we consider natural logarithm or lag of the data series for our analysis. But what is the reason behind this. Why we are not considering the raw data for analysis. Is their any paper regarding this. Please answer.
We use natural logarithm for a distribution of data where some values are too large for some periods and other values are too small for other periods. This situation gives rise to outliers in your data. For example, if you are examining the relationship between bank size and bank profit with this type of data. (Profit: Yr 1: 1,200. Yr 2:1300. Yr 3:2500. Total Asset: Yr 1: 120,000. Yr 2: 2,000,000. Yr 3: 250,000.) This type of data show that the significance of your result will be due to the large value of Year 2 asset values not necessarily due to a well-behaved data distribution. This problem is also known as scale effect. Some academics call it 'outliers'. Therefore, to reduce the impact of these outliers on your data, we take the natural log of the entire distribution (instead of the raw values) to normalize the distribution of the data.