Rolling Beta

[ehcpieterse]

Hi,

I have had a *very* rough attempt at estimating the beta between indices. I am convinced that someone on this list has had experience with this before, but my search yielded little information. I have daily log returns from 01/01/2000 on 13 different equity indices. I am just looking for an efficient way to produce a rolling beta for all index pairs.

I would appreciate any advice.

Thanks,
Eduard

CODE:

IndexLocal <- read.csv("IndexLocal.csv", header=TRUE, sep = ",")
MyLag <- 30
IndexDim <- dim(IndexLocal)
Results <- NULL
reg <- NULL
k <- 0

for (i in 2:14)
{
        for (j in 2:14)
        {
                        k <- k + 1
                        for (r in MyLag :IndexDim[1])
                        {
                                x. <- IndexLocal[(r-MyLag+1):r,i]
                                y. <- IndexLocal[(r-MyLag+1):r,j]
                                glm.linear<-coef(glm(y. ~ x.))
                                reg[r] <- glm.linear[2]
                        }
                                if (j >= i)
                                {
                                        Results <- cbind(Results,reg)
                                }
        }
}

[Brian G. Peterson]

see PerformanceAnalytics functions

CAPM.beta
and
chart.RollingRegression
and
charts.RollingRegression

Cheers,

  - Brian

ehxpieterse wrote:

> Hi,
>
> I have had a *very* rough attempt at estimating the beta between indices. I
> am convinced that someone on this list has had experience with this before,
> but my search yielded little information. I have daily log returns from
> 01/01/2000 on 13 different equity indices. I am just looking for an
> efficient way to produce a rolling beta for all index pairs.
>
> I would appreciate any advice.
>
> Thanks,
> Eduard
>
> CODE:
>
> IndexLocal <- read.csv("IndexLocal.csv", header=TRUE, sep = ",")
> MyLag <- 30
> IndexDim <- dim(IndexLocal)
> Results <- NULL
> reg <- NULL
> k <- 0
>
> for (i in 2:14)
> {
> for (j in 2:14)
> {
> k <- k + 1
> for (r in MyLag :IndexDim[1])
> {
> x. <- IndexLocal[(r-MyLag+1):r,i]
> y. <- IndexLocal[(r-MyLag+1):r,j]
> glm.linear<-coef(glm(y. ~ x.))
> reg[r] <- glm.linear[2]
> }
> if (j >= i)
> {
> Results <- cbind(Results,reg)
> }
> }
> }
>
>
>  

--
Brian G. Peterson
http://braverock.com/brian/
Ph: 773-459-4973
IM: bgpbraverock

_______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-sig-finance
-- Subscriber-posting only.
-- If you want to post, subscribe first.

[Brian G. Peterson]

As with most things in R, you can see the code by typing the function
name without the parentheses.

In the case of rolling beta, the important line is this:

rollapply(na.omit(merged.assets[,, drop = FALSE]),
                  width = width, FUN = function(x) lm(x[,1, drop = FALSE]
                  ~ x[, 2, drop = FALSE])$coefficients[2],
                  by = 1, by.column = FALSE, na.pad = na.pad,
                  align = "right")

Regards,

   - Brian


Eduard Pieterse (Macquarie Securities) wrote:

> Hi Brian,
>
> Chart.rollingRegression seems to be exactly what I'm after. How would I
> "disconnect" the chart output and be able to save/use the data behind
> it?
>
> I am still finding my feet with R and know what I need to do, but my R
> language knowledge is lacking somewhat.
>
> Thanks,
> Eduard
>
> -----Original Message-----
> From: Brian G. Peterson [mailto:[hidden email]]
> Sent: 19 October 2009 17:28
> To: Eduard Pieterse (Macquarie Securities)
> Cc: [hidden email]
> Subject: Re: [R-SIG-Finance] [R-sig-finance] Rolling Beta
>
> see PerformanceAnalytics functions
>
> CAPM.beta
> and
> chart.RollingRegression
> and
> charts.RollingRegression
>
> Cheers,
>
>   - Brian
>
> ehxpieterse wrote:
>  
>> Hi,
>>
>> I have had a *very* rough attempt at estimating the beta between
>> indices. I am convinced that someone on this list has had experience
>> with this before, but my search yielded little information. I have
>> daily log returns from 01/01/2000 on 13 different equity indices. I am
>>    
>
>  
>> just looking for an efficient way to produce a rolling beta for all
>>    
> index pairs.
>  
>> I would appreciate any advice.
>>
>> Thanks,
>> Eduard
>>
>> CODE:
>>
>> IndexLocal <- read.csv("IndexLocal.csv", header=TRUE, sep = ",") MyLag
>>    
>
>  
>> <- 30 IndexDim <- dim(IndexLocal) Results <- NULL reg <- NULL k <- 0
>>
>> for (i in 2:14)
>> {
>> for (j in 2:14)
>> {
>> k <- k + 1
>> for (r in MyLag :IndexDim[1])
>> {
>> x. <- IndexLocal[(r-MyLag+1):r,i]
>> y. <- IndexLocal[(r-MyLag+1):r,j]
>>    
>
>  
>> glm.linear<-coef(glm(y. ~ x.))
>> reg[r] <- glm.linear[2]
>> }
>> if (j >= i)
>> {
>> Results <- cbind(Results,reg)
>> }
>> }
>> }
>>
>>    

--
Brian G. Peterson
http://braverock.com/brian/
Ph: 773-459-4973
IM: bgpbraverock

_______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-sig-finance
-- Subscriber-posting only.
-- If you want to post, subscribe first.