecast end-2020 ICRs - Approach I: Regression-based method (sample: 29,161 firm-year observations, 2002−2019) forecasting ROA, EIR, LEV using lagged variables, country and industry fixed effects, macro and global variables; WEO October vintage used for 2020−21 projections. - Approach II: Direct method using consensus earnings forecasts (January 2 and June 30 vintages) to apply operating income shocks and exchange rate/interest shocks to 2019 figures; scaling factor θ set so ASEAN-6 median ICR in 2020 matches Approach I (~1.2). ### Cash position and balance-sheet adjustments (calibration) - cash2020 = cash2019 + EBIT shock × (cash flow from operations2019) − capital expenditure2020 − debt amortization2020 + net interest payment2019 − dividend payment2020. - Assumptions: - Capital expenditure2020 = 0.75 * capital expenditure2019. - Debt amortization2020 = 10 percent of maturing debt (debt rollover ratio = 90 percent). - Dividend payment2020 = 0. - Rationale: CAPEX scaling and rollover calibrated against historical GFC patterns and sample medians. --- ### Digitalization, fintech, and financial inclusion ### Fintech lending scale and trends - Accumulation of P2P loans grew more than 230 percent (y/y) to IDR 88.4 trillion in January 2020. - Total outstanding P2P loans stood at IDR 13.5 trillion in January 2020, increase of 137 percent (y/y). - Fintech lending volume remains small at less than 0.1 percent of GDP (pre-COVID). - Business loans constitute more than 85 percent of total fintech loans. - Accumulated/active fintech borrowers in September 2020: close to 29.2/9.3 million. - Average loan durations span from 10 days to one year. ### COVID-19 impact on fintech and digital financial inclusion - Outstanding P2P credit declined by 8 percent since December 2019. - NPL ratio in fintech increased from 3.7 percent at end-2019 to 8 percent in July 2020; including restructured loans, ratio is 10 percent. - Out of 1.96 million loan restructuring requests to 88 P2P platforms, 34 percent granted; amount requested to be restructured in May = IDR 1.08 trillion (8 percent of total loans), approved = IDR 236.9 billion (2 percent). - Digital landscape expansion: - More than 200 e-commerce players and more than 350 fintech players as of end-2019. - E-commerce transactions grew by 137 percent (CAGR) from 2017 to Aug-2019. - Transactions at four largest e-commerce sites estimated to double to US$29 billion in 2020 from US$14 billion in 2019. - E-money transaction values increased by 114 percent between 2015 and 2019; transaction volumes increased by 91.4 percent. - Digital inclusion gaps: - Almost half the population unbanked; close to 6 percent of the world’s unbanked adults reside in Indonesia. - Two-thirds of population financially illiterate as of 2019. - More than half of the population does not use the internet; more than half does not own a smartphone. - Java accounts for close to 85 percent of fintech lending. ### Policy priorities for digitalization - Focus on three pillars: - Ensure equal opportunity and access to avoid digital divide (supply-side). - Address constraints to adoption (demand-side). - High-quality supervision and regulation balancing innovation with financial stability, integrity, cyber-risk and consumer protection. - Recommended actions: - Invest in digital-friendly infrastructure (electricity, logistics, internet). - Expand financial digital literacy and worker skill upgrades. - Expand G2P reach and require more recipients to receive transfers via digital accounts. - Strengthen cybersecurity, data privacy, and expand regulatory perimeter to cover unlicensed nonbanks and fintechs. - Improve BI–OJK interagency coordination to supervise fintech activities. --- ### Climate policy, energy subsidies, and green finance ### Climate budgeting and green sukuk - Climate Budget Tagging (CBT) in place since 2016 to monitor climate expenditures; scope extended to adaptation since 2018. - Green sukuk international issuances: US$1.25 billion (2018), US$0.75 billion (2019), US$0.75 billion (June 2020). - Retail green sukuk: IDR 1.46 trillion ≈ US$0.10 billion (November 2019). - Green sukuk funds used for mitigation (sustainable transport, waste management, renewables) and adaptation (resilience projects). ### Energy subsidy context and reform proposals - Energy subsidies remain significant: "they still amount to over 10 percent of total government expenditure." - 2021 reform proposed (postponed due to COVID-19) shifts from commodity-based to people-based subsidies: - Reduce fixed diesel subsidy from IDR 1,000 to IDR 500 per liter. - Apply market price adjustments to premium gasoline and electricity for non-subsidized customers. - Gradual LPG price adjustments and integrate related subsidies into social assistance. - Feebate and carbon tax considerations: - US$25 per ton CO₂ could cut emissions by 13.0 percent and raise 0.7 percent of GDP in revenue. - Higher rates (US$50, US$75) imply larger cost increases and higher jobs-at-risk concentrated in mining and electric/gas sectors. - Recommended transition package: - Comprehensive transition plan toward greener economy including labor reallocation, compensation financing (carbon tax revenues), reforestation incentives (feebates or commodity tax), renewable generation feebates, and energy subsidy reform implementation. ### Adaptation and fiscal considerations - National Action Plan for Climate Change Adaptation (RAN-API, 2014) outlines resilience strategies across economy, living systems, ecosystems, urban/coastal areas. - Gaps: adaptation plan lacks dedicated fiscal costs; recommend integrating medium-term expenditure needs and strengthening monitoring. --- *Prepared by IMF staff; content from the IMF chapter and staff notes in the provided source (1idnea2021002).* ### References_______________________________________________________________________________ 19 ### INDONESIA’S GOVERNMENT BOND YIELDS AND NONRESIDENT PARTICIPATION IN GOVERNMENT BOND MARKETS ### Executive summary and objective - The note analyzes whether government bond yields in Emerging Markets (EMs) vary with (i) the share of nonresident participation in government bond markets and (ii) exchange rate volatility. - It documents recent evolution of Indonesian government bond yields, exchange rate volatility, and nonresident participation, compares Indonesia with other large EMs, and provides an econometric analysis of correlations between yields and market characteristics. - The analysis differentiates itself by controlling for domestic factors and time and country specific effects. ### Evolution of Indonesian government bond yields - Since the early 2000s Indonesian government bond yields declined; since the beginning of the 2010s Indonesian U.S. dollar bond yields have declined more than yields on local currency government bonds. - After the COVID-19 shock and the spike in global risk aversion (measured by the VIX), yields of both Indonesian local currency government and U.S. dollar government bonds increased sharply before declining as external pressures eased and liquidity was restored. - The decline in Indonesian U.S. dollar government bond yields exceeded that in local currency yields, increasing the yield differential between these bonds. - These differing dynamics reflect differences in market liquidity, perceptions of currency and credit risk, and local currency bond market characteristics such as the share of nonresident participation. ### Nonresident participation in government bond markets - The share of nonresident participation in Indonesia’s local currency bond markets increased more than in other EMs since the 2010s. - In 2019, Indonesia was among the EMs with the highest participation of nonresidents in local currency bond markets. - During the first semester of 2020: - The share of nonresident holdings in total outstanding debt securities in Indonesia declined from 50 percent to 45 percent. - The average for EMs declined from 36 percent to 34 percent. - Cross-country composition of the 2020 decline: - In Indonesia, Egypt, and Malaysia most of the decline was from outflows in local currency bond markets. - In Poland and Turkey the outflows were more evenly distributed between FX and local currency securities. - In Mexico the decline in local currency participation was the sharpest among EMs but was largely compensated by large issuances of FX debt. - In Indonesia, domestic banks substantially increased their holdings of local currency government bonds, more than compensating for reduced holdings by nonresident investors. ### Nonresident participation during episodes of global financial stress - Nonresident participation in EM local currency bond markets typically declines during global financial market turmoil, contributing to disorderly local currency bond market conditions and higher yields. - EM selloffs in the Global Financial Crisis (GFC) and 2018 were particularly acute for the EM government bonds asset class. - ASEAN countries experienced declines in nonresident participation across several stress episodes, including the 2015 China devaluation episode. - For Indonesia, the decline in nonresident participation in local currency bond markets during the COVID-19 crisis exceeded that of prior episodes. ### Key empirical findings reported - A larger share of nonresident participation in domestic bond markets appears associated with: - Lower levels of local currency government bond yields. - Higher sensitivity of local currency yields to changes in global risk aversion. - Both local currency and foreign currency bond yields are positively correlated with exchange rate volatility. ### Figures and tables referenced (content indicators) - Figures described include: Government Bond Local Currency and FX Bond Yields; Local Currency Bond Yields During the COVID-19 Crisis; Nonresident Participation in Local Currency Government Bond Markets; Nonresident Participation in Total Debt Government Bond Markets During COVID-19 Crisis; Nonresident Participation in Local Currency Government Bond Markets During Episodes of Global Financial Stress; Issuance of Hard Currency Bonds During Episodes of Global Financial Stress; Exchange Rate Market Pressures; Exchange Rate Volatility; Impact of Nonresident Participation and FX Volatility on Local Currency Government Bond Yields. - Tables described include: Determinants of Local Currency Government Bond Yields; Determinants of U.S. Dollar Government Bond Yields; Determinants of Country Credit Risk and Currency Risk. *Prepared by Francisco Arizala (SPR); content from the IMF chapter "INDONESIA’S GOVERNMENT BOND YIELDS AND NONRESIDENT PARTICIPATION IN GOVERNMENT BOND MARKETS."