## Macro-Financial Impacts of Foreign Digital Money — Working Paper No. WP/2023/249 ## Source details **Canonical URL:** [Macro-Financial Impacts of Foreign Digital Money — Working Paper No. WP/2023/249](https://www.imf.org/-/media/files/publications/wp/2023/english/wpiea2023249-print-pdf.pdf) ## Other formats - [Markdown version](/-/media/files/publications/wp/2023/english/wpiea2023249-print-pdf.pdf.md) - [Structured JSON version](/-/media/files/publications/wp/2023/english/wpiea2023249-print-pdf.pdf.json) --- ### Major themes and model setting - Two-country New Keynesian framework assessing risks from a foreign stablecoin for a small developing economy. - Domestic features: - Bank-based financial system with financial frictions following Gertler and Karadi (2011) and Aoki et al. (2018). - Households hold M_t, D_t, M_F,t, B_F,t, and SC_t; foreign bonds are a store of value only; M_t, D_t, M_F,t, and SC_t usable as means of payment. - Foreign features: - Large economy calibrated to the US (foreign economy is "67 times larger than our small economy"). - Stablecoin issuer transforms foreign cash and foreign bonds into a global crypto asset backed by foreign cash and bonds; issuer backing share of foreign cash set at 95%. - Key modelling elements: - Endogenous currency substitution (Özbilgin (2012) style). - Capital flow management measures (CFMs) modeled as a tax on return of foreign bonds with baseline φ_bf = 0.25. - Households choose payment instruments based on expected returns, transaction costs, inflation, exchange rates, and stablecoin price. - Banking sector includes foreign borrowing equal to 25% of bank assets; banking steady-state leverage ratio = 4; bankers’ survival rate = 0.94; proportional transfer to new bankers = 0.002; credit spread = 2% annually. ### Core mechanisms and analytical results - Currency substitution amplification: - Adding a global stablecoin loosens household trade-offs, allowing larger shifts away from domestic-currency assets and amplifying currency substitution. - Negative shocks lead to amplified capital outflows, magnified domestic output losses, larger deposit outflows, and larger reductions in bank net worth. - Contractionary foreign monetary policy shock produces particularly severe banking and output effects. - Monetary policy transmission: - Stablecoin-driven reallocation from domestic deposits reduces bank intermediation and the share of activity the central bank can influence via interest rates. - Transmission of policy to investment, output, and prices weakens; optimal central bank reacts more aggressively to inflation deviations. - Interaction with CFMs: - CFMs (tax on foreign bond returns) increase households’ responsiveness of stablecoin holdings to contractionary foreign monetary shocks: with CFMs restricting foreign bond diversification, households switch more into the stablecoin. - Stablecoins can be used to circumvent CFMs, increasing exposure to foreign shocks and worsening macrofinancial impacts relative to the no-stablecoin case. - Banking amplification channel (selected equations preserved conceptually): - Bank flow-of-funds: (1 + κ_b/2 x_t^2) Q_t K^b_t = N_t + D_t + s_t D*_t. - Bankers’ incentive constraint: V_t(N_t) ≥ Θ(x_t, x^c_t) Q_t K^b_t. - Tobin’s Q, leverage, and foreign borrowing enter net-worth dynamics and credit spreads (ψ_t, lev_t, μ_t, μ^*_t as defined in source). ### Policy responses evaluated — findings and comparative outcomes - Domestic CBDC: - A domestic CBDC can partially reduce equilibrium stablecoin holdings by offering a domestic alternative. - Cash-like CBDC scenarios (I.A and I.B) yield marginal reductions in stablecoin uptake and only slight mitigation of downturns in investment and output. - Stablecoin-like CBDC (II) with μ_DC^II = μ_SC = 1.65 and σ_DC^II = σ_SC = 2.85 yields larger steady-state CBDC holdings but under contractionary foreign shocks households reduce CBDC and increase stablecoin holdings; macro outcomes marginally worse. - Design features increasing CBDC attractiveness: legal tender status, universal acceptance, offline functionality, lower fees. - Comprehensive stablecoin ban: - Modeled as SC_t = 0 domestically (100% effective). - A comprehensive domestic ban largely returns domestic responses to foreign shocks toward the ‘no stablecoin’ path: reduces currency substitution, capital outflows, bank disintermediation, and output losses relative to the with-stablecoin case. - Practical enforcement is difficult given peer-to-peer transfers and informal-sector use; cross-country coordination can weakly improve enforceability, especially if issuer located in foreign economy. - Optimal monetary policy implications: - Presence of the stablecoin attenuates monetary policy transmission, compelling central bank to respond more assertively to inflation; ρ_π frequently reaches its upper limit. - Policy optimization imposes an upper limit of 5 on the inflation reaction to avoid corner solutions. - Central bank loss function used: L_CB = Var(π_t) + λ_y Var(ΔY_t) + λ_e Var(s_t) + λ_R Var(R_t). ### Quantitative simulations — shocks and amplification (selected calibrated experiments) - Baseline calibration highlights: - Banking sector ex-ante steady-state leverage ratio = 4. - Credit spread = 2% annually. - Foreign borrowing = 25% of bank assets. - Bankers’ survival rate = 0.94. - Stablecoin baseline holdings set so roughly 2% of payment assets are in the stablecoin; stablecoin holdings < 4% the size of deposits in baseline. - Liquidity usefulness ordering (given σ_SC = σ_D): μ_{M*} < μ_{SC} < μ_D with μ_{M^*} = 0.35, μ_{SC} = 1.65, μ_D = 2.65. - Domestic TFP shock (1% negative): - Without stablecoin: output, consumption, investment fall; inflation rises; central bank raises policy rate; domestic currency appreciates; bank net worth declines; credit spread increases. - With stablecoin: larger slump in output, consumption, investment; larger reduction in domestic