Search
× Search
Saturday, December 21, 2024

Archived Discussions

Recent member discussions

The Algorithmic Traders' Association prides itself on providing a forum for the publication and dissemination of its members' white papers, research, reflections, works in progress, and other contributions. Please Note that archive searches and some of our members' publications are reserved for members only, so please log in or sign up to gain the most from our members' contributions.

WRIGHT,J.H.,2000 Alternative Variance-Ratio Tests Using Ranks and Signs

photo

 Sercan K., Software Engineer at Foreks

 Wednesday, January 7, 2015

Hello there, I've got a system working with 20 mins data and I am trying to analyse my trading model's returns from last year' trades. Because it trades with 20 mins intervals I have made pretty much trade around 570 trades last year. So I list all them down, runned Auto.VR function and Wright's rank and sign tests from vrtest package of R. ( http://cran.r-project.org/web/packages/vrtest/vrtest.pdf ) Documentation isn't that clear on explaining what the outputs mean. Those are my outcomes of Automatic Variance Ratio Test and Wright's Rank and Sign test > Auto.VR(vec) $stat [1] -0.8288204 $sum [1] 0.908094 Wright(result,c(2,5,10)) R1 R2 R3 k=2 -1.757248 -1.726898 -0.94609983 k=5 -1.864323 -1.886648 -0.32387514 k=10 -1.949724 -2.032465 0.08406318 Do you have any idea about analysing these 'ranks and signs'


Print

3 comments on article "WRIGHT,J.H.,2000 Alternative Variance-Ratio Tests Using Ranks and Signs"

photo

 William K. Wong, Ph.D., Algorithmic Trader, HFT, FX, Global Futures

 Thursday, January 8, 2015



Just by looking at your results, I'm presuming R1 and R2 are your rank-based variance test statistics, and your R3 should probably be equivalent to S1 which is the sign-based variance test statistics. k is the number of lag here.



This technique is a null hypothesis to decide whether it succeed or fail to reject the Random Walk Hypothesis (RWH).



I'm presuming these numbers are your significance level. So, it looks like all of them failed to reject RWH. So, I'd probably say your weak form of Efficient Market Hypothesis is supported.



This paper would help you.


http://www.researchgate.net/profile/Jorge_BELAIRE-FRANCH/publication/228583649_Ranks_and_signs-based_multiple_variance_ratio_tests/links/00b4951920895f04f6000000


photo

 Sercan K., Software Engineer at Foreks

 Thursday, January 8, 2015



Mr William, thank you for your help and Can I ask you one more question; Is it statistically reliable to test my sequential returns with variance ratio test and sign test and would you suggest another approaches for me to test how random my trades are..


photo

 William K. Wong, Ph.D., Algorithmic Trader, HFT, FX, Global Futures

 Thursday, January 8, 2015



Well, it's kind of strange to determine the randomness of your trade. One would hope that one's trades are highly skewed. I guess it would depend on the purpose of what you are trying to achieve.

In terms of whether it's statistically reliable, it's fairly good, as least it overcomes some of the limitations from the conventional Lo/Mackinlay method. Frankly, I have not done extensive study on it, you would have to experiment with it to find out whether it's suitable for your purpose.

Please login or register to post comments.

TRADING FUTURES AND OPTIONS INVOLVES SUBSTANTIAL RISK OF LOSS AND IS NOT SUITABLE FOR ALL INVESTORS
Terms Of UsePrivacy StatementCopyright 2018 Algorithmic Traders Association