* ### 8. In the COVID-19 crisis, large ### 8. In the COVID-19 crisis, large nonresident outflows from EMs local currency bond markets have been compensated for by sustained issuance of hard-currency bonds by EMs ### A. Substitution toward hard-currency (U.S. dollar) bond issuance - Hard-currency bond issuance benefited from ample global liquidity after monetary easing by the U.S. Fed and other major central banks. - Indonesia experienced unprecedented hard-currency bond issuance during COVID-19 compared to other periods of global financial stress and outperformed issuances by other EMs. - Net substitution in 2020: - Nonresident outflows from local currency bonds: US$9 billion - Issuance of Indonesian government bonds denominated in U.S. dollar: US$10.9 billion ### B. Exchange rate volatility and nonresident investor behavior - Sharp exchange rate depreciations lower take-home returns for nonresident investors, affecting demand for local currency EMs bonds. - Nonresident investors optimize returns measured in foreign currency; exchange rate changes affect their investment decisions and can induce synchronized exit/entry that reinforces yield movements on local currency government bonds. - During global market turbulence, exchange rate volatility combined with a preference for FX assets by nonresident investors can increase the yield differential between local currency and FX government bonds. - Since the beginning of the COVID-19 crisis, the differential increased in Indonesia, Brazil, and Russia, but declined in the Philippines and South Africa (text chart referenced). - Historical episodes with large exchange rate volatility for Indonesia: GFC (2008−2009), Taper Tantrum (2013), China devaluation (2015), and COVID-19 crisis (2020). - Over 2004−2020, exchange rate volatility in Indonesia was close to the median across EMs but relatively higher than other ASEAN peers. ### C. Determinants of government bond yields — empirical findings - Model specification (quarterly data, 2004:Q1−2020:Q2, sample of 18 large EMs): - Dependent variables: government bond yields in local currency and in U.S. dollars. - Key regressors: share of nonresident holdings of total government bond securities (t-1), volatility of lagged daily changes in the exchange rate (t-1), monetary policy rate (t-1), real GDP growth (t-1), inflation (t-1), equity returns (t-1), VIX (t-1), US 10-year yield (t-1), plus interaction terms and time and country fixed effects. - Bond yields, monetary policy rate, and global factors enter in logs. - Local currency bond yields: - Nonresident in percent of total bonds (t-1): coefficient -0.009 *** - Nonresident in percent of total bonds (t-1) x VIX (t-1): coefficient 0.004 *** - FX daily volatility (t-1): coefficient 0.063 ** - Monetary policy rate (t-1): coefficient 0.334 *** - Interpretation: - A higher presence of nonresidents is on average associated with lower local currency yields (resource availability lowers funding costs). - The positive interaction with VIX implies that when global risk perceptions (VIX) increase, a higher share of nonresident participation is associated with higher local currency yields—revealing a tradeoff. - Quantitative impacts reported: - At the average sample level of the VIX, a one-standard deviation increase in the share of nonresident participation is associated with an increase of 1.0 percent in the average level of local currency bond yield. - An increase of one-standard deviation in exchange rate volatility is associated with an increase of 1.03 percentage points in local currency bond yields. - Local currency yields are positively associated with monetary policy rates. - U.S. dollar (FX) government bond yields: - Nonresident participation and the monetary policy rate do not have the same effects as for local currency yields (coefficients insignificantly different from zero for nonresident share and monetary policy rate). - Key positive associations: - FX daily volatility (t-1): coefficient 0.141 *** - VIX (t-1): coefficient 0.321 ** - US treasury 10-year bond yield (t-1): coefficient 0.656 *** - Real GDP growth (t-1): coefficient -1.279 *** - Interpretation: - FX-denominated bond yields respond to global factors (VIX, US 10-year yield) and decline with stronger economic activity. - FX bond yields are also positively associated with exchange rate volatility. - Yield differential between local currency (LC) and FX bonds ("currency spread"): - Positively correlated with: - Monetary policy rate (e.g., coefficient 0.518 *** in one specification) - Inflation (e.g., coefficient 6.498 *** in one specification) - FX daily volatility (e.g., coefficient 0.259 **) - Interpretation: - Higher monetary policy rates can operate as a floor for domestic interest rates, widening the spread. - Higher inflation raises investor concerns about real returns, widening the currency spread. - Higher exchange rate volatility leads nonresident investors to demand higher yields to compensate for uncertain home-currency returns. - Country risk premium (EMBI spread): - Positively correlated with global factors rather than with the currency risk premium variables. - Significant positive associations with: - VIX (e.g., coefficient 1.008 *** in one specification) - US treasury 10-year bond yield (e.g., coefficient 1.376 ***) - Interpretation: - The country risk premium compensates for risk differences among U.S. dollar–denominated assets and is driven by global risk aversion and alternative U.S. dollar investments. ### D. Policy implications and recommendations - Benefits and risks of nonresident participation: - Benefits: increased access to financing, enhanced efficiency, promoted financial sector competitiveness, and facilitation of productive investment and consumption smoothing. - Risks: greater exposure to downside risks and potential for capital outflows to amplify domestic market stress during crises. - Suggested measures to reduce vulnerabilities in domestic government bond markets: - Strengthen domestic market infrastructure, including hedging markets and liquidity. - Broaden inclusion in global bond indices. - Maintain a stable macroeconomic environment. - Keep inflation low to contribute to a lower currency risk premium. - For Indonesia specifically: - Increase revenue mobilization. - Deepen domestic bank and nonbank financial markets (e.g., increase saving rates, further develop pension funds’ investments in government bonds). - Attract more FDI to offer alternative financing sources and reduce reliance on volatile nonresident portfolio flows into government bonds. *Source: International Monetary Fund staff note.* ### 3. The U-shaped pattern generally also ### 3. The U-shaped pattern generally also holds for producer prices of ASEAN countries, which also saw a notable drop during the early phase of the pandemic ### Inflation phases and principal drivers during the COVID-19 pandemic - Pre-COVID-19 phase - Inflation in Indonesia was on a downward trend prior to the pandemic. - Bank Indonesia preannounced lowering of the inflation target by ½ percentage point at end-2017 and end-2019. - Volatile food: 16 percent of CPI basket. - Administered prices: 18 percent of CPI basket. - Lockdown phase - Aggregate demand plummeted, exerting downward pressure on prices; supply bottlenecks appear limited. - Some sectors saw increased demand (personal care and hygiene products). - Administered prices were little changed during the lockdown phase and into the early recovery phase, contributing to disinflationary pressure. - Note: There was a temporary increase in administered prices during the lockdown phase due to social distancing measures related to passenger travel. - Early recovery (around June/July 2020) - Q3 real GDP grew at 12.4 percent (q/q s.a.a.r). - Month-on-month consumer price changes continued to fall, dipping into negative territory. - Sectoral shocks began to reverse but slack remained significant. - Transition to new normal - Demand expected to recover and the output gap to narrow over time. - Inflation expected to pick up gradually; speed depends on economic recovery pace, food supply dynamics, and administered price path. ### Measurement complications during the pandemic - Consumption baskets changed substantially during the pandemic, especially in the lockdown phase (papers cited: Cavallo, 2020; Reinsdorf, 2020; Seiler, 2020). - Standard fixed-basket inflation measures may understate “true” price increases because items in higher demand during the pandemic generally see higher price increases (Reinsdorf, 2020). - Jaravel and O’Connel (2020) find evidence of an initial spike in consumer prices during the lockdown based on high-frequency data for advanced economies. - International evidence suggests mismeasurement from fixed baskets can be significant but mostly temporary, arguing against nonstandard changes to inflation baskets (Reinsdorf, 2020). ### Disaggregate analysis of CPI and PPI components - CPI and PPI disaggregation approach - CPI basket: 11 components and 43 subcomponents. - PPI basket: 9 components and 35 subcomponents. - Indonesian statistics office does not publish more disaggregated data. - CPI component contributions - Most CPI components contributed to the disinflationary trend; food prices accounted for the largest share of the decline. - Exceptions with increased demand: personal care products and health-related items. - Subcomponent shifts - Downward shift in inflation rates across most goods and services subcomponents. - Upper-end of distribution saw pronounced declines (education, restaurants, other services: declines of 2 to 5 percentage points in year-on-year inflation). - These components are contact-intensive sectors, consistent with disproportionate pandemic effects. - Inflation-at-risk and distributional analysis - Used monthly data on 90 CPI subcomponents and fitted a skewed-t probability distribution (Azzalini and Capitanio, 2003) characterized by mean, variance, skewness, kurtosis. - Inflation densities shifted to the left in 2020:Q2 (lockdown quarter) and narrowed compared to the prior quarter—initial shock relatively uniform across components. - Density widened in 2020:Q3, indicating greater dispersion across sectors during early recovery. ### Regional CPI: broad-based disinflation with geographic differentiation - CPI data for 90 cities show disinflation in most regions of Indonesia. - Larger cities generally had more elevated inflation rates than smaller cities; five cities with biggest CPI weights: Jakarta, Bekasi, Surabaya, Depok, and Tangerang. - Cities below the 10th percentile were mostly outside Java (particularly Sumatra); common drivers included significant declines in air transport and the food, beverages and tobacco component. - Cities above the 90th percentile include East Nusa Tenggara province and Papua province; elevated inflation driven by food commodities due to reliance on supplies from other parts of Indonesia. - Bali (Denpasar): inflation fell from 2.2 percent in June to 0.8 percent in September; regional GDP contracted by 12 percent y/y in 2020:Q3. ### Producer prices: sectoral drivers and trends - PPI headline movements - PPI inflation fell from 4.2 percent y/y in 2018:Q3 to 0.7 percent in 2020:Q1 and −0.7 percent in 2020:Q2. - PPI inflation recovered to 0.3 percent in 2020:Q3. - Sectoral deflation - Mining deflation: 10.8 percent y/y as of September 2020. - Utilities deflation: 0.6 percent y/y as of September 2020. - Coal exports account for 20.9 percent of total exports. - Coal prices dropped 51 percent from their peak in September 2018. - Dispersion - Most PPI sectors experienced limited disinflation; overall PPI well below the median of the distribution due to heavy weights of sectors in deflation. ### Property prices - Property price increases averaged around 2 percent for much of 2019, dipping to 1.2 percent in 2020:Q3. - Regional large contractions in 3Q observed in Batam, Balikpapan, and Banjarmasin; larger contractions mostly in large and medium properties. - Batam: residential property price inflation fell 3.8 percent in March and 1.6 percent in September. - Balikpapan: residential property prices deflation of 0.4 percent y/y in September, down from 0.1 percent deflation in March. - Balikpapan regional context: mining sector 33 percent of regional GDP; regional GDP contracted 6.2 percent y/y in Q2. - Banjarmasin: residential property price inflation fell from 0.5 percent in March to −0.4 percent in September. - Batam affected by declining tourism and trade due to proximity to Singapore; Balikpapan and Banjarmasin affected by weaker coal demand and pandemic-related mining weakness. ### Takeaways and outlook - Drivers and expected path - Disinflationary pressures arise from a complex interplay of pandemic-related and non-pandemic shocks. - Most shocks expected to dissipate gradually, with inflation returning toward pre-COVID-19 levels. - Staff projections for gradual return of inflation over 2021: - Headline inflation projected to climb from 1.7 percent in December 2020 to 3 percent next year. - Core inflation forecasted to climb from 1.6 percent in December 2020 to 2.6 percent at end-2021. - Bank Indonesia’s forecast envisions inflation gradually returning to the 3.0%±1% target range in 2021. - Factors affecting outlook - Aggregate demand recovery: Q3 expansion; high-frequency data suggest growth slowed in 2020:Q4 but recovery expected to regain traction. - Unwinding sectoral demand shocks: sectors hardest hit in Q2 showed strongest Q3 rebounds (transportation; travel and accommodation). - Mining sector idiosyncrasies: large shocks from coal and commodity price declines since late 2018, reinforced by the pandemic; commodity prices recovered from spring trough, supporting a pickup in producer prices in Q3. - Risks - Downside: pandemic weighs on activity longer than anticipated; disinflationary forces become entrenched; sustained undershooting of Bank Indonesia’s inflation target. - Upside: more vigorous rebound could build price pressures earlier than expected, lifting inflation more quickly. - Research note - The COVID-19 pandemic is an exceptional event with complex effects on output and inflation dynamics; additional research will be needed to investigate interacting shocks during and after the pandemic. *International Monetary Fund — Indonesia country chapter (excerpts provided in source content).