cash and increased stablecoin holdings; slightly larger fall in domestic deposits; stablecoin price deviates substantially due to exchange-rate movements. - Domestic monetary contraction (contractionary domestic policy shock): - With stablecoin: more severe declines in investment and consumption; larger reduction in cash holdings; increased stablecoin holdings; banks’ net worth falls more and credit spread widens more relative to no-stablecoin. - Foreign monetary contraction (contractionary foreign monetary shock): - Real exchange rate depreciation bolsters exports but raises import prices and inflation; central bank raises interest rates, reducing consumption. - Stablecoin availability produces substantially larger spillovers: larger declines in output, consumption, investment; greater currency substitution; larger increases in holdings of stablecoin and foreign bonds; more pronounced depreciation; greater banking stress (declines in bank net worth and wider credit spreads). - CFMs (φ_bf = 0.25 baseline) redirect resources toward domestic assets but, because CFMs do not apply to stablecoins, households reallocate into stablecoin when CFMs are active, accelerating adoption and worsening macro outcomes relative to no-stablecoin case. ### CBDC scenarios and numeric outcomes (Table 2 values preserved) - Scenarios and steady-state CBDC holdings and responses to ↑R*: - I.A Cash-like: μ_DC = 1, σ_DC = -1.85, CBDC Holdings Steady State = 8.12%, Response to ↑R* = Large increase. - I.B Cash-like: μ_DC = 1, σ_DC = -5.85, CBDC Holdings Steady State = 10.69%, Response to ↑R* = Small increase. - II Stablecoin-like: μ_DC = 1.65, σ_DC = 2.85, CBDC Holdings Steady State = 16.49%, Response to ↑R* = Decrease. - Interpretation: - Cash-like CBDCs (I.A, I.B) only marginally reduce stablecoin uptake and marginally mitigate downturns. - A CBDC matching stablecoin liquidity parameters reduces steady-state stablecoin holdings but cannot hedge against domestic inflation or depreciation because it is denominated in domestic currency; under foreign contractionary shocks households may reallocate away from the CBDC toward the stablecoin. ### Holdings and steady-state distributions (Table 4 values preserved) - Representative domestic household steady-state distribution (% by domestic currency value): - No SC: Domestic Cash 35.01%; Deposit 54.31%; Stablecoins 0%; Foreign Cash 10.68%; Domestic CBDC 0%. - With SC: Domestic Cash 35.68%; Deposit 50.85%; Stablecoins 1.97%; Foreign Cash 11.49%; Domestic CBDC 0%. - Ban SC: Domestic Cash 34.56%; Deposit 54.86%; Stablecoins 0%; Foreign Cash 10.57%; Domestic CBDC 0%. - CBDC I.A: Domestic Cash 31.68%; Deposit 47.28%; Stablecoins 1.85%; Foreign Cash 11.05%; Domestic CBDC 8.12%. - CBDC I.B: Domestic Cash 31.08%; Deposit 45.47%; Stablecoins 1.77%; Foreign Cash 10.97%; Domestic CBDC 10.69%. - CBDC II: Domestic Cash 33.04%; Deposit 41.06%; Stablecoins 1.51%; Foreign Cash 7.89%; Domestic CBDC 16.49%. ### Key calibrated parameters (selected from Table 3) - β 0.995; β_E 0.975; θ 0.501; σ 0.94; ξ 0.0045; κ_b 0.0219. - α_K 0.30; δ 0.025; η_d 8; κ_I 2.48; ε_m 1. - μ_{M^*} 0.35; μ_D 2.65; μ_{SC} 1.65; μ_{DC} 1. - σ_{M^*} 1.35; σ_D 2.85; σ_{SC} 2.85. - ρ_i 0.82; φ_π 3; φ_π^D 1.5; φ_Y 0.09; φ_bf 0.25. - n 0.25; ς 67; b 0.95. - Steady-state targets: TB/Y 0.0037; C/Y 0.65; K/Y 7.; I/Y 0.18; J 0.23; Π 4%. ### Policy implications and recommendations (model-driven) - Recognize stablecoins as an additional cross-border transmission channel that can amplify foreign monetary shocks, weaken domestic monetary transmission, and increase banking-sector stress. - CFMs targeting foreign bond returns can be undermined by stablecoins because stablecoins are not covered by the CFM tax and thus become an alternative channel for households to access foreign-denominated liquidity. - Domestic CBDC may modestly reduce stablecoin holdings but cannot substitute fully for the hedge value of a foreign-currency stablecoin; CBDC design should consider adding legal tender status, universal acceptance, offline functionality, and low transaction fees to maximize substitution away from foreign stablecoins. - A comprehensive domestic ban can, if enforceable, largely restore outcomes toward the no-stablecoin baseline; however, enforcement is challenging and benefits from cross-country coordination, especially when the issuer is located in the foreign economy or offshore. - Central banks should anticipate weaker monetary transmission in the presence of foreign stablecoins and may need to respond more aggressively to inflation deviations (policy rule reaction constrained by an upper limit of 5 on inflation reaction in optimization exercises). *Source: https://www.imf.org/-/media/files/publications/wp/2023/english/wpiea2023249-print-pdf.pdf* ### introduction amplifies currency substitution, reducing bank intermediation and ### introduction amplifies currency substitution, reducing bank intermediation and ### JEL Classification and Keywords - JEL Classification Numbers: E50, F30, F31, G15, G18, G23 - Keywords: Cryptocurrency; Open Economy; Financial Frictions; Optimal Policy - Author’s E-Mail Address: leanh1796@gmail.com; acopestake@imf.org; btan2@imf.org; epapageorgiou@imf.org; speiris@imf.org; urawat@imf.org ### Major themes and setting - Scope: - Two-country New Keynesian model assessing risks from a foreign stablecoin for a small developing economy. - Domestic economy: bank-based financial system with financial frictions following Gertler and Karadi (2011) and Aoki et al. (2018). - Foreign economy: large economy (calibrated to the US following Adrian et al. (2020)) hosting a stablecoin issuer producing a global crypto asset backed by foreign cash and bonds following Cova et al. (2022). - Key modelling elements: - Endogenous currency substitution following Henriksen and Kydland (2010) and Özbilgin (2012). - Capital flow management measures (CFMs) modeled as a tax on the