They can simplify the data preprocessing phase by reducing the amount of data that needs to be processed.

They can lead to inaccurate and biased predictions, affecting the model's reliability and compliance with financial regulations.

They can enhance the model's complexity unnecessarily, making it harder to interpret and increasing computational costs.

They can reduce the computational cost by eliminating the need for extensive data cleaning and preprocessing steps.

They can improve the model's scalability by automatically adjusting to the quality of input data without additional preprocessing.