Purpose: The paper aims to examine whether the daily conventional money market overnight rate influences the monthly investment rate of Islamic deposits in Malaysia. The traditional approach, which averages the high-frequency data to match the low-frequency data, results in information loss for the high-frequency data. Design/methodology/approach: The paper uses the mixed data sampling (MIDAS) model to study the relationship between Islamic banking and conventional banking. The Malaysian data are used for the analysis as Malaysia has one of the most developed Islamic financial industries in the world, and it is well-known for its dual banking system. Findings: The evidence shows that the conventional overnight rate has a positive effect on the Islamic deposit rate. The results are consistent for Islamic deposit rates with different maturities. The positive aggregate effect holds when the lag length of the daily conventional overnight rate goes up to 90 days. Additional evidence shows that the daily conventional overnight rate has a similar effect on the conventional deposit rate. Originality/value: This paper documents that the relationship between Islamic banking and conventional banking is not monotonous. When high-frequency data is averaged with low-frequency data, the non-linear relationship will be masked. It highlights the importance of using high-frequency data to get a detailed picture.