Two-Sample t-Test: paired vs. unpaired

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

Two-Sample t-Test: paired vs. unpaired

Andreas Klein-4
Hello.

I got stuck with a simple decision problem:

I have two time series of utilities of portfolio returns. The first portfolio consists of 100 stocks and the second portfolio consists of 10 of the 100 stocks.

The correlation coefficient between the portfolio returns is almost zero.

When I produce idependently from each other the utility time series of each of the two portfolios I get a correlation coefficient of around 0.8

=> t.test(utility_series1, utility_series2, paired=???)

The facts: Both are independent at the beginning, but get highly correlated in the end. Therefore the results of t.test() strongly depends on the choice of paired vs. unpaired.

But I cannot decide whether I should treat the series as independent or not.

Any suggestions?


Regards,
Andreas




_______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-sig-finance
-- Subscriber-posting only. If you want to post, subscribe first.
-- Also note that this is not the r-help list where general R questions should go.
Reply | Threaded
Open this post in threaded view
|

Re: Two-Sample t-Test: paired vs. unpaired

Arun.stat
Do I really need I assume independent series for doing paired t-test? Independence among the paired observation should be sufficient for a valid paired t-test.
Reply | Threaded
Open this post in threaded view
|

Re: Two-Sample t-Test: paired vs. unpaired

Adams, Zeno
In reply to this post by Andreas Klein-4
Since you are interested in testing the utilities you should look at the correlation of the utilities. If you do not use the paired version you ignore this information. I recommend using the paired t-test.

Zeno


-----Original Message-----
From: [hidden email] on behalf of Andreas Klein
Sent: Sat 12/11/2010 11:33 AM
To: [hidden email]
Subject: [R-SIG-Finance] Two-Sample t-Test: paired vs. unpaired
 
Hello.

I got stuck with a simple decision problem:

I have two time series of utilities of portfolio returns. The first portfolio consists of 100 stocks and the second portfolio consists of 10 of the 100 stocks.

The correlation coefficient between the portfolio returns is almost zero.

When I produce idependently from each other the utility time series of each of the two portfolios I get a correlation coefficient of around 0.8

=> t.test(utility_series1, utility_series2, paired=???)

The facts: Both are independent at the beginning, but get highly correlated in the end. Therefore the results of t.test() strongly depends on the choice of paired vs. unpaired.

But I cannot decide whether I should treat the series as independent or not.

Any suggestions?


Regards,
Andreas




_______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-sig-finance
-- Subscriber-posting only. If you want to post, subscribe first.
-- Also note that this is not the r-help list where general R questions should go.



EBS European Business School gemeinnuetzige GmbH, Universitaet fuer Wirtschaft und Recht i.Gr. - Amtsgericht Wiesbaden HRB 19951 - Umsatzsteuer-ID DE 113891213 Geschaeftsfuehrung: Prof. Dr. Christopher Jahns,  President; Prof. Dr. Rolf Tilmes, Dean Business School; Sabine Fuchs, CMO; Prof. Dr. Dr. Gerrick Frhr. v. Hoyningen-Huene, Dean Law School; Aufsichtsrat: Dr. Hellmut K. Albrecht, Vorsitzender
        [[alternative HTML version deleted]]

_______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-sig-finance
-- Subscriber-posting only. If you want to post, subscribe first.
-- Also note that this is not the r-help list where general R questions should go.