Understanding how monetary policy affects the real economy requires analyzing banks' responses. Since the 2008 crisis, the role of banks in monetary policy transmission has gained renewed attention, particularly through the bank lending channel (Bernanke and Blinder, 1988), which posits that monetary policy shocks alter banks' loan supply. The crisis also highlighted the link between monetary policy and excessive bank risk-taking (Altunbas et al., 2012), leading to the concept of the bank risk-taking channel (Borio and Zhu, 2012), where low interest rates increase risk tolerance. Given banks' central role in funding economic activity, the monetary policy-bank output nexus is crucial for policy design and financial stability. This study focuses on the "search-for-yield" incentive as a mechanism for monetary policy transmission. Theoretically, lower interest rates reduce bank profitability, prompting a search for higher-yielding, and thus riskier, assets (Rajan, 2006). This "search-for-yield" effect is particularly pronounced when banks have rigid return targets. This research empirically tests whether this incentive underlies the bank lending and risk-taking channels of monetary policy. This paper differs from existing research in several ways. While previous studies examined how banks' responses to monetary policy shocks vary based on bank-specific characteristics (size, liquidity, capitalization, ownership) or macroeconomic factors (financial deregulation, competition, integration), this study specifically investigates the "search-for-yield" mechanism as an underlying supply-side driver. Furthermore, beyond the bank lending and risk-taking channels, this research explores the role of the bank liquidity creation channel, examining how monetary policy impacts banks' core function of creating liquidity for the real economy. Analyzing these three channels provides a comprehensive understanding of monetary policy transmission. The empirical analysis uses data from the Vietnamese banking sector (2007-2019), focusing on loan growth (bank lending channel), the Z-score index (bank risk-taking channel), and liquidity creation measures (bank liquidity creation channel) from Berger and Bouwman (2009). Three interest rate indicators (refinancing, rediscounting, lending rates) are employed to provide a broader perspective. The generalized method of moments (GMM) estimator is the primary econometric technique, with robustness checks using fixed effects models and Driscoll-Kraay standard errors. Vietnam's relatively underdeveloped capital market and reliance on bank financing make it a suitable context for this study, particularly given the SBV's diverse monetary policy tools and banking sector reforms.

This research builds upon the literature analyzing the bank lending and risk-taking channels of monetary policy transmission. The bank lending channel suggests that monetary policy tightening reduces loanable funds, potentially contracting lending (Bernanke and Blinder, 1988; Kashyap and Stein, 1995). Alternatively, monetary policy impacts banks' external finance premium based on their balance sheet strength (Disyatat, 2011). Monetary expansion, conversely, may boost loan supply due to improved bank balance sheets (Maddaloni and Peydró, 2011). Existing research explores the influence of bank-specific characteristics (size, capital, liquidity, risk) and macroeconomic factors (prudential policies, market development) on the bank lending channel (Kashyap and Stein, 2000; Kishan and Opiela, 2006; Altunbas et al., 2010; Sáiz et al., 2018; Hussain and Bashir, 2019; Zhan et al., 2021; Fiador et al., 2022; Fabiani et al., 2022; Cheng and Wang, 2022). The bank risk-taking channel suggests that monetary policy affects bank risk-taking through interest rate changes impacting risk perception and tolerance, leverage adjustments altering risk pricing, and communication policies influencing agents' perceptions (Borio and Zhu, 2012; Dell'Ariccia et al., 2014; Angeloni et al., 2015). Empirical research confirms this channel's presence, often conditioned on bank-specific factors (Altunbas et al., 2014; Jiménez et al., 2014; Drakos et al., 2016; Dang and Huynh, 2022b). The "search-for-yield" effect is a key theoretical underpinning for both channels. Lower interest rates, coupled with sticky return targets, can increase risk tolerance (Rajan, 2006). The Samuelson effect (Samuelson, 1945) and the lending rate elasticity exceeding deposit rate elasticity (Hancock, 1985) further contribute to this mechanism. Orzechowski (2017) and Dang and Dang (2020) provide some related findings but lack a direct examination of the "search-for-yield" incentive. This study addresses this gap by incorporating multiple channels and comprehensive "search-for-yield" measures.