return of foreign bonds following Davis and Presno (2017). - Households hold domestic cash (M_t), domestic deposits (D_t), foreign cash (M_F,t), foreign bonds (B_F,t), and stablecoins (SC_t). Foreign bonds are a store of value only; the first three and stablecoins can be used as means of payment. - Payment instrument choice depends on expected returns, transaction costs, inflation, exchange rates, and stablecoin price. ### Core mechanisms and analytical results - Currency substitution and amplification of shocks: - Addition of a global stablecoin to payment asset menu loosens households’ trade-off between reducing domestic-currency exposure and meeting liquidity needs, allowing larger shifts away from domestic currency assets. - Currency substitution and capital outflows are amplified in response to negative shocks, magnifying domestic output losses. - Banks experience larger deposit outflows and larger reductions in net worth, with particularly severe effects under a contractionary foreign monetary policy shock. - Presence of a global stablecoin can amplify international transmission of shocks, akin to effects found for a foreign CBDC. - Monetary policy transmission: - Stablecoin-driven reallocation from domestic deposits reduces bank intermediation and the share of economic activity the central bank can influence via interest rates. - Transmission of monetary policy to investment, output, and prices weakens, leading an optimal central bank to react more aggressively to deviations of inflation from target. - Interaction with capital controls (CFMs): - A tax on returns to foreign bonds increases households’ responsiveness of stablecoin holdings to a contractionary foreign monetary policy shock—households switch more into the stablecoin when CFMs restrict foreign bond diversification. - Stablecoins act as a channel for circumvention of CFMs, increasing exposure to foreign shocks and worsening macrofinancial impacts relative to the no-stablecoin case. - Implication: crypto-asset-based circumvention could undermine attempts to use capital controls to insulate small developing economies from foreign spillovers. ### Policy responses evaluated - Domestic CBDC: - A domestic CBDC can partially reduce stablecoin holdings in equilibrium by substituting an alternative asset. - However, because the CBDC is denominated in domestic currency, it does not hedge against domestic inflation or depreciation and therefore does not mitigate the stablecoin’s role in transmitting foreign shocks to the domestic economy. - Comprehensive stablecoin ban: - A complete domestic ban on holding and using the stablecoin (while it remains legal in the foreign economy) returns the economy’s response to foreign shocks almost to the baseline ‘no stablecoin’ path. - A comprehensive ban almost entirely alleviates currency substitution, capital outflows, bank disintermediation and larger output losses. - Practical limitations: policing a complete prohibition is difficult given decentralized technology, peer-to-peer transfers, and informal-sector use; cross-country coordination (or multilateral coordination if issuer is in a third country) can weakly improve outcomes. ### Model structure and household decision rules (high-level) - Representative household maximizes expected utility U_t = E_t Σ_{t=0}^∞ β^t [log(C_t) − Ψ(1−L_t)^{1+φ}/(1+φ)] subject to a detailed budget constraint including holdings of M_t, M_F,t, SC_t, D_t, and B_F,t, and costs such as storage cost ε_m and transaction costs τ(1−j_t). - Stablecoin price is P_sc,t and real exchange rate is s_t; foreign bond nominal rate R_z,t includes a dollar borrowing risk premium R_z,t = R^*_t − φ(e^{(b_F,t − b_F) − 1}). - Currency substitution condition (payment instrument choice) generalizes Özbilgin (2012), with liquidity composite Ω(D_t, s_t M_F,t, s_t SC_t) = μ_M* (s_t M_F,t)^{σ_{M*}−1} + μ_SC (s_t SC_t)^{σ_{SC}−1} + μ_D (D_t)^{σ_D−1} (exponents and functional form preserved as in source). - Households choose frequency n_t of asset market visits at cost κ (Freeman and Kydland (2000) style), and trade off inflation, returns, exchange rate depreciation, stablecoin price and transactions costs when choosing payment assets. - Consumption aggregate: C_t = [(1−γ)^{1/η} C_{H,t}^{(η−1)/η} + γ^{1/η} C_{F,t}^{(η−1)/η}]^{η/(η−1)} with consumption price index P_t = [(1−γ)(P_{H,t})^{1−η} + γ(P_{F,t})^{1−η}]^{1/(1−η)}; producer currency pricing (PCP) for domestic goods and dominant currency pricing (DCP) for exports. ### Contributions relative to existing literature - Focus on small emerging market and developing economy exposure to a foreign stablecoin; large economy is calibrated to be "67 times larger than our small economy" (roughly USA vs. Malaysia). - Incorporates a full banking sector with financial frictions (Gertler and Karadi (2011) framework) to capture stablecoin-driven bank disintermediation and stress—distinct from Cova et al. (2022) and Minesso et al. (2022). - Allows rich currency substitution dynamics following Özbilgin (2012), and models CFMs interacting with stablecoin adoption, illustrating circumvention risks documented empirically (e.g., Alnasaa et al., 2022; Graf von Luckner et al., 2023). *Source: https://www.imf.org/-/media/files/publications/wp/2023/english/wpiea2023249-print-pdf.pdf* ### 2.2 Production Sectors ### 2.2 Production Sectors ### Aggregation and final demand - Total output Yt is a CES aggregate of differentiated intermediate inputs indexed by i∈[0,1]: - Yt = [∫0 1 Yt(i)^(ε−1)/ε di]^(ε/(ε−1)) (Equation 14) - Demand for each intermediate good: - Yt(i) = [Pt(i)/Pt]^(−ε) Yt (Equation 15) - Retail price index: - Pt = [∫0 1 Pt(i)^(1−ε) di]^(1/(1−ε)) (Equation 16) - Under symmetry, aggregate output can be written: - Yt = At [Kt−1^(αK)]^(αK) [Ht^(1−αK)]^(1−αK) (Equation 17) - Total factor productivity At follows an AR(1) with technology shock vat ∼ N(0,σa^2): - log(At) = (1−ρa) log(A) + ρa log(A) + vat ### Pricing, market structure, and price adjustment costs - Firms operate in monopolistic competition and set prices for their own differentiated goods with elasticity ε between retail products. - Dominant currency pricing (DCP) is assumed: domestic prices set in Home currency; export prices set in US dollars. - Retailer i chooses domestic price P_Ht(i) (in domestic CPI terms) to maximize expected discounted profits: - Objective includes markup over marginal cost MCt, domestic consumption and investment C_Ht + I_Ht, and price adjustment costs ACt(i) (Equation 18). - Price adjustment costs (Rotemberg quadratic form) in nominal terms: - ACt(i) = ΩP/2 [ P_Ht(i)/P_Ht(i−1) − ̄π ]^2 P_Ht (C_Ht + I_Ht) (Equation 19) - New Keynesian Phillips Curve (Rotemberg form) linking real marginal cost mct and inflation π_Ht: - mc_t = p_Ht ((ε−1)/ε) + (ΩP/ε) (π_Ht − ̄π) π_Ht − [ β λt+1/λt πH,t+1 (πH,t+1 − ̄π) pH,t+1 (C_H,t+1 + I_H,t+1) / ( pH,t (C_Ht + I_Ht) ) ] (Equation 20) - ΩP is the price adjustment cost parameter. ### Export pricing under DCP - Exporter i chooses p*_Ht(i) (denominated in US dollars) maximizing discounted profits, taking exchange rate s_t into account (Equation 21). - Adjustment costs and Philips Curve for exports are analogous to domestic definitions (Equations 19 and 20 analogues). ### Capital producers and investment dynamics - Capital producers purchase final goods and non-depreciated capital to produce capital goods; they maximize expected profits: - max E_t ∑_{t=0}^∞ β^t λ_t/λ_0 [ Q_t K_t − (1−δ) Q_t K_{t−1} − I_t ] (Equation 22) - Capital accumulation with investment adjustment costs: - K_t = (1−δ) K_{t−1} + [ 1 − κ_I/2 ( I_t / I_{t−1} − 1 )^2 ] I_t (Equation 23) ### Market-clearing and central bank rule (contextual to production) - Domestic central bank follows a Taylor rule: - ln( R_t / R_ss ) = ρ_r ln( R_{t−1} / R_ss ) + (1−ρ_r)[ ρ_π ln( π_t / π_ss ) + ρ_y ln( Y_t / Y_{t−1} ) ] (Equation 34) - Goods market clearing requires that total output equals domestic and foreign expenditure plus adjustment and transaction costs and banker/household costs χ_h and χ_b (Equation 35). ### Key model mechanisms linking production to finance and external sector - DCP implies domestic price-setting and export pricing differ in currency denomination; exchange rate movements affect domestic valuation of US-dollar stable assets and exports. - Price adjustment costs convert nominal price-setting frictions into real effects through the Rotemberg Philips Curve. - Investment adjustment costs and capital producers’ optimization transmit shocks to capital accumulation and Tobin’s Q. --- ### 2.3–3.4 Financial Intermediaries, Market Clearing, and the Foreign Economy (context necessary for production outcomes) ### Banking sector funding and constraints - Domestic banks fund capital loans with net worth Nt, domestic deposits Dt, and foreign deposits D*_t; fraction x_t financed by foreign borrowing; foreign deposits converted at exchange rate s_t and incur a risk premium (1 + κ_b/2 x_t^2). - Flow of funds constraint for a representative bank: - (1 + κ_b/2 x_t^2) Q_t K^b_t = N_t + D_t + s_t D*_t (Equation 24) - Banks’ net worth: - N_t = (Z_t + Q_t λ) K^b_{t−1} − D_{t−1} R_{t−1}/π_t − s_t D^*_{t−1} R^*_{t−1}/π^*_t (Equation 25) - Incentive (participation) constraint for bankers: - V_t(N_t) ≥ Θ(x_t, x^c_t) Q_t K^b_t (Equation 26) - Recursive maximization of terminal wealth for bankers: - V_t(N_t) = max E_t [ Λ_{t;t+1} [ (1−σ) N_{t+1} + σ V_{t+1}(N_{t+1}) ] ] (Equation 27) - Tobin’s Q and leverage representations: - ψ_t = V_t / N_t, lev_t = Q_t K^b_t / N_t - ψ_t = max_{lev_t, x_t} [ μ_t lev_t + (1 − κ_b/2 lev_t^2 x_t) v_t + μ^*_t lev_t x_t ] subject to ψ_t ≥ Θ(x_t, x^c_t) lev_t (Equations 28–29) - Definitions: - μ_t = E_t[ Λ_{t;t+1} ( Z_{t+1} + λ Q_{t+1}/Q_t − R_{t,t+1} ) ] (Equation 30) - μ^*_t = E_t[ Λ_{t;t+1} ( R_{t+1} − s_{t+1}/s_t R^*_{t,t+1} ) ] (Equation 31) - v_t = E_t[ Λ_{t;t+1} R_{t+1} ] (Equation 32) - Ω_{t+1} = Λ_{t;t+1} (1−σ + σ ψ_{t+1}) (Equation 33) ### Foreign economy and stablecoin issuer (relevant for trade and liquidity) - Foreign economy calibrated to the US; foreign households can hold foreign cash M^*_t and stablecoin SC^*_t which provide utility (Equation 37). - Foreign production mirrors domestic structure with CES aggregates and Rotemberg pricing (Equations 40–43). - Stablecoin issuer transforms foreign cash and foreign bonds into global stablecoin supply: - SC^s_t = [ b^{1/ρ} (M^*_{SC,t})^{ρ−1)/ρ} + (1−b)^{1/ρ} (B^*_{SC,t})^{ρ−1)/ρ} ]^{ρ/(ρ−1)} (Equation 44) - Issuer entrepreneur maximizes discounted profit; first-order conditions link stablecoin price P_SC_t, issuer multipliers λ_sc_t, and holdings M^*_SC,t and B^*_SC,t (Equations 45–49). - Market clearing for stablecoin supply: - SC^s_t = n/(1−n) SC_t + SC^*_t (Equation 50) - Dollar and US bond market clearing conditions (Equations 51–52) and foreign government constraint (Equations 53–54). - Foreign Taylor rule analogous to domestic (Equation 55). --- ### 4 Simulations — calibration and key quantitative implications for production and financial amplification - Baseline calibration notes (Section 4.1): - Banking sector ex-ante steady-state leverage ratio = 4. - Credit spread = 2% annually. - Proportional transfer to new bankers = 0.002. - Foreign borrowing = 25% of bank assets. - Bankers’ survival rate = 0.94. - Stablecoin calibration: - Main parameters set so approximately 2% of payment assets are held in the stablecoin (Table 4 referenced). - Given σ_SC = σ_D, the liquidity usefulness ordering implied: μ_{M*} < μ_{SC} < μ_D. - Share of foreign cash in issuer’s backing technology set at 95%. - In baseline, stablecoin holdings are less than 4% the size of deposits. ### Domestic TFP shock (Section 4.2) - Shock: 1% negative shock to domestic TFP. - Baseline without stablecoin (black line in Figure 1): - Output, consumption, and investment fall. - Inflation rises; central bank raises policy