* ### 1. Bank lending has decelerated rapidly in ### 1. Bank lending has decelerated rapidly in ### Recent lending trends - Loan growth fell from 8 percent (y/y) at end-2019 to below 2 percent (y/y) in July 2020, a record low in recent history. - Private and foreign banks’ lending contracted in June and July 2020; public banks (including SOEs) and regional banks maintained positive but declining loan growth. - Historical stress episodes (taper tantrum, 2015−16 China stock market turbulence, 2018 EM sell-off) showed a widening gap in loan growth between public and nonpublic banks, with public banks providing a cushion. ### Deposits and public banks - Deposit growth rose from around 6 percent (y/y) at end-2019 to about 9 percent by July 2020. - SOE banks recorded deposit growth of 13 percent (y/y) by July 2020. - Example from end-2018: public banks recorded deposit growth of 24 percent (y/y) versus 8 percent for private and joint banks. - A stronger deposit base for public banks may support their lending through safe-haven depositor behavior and fiscal resources tied to pandemic recovery programs. ### Monetary policy transmission and lending - Bank lending dynamics have not closely followed changes in monetary policy rates in recent years; credit growth and policy rate changes have been asynchronous at times, especially since 2018. - Pre-COVID-19 cuts in the BI policy interest rate in 2019 had yet to transmit to loan growth, which was on a sustained declining trend. - Other factors, such as retrenchment in credit demand during COVID-19, have affected bank credit growth. ### The role of bank characteristics (overview) - Three potential drivers of loan growth examined: size, liquidity, and capitalization. - Size: small banks may face funding constraints; larger banks may show lower loan growth due to maturity. - Liquidity: higher liquidity allows drawing down liquid assets to maintain and expand lending. - Capitalization: high capital ratios could support lending, but well-capitalized banks may also constrain lending to meet regulatory objectives. ### Data and empirical strategy - Monthly bank balance sheet data from CEIC (collected by OJK) covering 86 commercial banks, representing more than 80 percent of loans by commercial and rural banks in Indonesia. - Sample covers January 2012 to July 2020. - Estimation equation (monthly loan growth Δloan_it): - Δloan_it = α + γ X_it−1 + δ Stress_it + φ b_i + θ_t + ε_it - X_it−1 includes bank-level characteristics (size measured by share of bank assets in total bank assets; liquidity measured by liquid assets to total assets) lagged one period. - Stress dummy = 1 during 12 months following onset of shocks (taper tantrum, 2015−16 China turbulence, 2018 EM sell-off, 2020 COVID-19 pandemic). - b_i = bank fixed effects; θ_t = monthly time fixed effects. ### Key empirical findings - Liquidity importance: - More liquid banks have higher loan growth; liquidity provides a cushion during stress episodes. - Deposit growth is positively associated with loan growth (pass-through effect). - Public vs nonpublic banks: - Stress episodes are associated with lower loan growth from nonpublic banks but not materially lower lending from public banks. - Higher lending from public banks provided a cushioning effect during difficult times. - Robustness: - Results hold with additional controls including capital adequacy ratio and gross nonperforming loans (noting data limitations: monthly dataset lacks total equity; quarterly data reduce sample size). - No evidence that larger banks are associated with relatively larger loan growth on average. - Liquidity matters equally for loan growth during 2012−2017 and afterwards (per Table 7). - Deposit growth appears to play a lesser role in supporting loan growth after 2017. - Non-linearity and thresholds: - Spline fixed-effect regressions indicate the marginal positive effect of liquidity on loan growth is significantly higher when a bank’s liquidity ratio is below 34 percent of total assets (Table 8). - Distributional concern: - The 10th percentile liquidity ratio declined further in 2020 to 15 percent (Table 1), highlighting a decline in liquidity for banks already low on liquidity. ### Policy implications and recommended measures - Strengthen liquidity of solvent banks with sound fundamentals, especially those with relatively low liquidity that have declined further since the pandemic onset, to support lending and economic recovery. - Monitoring and providing funding to banks would facilitate reductions in banks’ interest rates by containing concerns about loss of funding or deposit bases. - Complementary government measures already noted: - SME loan guarantees (insurance covered loan with premium paid by the government). - SME temporary interest payment subsidies. - State fund deposits with low interest (conditional on further lending to MSME). - If credit growth does not recover in early 2021 (when weak borrower demand and lender risk aversion subside), consider additional support such as funding-for-lending programs: - Central bank provides relatively cheap funding to eligible banks with explicit lending requirements (e.g., to SMEs). - Important design elements: (i) incentives (favorable interest rates or higher funding caps for new lending); (ii) borrowing caps to limit bank borrowing and facilitate exit strategy. - Considering BI’s legal limitation to provide direct lending to banks, expanding fiscal resources to all eligible banks (beyond SOE banks) with appropriate safeguards and incentives to lend could further support credit creation. ### Nonfinancial corporate (NFC) vulnerabilities (summary introduction) - Sample of 459 Indonesian NFC firms at end-2019 showed comparatively weak debt service capacity and modest liquidity buffers versus regional peers: - About half of the firms used more than half of their operating income to cover interest payments. - Cash holdings covered less than 20 percent of outstanding current liabilities for these firms. - Without adequate policy support, about half of these firms might have been unable to cover interest payments and may have faced cash shortages by end-2020. - Their liabilities are expected to account for about one-third of total outstanding NFC debt, more than double the end-2019 share, posing potential strains on the banking system. - Policymakers face a tradeoff between preserving policy space and continuing support to save firms and jobs; understanding NFC scale and nature relative to regional peers can inform policy choices. *Prepared by Tidiane Kinda and Agnes Isnawangsih (APD).* ### 2. The challenge for Indonesian firms has been compounded by their financial ### 2. The challenge for Indonesian firms has been compounded by their financial ### Data and methodology - Dataset constructed from S&P’s Capital IQ database comprising 459 nonfinancial firms in Indonesia as of end-2019. - Sample accounts for about 62 percent of total NFC debt in the economy and 22 percent of GDP in terms of revenue. - Sample includes 439 publicly listed and 20 non-listed private firms; 23 of the 459 sample firms are state-owned enterprises (SOEs). - Dataset contains currency composition of individual firms’ outstanding debt. - ICR defined as EBITt / INTPt and decomposed as ICRt = ROAt / (EIRt × LEVt−1). - ROA = return on assets; EIR = effective interest rate; LEV = leverage. - Two approaches used to estimate end-2020 outcomes under the October 2020 IMF WEO: - Direct shock approach: apply operating income shocks (from Consensus earnings forecasts) and exchange rate and interest payment shocks (in line with October 2020 IMF WEO) to end-2019 operating income and interest payments. - Regression-based approach: predict ICRs at end-2020 using macroeconomic and global explanatory variables. - Cash position at end-2020 estimated by adding expected 2020 net cash flow to end-2019 cash balance; cash flow from operations assumed to decline in line with operating income shock; capex and debt refinancing set broadly consistent with past crisis magnitudes. ### Pre-pandemic financial vulnerabilities and firm characteristics - Long-run context: - Sustained decline in commodity prices since the Global Financial Crisis (GFC) in 2008 and heightened global trade tensions weighed on corporate performance. - Increased reliance on external (FX) debt in recent years contributed to persistently high currency risk premium and exchange-rate exposure. - Prudential requirements introduced in 2014 on hedging, liquidity, and minimum credit rating mitigated some risks. - Solvency (ICR) trends: - Since 2015, the median ICR of sample Indonesian firms remained relatively stable within a narrow range of 2 and 2.5. - Median ICR of 2 as of end-2019 was among the lowest in the region. - SOEs (“IDN-SOE”) experienced a noticeable fall in ICR over 2015−2019. - Profitability: - Median ROA of sample Indonesian firms was about 5.1 percent as of end-2019. - Comparable figure for other firms in ASEAN was 3.4 percent. - Private sector Indonesian firms (“IDN-Private”) profitability in 2019 comparable to 2015 despite slight declines in 2018−2019. - Profitability of Indonesian SOEs dropped sharply from a relatively high level in 2015. - Leverage and financing costs: - Median debt-to-assets ratio of sample Indonesian firms stood at 27 percent as of end-2019 (compared with 22 percent for other NFCs in the region). - SOE leverage rose from about 20 percent in 2015 to 36 percent in 2019. - Private sector firms reduced leverage over 2015−2019. - Effective interest rates increased since 2015 across the region; private sector Indonesian firms had markedly higher effective interest rates than peers; SOE effective interest rates declined over 2015−2019 indicating subsidization. ### Debt-at-risk and industry concentration - Debt-at-risk (share of outstanding NFC debt held by firms with ICR below 2): - As of end-2019, debt-at-risk share in Indonesia was about 56 percent. - Singapore comparison: 35 percent. - Indonesia’s debt-at-risk share higher than other major ASEAN economies and the greatest among them. - Industry-level vulnerabilities: - Utilities, materials, energy, and consumer staples and discretionary sectors had weakest debt service capacity in 2019. - Debt of low-ICR firms in utilities accounted for over 20 percent of total outstanding NFC debt in Indonesia in 2019; materials about 10 percent. - High share of risky firms in consumer staples and discretionary raises employment-at-risk concerns due to high labor intensity. ### FX debt and net FX exposure - FX debt statistics and patterns (pre-pandemic): - As of end-2019, sample Indonesian firms held about 39 percent of their outstanding debt in foreign currencies. - Average share for other sample firms in major ASEAN economies was about 16 percent. - FX debt share increased over 2016−2019, especially among small-sized firms; median