That assumption is not embedded in the test, which was measuring relative moves, not absolute share price. We have reasonable control group here, which is all the other days Elon
didn't sell.
Here's the entire time series of TSLA daily % change I looked at. The average daily TSLA change was -0.1%. Can you point out for us which of these days Elon sold stock on this chart?
View attachment 870186
If Elon's individual sales have any material effect, then it will show up as a tendency to have a different % change for TSLA on those days compared to the baseline expectation (-0.1%). Or alternatively, if the news of Elon selling has this effect, then it should show up a couple days after the Form 4 confirms the sale occurred.
That could be, but where is your supporting evidence? Personally I haven't thought of any actual way to test whether that's true.
Depending on what evidence you have, the "I think" preface suggested by
@2daMoon might be appropriate.