rate; domestic currency appreciates (real exchange rate falls). - Banks’ net worth declines; credit spread increases; reduced demand for domestic and foreign deposits. - With stablecoin available (red line in Figure 1): - Larger slump in output, consumption, and investment relative to baseline without stablecoin. - More pronounced reduction in domestic cash holdings and increased holdings of stablecoin (marginal currency substitution). - Slightly larger fall in domestic deposits — potential bank disintermediation. - Stablecoin price in domestic currency deviates substantially due to exchange rate movements; domestic households’ forward-looking expectations influence asset reallocation. - Interpretation: - Stablecoin availability amplifies negative supply shocks through currency substitution, deposit runs toward stablecoin, and tighter bank funding conditions. ### Domestic monetary policy shock (Section 4.3) - Shock: contractionary shock to domestic monetary rate (illustrated in Figure 2). - Baseline without stablecoin: - Immediate declines in consumption, investment, and output; inflation falls; domestic currency appreciates. - Tobin’s Q drops; banks’ net worth declines; credit spread rises initially then declines as leverage falls. - With stablecoin available: - More severe declines in investment and consumption; larger reduction in cash holdings and increased holdings of stablecoin. - Banks’ net worth falls by more and credit spread widens by more. - Overall slightly more severe recession with heightened currency substitution and larger decrease in domestic cash. - Interpretation: - Stablecoin availability amplifies contractionary monetary shocks via balance-sheet channels and currency substitution, increasing stress on banks and investment demand. --- *Italic: Content excerpted from the provided PDF content unit.* ### 4.4 Foreign Monetary Rate Shock ### 4.4 Foreign Monetary Rate Shock ### Transmission of a contractionary foreign monetary shock - The shock triggers a contraction in domestic output as the appeal of the domestic currency wanes. - Real exchange rate depreciation: - Bolsters exports through an expenditure-switching effect and supports aggregate demand. - Elevates prices of imported goods, prompting higher inflation. - Central bank response: - Raises interest rates in response to inflationary pressures, which raises savings rates and reduces consumption. - Financial amplification: - High inflation alleviates the real burden of debt denominated in the home currency but deteriorating bank balance sheets are associated with declines in investment and asset prices (Tobin’s Q), aligning with Kiyotaki and Moore (1997) and Gertler and Karadi (2015). ### Effects of stablecoin availability - Stablecoin as an additional cross-border channel: - Stablecoin can be traded between domestic households, foreign households and issuers; its first-order condition (FOC) for holdings involves the exchange rate, creating an additional cross-border transmission channel. - The presence of the stablecoin produces a substantially larger spillover effect: responses of output, consumption, and investment are more pronounced than without stablecoins. - Cash and deposits decrease more in response to the foreign shock. - Domestic banking sector experiences stress, evident in declines in bank net worth and widening credit spreads. - Portfolio reallocation: - Holdings of the stablecoin and foreign bonds increase as domestic households reallocate away from domestic cash and deposits and the exchange rate depreciates substantially. - Net effect: - Stablecoins magnify currency substitution and intensify macroeconomic responses to contractionary foreign shocks. - Stablecoins contribute to a minor form of bank disintermediation via a more pronounced decrease in domestic deposits. - The largest impacts result from the foreign monetary policy shock, introducing an additional international linkage channel similar to findings for CBDCs in Minesso et al. (2022). ### Quantitative and model-specific details - Credit spread interpretation: - The credit spread measures the risk in the credit market; a higher spread usually reflects a higher lending rate to firms in compensation for the risk premium. - Policy rule and optimization setup: - Central bank loss function: L_CB = Var(π_t) + λ_y Var(ΔY_t) + λ_e Var(s_t) + λ_R Var(R_t) (Equation 57) - The monetary policy optimization minimizes L_CB subject to the model equilibrium and a Taylor-type rule specification. - Upper limit imposed on the inflation reaction to avoid corner solutions: an upper limit of 5 on the inflation reaction is imposed. - Shocks considered in optimization: the foreign interest rate and domestic TFP shock; the domestic interest rate shock is excluded. - Capital controls (CFMs) formulation: - Tax on the return of foreign bonds: τ_d,t = φ_bf (R_z,t − R_t) (Equation 59) - Calibration in baseline: φ_bf set to 0.25 in the baseline calibration (in line with Davis and Presno (2017)). - CFMs do not apply to the stablecoin; stablecoin can be used to circumvent capital controls, increasing its utility to households. - Impacts of CFMs: - Imposing CFMs (φ_bf > 0) redirects resources toward domestic assets, dampens attractiveness of foreign bonds, reduces capital outflows and ameliorates exchange rate depreciation. - With CFMs active, households reallocate partially into the stablecoin (since CFMs do not apply to it), accelerating adoption of the stablecoin. - When CFMs are active in both cases (with and without stablecoin), macroeconomic outcomes are generally worse in the presence of the stablecoin: larger exchange rate depreciation, higher policy rate hikes, worse investment, output and consumption, lower bank net worth and wider credit spreads. - Empirical confirmation: recent empirical work (e.g., Graf von Luckner et al., 2023) finds circumvention is occurring at small scale; model suggests larger adoption