Statistics Question
Barry writes:
Hello Victor,
Hope all is well. I have been enjoying the YouTube videos you have been putting out.
I was wondering if the rules that apply to normal statistical tests should be as constrained when applied for financial data?
For example when comparing the mean returns of weeks following drops greater than 30% in something like bitcoin compared with it's weekly returns excluding those weeks if a tests value of 20% would be consistent with random as one might conclude with most other sciences or is that an indicator there could be something more meaningful to investigate?
Victor Niederhoffer writes:
Thats a very good question and related to it is the supposition that a 20% prob of chance occurrence mite be more p reductive of a regularity in the future than a 5% chance statistics I'm posting your query to the spec list since it doesn't contain any meals for a day but is a meal for a life time.