FX debt share rose more than asset-weighted share. - FX debt share increased relatively more for 75th percentile firms (still comparable to post-GFC levels). - Private sector firms and firms in utilities had relatively higher FX debt shares in 2016 and their shares increased relatively more in 2016−2019. - Net FX exposure gauge: - Industry-level comparison of average foreign sales (proxy for natural hedges) with average FX debt share shows Indonesian firms generally have much higher FX debt shares relative to foreign sales than ASEAN peers. - Utilities’ high FX debt share and lack of foreign-sales natural hedges warrant caution. - Hedging regulation note: - Regulation (No. 16/21/PBI/2014) requires non-bank corporations with external debt to meet a minimum hedging ratio of 25 percent of the negative difference between maturing foreign currency assets and foreign currency liabilities in the next three months and in the next three to six months. - Bank Indonesia’s March 2020 Financial Stability Review: about 93 percent of 3,807 nonbank corporations with external debt met the regulatory minimum hedging ratio at end-2019:Q4. ### Liquidity buffers - Cash ratio (cash and cash equivalents relative to total current liabilities) as of end-2019: - Median cash ratio for Indonesian firms: about 20 percent. - Asset-weighted average cash ratio for Indonesian firms: about 47 percent (higher than Vietnamese firms about 20 percent among ASEAN-6). - Median cash ratio private sector firms: 19 percent; SOEs: 34 percent. - Median cash ratios by industry: materials 13 percent; consumer staples 12 percent. - Comparison context: - Firms in other ASEAN economies exhibited combinations of low cash buffers with high ICRs (e.g., Vietnam) or low ICRs with large cash buffers (e.g., Singapore). ### Expected impact of COVID-19 (under October 2020 IMF WEO) - ICR and liquidity projections at end-2020 (without ample policy support): - Share of sample firms with ICRs below one could reach between 47−50 percent by end-2020. - About 45 percent of sample firms could face cash shortages at end-2020 without liquidity support measures. - Q2 2020 actual data for a sub-sample of 397 firms: - Share of firms at risk (ICR below one) about 49 percent. - Median ICR about 0.8. - Cash ratio fell to 19 percent from 20 percent at end-2019. - Decline in ICR primarily reflected lower corporate sales and profitability; Bank Indonesia policy rate reduction provided some relief on interest payments. - Industry heterogeneity in COVID-19 impact: - Consumer discretionary expected to be hardest hit in terms of share of firms facing interest payment difficulties and cash shortages. - Energy sector expected to have relatively small share cash-strapped due to strong 2019 cash position (median cash ratio of 43 percent). - Materials and industrials identified as vulnerable; together account for a significant share of Indonesia NFC debt-at-risk in 2019 (28 percent). - Debt-at-risk escalation by end-2020: - Share of NFC debt held by firms with ICR below one expected to rise to 32 percent by end-2020. - Share with ICR below two expected to rise to 88 percent by end-2020. - These levels would be 2½ and 1½ times higher than 2019 levels, respectively. - Further increases during remainder of 2020 expected to be relatively small given mobility and activity improvements. ### Policy implications and recommendations - Immediate and sustained measures: - Authorities have implemented emergency support measures: interest rate subsidies, bank loan moratoria for MSMEs, credit guarantees, corporate income tax reduction, liquidity support, and regulatory easing to facilitate loan restructuring. - Given the magnitude and persistence of the COVID-19 crisis, some targeted assistance measures should be sustained or scaled up as needed until recovery firmly takes hold. - Authorities’ plan to prioritize support for high value-added sectors in 2021, including through an ambitious vaccination program, would help accelerate corporate recovery. - Structural and social measures: - Strengthen the insolvency framework. - Strengthen social safety nets. - Implement active labor market policies to increase employment opportunities and minimize long-term scars on the economy. *Source: IMF chapter text (firm-level analysis using S&P Capital IQ; estimates under October 2020 IMF World Economic Outlook).* ### 2. In light of the large uncertainty surrounding the near-term economic prospects and the ### 1idnea2021002 - 2. In light of the large uncertainty surrounding the near-term economic prospects and the ### Approach I. Regression-Based Method — design and implementation - Sample: 29,161 firm-year observations over the period of 2002−2019 from the sample ASEAN-6 economies. - Objective: obtain alternative forecasts of ICRs for end-2020 (ICR2020,j) by forecasting subcomponents and constructing firm-specific ICR projections. - ICR construction: ICR = ROA / (EIR × LEV) - ROA = return on assets (firm-level). - EIR = effective interest rate. - LEV = debt-to-assets ratio. - Regression specification (for each subcomponent y = ROA, EIR, LEV): - Dependent variable: yi,j,k,t (subcomponent for firm i, industry j, country k, year t). - Includes: lagged dependent variables, country-level domestic variables Xk,t (including real GDP growth and lending rates), external sector variables Ek,t (trade partners’ GDP growth weighted by country k’s exports, bilateral exchange rate to USD), global variables Wt (commodity prices and LIBOR), a dummy for year 2010 onward, industry fixed effects δj, country fixed effects μk, and an error term εi,j,k,t. - Macroeconomic inputs: IMF WEO (including global assumptions) for 2002−2019 and IMF International Financial Statistics for lending rates; WEO October vintage used for 2020−21 projections. - Model choices and robustness: - Different specifications and alternative measures considered (country- and industry-level regressions, variable measures). - Prioritized out-of-sample forecasting performance: specification chosen to provide best fit to actual data until 2019. - Model predictions are robust to different regression specifications and provide good fits for the ICR. ### Approach II. Direct Method Using Consensus Forecast Earnings — design and implementation - Strategy: apply shocks directly to ICR subcomponents — operating income (numerator) and interest payment (denominator). - Operating income shock: - Use consensus earnings forecasts for individual firms for FY 2020 from the January 2 vintage and the June 30 vintage. - For each firm, compute percentage change of earnings forecasts between these vintages; where forecasts unavailable, apply industry-median earning shock to individual firms’ 2019 earnings: - EBIT2020,j = (1 + θ × EBIT shocki / 100) × EBIT2019,j - θ is a scaling factor set so that the ASEAN-6 median ICR in 2020 from Approach I and Approach II are equal to each other (at about 1.2). - Note: June 30 vintage is the reference vintage as it is closest to the magnitude of downturn implied by the October 2020 IMF WEO; results robust to alternative vintages. - Interest payment shocks: - Two shocks applied: exchange rate shock and interest payment shock. - Exchange rate shock = percent change of bilateral exchange rate vis-à-vis USD between end-2019 level and projected end-2020 level in October 2020 IMF WEO. - Exchange rate shock applied to actual foreign-currency-denominated portion of each firm’s outstanding debt as of end-2019 (not imputed amounts). - Interest payment shock set at 0 (conservative relative to scenarios with slightly declining interest rates). - Expected interest payment formula: - INTP2020,j = [FX debt share2019,j × (1 + FX shock2020 / 100) + local currency debt share2019,j] × (1 + INTP shock2020 / 100) × INTP2019,j - Calibration note: - θ scaling used because without it Approach II would yield a much higher ASEAN-6 median ICR than Approach I. ### Cash Flow Analysis and balance-sheet adjustments for end-2020 cash position - End-2020 cash position specification: - cash2020 = cash2019 + EBIT shock × (cash flow from operations2019) − capital expenditure2020 − debt amortization2020 + net interest payment2019 − dividend payment2020 - Assumptions and parameter values: - Capital expenditure2020 = 0.75 * capital expenditure2019 - Debt amortization2020 = 10 percent of maturing debt (i.e., debt rollover ratio = 90 percent) - Dividend payment2020 = 0 - Rationale and historical calibration: - CAPEX: sample median during the GFC in 2009 was about 80 percent of the level in 2008; 0.75 scaling chosen as plausible. - Debt rollover ratio: set at 90 percent of maturing debt, slightly lower than the median level observed in 2009 (about 100 percent) in the ASEAN-6 sample. - In normal times, sample median rollover ratio estimated at about 110 percent, implying a nominal debt growth of about 10 percent. ### Foreign Exchange Markets and Instruments in Indonesia — overview and recent developments - Market growth and composition: - Average daily turnover of the Indonesian rupiah (IDR) increased from US$4 billion to US$27 billion during the past decade (BIS, 2019). - Gross capital flows grew by over 200 percent between 2007−2019; gross trade flows increased by 34 percent in same period. - Indonesia recorded an increase in FX turnover of over 1,200 percent over the decade, similar to Brazilian real and Indian rupee. - Main instruments for IDR transacting: FX swaps and outright spot deals; FX swaps trading volume exceeds outright forwards substantially. - Offshore trading has grown strongly and now accounts for the bulk of IDR turnover volumes (offshore trading has grown over six times). - Offshore growth accompanied by three-fold increase in nonresident participation in local bond markets, raising hedging needs. - FX market liquidity and stress episodes: - FX funding and market liquidity pressures spike during stress episodes (global financial crisis, taper tantrum, COVID-19 selloff). - COVID-19 globalized market selloff resulted in a tripling of both market and funding liquidity measures. - Factors contributing to moderated liquidity pressures since the GFC include: changing role of liquidity providers, growth of offshore trading reducing transaction costs, participation of foreign intermediaries, and improvements in technology. - Bank Indonesia (BI) FX policy and instruments: - Regime: flexible exchange rate; BI aims to mitigate excessive rate volatility and sustain adequate market liquidity. - Three main FX instruments BI deploys: - Direct spot interventions (buying and selling FX). - Simultaneous spot and forward market interventions via FX swaps (selling rupiah in spot and selling FX in forward market). - Forward market interventions via selling non-deliverable FX forward settled in rupiahs (DNDF). - Size and usage of FX swaps and DNDF: - Outstanding stock of FX Swaps (DNDF) was US$1.5 (3.8) billion at end 2020:Q3, with total issuance of US$123 (58) billion. - Average daily issuance: FX swaps US$290 million; DNDF US$98 million. - These represent about 17 percent of the onshore FX swap and outright forward market turnover respectively. - FX swaps are used primarily as a domestic monetary operations instrument for both liquidity absorption and injection; on average represent approximately 25−35 percent of total liquidity injection together with bonds and term repos. - First BI FX swap auction held on July 18, 2013, offering an oversubscribed target of US$500 million on 1-, 3- and 6-month tenors. - Costs and reserve use: - Estimated size of spot interventions during stress episodes proxied by change in reserve holdings: interventions in the range of 1−2 percent of GDP. - Total cost of holding reserves (cost of rolling over FX position) estimated at around 0.57 percent of GDP (monthly average, 2016−2020). - Marginal cost of holding reserves (economic opportunity cost of increasing central bank’s FX position) estimated at 5 percent per unit (monthly average, 2016−2020). - DNDF characteristics