could have harmful macroeconomic effects. ### Optimal monetary policy findings (Table 1 and interpretation) - Main conclusion: - In the model incorporating the stablecoin, the central bank responds notably more assertively to inflation, with ρ_π frequently reaching its upper limit. - The presence of the stablecoin attenuates monetary policy transmission, compelling the central bank to adopt a more forceful stance to fulfill its mandate. - Supporting mechanism: - Monetary policy transmission in the model operates primarily through saving and investment decisions. - Deposits are assumed to be the only domestic asset with a policy rate; lower holdings of deposits could weaken monetary policy transmission. - (Table 1 reproduced in the source shows optimal Taylor rule parameters under various relative weights with and without stablecoin; the qualitative result is consistent across weight variations.) ### CBDC as a policy response (Section 5.1) — scenarios and outcomes - CBDC modeling setup: - CBDC holdings DC_t are included in the household liquidity bundle and in the household budget constraint (Equation 61). - Three CBDC characterizations considered (Table 2): two cash-like variants (I.A, I.B) and one stablecoin-like variant (II). - Table 2 (CBDC Scenarios) — values preserved from source: - Scenario | μ_DC | σ_DC | CBDC Holdings Steady State | Response to ↑R* - I.A Cash-like | 1 | -1.85 | 8.12% | Large increase - I.B Cash-like | 1 | -5.85 | 10.69% | Small increase - II Stablecoin-like | 1.65 | 2.85 | 16.49% | Decrease - CBDC quantitative findings: - Cash-like CBDC (cases I.A and I.B): - Increasing CBDC adoption slightly reduces uptake of the stablecoin and marginally mitigates the downturn in investment and output. - Effects are quantitatively marginal and rely on extreme assumptions for σ_DC^I (specifically σ_DC^I set equal to -1.85 and -5.85 in cases I.A and I.B, versus 1.35 for cash or 2.85 for deposits and the stablecoin). - Stablecoin-like CBDC (case II): - With μ_DC^II = μ_SC = 1.65 and σ_DC^II = σ_SC = 2.85, the CBDC is more attractive in steady state and reduces equilibrium stablecoin holdings further. - However, under a contractionary foreign monetary policy shock, households reduce CBDC holdings and increase reallocation into the stablecoin; macro outcomes are marginally worse. - Overall interpretation: - A cash-like CBDC provides little support for mitigating spillovers from a foreign stablecoin. - An improved CBDC that shares stablecoin advantages (higher μ_DC) still remains a domestic-currency asset and cannot fully substitute for stablecoins as a hedge against domestic inflation or depreciation. - CFMs that stablecoins can evade would further limit a CBDC’s ability to neutralize stablecoin-driven spillovers. - Design considerations noted: - Features that could make a CBDC more attractive than the stablecoin include legal tender status, universal acceptance for payments, offline functionality, and lower transaction fees; as a public-sector offering a CBDC need not generate profit and could be cheaper than other digital payment offerings. ### Policy-relevant numeric and calibration notes - Upper limit on inflation reaction in policy optimization: 5. - CFMs calibration: φ_bf = 0.25 in baseline. - CBDC steady-state holdings (share of total payment assets) from Table 2: - I.A: 8.12% - I.B: 10.69% - II: 16.49% *Italic: Source — IMF Working Paper (section 4.4 and related sections reproduced from the provided PDF content).* ### 5.2 Banning the Stablecoin ### 5.2 Banning the Stablecoin ### Policy setup and modeling assumption - A comprehensive unilateral ban is modeled as 100% effective at preventing domestic households from accessing the stablecoin, equivalent to setting SC_t = 0 in the non-cash liquidity bundle (Equation 4). - In steady state with the ban, domestic holdings of the stablecoin are zero, but the allocation across other payment assets differs from the no-stablecoin baseline because the stablecoin continues to circulate in the foreign economy. ### Macroeconomic responses to a foreign contractionary monetary shock - Figure 8 (no CFMs) and Figure 9 (with CFMs) compare three cases: no stablecoin, with stablecoin, and after the domestic ban is imposed. - Key qualitative outcomes of imposing the ban: - The ban largely returns the domestic economy to the ‘no stablecoin’ path, though not identically because the stablecoin still circulates abroad. - The ban tempers the amplification effect of the stablecoin, reducing the negative effects on consumption, investment and output. - The ban reduces stress in the banking sector relative to the case with the stablecoin available domestically. - When CFMs are present, the ban helps preserve the effectiveness of the pre-stablecoin CFM regime. ### Feasibility and enforcement considerations - Practical enforcement of a fully comprehensive ban is an open question and likely varies across countries. - A fully comprehensive ban would be difficult to enforce: households and merchants in the informal sector could evade a legal prohibition by using the stablecoin if it confers advantages such as anonymity coupled with low transaction costs. - Cross-country coordination can play a role in enforcement: - If the stablecoin issuer is headquartered in the foreign economy, that foreign government may have greater ability to enforce compliance with regulation. - The foreign stablecoin issuer may be more likely to impose restrictions on domestic household use if instructed by the legal authority within whose jurisdiction the firm and its workers are located. - If the stablecoin issuer is footloose and bases itself ‘offshore’ in third countries, broader multilateral coordination could be required. ### Implications summarized in the paper’s broader context - The paper’s model shows that availability of the stablecoin can amplify currency substitution and weaken monetary policy transmission, worsening recessionary shocks and increasing banking stress. - CFMs on domestic holdings of foreign bonds increase responsiveness of stablecoin holdings to a contractionary foreign monetary policy shock because households reallocate toward the stablecoin when other diversification channels are obstructed. - A domestic CBDC can reduce stablecoin holdings in equilibrium but does not counteract the stablecoin’s role in transmitting foreign shocks because the CBDC remains denominated in domestic currency. - A comprehensive unilateral ban on holding and using the stablecoin domestically could help mitigate these transmission and amplification effects, with cross-country cooperation potentially improving enforceability. - The paper notes further policy options and differences in the case of a foreign CBDC vs a foreign stablecoin as areas for future research. *Section 5.2 — Banning the Stablecoin.