KPSS test explained: Time series stationarity (Excel)

[@NEDLeducation]

You can find the spreadsheets for this video and some additional materials here: drive.google.com/drive/folders/1sP40IW0p0w5IETCgo464uhDFfdyR6rh7 Please consider supporting NEDL on Patreon: www.patreon.com/NEDLeducation

[@judgewooden]

Great video, thanks for making it. You present it succinctly, which is very rare on KZbin.

[@gaborszigel5804]

Very useful, understandable and user-friendly video, thanks!

[@shayansalesi3851]

Love it, these keep getting better!

[@thegoru0106]

Great explanation!

[@NWKastaways]

To avoid look-ahead bias, would we want to calculate intercept and all that as a rolling period?

[@NEDLeducation]

Hi Juan, and thanks for the excellent question! It depends on the purpose of the test. If you undertake it for pair-trading, for example. applying it on a rolling window would be best for a realistic simulation.

[@NWKastaways]

@@NEDLeducationThanks for the reply. I coded your Engle Grenger method to ThinkOrSwim and did a rolling period from your other video. Just thought I'd ask and get your input. Thanks!

[@pberent]

Great explanation. Many thanks

[@dudiaharon9800]

thanks for your videos! you are awesome!! do you intend to show in the future videos about pooled/panel regressions? fixed effects and random effects models?

[@NEDLeducation]

Hi Dudi, and glad you are enjoying the channel! Yes, these are definitely in my long-term plans when I return to recording econometrics videos.

[@Trivitu]

Great video! I just have one question, I did this KPSS test in Excel, following your instructions step by step, the problem is that when I used Eviews software, the KPSS statistics were diferent, do you know why?

[@r327079]

Hi Sava, Great VDO. Can we look into the time series data of stocks pair in similar manner (for checking cointegration). What will X, Y for stocks Pair evaluation?

[@NEDLeducation]

Hi Robin, and thanks for the excellent question! Yes, this is a valid alternative (or an addition to) Dickey-Fuller test for pair trading applications.

[@KARE19269]

Plz do a video on PP test and break unit root tests.

[@krgonline]

nice one... One query.. how to interpret KPSS result if the KPSS statistic is in between the 1% or 5% or 10% level. The higher level critical value at 1% and the rejection of a higher statistic creates confusion whenever the statistic is in between

[@gaborszigel5804]

I am sorry, I am back again. I tried to estimate the KPSS test statistic based on this video in Excel, but it seems to me, that the results are different than in Gretl. I think, Gretl does the estimation the standard error differently (based on "robust estimation of variance" - whatever that shall mean). Do you have an idea, how this robust estimation is different? Thanks in advance!

[@gahinberwary507]

hi thank you for your videos why the results of this doesn’t meet the same as in Eveiw?

[@Yearjohn]

Savva, hi, thank for the great video, offtopic, but meant to ask for long time, is it possible to run quantile regression in excel? can you do it?

[@NEDLeducation]

Hi Yerzhan, and glad you liked the video! As for your suggestion, I was actually planning to cover quantile regressions for quite some time, it is not straightforward, but certainly doable in Excel so stay tuned :)

[@Yearjohn]

@@NEDLeducation спасибо за ответ! Будем ждать ! you rock !

[@NEDLeducation]

Как и обещал, вот видео по квантильной регрессии :) kzbin.info/www/bejne/qoaaZ5WXm8uihM0

[@Yearjohn]

@@NEDLeducation super, you are the best!

[@Regular.Biceps]

Sir, advice and a request - you should be on Coursera, believe me... I have studied these topics there but they are nowhere close to the practical application and practical understanding that I've gained through your videos... Maybe the better resources will help you develop more quality videos and more detailed conceptual and practical use of econometrics models.. you're the teacher everyone needs

[@NEDLeducation]

Hi, and thanks so much for these kind words. Who knows, I might consider branching out someplace like Coursera one day :)

[@Regular.Biceps]

@@NEDLeducation would love to be enrolled in your courses

[@jaysonwang9692]

Can this be used for daily data?

[@NEDLeducation]

Hi Jayson, yes, it can! The application will be the same for any frequency of the time series.