and market integration: - DNDFs launched since November 2018; settlement paid in rupiah; largely issued at 1-month tenor (3-month tenor offered since February 2019). - DNDF market is dominated by onshore banks and liquidity is limited. - Onshore DNDF pricing differs from offshore NDF; DNDF typically below NDF market rate; DNDF fixed daily by BI using Jakarta Interbank Spot Dollar Rate. - DNDF market remains a tiny fraction of the NDF market; participation restrictions eased following the COVID-19 outbreak and global liquidity stress. - Information spillovers and price discovery: - Some evidence of one-directional spillover from offshore NDF to onshore spot and forward rupiah markets. - IDR is the most traded NDF among South East Asian currencies; empirical evidence points to information transmission effects from NDF to onshore forward and spot rates. *Source: 1idnea2021002 - 2. In light of the large uncertainty surrounding the near-term economic prospects and the* ### 11. The issuance of FX swaps and DNDF in the forward market typically coincides with ### 11. The issuance of FX swaps and DNDF in the forward market typically coincides with ### C. Effects of FX Intervention Instruments on FX Liquidity and Forward Exchange Rate - Empirical background: - Currency swaps and auctions of derivative contracts have material effects on exchange rates; examples cited include Kohlscheen and Andrade (2014), Chamon and others (2015), and Nedeljkovic and Saborowski (2017). - Domanski and others (2016) show FX spot intervention increases FX market liquidity for Brazil, Peru, Russia, and Turkey. - Data and measurement approach: - Daily data used on forward and spot exchange rates, bid-ask spreads, and daily issuance of FX Swap and DNDF. - Two price-based FX liquidity indicators constructed: - Funding liquidity = spread between the forward and spot exchange rates (IDR against the USD) normalized by the spot rates. - Market liquidity = spread between the ask and bid spot rates (IDR against the USD) normalized by the mid-price. - Simple OLS regressions used to assess association between FX swaps, DNDF, and FX liquidity, with interaction for stress episodes proxied by the VIX volatility index. - Usage statistics since activation: - Activation years: 2013 (FX swaps) and 2018 (DNDF). - Total intervention days: 719. - Total amounts issued: approximately US$150 billion. - Average daily issuance: FX swap = US$290 million; DNDF = US$98 million. - Combined issuance represents 17 percent of the daily onshore non-spot trading volume. - DNDFs have been used more frequently, with intervention carried out every other day. - Stylized facts on liquidity pressures: - FX liquidity pressures spike during periods of global stress; highest stress during the global financial crisis followed by the taper tantrum. - During the COVID-19 crisis, both funding and market illiquidity rose relative to the EM selloff period. - The introduction and availability of the FX swap instrument is associated with higher FX liquidity (both market and funding). - The introduction and availability of DNDF is associated with lower FX liquidity (both market and funding). - Caution: availability dummies mask actual amount and timing of interventions and localized stress episodes. - Interventions' effects during stress (VIX interactions; amounts measured in billions of U.S. dollar): - FX liquidity tightens when the VIX rises. - Use of DNDFs and FX swaps mitigates VIX-induced FX funding liquidity stress: - An intervention of US$250 million is associated with reducing VIX induced liquidity stress by approximately 18 percent (FX swaps) and 27 percent (DNDF). - Effects on market liquidity are generally not significant, though DNDF interventions are associated with reducing FX market illiquidity during times of stress. - Results robust to dropping zero-issuance days when instruments available. ### Effects on forward exchange rate (GARCH analysis) - Model specification: - Daily multivariate GARCH (1,1) with forward return equation and conditional variance equation; explanatory variables include spot returns, DNDF / FX swap quantity or existence, VIX, and interactions. - Main findings: - Uninteracted effects of forward interventions are positive on both the value and volatility of the forward exchange rate—consistent with endogeneity where volatility may cause intervention. - Interaction terms with VIX are the coefficients of interest. - Interaction results: - Interactions of DNDFs and FX swaps with VIX are negatively associated with the value of the forward exchange rate, but these effects are imprecise (insignificant). - In the conditional volatility equation, interactions of DNDFs and FX swaps with VIX are negative and significant, indicating that increasing the size of DNDF and FX swap interventions dampens forward exchange rate volatility during times of stress. - An intervention of US$50 million is associated with reducing VIX induced volatility stress by approximately 65 percent (FX swaps) and 21 percent (DNDF). - Context: Indonesia’s onshore market where interventions are carried out represents 20 percent of the trading market; despite small market size, interventions have economically sizeable effects on offshore forward rates. ### D. Further Considerations on FX Forward Intervention Strategies - Theoretical and practical implications: - Eaton and Turnovsky (1984): spot and forward interventions have equivalent effects on the spot exchange rate absent convertibility risk; with convertibility risk, forward interventions may have asymmetric effects. - Walker (2019): a change in central bank forward position has a greater effect on the forward rate than the spot rate. - Forward sale of U.S. dollar can reduce the basis spread for given interest rates, affecting divergence from covered interest rate parity (CIP) and potentially encouraging arbitrage during stress. - Supply of currency swaps provides an alternative for traders demanding currency for speculative or hedging reasons (Frankel and Froot (1990); Taylor and Allen (1992); Menkhoff and Taylor (2007)). - Covered Interest Parity (CIP) and arbitrage: - Under CIP (absent counterparty risk), forward premium = nominal interest-rate advantage; synthetic dollar rate from swap market should equal direct dollar rate (Du and others, 2018). - If synthetic dollar rates are cheaper than direct dollar rates (positive CIP deviation), arbitrage strategies can yield annualized risk-free profit. - Interventions that reduce the forward premium can narrow positive CIP divergence or widen negative CIP divergence, thereby affecting arbitrage opportunities. ### CIP deviations and episodes for Indonesia - Historical observations: - CIP deviations emerged during the GFC among advanced countries and persisted afterward in some cases (Du and others, 2018; Cerutti, Obstfeld, and Zhou, 2020). - At onset of the GFC in Indonesia: - Synthetic dollar rates turned negative. - Banks could earn substantial risk-free arbitrage gains by borrowing in IDR and investing in dollar. - JIBOR interbank borrowing rates were almost six times the LIBOR rates. - Expected Rupiah depreciation at annualized 25−35 percent in 2008:Q4 contributed to profits. - Since the GFC, CIP deviations in Indonesia have narrowed but re-appear during acute stress episodes and may not fully close after opening, possibly due to limits on arbitrage. - COVID-19 crisis: - CIP deviations widened reflecting increased funding liquidity pressures. - An increase in the forward premium increases the cost of hedging and can widen CIP deviations absent similar movement in interest rate differentials. - During COVID-19, CIP deviations (direct vs synthetic dollar rates) turned negative such that it was more profitable to fund in the synthetic dollar risk-free rate and invest in the direct dollar risk-free rate. - A hedging-based investment into domestic bonds offered lower risk-adjusted returns when synthetic domestic yields became higher than direct domestic yields, signaling increased volatility of holding such investments. - These deviations persisted for several months after COVID-19 market pressures dissipated. - Capital flow implications: - Pricing issues around obtaining FX during stress can put further pressure on capital flows. - Example: 2008:Q2 onset of GFC—synthetic dollar rates turned negative, acquisition of external financial assets rose, and capital inflows reduced. - Possible channels: - Arbitrage trading using cheap rupiah liquidity to invest abroad for short-term profits. - Lower returns from hedging-based investment strategies into Indonesia relative to direct dollar investments, potentially reducing capital inflows. - Similar dynamics documented for other emerging economies (e.g., Russia). ### E. FX Intervention During Crises - Common FX easing measures used by emerging market economies in crises: - Direct supply of FX through FX swaps. - Lowering macroprudential limits relating to foreign currencies. - Expanding counterparties (including nonbanks) for FX provision. - Other measures referenced across episodes include: - GFC: easing terms of existing FX facilities (extending maturities, broadening collateral), introducing new FX liquidity facilities (dollar repo and swap facilities), widening counterparties to nonbank financial institutions and key nonfinancial institutions, relaxing foreign exchange liquidity limits, transferring foreign currency deposits held overseas to domestic banks, guaranteeing FX liabilities, lowering taxes on FX transactions, lowering required reserve ratio for bank FX liabilities, and shifting currency structure of required reserves away from foreign exchange. - COVID-19: lifting levies on non-deposit liabilities of FIs in FX (Korea), lowering FX liquidity coverage ratio (Korea), foreign currency sales from proceeds of wealth funds (Russia), expanding list of foreign currencies acceptable as FX reserves (Chile, Russia), sale of FX through FX swaps or dollar repos (Chile, India, Russia, Mexico, England, Australia), lowering FX reserve requirement ratios (Turkey, Indonesia). ### F. Summary and Policy Implications - Key empirical findings: - Both FX swaps and DNDF have been actively used by Bank Indonesia since activation (2013 for FX swaps; 2018 for DNDF); total issuance across 719 intervention days is approximately US$150 billion, average daily issuance US$290 million (FX swaps) and US$98 million (DNDF). - FX liquidity pressures spike during global stress; interventions mitigate funding-liquidity stress induced by global volatility (VIX). - An intervention of US$250 million reduces VIX-induced funding liquidity stress by approximately 18 percent (FX swaps) and 27 percent (DNDF). - An intervention of US$50 million reduces VIX-induced forward volatility by approximately 65 percent (FX swaps) and 21 percent (DNDF). - Interventions have limited and imprecise effects on the level (value) of the forward exchange rate but significant dampening effects on forward volatility. - DNDFs have become more frequently used than FX swaps (intervention roughly every other day for DNDF). - Policy considerations: - Forward-market interventions can reduce FX funding stress and forward volatility during global episodes but may have asymmetric effects versus spot interventions and can influence arbitrage opportunities through CIP deviations. - Intervention design should account for: - The asymmetric impact of forward vs spot interventions (convertibility/counterparty risk). - Potential to narrow positive CIP deviations or widen negative CIP deviations, thereby altering arbitrage incentives. - The small size of the onshore market (20 percent of trading market) relative to offshore markets where effects are observed. - Trade-offs between providing FX liquidity, influencing hedging costs, and possibly encouraging arbitrage or affecting capital flows. - During crises, a broader menu of FX easing measures (direct FX supply, macroprudential easing, expanding counterparties, FX swaps/dollar repos, reserve requirement adjustments) have been used by other countries and may be considered as complements to forward interventions. *Source: IMF staff analysis in "Indonesia" chapter on FX swap and DNDF interventions (text and figures cited within the provided content).