* ### References ### References (wpiea2023249-print-pdf - References) ### Major themes in the references - Literature on central bank digital currency (CBDC), stablecoins (SC), crypto-assets, and digital money: multiple IMF papers, BIS Working Papers, academic journals, and working papers cited. - Quantitative macro-finance frameworks and DSGE models underpinning analysis: Adrian et al. (2020, 2021), Gerali et al. (2010), Christiano et al. (2005), Gali and Monacelli (2005), Benigno (2009), Gertler and Kiyotaki (2010), Kiyotaki and Moore (1997). - International spillovers and global financial cycle: papers by Miranda-Agrippino and Rey (2020), Bernanke et al. (1999), Rey (2013), Gopinath et al. (2020). - Policy-focused work on crypto-assets, capital flows, and CFMs: IMF.2020a; IMF.2023; IMF-FSB.2023; He et al. (2022). - Empirical and theoretical work on currency competition, currency substitution, and exchange-rate pass-through: Eichengreen (2012), Fern ández-Villaverde and Sanches (2019), Gopinath, Itskhoki, and Rigobon (2010), Ozbilgin (2012). - Banking, financial intermediation, and credit frictions: Bernanke, Gertler, and Gilchrist (1999); Gertler and Karadi (2011, 2015); Gerali et al. (2010); Carlstrom and Fuerst (1997). --- ### Appendix (Model details and equilibrium conditions) ### A.1 Intratemporal allocation decisions (home and foreign) - Consumption bundle (home): - C_t = [(1−γ)^{1/η} C_{Ht}^{(η−1)/η} + γ^{1/η} C_{Ft}^{(η−1)/η}]^{η/(η−1)} (Equation 62) - Between-good optimization (home): - Budget constraint: P_{Ht} C_{Ht} + P_{Ft} C_{Ft} = Z_t (Equation 64) - First-order conditions: - C_{Ht} = (1−γ)(ζ P_{Ht})^{−η} C_t (Equation 65) - C_{Ft} = γ(ζ P_{Ft})^{−η} C_t (Equation 66) - Demand functions in terms of domestic CPI P_t: - C_{Ht} = (1−γ) (P_{Ht}/P_t)^{−η} C_t (Equation 67) - C_{Ft} = γ (P_{Ft}/P_t)^{−η} C_t (Equation 68) - Domestic CPI: - P_t = [(1−γ) P_{Ht}^{1−η} + γ P_{Ft}^{1−η}]^{1/(1−η)} - Normalized relative prices: p_{Ht} = P_{Ht}/P_t, p_{Ft} = P_{Ft}/P_t - C_{Ht} = (1−γ) p_{Ht}^{−η} C_t (Equation 69) - C_{Ft} = γ p_{Ft}^{−η} C_t (Equation 70) - 1 = [(1−γ) p_{Ht}^{1−η} + γ p_{Ft}^{1−η}]^{1/(1−η)} (Equation 71) - Investment bundle follows analogous structure: - I_t = [(1−γ)^{1/η} I_{Ht}^{(η−1)/η} + γ^{1/η} I_{Ft}^{(η−1)/η}]^{η/(η−1)} (Equation 72) - I_{Ht} = (1−γ) p_{Ht}^{−η} I_t (Equation 73) - I_{Ft} = γ p_{Ft}^{−η} I_t (Equation 74) - Foreign economy mirror (asterisked variables) with: - C^*_t = [γ^*^{1/η} (C^*_{Ht})^{(η−1)/η} + (1−γ^*)^{1/η} (C^*_{Ft})^{(η−1)/η}]^{η/(η−1)} (Equation 75) - C^*_{Ht} = γ^* (p^*_{Ht})^{−η} C^*_t (Equation 76) - C^*_{Ft} = (1−γ^*) (p^*_{Ft})^{−η} C^*_t (Equation 77) - 1 = [γ^* (p^*_{Ht})^{−1/η} + (1−γ^*) (p^*_{Ft})^{−1/η}]^{−η} (Equation 78) - Investment analogues: Equations 79–81 - Real exchange rate definitions: - s_t = e_t P^*_t / P_t (Equation 82) - s_t / s_{t−1} = Δ e_t π^*_t / π_t, where Δ e_t = e_t / e_{t−1} (Equation 83) ### A.2 Domestic household: preferences, budget constraint, and FOCs - Utility: - U_t = E_t Σ_{t=0}^∞ β^t [log(C_t) − Ψ (1−L_t)^{1+φ} / (1+φ)] (Equation 51 in appendix; labeled U_t) - Budget constraint (domestic currency terms): - C_t + Q_t K^h_t + D_t / P_t + s_t b_{Ft}/P_t + M_t/P_t + s_t M^*_{H,t}/P_t + P_{sc,t} s_t SC_t / P_t + χ_h(K^h_t, K_t) + τ(1−j_t) = w_t H_t + (Z_t + λ Q_t) K^h_{t−1} + R^D_{t−1} D_{t−1}/P_t + ε_m M_{t−1}/P_t + s_t M^*_{H,t−1}/P_t + P_{sc,t} s_t SC_{t−1}/P_t + s_t (1−τ_{d,t−1}) R_{z,t−1} b_{F,t−1}/P_t + Π_t (Equation 84) - Aggregation and cash-in-utility structures: - M_t / P_t = (1/n_t) ∫_0^{n_t} c_t(j) dj (Equation 85) - Z_t = (1/n_t) ∫_0^{n_t} c_t(j) dj = Ω(D_t, e_t M^*_{H,t}, e_t SC_t) / P_t (Equation 86) - L_t + H_t + κ n_t = 1 (Equation 87) - First-order conditions (selected): - FOC wrt consumption: λ_t = 1 / C_t (Equation 88) - FOC wrt deposits: λ_t = β E_t [λ_{t+1} R^D_t / π_{t+1}] + λ Ω_t n_t μ_D D^{−σ_D}_t (Equation 89) - FOC wrt stablecoins: λ_t P_{SC,t} s_t^{−1} = E_t [β λ_{t+1} P_{SC,t+1} / π_{t+1}] + λ Ω_t n_t μ_{SC} SC^{−σ_{SC}}_t (Equation 90) — exact algebra as in source - FOC wrt foreign cash: λ_t s_t^{−1} = E_t [β λ_{t+1} / π_{t+1} s_{t+1}] + λ Ω_t n_t μ_{M^*} M^{−σ_{M^*}}_{F,t} (Equation 91) - FOC wrt domestic cash: λ_t = E_t [β λ_{t+1} / π_{t+1}] + λ m_t n_t (Equation 92) - FOC wrt household capital supply: λ_t = β E_t [λ_{t+1} (Z_{t+1} + λ Q_{t+1}) / Q_t + χ_h(K^h_t, K_t)] (Equation 93) - FOC wrt labor: w_t = Ψ (1−L_t)^{φ} λ_t (Equation 94) - FOC wrt foreign bonds with capital outflow control: λ_t = β E_t [λ_{t+1} (1−τ_{d,t}) R_{z,t} / π^*_{t+1} (s_{t+1}/s_t)] (Equation 95) - FOC wrt investment (capital producer): 1 = Q_t [1 − Ω_k/2 (I_t/I_{t−1} − 1)^2 − Ω_k (I_t/I_{t−1})(I_t/I_{t−1} − 1)] + β E_t [Q_{t+1} λ_{t+1}/λ_t Ω_k (I_{t+1}/I_t)^2 (I_{t+1}/I_t − 1)] (Equation 96) ### A.3 Foreign household: preferences, budget constraint, and FOCs - Utility (foreign): - U^*_t = E_t Σ_{t=0}^∞ β^t [ (C^*_t)^{1−σ^*} / (1−σ^*) − κ L_H^{*1+φ^*}_t / (1+φ^*) + μ_{M^*} (M^*_t)^{1−σ_{M^*}}/(1−σ_{M^*}) + μ_{SC^*} (SC^*_t)^{1−σ_{SC^*}}/(1−σ_{SC^*}) ] (Equation 