Quantitative Risk Management, Log returns on Parametric Value at Risk

Hello guys, i have this huge coursework which i have been working for a long time and the deadline is tomorrow, so far i have been able to answer everything but one question on which i would really appreciate the help.

Basically one part of the coursework is to acquire historical data for a traded financial asset and suppose that you had invested 1 million pounds in this asset at the date given by the earliest date in your data. Using the data up to but not including 20th of february, calculate the simple daily returns for the asset and then calculate the value at risk and expected shorfall using several methods:
Basic Historical Simulation
Age-Weighted Historical Simulation
Parametric, using normal distribution without volatility adjustment
Parametric, using normal distribution, with volatility adjustment
Parametric Using an appropiate distribution, without volatility adjustment
Parametric, Using appropiate distribution and with volatility adjustment

After the calculation there are several questions but there is one that i havent been able to answer, its the following:

Explain why it would be problematic to have used log returns to calculate VaR and ES for any of the parametric methods in the previous questions.

all i have been able to find about log returns is that they are smaller than simple returns so i guess it could make the value of the risk lower than it really is, but i feel there is more to it. I would appreciate any help !!