* ### 25. The main points from the analysis in this paper are the following: ### 25. The main points from the analysis in this paper are the following: ### FX market structure and interventions - The FX market in Indonesia is large and active and has grown considerably since the GFC; the market is dominated by offshore trading activity. - Onshore activity has become more diversified and the use of FX swaps has gained prominence. - BI uses a variety of instruments to intervene in FX markets: - Direct intervention in the spot market. - FX swaps (considered a domestic OMO tool). - Non-deliverable forwards (DNDFs). - The daily size and frequency of interventions is nonnegligible: - DNDF are carried out almost every other day. - The total daily volume of interventions together represent 17 percent of the onshore non-spot trading volume. ### Evidence on effectiveness and trade-offs of instruments - Suggestive evidence indicates interventions in the forward market are easing pressures on FX liquidity and forward exchange rate volatility. - Instrument-specific effects: - DNDFs can alleviate pressures on both funding and market liquidity. - FX swaps have similar effects on funding liquidity. - Both DNDFs and FX swaps have some effects on reducing forward exchange rate volatility (over and above the spot) during stressed times, with the effect relatively larger for FX swaps. - Trade-offs and caveats: - Use of FX swaps must be balanced considering simultaneous effects on domestic liquidity expansion and the supply of dollar collateral in episodes of stress. ### Monetary stance, arbitrage, and crisis context - A highly accommodative monetary stance can open arbitrage opportunities that may further pressure capital flows: - During the GFC, the synthetic dollar funding rate turned negative (and much lower than the direct dollar rate), which may have induced further outflows based on short-term, arbitrage-based profit seeking. - Increased uncertainty and counterparty risks are prevalent in such episodes. - Identifying the nature of a shock is important when evaluating policy options: - The GFC was a financial crisis stemming from endemic vulnerabilities in advanced-country financial sectors. - The EM selloff and taper tantrum episodes were localized at the EM level and mainly a consequence of U.S. monetary tightening. - The COVID-19 shock has materialized more as a global shock, with a generalized sell-off at the peak of uncertainty during March and early April, followed by a more heterogeneous recovery. ### Digitalization, financial inclusion, and MSME financing (related findings from the chapter) - Structural constraints and financing gaps: - Almost half of the population in Indonesia is still without a bank account; close to 6 percent of the global unbanked adults reside in Indonesia. - The ratio of bank credit to GDP remains low at 35 percent (compared with 60 percent of GDP in mid-1997). - In 2017, 17 percent of Indonesians borrowed from a formal bank or microfinance institutions, while 36 percent borrowed from informal sources. - There are about 63 million MSMEs in Indonesia (99 percent of all firms), accounting for 97 percent of employment and contributing close to 60 percent of GDP. - Only close to one-fifth of bank loans are to MSMEs; more than 70 percent of MSMEs lack access to credit. - The MSME financing gap is estimated to be US$165 billion, which equates to approximately 19 percent of Indonesia’s GDP in 2017. - COVID-19 impact on MSMEs and access to finance: - Around 75 percent of Indonesia’s MSMEs reported a lack of working capital as the main concern to retain their business during COVID-19. - Only 1 percent of MSMEs surveyed state that they borrowed from banks to cope with the COVID-19 shock; a larger proportion utilized nonbank financial institutions and digital finance platforms. - Digitalization trends and scale: - Indonesia’s digital landscape has expanded rapidly, concentrated in e-commerce and digital financial services (DFSs). - As of end-2019, there were more than 200 e-commerce players and more than 350 fintech players. - From 2017 to August 2019, e-commerce transactions grew by 137 percent (CAGR); penetration of e-commerce users increased to 44 percent in 2019 from 35 percent in 2018. - Transactions at the four largest e-commerce sites are estimated to double to US$29 billion in 2020 from US$14 billion in 2019. - E-money transaction values increased by 114 percent between 2015 and 2019; transaction volumes increased by 91.4 percent. - During the pandemic (May 2020), e-money transactions and digital banking transaction values grew at 14.3 percent (y/y) and 30.3 percent (y/y), respectively (declining later to 14.8 percent and 2 percent, respectively, in October 2020); electronic card transaction values declined. - Digital lending expansion: - At end-October 2020, there were 155 registered P2P lending firms (144 conventional and 11 sharia lenders) with total assets of IDR 3.4 trillion; 36 have been licensed by OJK. - There are 3 licensed equity crowdfunding platforms and 86 digital financial innovation organizers (IKD). - The expansion of “alternative data” from e-commerce and digital payments is being used by digital lenders to identify creditworthy clients and help increase access to loans for those not included in the credit registry. *Source: 25. The main points from the analysis in this paper are the following:* ### 13. Fintech lending remains small but was growing at a rapid pace prior to the COVID-19 ### 13. Fintech lending remains small but was growing at a rapid pace prior to the COVID-19 ### Fintech lending growth and scale - The accumulation of P2P loans in Indonesia grew at more than 230 percent (y/y) to IDR 88.4 trillion in January 2020. - Total outstanding P2P loans stood at IDR 13.5 trillion in January 2020, with an increase of 137 percent (y/y). - The volume of P2P lending remains small at less than 0.1 percent of GDP, and has no systemic relevance so far. - Business loans constitute more than 85 percent of total fintech loans. - The number of accumulated/active fintech borrowers in September 2020 are at close to 29.2/9.3 million. - The average duration of loans from P2P lending platforms span from 10 days to one year. ### COVID-19 impact on fintech lending - Outstanding P2P credit has declined by 8 percent since December 2019, in line with a decline in overall bank credit growth. - Demand weakened and fintech firms scaled down new P2P lending to preserve liquidity and manage credit risks. - Much fintech lending targeted small borrowers who are disproportionately affected by the crisis, risking a sharp rise in credit costs. - NPL ratio in the fintech sector increased from 3.7 percent at end-2019 to 8 percent in July 2020; if restructured loans are included, this ratio is 10 percent. - Out of the 1.96 million loan restructuring requests received by 88 P2P lending platforms, only 34 percent were granted. - The total amount of reported outstanding loans requested to be restructured in May stood at IDR 1.08 trillion (8 percent of total loans), with just IDR 236.9 billion (2 percent of loans) approved by the lenders. ### Digitalization in the traditional financial sector - Indonesian banks continue to dominate deposits and lending while increasing digitalization efforts to raise efficiency and improve service. - Current regulations allow traditional brick-and-mortar banks to offer digital products and services, but they do not allow for a completely “virtual/digital bank.” - Banks and multi-finance companies are investing funds in P2P lending platforms to access MSMEs and reduce credit assessment costs through outsourcing. - P2P lending platforms are not allowed to provide on-balance sheet loans; banks and multi-finance companies are one of their funding sources. - A BI survey (2018) indicated the majority of banks are at low levels of digitalization; not a single bank had adopted biometric verification, blockchain, big data analysis, or artificial intelligence. - Legacy systems, lack of technically skilled staff, and high investment costs for small banks create a digital divide between BUKU 4/3 banks and smaller banks. ### Digital financial inclusion and usage - A “digital” financial inclusion index (Sahay and others, 2020) combines access and usage of mobile/internet-enabled payment services; a “traditional” financial inclusion index captures bank-driven access and usage. - Indonesia’s improvement in financial inclusion between 2014−17 is driven by both digital and traditional financial services. - Rapid growth in access to DFSs agents, with over 500,000 agents across the country as of June 2020, high mobile subscription and improvements in internet penetration are main drivers of digital financial inclusion. - Usage of mobile and internet banking transactions remains low in volume and value. - Two-thirds of the population was financially illiterate as of 2019. - More than half of the overall population does not use the internet; Indonesia has one of the lowest internet penetration rates in the ASEAN region. - More than half of the population does not own a smartphone; two-thirds of the population remains financially and digitally illiterate. - Mobile phone ownership is at 81 percent among higher-income population and 63.5 percent among low-income individuals; the rural-urban mobile phone ownership divide is close to 12 percent. ### Regional concentration and market dynamics - Regions with higher access to bank loans tend to have higher P2P lending. - Fintech loans are mostly small, uncollateralized, short duration, and to small clients, while traditional institutions focus on big clients and larger collateralized loans of longer duration. - Fintech lenders compete mainly with informal money lenders, microfinance institutions, and small banks. - Fintech lending is geographically concentrated in Java, accounting for close to 85 percent of fintech lending. - Uneven growth attributed to Java’s larger economy, higher financial literacy levels, and superior ICT infrastructure outside Java. ### Economic impact of digitalization - McKinsey and Company (2016): digitalization could expand Indonesia’s economy by 10 percent by 2025 and add 3.7 million jobs. - Kinda (2019): firms engaged in e-commerce in Indonesia are at least 30 percent more productive and export at least 50 percent more on average. - Asian Development Bank (2020): improvement in productivity through adoption of digitalization, robotization and artificial intelligence technologies could add an additional accumulated growth of 11 percent of GDP during the period 2020−40. - An increase in Indonesia’s adoption of digital financial payments to the level in China could raise Indonesia’s real GDP growth rate by 5 percentage points (instrumental variable regression using data for 52 EMDEs, relating DFS usage to average growth over 2011−18). ### Role of digitalization during the COVID-19 pandemic - Digitalization provided resilience by: - enabling firms and workers to maintain some operations during lockdowns; - supporting consumption through online purchases amidst social distancing; - ensuring continued access to financial services, particularly digital payments; - facilitating delivery of government social assistance disbursements; - providing alternative sources of financing for small businesses and borrowers where traditional credit declined. - Digital payments data allow authorities to better track consumer spending patterns in real-time and target support measures. ### Key challenges and risks - Indonesia ranks low in the World Economic Forum’s Network Readiness Index. - Digital infrastructure gaps, rural-urban digital divides, lack of skilled