As Korea moves toward regulating won-denominated stablecoins, concern has grown that a shift of transaction deposits from banks into stablecoins could weaken commercial banks' credit intermediation, so-called disintermediation. This study quantitatively assesses the size and nature of such disintermediation and whether its effect is closer to a full-blown credit crunch or to a structural transition in which funds simply change location. To this end, we develop an agent-based model (ABM) that captures the interactions and fund flows among households and firms, commercial banks, and stablecoin issuers. In addition, rather than assuming that a fall in deposits translates directly into a fall in loans, we analyze a benchmark interacting model that reflects how individual banks respond under loan-to-deposit ratio (LDR) regulation. At 10 percent adoption, net deposit outflows reach KRW 66 trillion (2.49 percent of total deposits) and the associated credit contraction KRW 83 trillion; at 15 percent, these rise to KRW 101 trillion (3.81 percent) and KRW 127 trillion. Still, more than 96 percent of total deposits remain within the banking sector, as issuer reserves held as bank deposits and user-side recycling appear to offset much of the outflow pressure. Sensitivity analysis shows that the share of issuer reserves held as bank deposits has a large effect on the degree of disintermediation, while in the benchmark model deposit outflows closely match the baseline, confirming robustness, and credit contraction, shaped by banks' LDR headroom, is smaller than in the baseline. These results suggest that won-stablecoin disintermediation is unlikely to amount to a systemic collapse and is better understood as a conditional phenomenon that can be managed through institutional design, with issuer structure, the holding of reserves as bank deposits, and LDR buffers as important policy priorities.