Hi guys,
Anyway I came across chebyshev's theorem during my statistic lessons and it suggests that for all data regardless of distribution, (1-1/k^2)*100% would be within k standard deviations regardless of any distribution.
I was wondering if this law applies to time series data (such as price), and how it can be used. Because bollinger bands only sample from the latest x number of bars, the sampled data always changes with every tick/period. In that case, how can we utilise this law to gain an edge in terms of expected value and expected losses (regression of price within a defined number of standard deviations)?
P.S I have just started on my statistic classes so pls be patient haha!
Anyway I came across chebyshev's theorem during my statistic lessons and it suggests that for all data regardless of distribution, (1-1/k^2)*100% would be within k standard deviations regardless of any distribution.
I was wondering if this law applies to time series data (such as price), and how it can be used. Because bollinger bands only sample from the latest x number of bars, the sampled data always changes with every tick/period. In that case, how can we utilise this law to gain an edge in terms of expected value and expected losses (regression of price within a defined number of standard deviations)?
P.S I have just started on my statistic classes so pls be patient haha!