workers, financial constraints for firms, technical uncertainties, and resistance to change hinder technology adoption. - Increased digital use heightens vulnerability to data and privacy risks, loss of connectivity due to natural disasters, cyber-attacks, money laundering and terrorist financing. - Illegal fintech lenders proliferated during the pandemic; OJK closed down close to a thousand illegal P2P lending companies between January-September 2020, totaling 2,840 illegal entities closed since 2018. - Fintech retrenchment could curtail access to finance for SMEs and low-income households; regulatory support measures channeled mainly through banks could exacerbate procyclicality in fintech lending. - Digital divide among banks may shift deposits to larger banks, increasing risk of financial exclusion if smaller banks scale back operations. - Excessive financial inclusion without sufficient regulation could expand risky credit access during downturns. ### Policy priorities and recommendations - Focus policy on three pillars: - (i) ensuring equal opportunity and access to avoid a digital divide and rise in inequality (supply-side); - (ii) addressing constraints to adoption (demand-side); - (iii) high-quality supervision and regulation balancing innovation and competition with risks related to financial stability, integrity, cyber-risk and consumer protection. - Invest in and expand digital-friendly infrastructure: - reliable electricity and logistics for e-commerce pickups/delivery; - expand access to quality internet coverage and high speed mobile internet, particularly in rural areas; - expand reach of DFSs agents through appropriate incentives. - Address barriers to technology adoption across firms and financial services: - expand and strengthen programs investing in technology awareness, financial digital literacy and worker skill upgrades; - encourage entry of foreign skilled labor to fill gaps and stimulate innovation; - further expand the reach of G2P to increase account ownership, including collaboration with nonbank financial service providers; - require more recipients to receive government assistance via digital transfers to their own accounts to boost account ownership; - expand noncash G2P beyond state-owned banks to include fintech service providers; - build strong cybersecurity policy and regulations for consumer data privacy and protection. - Strengthen regulation and interagency cooperation: - expand the regulatory perimeter to cover unlicensed nonbanks, cooperatives and fintech companies currently outside OJK and BI oversight; - improve interagency coordination between BI and OJK to regulate, license and supervise fintech activities explicitly; - ensure the e-commerce and fintech landscape remains sufficiently competitive post COVID-19 to continue providing low-cost services. *Source: 1idnea2021002 - 13. Fintech lending remains small but was growing at a rapid pace prior to the COVID-19 (IMF).* ### Box 1. Recent Policy Efforts to Support Integration of the Digital Economy and Finance: Before and ### Box 1. Recent Policy Efforts to Support Integration of the Digital Economy and Finance: Before and During COVID-19 ### Policy framework and pre-COVID initiatives - Development of DFSs is embedded in the National Financial Inclusion Strategy (2016). - Reforms allowed e-money issuers (banks and nonbanks) and banks to engage agents to expand service delivery outreach; agents are now present in all provinces and in 99 percent of the districts. - Electronification of central government social aid program disbursements and local government financial transactions has been advanced. - In end-2019, BI formulated the Indonesia Payment System Blueprint 2025 with initiatives to: - support the integration of digital economy and finance through digital banking transformation and interlink between banks and fintech using open Application Programming Interface (API) standards; - strengthen the configuration of retail payment systems to make it more efficient and safer; - strengthen financial market infrastructures through modernizing the infrastructure and in accordance with international best practices; - establish a public data infrastructure that will connect all payment transactions and manage the flow of payment data, including digital ID; - strengthen the current regulatory, licensing and supervisory framework on the payment system to improve efficiency, market discipline, integrity, risk management, and consumer protection. - Regulatory scope: BI regulates e-money, payment processing and fund transfers in the e-retail and marketplace industries; OJK regulates P2P lending and equity crowdfunding. Regulatory status of “marketplace lending platforms” is the same as nonbank financial institutions. - Both BI and OJK adopted the regulatory sandbox, preceded by mandatory registration of fintech business. The registration is a two-stage process: (1) registration with intent to operate; (2) licensing requiring proof of operational reliability (platform risk mitigation, customer data safety, fraud detection and prevention). A sandbox system allows operational reliability testing. ### Measures introduced during COVID-19 to encourage digital payments - April 2020: BI relaxed rules on credit cards, reduced the costs of the National Clearing System (SKNBI), and supported electronic disbursements of various government social assistance programs. - BI promoted MSME adoption of cashless digital payments: - August 2019: BI introduced the Quick Response Indonesian Standard (QRIS) as the national standard QR code payment to facilitate interoperable payments for MSMEs. - More than 5.46 million merchants have adopted the QRIS since it came into effect; close to 43.6 percent (2.4 million) of adopters have joined during the pandemic since March 2020. - April 2020: BI temporarily reduced the Merchant Discount Rate (MDR) on QRIS to 0 percent for micro-businesses. - June 2020: The government launched a digital credit program (DigiKu) for MSMEs under the “Proudly Made in Indonesia” movement to facilitate low-interest loans to MSMEs through digital platforms, channeled via collaboration between the government, state-owned banks, and tech-based companies. - September 2020: OJK launched the Online Banking Provisions Information System (SIKEPO) mobile application to increase compliance and disseminate information on banking regulations. ### International cooperation - August 2020: OJK and the Securities Commission Malaysia signed a financial technology cooperation agreement to facilitate information exchange on new technology developments, trends, and regulatory aspects. - September 2020: OJK signed an agreement with Bangko Sentral ng Pilipinas to deepen financial inclusion, information sharing on digital financial innovation, and improve the supervisory framework. ### Key statistics and empirical findings (figures cited in the source) - QRIS adoption: more than 5.46 million merchants; close to 43.6 percent (2.4 million) joined during the pandemic since March 2020. - April 2020 MDR on QRIS for micro-businesses: 0 percent (temporary reduction). - P2P lending growth: accumulation of outstanding P2P loans grew at more than 230 percent (y/y) to IDR 88.4 trillion in Jan-20. - Fintech lending share: less than 0.1 percent of GDP (pre-COVID level). - New P2P loan issuances: sharp decline since onset of the pandemic; decline occurred across most regions. - ADB survey sample sizes: 1,046 households between May-July 2020 and 525 MSMEs between April-May 2020. - MSME liquidity: More than 85 percent of MSMEs reported having no cash/savings or funds that would run out in a month. ### Observed gaps and structural constraints - Despite progress, close to half the population still remains unbanked; 6 percent of the world’s unbanked reside in Indonesia. - Majority of middle to lower segment population and MSMEs lack access to finance from traditional financial institutions, constraining MSME productivity and competitiveness. - Usage of DFSs remains considerably lower in comparison to other EMEs, owing to low levels of financial digital literacy and a small base of internet users. - Fintech and bank credit concentration: geographic concentration in the Java region, leading to geographical disparities. ### Climate- and energy-related context linked to transition risk (selected figures relevant to financial stability and digital-finance policy environment) - The oil, gas and coal related sector accounts for 7.2 percent of total GDP. - Over 1.4 million workers (1.1 percent of total) are employed in the mining and quarrying sector. - Coal exports account for around 10 percent of total exports. - Government receives royalties from oil, gas and coal mining equivalent to 7.5 percent of total government revenue. - Sector GDP shares in 2019: oil, gas and geothermal mining sector 2.8 percent; coal and lignite coal mining sector 2.3 percent; coal and gas refining sector 2.1 percent. - Over 60 percent of total coal production in Indonesia was exported in 2019. - Oil, gas, and coal mining royalties in 2019 were 4.3 percent, 1.9 percent, and 1.3 percent of total revenue, respectively. - Indonesia ranked as the eighth largest GHG emitter in the world in 2017 (based on total emissions excluding Land Use Change and Forestry (LUCF)). - LUCF accounts for more than half of Indonesia’s emissions increase over the past three decades. - Coal-based power generation share reached about 57 percent in 2018. - Indonesia’s NDC targets: unconditional GHG reduction target of 29 percent and conditional reduction target up to 41 percent. *Source: 1idnea2021002 - Box 1. Recent Policy Efforts to Support Integration of the Digital Economy and Finance: Before and During COVID-19* ### 8. Indonesia is also a member of the Coalition of Finance Ministers for Climate Action ### 8. Indonesia is also a member of the Coalition of Finance Ministers for Climate Action ### Climate budgeting, green sukuk, and disaster finance - Since 2016 the government has implemented a Climate Budget Tagging (CBT) process to monitor and track expenditures on climate change actions in the budget system—focusing first on mitigation and since 2018 extending the scope also to adaptation expenses. - Green sukuk issuance supported by CBT: - International issuances: US$1.25 billion in 2018, US$0.75 billion in 2019, and US$0.75 billion in June 2020. - Retail issuance for local investors: IDR 1.46 trillion, approximately US$0.10 billion (November 2019). - Green sukuk funds are used for: - Mitigation: sustainable transport, waste management, waste to energy management, and renewable energy. - Adaptation: resilience to climate change for highly vulnerable areas and sectors. - The government has developed a Disaster Risk Financing and Insurance strategy aiming to protect state-owned assets (national and subnational), key public infrastructure, households and communities, recover social aspects of communities, and support the development of the domestic insurance industry; financing schemes such as pooling fund and insurance are tailored to fit respective disaster risk. ### Financial sector and regulatory support for green investment - Bank Indonesia (BI) has adopted environmentally sound policies, including adoption of Sustainable and Responsible Investment in foreign exchange management. - Financial Services Authority of Indonesia (OJK) regulation (effective 2020): financial services institutions shall apply sustainable finance principles to their business activities and shall submit their sustainable finance action plan to OJK on annual basis. ### Paris Agreement, NDC, and potential for stronger commitments - Indonesia is projected to overachieve on its commitment under the Paris Agreement, excluding LUCF, on the basis of currently implemented policies (source: Climate Action Tracker). - Climate Action Tracker notes Indonesia’s climate commitments are not consistent with either limiting global temperature increase to 2°C or