Deterministic vs stochastic trends

[@abheeksinha9334]

Thanks a lot for the explanation. Here the the simulated R code if anyone needs it. n=10000 b=1 rho=1 alpha=0.05 #Initialise x and y x

[@wexer82]

i wish i could understand this :(

[@frankthetank8216]

Is there a video which explains te concepts that lead to stochastic/deterministic models?I also don't understand the error parameter. Is this a constant value? If it's a variable, on what is it dependent? What does iid mean?

[@chaibennett8464]

thank you. a very clear lecture

[@dunwantoknow]

Thanks a lot, it help to clarify my confusion (:

[@MFper4mance]

Just to be sure (I may have skipped some important part), the deterministic trend is still a stochastic process right?

[@SpartacanUsuals]

+MFper4mance Thanks for your comment. Yes, it is still stochastic, due to the error term. Apologies if this was unclear. Best, Ben

[@MFper4mance]

+Ben Lambert Thanks!

[@IdoRosen]

This was what I was getting at in my other comment (now deleted): The linear model in the video is stationary, homoscedastic, and non-Markov, and the AR(1) model in the video is non-stationary, heteroscedastic, and first-order Markov; but, both are still stochastic processes (and not deterministic) because they have a random variable (the error term $\epsilon_t$). Using "deterministic" vs "stochastic" seemed confusing. (It may be a case of econometrics vs statistics terminology... In any case, thank you for making the video.) PS: I came here by way of answering a question about this video on Cross-Validated: stats.stackexchange.com/questions/273161/what-is-the-difference-between-deterministic-and-stochastic-model

[@felixa.wayland8173]

nooo

[@ivotavares6576]

+Ben Lambert , when you do the variance of the r.w. with drift, you're assuming that Var(X_0)=0, i.e. that the series has an initial fixed value? Otherwise, with an infinite past it would just diverge. Am I right?

[@snehalbhartiya6724]

Good stuff man!

[@AbdirahmanSuleiman-xf5nd]

thanks for your helping us

[@Raven-bi3xn]

Amazing! Thank you so much.

[@StephenGillie]

These videos are good, but very hard to hear. At maximum volume, this video is much quieter than most music videos. Can you increase the loudness?

[@odiseeacarmen1740]

homoskedastic and conditional and non-conditiional heteroskeastic nonstationarity means actually the stochastic trend? or am i wrong?

[@krabbypatty6896]

Thank youu. It helps me a lot.

[@tiggersmith7393]

I was with you for the first two seconds.....I think your brain is a different shape to mine.

[@marciachumpitazi8301]

love you benny

[@smsm314]

Hello sir, good job, good luck Please, I have a series it is DS (of model 3: with constant and trend) at the level, and TS at the first difference; are we going to the second difference or not? Ie is this series integer of order 2 or 1? Cordially

[@yul7960]

Clear, thanks!

[@littlerainyone]

I can't read the Matlab code. Do you offer a download elsewhere?

[@SpartacanUsuals]

Hi, here you are: clear; close all; clc; n=10000; % Number of time periods b=1; rho=1; %The parameter on the lagged value of x alpha=0.05; % The slope parameter % Initialise x and y x=zeros(n,1); x(1)=0; y=zeros(n,1); y(1)=0; % Generate the x and y series for i = 2:n x(i)=alpha+rho*x(i-1)+b*randn(); y(i)=alpha*i+b*randn(); end % Plot the x and y series zoom=1.0; FigHandle = figure('Position', [750, 300, 1049*zoom, 895*zoom]); plot(x, 'LineWidth', 1.4) hold on plot(y, 'm','LineWidth', 1.4) Hope that helps! Ben

[@littlerainyone]

Ben Lambert It does! Thanks a million. Please consider posting them all! I just finished the entire course. What a resource!

[@th4ros96]

I love you

[@jamesmayhew4387]

When?........when did i ask?