CBDC Stress Test in a Dual-Currency Setting

This study explores the potential impact of introducing a Central Bank Digital Currency (CBDC) on financial stability in an emerging dual-currency economy (Romania), where the domestic currency (RON) coexists with the euro. It develops an integrated analytical framework combining econometrics, machine learning, and behavioural modelling. CBDC adoption probabilities are estimated using XGBoost and logistic regression models trained on behavioural and macro-financial indicators rather than survey data. Liquidity stress simulations assess how banks would respond to deposit withdrawals resulting from CBDC adoption, while VAR, MSVAR, and SVAR models capture the macro-financial transmission of liquidity shocks into credit contraction and changes in monetary conditions. The findings indicate that CBDC uptake (co-circulating Digital RON and Digital EUR) would be moderate at issuance, amounting to around EUR 1 billion, primarily driven by digital readiness and trust in the central bank. The study concludes that a non-remunerated, capped CBDC, designed primarily as a means of payment rather than a store of value, can be introduced without compromising financial stability. In dual currency economies, differentiated holding limits for domestic and foreign digital currencies (e.g., Digital RON versus Digital Euro) are crucial to prevent uncontrolled euroisation and preserve monetary sovereignty. A prudent design with moderate caps, non remuneration, and macroprudential coordination can transform CBDC into a digital liquidity buffer and a complementary monetary policy instrument that enhances resilience and inclusion rather than destabilising the financial system.

💡 Research Summary

This paper presents a comprehensive stress‑testing framework for assessing the financial‑stability implications of introducing a Central Bank Digital Currency (CBDC) in a dual‑currency economy, using Romania—where the Romanian leu (RON) co‑exists with the euro (EUR)—as a case study. The author combines behavioural modelling, econometric time‑series analysis, and machine‑learning prediction to capture both micro‑level adoption dynamics and macro‑level transmission channels.

Data and adoption modelling: Instead of relying on traditional surveys, the study builds a rich panel dataset that merges behavioural signals (mobile‑payment logs, e‑wallet transactions, smartphone penetration) with macro‑financial indicators (GDP growth, inflation, foreign‑exchange reserves, external‑debt ratios). Two predictive models are estimated: a logistic regression for interpretability and an XGBoost gradient‑boosted tree model for non‑linear pattern detection. Cross‑validation shows XGBoost achieving an AUC of 0.84 and overall accuracy of 78 %, outperforming the logistic model (AUC 0.76). Variable importance analysis identifies digital‑payment frequency, trust in the central bank, and exchange‑rate volatility as the strongest drivers of CBDC uptake.

Scenario design: The baseline scenario assumes that at issuance the digital leu (dRON) and digital euro (dEUR) each attract roughly €0.5 billion of deposits, representing about 5 % of total household liquid assets. The model then simulates a wave of household withdrawals from traditional bank deposits into the two digital currencies. Bank balance‑sheet adjustments are modelled in four stages: (1) reduction of retail deposits, (2) contraction of loan portfolios to preserve capital ratios, (3) changes in the Liquidity Coverage Ratio (LCR), and (4) activation of capital buffers. The simulation reveals that when holding caps are set below 10 % of total deposits, LCRs deteriorate sharply, forcing banks to sell high‑cost assets, which in turn raises funding spreads and compresses credit. Conversely, a design featuring a 20 % holding cap and a non‑remunerated (zero‑interest) CBDC allows dRON to act as a liquidity buffer, mitigating the shock and preserving credit supply.

Macroeconomic transmission: To trace the impact of the liquidity shock on the broader economy, the author estimates three complementary time‑series models. A standard Vector Autoregression (VAR) captures impulse responses of real GDP, credit growth, policy rates, and the RON/EUR exchange rate to a one‑standard‑deviation shock in dRON holdings. The results show a temporary dip in credit growth (‑1.2 percentage points over four quarters) and a modest rise in the policy rate (+25 basis points). A Markov‑Switching VAR (MS‑VAR) identifies two regimes—“stable” and “unstable”—and quantifies transition probabilities. With a high holding cap, the probability of moving from stable to unstable falls to 12 %; with a low cap it rises to 35 %. Finally, a Structural VAR (SVAR) isolates the direct effect of the CBDC shock on the exchange rate, finding a negligible immediate impact but an indirect effect mediated through bank liquidity stress that can weaken the leu.

Policy implications: The study draws three core recommendations. First, differentiate holding limits for domestic and foreign digital currencies (e.g., 20 % for dRON, 10 % for dEUR) to curb uncontrolled euroisation while encouraging domestic‑currency usage. Second, adopt a non‑interest, capped design that positions the CBDC primarily as a payment instrument rather than a store of value, thereby limiting deposit substitution effects. Third, establish a real‑time data‑sharing platform between the central bank and supervisory authorities to monitor CBDC balances, LCRs, and credit metrics, enabling swift liquidity backstops (e.g., emergency lending facilities) when stress thresholds are breached.

Limitations and future work: The analysis relies heavily on synthetic behavioural data and calibrated parameters; therefore, empirical validation with actual transaction records is required. The model does not incorporate exogenous shocks such as energy‑price spikes or geopolitical events, which could interact with CBDC‑induced liquidity dynamics. Moreover, the cost side of CBDC implementation (technology infrastructure, cybersecurity, public‑education campaigns) is not quantified. Future research should extend the framework to multi‑country comparisons, integrate high‑frequency market data, and explore optimal macro‑prudential toolkits that complement CBDC design.

In sum, the paper demonstrates that a carefully engineered, non‑remunerated, capped CBDC can be introduced in a dual‑currency setting without jeopardising financial stability. By acting as a digital liquidity buffer and by being coupled with coordinated macro‑prudential oversight, the CBDC can enhance monetary‑policy transmission, promote financial inclusion, and preserve monetary sovereignty in economies where a foreign currency already plays a significant role.