the Paris Agreement goal of 1.5°C. - Risks: - Wildfires create an inevitable and imperfectly controllable risk of missing targets. - Policy consideration: - Update NDC ahead of COP 26 in November 2021. - Consider carbon neutrality pledge (examples cited: 2050 for EU, Japan, Korea, UK; 2060 for China) to signal commitment and boost green investment. ### Policy options for mitigation — overview - Need to balance promoting development and lowering GHG emissions. - Recommend a comprehensive transition plan toward a greener economy to facilitate policy design (especially energy use), discussed at national level with transparent public communication. - The plan could include incentives and disincentives using expenditure, tax and regulatory measures, and provide compensatory support for parties negatively affected. - Consider transition risks affecting the coal sector; suggested measures: - Facilitate employment in other sectors. - Reduce financial sector risks through financial regulations including additional bank capital buffer requirement. ### Land use, forestry, and feebate/tax options - Policy actions taken: moratorium on peatland development, social forestry programs, reinforcement of wildfire monitoring, restoration emphasis on mangrove and peatland. - Social forestry: in 2014 a project allows forest-dependent communities access to 12.7 million hectares; as of June 2020 around 4.2 million hectares distributed. - Suggested instruments: - Feebate scheme: levies on landowners who reduce carbon storage and subsidies to landowners who increase carbon storage. - Alternative: tax on commodities from plantations to support reforestation. - These measures could reduce GHG emissions without adding fiscal burden. ### Energy sector challenges and feebate proposals - National Energy Policy (KEN, 2017): increase renewable energy in electricity to 23 percent in 2025. - As of 2018: - Power plant installed capacity of new renewable energy is 14 percent of total. - Power production of new renewable energy is 17.1 percent. - Concern: planned increase in electricity generation capacity relies substantially on coal, making the 23 percent target uncertain under current plans. - Proposed instrument: feebate for power generators—levies for above average emissions per kWh and rebates for below average emissions per kWh, potentially revenue-neutral. - 2020 fiscal stimulus largely climate neutral but includes climate-negative measures (notably fossil fuel subsidies) that dominate climate-positive measures including biofuel subsidies. ### Energy subsidy reform and pricing mechanisms - Energy subsidies remain significant: "they still amount to over 10 percent of total government expenditure." - 2021 budget proposed an energy subsidy reform to transform energy subsidies into social assistance (Box 1). The reform was postponed due to the COVID-19 crisis. - Rationale for reform: - Generalized energy subsidies are regressive; richer households consume more energy. - Targeted social assistance is more effective and allows reallocating scarce budget resources. - Broader reform options: - More frequent adjustments in domestic retail prices with global market prices. - Relaxation of price controls to promote competition in fuel retail sales and power generation. - Ex post regulation (price notification system) preferable to ex ante regulation for efficiency while ensuring regional pricing equality. - More granular and targeted regional support where energy costs are high. ### Regulatory frameworks and sectoral initiatives (EVs, mining, omnibus bill) - 2019 regulation to accelerate the battery electric vehicle (EV) program: - Policy goal: become an EV hub for Asia with a 20 percent target share in total car production by 2025. - Indonesia has nickel resources important for batteries. - Proposed complement: vehicle feebate scheme—sliding fees on high-emission cars and rebates for low-emission vehicles. - Concerns: - The omnibus bill on job creation relaxed environmental assessment and included royalty incentives to promote investment in mining (which might include coal mining). - Recommendation: harmonize implementing rules to incentivize green investment and disincentivize fossil fuel investment. ### Carbon tax analysis and impacts - Carbon tax could be an effective mitigation measure. - Estimated effects (Table 3 & Table 4 references): - US$25 per ton CO₂: - Could cut emissions by 13.0 percent. - Could raise 0.7 percent of GDP in revenue. - Higher rates (US$50, US$75 per ton CO₂) imply larger cost increases and higher jobs-at-risk, with concentrated impacts in mining and quarrying and electric and gas sectors. - Labor and adjustment considerations: - Overall jobs at risk are limited in the medium term at US$25 rate, but mining and electric/gas sectors may not absorb impacts in short term. - Gradual implementation over several years can spread out jobs-at-risk. - Carbon tax revenues could finance compensation, job reallocation support, and workers retraining programs. ### Adaptation — plans, fiscal needs, and monitoring - National Action Plan for Climate Change Adaptation (RAN-API, 2014) identifies strategies addressing: - Rise in land and sea surface temperature and sea level, rainfall pattern changes, and extreme weather events. - RAN-API objectives: build resilience of (i) economy (food security, energy independence), (ii) living systems (public health, housing, infrastructure), (iii) environmental ecosystems (forest, biodiversity), (iv) urban areas, coastal areas, small islands. - Gaps and recommendations: - The adaptation plan should include fiscal costs; currently it lacks dedicated fiscal backing. - Crucial to determine necessary medium-term expenditure given Indonesia’s limited tax base. - Strengthen monitoring and review processes; ensure nationwide capacity development for adaptation. - Integrate capital movement plan (including fiscal aspects) into the adaptation plan for holistic perspective. ### Conclusion and recommended policy framework (summary of recommended steps) - Climate-related risks, including transition risk, have been increasing due to more frequent natural disasters and reliance on coal; preparations should continue. - Recommended policy steps: - A comprehensive transition plan toward a greener economy: include labor reallocation steps, capacity building, a road map of alternative revenue mobilization (such as carbon tax), and additional financial regulation for bank capital buffers to address stranded asset risks. - Reforestation incentive scheme: consider feebate on landowners or a commodity tax to finance reforestation. - Renewable energy generation incentive scheme: feebate for power generators; update National Energy Policy to support renewables and attract ESG investment. - Reforms of energy subsidies and pricing mechanisms: implement 2021 energy subsidy reform as soon as possible; move to market-based price adjustments and strengthen targeted direct support for low-income households. - Vehicle incentive scheme: feebate to support EV market share. - Harmonize implementing rules for the Omnibus Bill on Job Creation with green investment incentives and fossil disincentives. - Introduction of a carbon tax (or coal tax) as an effective mitigation measure and source of revenue for compensation and transition support. - Strengthen monitoring and review of adaptation measures and integrate fiscal costs and financing into the adaptation plan. ### Box 1 — Indonesia’s Energy Subsidies and the 2021 Reform Proposed by the Government (highlights) - Indonesia uses ex ante and ex post energy subsidies and price controls to support low-income households and ensure price equality across provinces. - Subsidized fuels: premium gasoline, diesel, kerosene, and 3 kg LPG cylinders. - Diesel receives a fixed rate subsidy of IDR 1,000 per liter. - Retail prices of LPG have been constant since 2008; price adjustments for fuels and electricity are infrequent in practice. - Electricity: PLN monopoly for distribution; subsidy targets customers purchasing 450VA and 900VA (assumed poor households). Compensation subsidies to PLN exist due to government-controlled retail prices; targeting is weak. - 2021 reform measures (paradigm shift from commodity-based to people-based subsidies): - Reduce fixed diesel subsidy from IDR 1,000 to IDR 500 per liter. - Apply market price adjustments to premium gasoline and electricity for non-subsidized customers, eliminating relevant compensation subsidies. - Gradual price adjustment of LPG and integrate related subsidies into social assistance transfers. *From the IMF Indonesia chapter.* ### Box 1. Indonesia’s Energy Subsidies and the 2021 Reform Proposed by the Government ### Box 1. Indonesia’s Energy Subsidies and the 2021 Reform Proposed by the Government ### References - Agarwal, Ruchir, Vybhavi Balasundharam, Patrick Blagrave, Eugenio Cerutti, Ragnar Gudmundsson, and Racha Mousa, 2020, Climate Change in South Asia: Further Need for Mitigation and Adaptation, IMF Working Paper, forthcoming (Washington: International Monetary Fund). - Buckley, Tim, 2019, Over 100 Global Financial Institutions Are Exiting Coal, With More to Come, IEEFA Report, February (Lakewood, Ohio: Institute for Energy Economics and Financial Analysis). - Coady, Parry, Le, and Shang, 2019, Global Fossil Fuel Subsidies Remain Large: An Update Based on Country-Level Estimates, IMF Working Paper (Washington: International Monetary Fund). - Coalition of Finance Ministers for Climate Action, The, 2019, An Overview. Available via the Internet: https://www.financeministersforclimate.org/about-us. - Government of Indonesia, 2016, First Nationally Determined Contribution Republic of Indonesia. - International Energy Agency, 2020, Coal Information: Overview, IEA Statistics report, July. Available via the Internet: https://www.iea.org/reports/coal-information-overview. - International Monetary Fund, 2021, Fiscal Policies to Address Climate Change in Asia and the Pacific: Opportunities and Challenges, IMF Departmental Paper, forthcoming (Washington). - ————, 2020a, Fiscal Monitor, October 2020: Policies for the Recovery (Washington). - ————, 2020b, World Economic Outlook, October 2020: A Long and Difficult Ascent (Washington). - ————, 2020c, Global Financial Stability Report, October 2020: Bridge to Recovery (Washington). - ————, 2020d, “Vulnerability to Climate Change and Natural Disaster,” Chapter 2 in Philippines: Selected Issues, IMF Country Report No. 20/37 (Washington). - ————, 2019a, Fiscal Monitor, October 2019: How to Mitigate Climate Change (Washington). - ————, 2019b, Fiscal Policies for Paris Climate Strategies—from Principle to Practice, IMF Policy Paper (Washington). - Ministry of Energy and Mineral Resources, 2018, Handbook of Energy & Economic Statistics of Indonesia (Republic of Indonesia). - Ministry of Finance, 2020, Green Sukuk Allocation and Impact Report (Republic of Indonesia). - Ministry of National Development Planning/National Development Planning Agency (BAPPENAS), 2014, Indonesia’s National Action Plan for Climate Change Adaptation: Synthesis Report (Republic of Indonesia). - Nishizawa, Hidetaka, Scott Roger, and Huan Zhang, 2019, “Fiscal Buffers for Natural Disaster in Pacific Island Countries,” IMF Working Paper No. 19/150 (Washington: International Monetary Fund). - Peh, Ghee, 2020a, Can the Indonesian Coal Industry Survive COVID-19?, IEEFA Report, May (Lakewood, Ohio: Institute for Energy Economics and Financial Analysis). - ————, 2020b, Indonesian Coal: No Bailout, Don’t Throw Good Money After Bad, IEEFA Report, September (Lakewood, Ohio: Institute for Energy Economics and Financial Analysis). - Secretariat General of the National Energy Council, 2019, Indonesia Energy Outlook 2019 (Republic of Indonesia). *Source: Box 1. Indonesia’s Energy Subsidies and the 2021 Reform Proposed by the Government (concluding references).* --- _Source: https://www.imf.org/-/media/files/publications/cr/2021/english/1idnea2021002.pdf_