97) - Budget constraint (foreign currency terms): - C^*_t + I^*_t + B^*_{Ft}/P^*_t + D^*_t/P^*_t + M^*_t/P^*_t + P_{sc,t} SC^*_t / P^*_t = w^*_t H^*_t + r^{k*}_t K^*_{t−1} + r^*_{t−1} D^*_{t−1}/P^*_t + R^*_{t−1} B^*_{F,t−1}/P^*_t + P_{sc,t} SC^*_{t−1}/P^*_t + R^*_{t−1} bF^*_{t−1}/P^*_t + Γ^*_t (Equation 98) - Capital accumulation: - K^*_t = (1−δ^*) K^*_{t−1} + [1 − Ω^*_k / 2 (I^*_{t+1}/I^*_t − 1)^2] I^*_t (Equation 99) - Selected FOCs: - Consumption: λ^*_t = (C^*_t)^{−σ^*} (Equation 100) - Deposits and internationally traded bonds: λ^*_t = β E_t [λ^*_{t+1} r^*_t / π^*_{t+1}] (Equation 101) - Cash: μ_{M^*} M^{*−σ_{M^*}} = λ^*_t − β E_t λ^*_{t+1} (Equation 102) - Stablecoins: μ_{SC^*} M^{*−σ_{SC^*}} = λ^*_t P_{sc,t} − β E_t λ^*_{t+1} P_{sc,t+1} (Equation 103) - Labor: w^*_t = κ L_H^{*φ^*}_t λ^*_t (Equation 104) - Capital: 1 = β E_t [λ^*_{t+1}/λ^*_t R^{k*}_{t+1} + (1−δ^*) q^*_{t+1}/q^*_t] (Equation 105) - Investment: analogous adjustment-cost condition (Equation 106) --- ### Calibrated parameter values (Table 3) - β 0.995 — Discount rate of household (IPF) - β_E 0.975 — Discount rate of entrepreneur (Gerali et al. (2010)) - θ 0.501 — Elasticity of leverage wrt foreign borrowing - σ 0.94 — Survival probability - ξ 0.0045 — Fraction of total assets brought by new banks - κ_b 0.0219 — Management cost for foreign borrowing - ζ_1 — Inverse of Frisch elasticity of labor supply (value not shown in excerpt) - ζ_0 7.1463 — Labor disutility (labor in steady state around 1/3) - κ_h 0.0197 — Cost parameter of direct finance (ABK) - α_K 0.30 — Cost-share of capital (IPF) - δ 0.025 — Depreciation rate - η_d 8 — Elasticity of demand - ω_c 0.4 — Calvo parameter of price stickiness - κ_I 2.48 — Investment adjustment cost (Christiano et al. (2005)) - ε_m 1 — Storage cost - μ_{M^*} 0.35 — Weight of USD - μ_D 2.65 — Weight of deposit - μ_{SC} 1.65 — Weight of SC - μ_{DC} 1 — Weight of CBDC - σ_{M^*} 1.35 — Elasticity of cash - σ_D 2.85 — Elasticity of deposit - σ_{SC} 2.85 — Elasticity of SC - Ā 1.000 — Steady state productivity - ρ_i 0.82 — Taylor rule persistence (IPF) - φ_π 3 — Taylor rule response to CPI inflation (IPF) - φ_π^D 1.5 — Taylor rule response to domestic inflation (IPF) - φ_Y 0.09 — Taylor rule response to output (IPF) - φ_bf 0.25 — CFMs rule response - γ 0.297 — Home bias - n 0.25 — Relative size of the population of the home economy - ς 67 — Relative size of the foreign economy to domestic economy - η 0.8 — Elasticity of substitution (IPF) - b 0.95 — Share of foreign cash in SC issuer tech - ω −1.5 — Leontief utility parameter - TB/Y 0.0037 — Steady state of the trade balance to GDP (IPF) - C/Y 0.65 — Steady state of consumption to GDP - K/Y 7. — Steady state of consumption to GDP - I/Y 0.18 — Steady state of investment to GDP - J 0.23 — Share of cash payment - Π 4% — Steady state of annualized inflation Notes: Table shows main parameters used in the model and sources. ABK refers to Aoki et al. (2016); IPF refers to Adrian et al. (2021). --- ### Holdings by payment asset (Table 4) — representative domestic household steady-state distribution (% by domestic currency value) - Scenario: No SC - Domestic Cash 35.01% - Deposit 54.31% - Stablecoins 0% - Foreign Cash 10.68% - Domestic CBDC 0% - Scenario: With SC - Domestic Cash 35.68% - Deposit 50.85% - Stablecoins 1.97% - Foreign Cash 11.49% - Domestic CBDC 0% - Scenario: Ban SC - Domestic Cash 34.56% - Deposit 54.86% - Stablecoins 0% - Foreign Cash 10.57% - Domestic CBDC 0% - Scenario: CBDC I.A - Domestic Cash 31.68% - Deposit 47.28% - Stablecoins 1.85% - Foreign Cash 11.05% - Domestic CBDC 8.12% - Scenario: CBDC I.B - Domestic Cash 31.08% - Deposit 45.47% - Stablecoins 1.77% - Foreign Cash 10.97% - Domestic CBDC 10.69% - Scenario: CBDC II - Domestic Cash 33.04% - Deposit 41.06% - Stablecoins 1.51% - Foreign Cash 7.89% - Domestic CBDC 16.49% Notes: Distribution measured in % by domestic currency value. --- ### Figures and impulse-response summaries (Figures 10–12) - Figure 10: Responses to a positive cost-push shock (domestic economy) - Time in quarters up to 12 (labels show 4, 8, 12). - Comparison: model without stablecoin (black) vs. with stablecoin (red). - Variables shown (impulse responses in % deviation from steady state; inflation, credit spread, and interest rate are annualized): - Output, Inflation (Annualized), Nominal Interest Rate (Annualized), Investment, Consumption, Domestic Deposit, Foreigner Deposit, Networth of Banks, Credit Spread (Annualized), Holding of Cash, Holding of Stablecoin, Price of Stablecoin (domestic currency), Trade Balance (share of GDP), Real Exchange rate, Holding of U.S Gov Bond. - Figure 11: Responses with CFMs to a contractionary foreign monetary policy shock (domestic economy) - Comparison: no CBDC (red), CBDC type I.A (black dashed), CBDC type I.B (dashed blue). - Variables shown (same units conventions as Figure 10): - Output, Inflation (Annualized), Nominal Interest Rate (Annualized), Investment, Consumption, Domestic Deposit, Holding of CBDC, Networth of Banks, Credit Spread (Annualized), Holding of Cash, Holding of Stablecoin, Price of Stablecoin (domestic currency), Trade Balance (share of GDP), Real Exchange rate, Holding of U.S Gov Bond. - Figure 12: Responses with CFMs to a contractionary foreign monetary policy shock (domestic economy) - Comparison: no CBDC (red) vs. CBDC type II (dashed blue). - Variables shown (same units conventions): - Output, Inflation (Annualized), Nominal Interest Rate (Annualized), Investment, Consumption, Domestic Deposit, Holding of CBDC, Networth of Banks, Credit Spread (Annualized), Holding of Cash, Holding of Stablecoin, Price of Stablecoin (domestic currency), Trade Balance (share of GDP), Real Exchange rate, Holding of U.S Gov Bond. --- *Macro-Financial Impacts of Foreign Digital Money — Working Paper No. WP/2023/249* --- _Source: https://www.imf.org/-/media/files/publications/wp/2023/english/wpiea2023249-print-pdf.pdf_