0
$\begingroup$

I'm statistically analyzing the bitcoin prices. We have the ability to analyze them based on hourly, daily, monthly etc. So based on the requirement it is possible to select the period. Consider if we are looking at all hourly prices or longer intervals (i.e. daily or weekly prices) What is the most suitable time interval to select ? (hourly or longer). Please explain

And what is the most widely analyzed time? (hourly/daily)? and why?

P.S

I have to estimate the risk associated to the cryptocurrency asset. I must charac-terize/describe the data series statistically estimating the kind of probability distribution and the associated parameters; estimate risk measures such as VaR, CVaR associated with your portfolio at dierent time horizons in both parametric and non-parametric approaches; estimating extreme event tail properties of the distribution.

Improve this question
$\endgroup$
2
  • 1
    $\begingroup$ It depends on the purpose of the analysis. For instance, if it is to create a trading strategy and prices vary rapidly, a daily or hourly analysis could lead to terrible results; but if it is to help understand what factors might be related to the prices, then perhaps an hourly analysis would help smooth out some of the "noise" and lead to better insight. I hope this makes it clear that your questions about "most suitable" and "most widely analyzed time" really aren't answerable because the meaning of "statistically analyze" isn't clear. $\endgroup$
    – whuber ♦
    Commented Feb 10, 2019 at 13:04
  • $\begingroup$ Hi Thanks alot for replying. I updated my answer. There I have mentioned what exactly I need to do with the dataset. If it is the requirement, what do you think? $\endgroup$
    – Hiru
    Commented Feb 10, 2019 at 13:13

0

Reset to default

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.