The study uses panel data from 31 Vietnamese commercial banks (2007-2019), representing over 90% of the sector's assets. Data on loan growth, the Z-score index, and liquidity creation measures (from Berger and Bouwman (2009)) are collected from bank financial reports. Monetary policy indicators (refinancing, rediscounting, lending rates) and macroeconomic data (inflation, GDP growth, VNindex) are sourced from the SBV, World Bank, and Vietstock databases, respectively. Outliers are addressed through winsorization. The main empirical model is: Y = α₀ + α₁ × Yᵢ,_ₜ₋₁ + α₂ × MPIᵢ,_ₜ₋₁ + α₃ × Search for yieldᵢ,_ₜ₋₁ + α₄ × MPIᵢ,_ₜ₋₁ × Search for yieldᵢ,_ₜ₋₁ + α₅ × Zᵢ,_ₜ₋₁ + α₆ × Xᵢ,_ₜ₋₁ + εᵢₜ where Y represents bank output (loan growth, InZscore, liquidity creation growth), MPI denotes monetary policy indicators, "Search for yield" proxies bank incentives, Z includes bank-specific controls (size, capital, liquidity), X represents macroeconomic controls (stock return, economic cycle), and ε is the error term. Lagged variables address endogeneity. The two-step system GMM estimator (Blundell and Bond, 1998) is used, with instrument collapse (Roodman, 2009) and diagnostic tests (Hansen test, AR tests) ensuring validity. The Z-score index (equation 2) measures bank risk-taking. Short-term interest rates serve as monetary policy indicators. The "search-for-yield" proxies are the differences between ROA/ROE and their three-year averages (SFYroa, SFYroe). The interaction term MPI × Search for yield assesses the conditional role of "search-for-yield" incentives. Beyond loan growth and the Z-score, the study examines the impact on bank liquidity creation, using both "cat fat" and "cat nonfat" measures (Berger and Bouwman, 2009) to capture on and off-balance-sheet liquidity. Robustness checks include using alternative monetary policy indicators (average short-term lending rates), employing a static fixed-effects model with Driscoll-Kraay standard errors to account for heteroskedasticity and cross-sectional dependence, and using an alternative Z-score calculation.

The preliminary analysis reveals high loan growth volatility and a wide range of Z-scores and "search-for-yield" measures across banks. Correlation analysis supports the bank lending and risk-taking channels, with negative correlations between loan growth and interest rates, and positive correlations between bank stability and interest rates. Diagnostic tests validate the dynamic GMM estimations. **Bank Lending Channel:** Results consistently show a significant negative relationship between monetary policy indicators (refinancing, rediscounting rates) and loan growth, confirming the bank lending channel. Importantly, the interaction term between interest rates and "search-for-yield" measures is significantly positive, indicating that banks with stronger incentives to search for yield (i.e., those with lower profitability relative to past performance) exhibit a more pronounced response to monetary policy changes. A one percentage point decrease in refinancing rates increases loan growth by 2.188%, amplified by 0.290% for a one standard deviation decrease in SFYroa (Table 3). **Bank Risk-Taking Channel:** The study finds a significant positive relationship between monetary policy indicators and the InZscore, signifying that lower interest rates increase bank risk-taking (Table 4). The interaction term between interest rates and "search-for-yield" measures is significantly negative in most regressions, showing that the impact of monetary policy on risk-taking is stronger for banks with greater search-for-yield incentives. A one percentage point decrease in rediscounting rates increases the InZscore by 0.039% (Table 4). **Bank Liquidity Creation Channel:** The analysis confirms the bank liquidity creation channel, with significant negative relationships between monetary policy indicators and liquidity creation growth (Tables 5 and 6). The interaction term between interest rates and "search-for-yield" measures is consistently positive, demonstrating that banks with weaker search-for-yield incentives show a weaker response to monetary policy in terms of liquidity creation. A one percentage point decrease in refinancing rates increases "cat fat" liquidity creation by 4.006%, amplified by 2.809% for a one standard deviation increase in SFYroe (Table 5). **Robustness Checks:** The findings are robust across alternative monetary policy indicators (average short-term lending rates), econometric methodologies (fixed-effects with Driscoll-Kraay standard errors), and alternative Z-score measures (Table 7, Table 8, Table 9, and Table 10).

The findings strongly support the hypothesis that banks' "search-for-yield" incentives significantly influence the monetary policy transmission mechanism across the bank lending, risk-taking, and liquidity creation channels. Lower interest rates during monetary easing reduce bank profitability, prompting a more aggressive search for yield, leading to increased lending, higher risk-taking, and enhanced liquidity creation (or reduced impact when incentives are weaker). This confirms the presence of a supply-side effect driving the monetary policy-bank output nexus. The quantitative impacts across all channels are economically significant. The robustness checks reinforce the findings, highlighting the consistency of the results across different methodologies and alternative measures. This demonstrates the importance of considering the search-for-yield incentive when analyzing monetary policy transmission and designing policy interventions.

This study provides strong evidence that banks' "search-for-yield" incentives play a crucial role in the monetary policy transmission mechanism, affecting bank lending, risk-taking, and liquidity creation. This supply-side effect necessitates that monetary authorities incorporate this mechanism into their policy decisions. Future research could extend this analysis to other markets and investigate the demand-side effects in greater detail. Further exploration into interaction effects between the supply-side and demand-side effects of monetary policy could offer a comprehensive understanding of the phenomena.

The study's scope is limited to the Vietnamese banking sector, which may restrict the generalizability of the findings to other contexts. While the study provides evidence for a supply-side effect, a more thorough investigation into demand-side mechanisms is warranted. Additionally, other potential factors beyond search-for-yield that may impact monetary policy transmission were not explored in detail.