If I use log scale comparing two stocks on a line chart, should both stocks start from the same point? I'm interested in comparing the appreciation of the two stocks over time.
But is it important that the two lines start from the same point of origin? Does it change the result if they start from different levels?
Another way of comparing two stocks is to look at the relative returns as percentage of the price. Suppose you do this for daily close prices: Relative return [%] = (Close(day n)-Close(day (n-1)))/Close(day n) * 100% This formula looks at the price increase/decrease on day n versus the previous day (day (n-1)) and divide that to day n's close price. Multiply this by 100% to convert to a percentage. Do this for two (or more) stocks and you can compare their relative performance.
I'm accustomed to doing that but charting price that way is misleading. Using a log scale means that the same slope equates to the same growth rate for each stock. Endpoint to endpoint, percent change is okay but it distorts the growth path the stock actually took. It flattens the early years and makes the recent years look hyperbolic.
I am not charting the normalized price. I am charting the normalized growth rate. 0.01 USD price change on a 1 USD stock is the same percentage as 1 USD price change on a 100 USD stock. A logarithmic comparison does exactly the opposite of what you describe: it flattens the recent years of a growth in price. The analysis method you choose depends on what kind of visualization you are looking for.
You're over complicating what you are trying to find, IMHO. It has NOTHING to do with the PA/movement of the stock, but the fundamentals of the company that underlies the stock.