Vanguard VIPER ETF launch

Financial Independence/Retire Early -- Learn How!
Post Reply
peteyperson
**** Heavy Hitter
Posts: 525
Joined: Tue Nov 26, 2002 6:46 am

Vanguard VIPER ETF launch

Post by peteyperson » Tue Feb 03, 2004 8:31 pm

Vanguard has now extended their ETFs to small cap with style segmentation and sector funds. They've also filed for three international ETFs which are expected to be launch in the Spring.

Interestingly, they are using the MSCI® US Small Cap Value Index instead of Russell or S&P for small cap. I wonder how that will play out and how the selection process compares to the Russell or S&P. They've also undercut the Barclays ETFs by offering the small cap value at 0.22%, undercutting by 0.03%. I wonder how the different benchmark will affect investor interest.

http://flagship2.vanguard.com/VGApp/hnw ... IPERByName

NB: Initial information came from Bylo's excellent site.

Petey

User avatar
aedelswil2429
* Rookie
Posts: 6
Joined: Tue Nov 26, 2002 3:42 am
Location: Switzerland
Contact:

Post by aedelswil2429 » Wed Feb 04, 2004 1:31 am

peteyperson wrote:...
NB: Initial information came from Bylo's excellent site.
...
Petey


And for those that haven't got the link to ByLo:

http://www.bylo.org/

TCAIO

Philipp

User avatar
ElSupremo
Admin Board Member
Posts: 343
Joined: Thu Nov 21, 2002 12:53 pm
Location: Cincinnati, Ohio

Post by ElSupremo » Wed Feb 04, 2004 7:22 am

Greetings Phillip :)
And for those that haven't got the link to ByLo:


And just as another reminder, Bylo's link as well as many other important links can be found when needed in our "Links We Like" section by clicking on the "Links" selection on the NFB home page main menu.

http://nofeeboards.com/files/links.html

:D
"The best things in life are FREE!"

www.nofeeboards.com

raddr
*** Veteran
Posts: 265
Joined: Mon Nov 25, 2002 3:25 am
Contact:

Post by raddr » Wed Feb 04, 2004 7:56 am

Interestingly, they are using the MSCI® US Small Cap Value Index instead of Russell or S&P for small cap. I wonder how that will play out and how the selection process compares to the Russell or S&P.



This table is hard to read but I did a Fama/French 3 factor analysis on the MSCI and S&P600 value indices (monthly data since 12/93) and it would appear that value loading (HML) is about the same but small cap loading (SMB) is slightly greater for the S&P index (0.642 vs. 0.413 - statistically significant). Interestingly, "market loading" is also a little greater for the S&P index though I doubt that it means anything. :?

Thus it would appear that you might capture the small cap effect with a bit more efficiency with the S&P600 index but I don't know if that will make much difference down the road. Returns were almost the same for the two indices over this time period (R-squared = 0.90).


Code: Select all

MSCI ScV                         
                         
Regression Statistics                         
Multiple R   0.939271901                     
R Square   0.882231703                     
Adjusted R Square   0.879185971                     
Standard Error   1.443985015                     
Observations   120                     
                         
ANOVA                         
   df   SS   MS   F   Significance F         
Regression   3   1811.914198   603.9713993   289.6616501   1.07125E-53         
Residual   116   241.8707561   2.085092725               
Total   119   2053.784954                   
                         
   Coefficients   Standard Error   t Stat   P-value   Lower 95%   Upper 95%   Lower 95.0%   Upper 95.0%
Intercept   -0.123102648   0.136368956   -0.902717534   0.368546521   -0.393198385   0.146993089   -0.393198385   0.146993089
Mkt-RF   0.980559664   0.034248328   28.63087688   2.0366E-54   0.912726576   1.048392752   0.912726576   1.048392752
SMB   0.413283522   0.036306224   11.38326924   1.29377E-20   0.341374515   0.48519253   0.341374515   0.48519253
HML   0.80421496   0.04649734   17.29593462   6.62564E-34   0.712121174   0.896308745   0.712121174   0.896308745
                         
S&P600v                         
                         
Regression Statistics                         
Multiple R   0.943096601                     
R Square   0.889431199                     
Adjusted R Square   0.886571661                     
Standard Error   1.706792539                     
Observations   120                     
                         
ANOVA                         
   df   SS   MS   F   Significance F         
Regression   3   2718.311472   906.1038239   311.0401778   2.77118E-55         
Residual   116   337.9243296   2.913140773               
Total   119   3056.235801                   
                         
   Coefficients   Standard Error   t Stat   P-value   Lower 95%   Upper 95%   Lower 95.0%   Upper 95.0%
Intercept   -0.240148835   0.161188319   -1.489865001   0.138973751   -0.559402417   0.079104747   -0.559402417   0.079104747
Mkt-RF   1.121170519   0.040481577   27.69582111   5.88809E-53   1.040991703   1.201349335   1.040991703   1.201349335
SMB   0.6416535   0.042914013   14.95207396   7.9931E-29   0.556656941   0.726650059   0.556656941   0.726650059
HML   0.793522552   0.054959929   14.43820195   1.13215E-27   0.684667554   0.90237755   0.684667554   0.90237755

peteyperson
**** Heavy Hitter
Posts: 525
Joined: Tue Nov 26, 2002 6:46 am

Post by peteyperson » Wed Feb 04, 2004 4:26 pm

Thanks Raddr.

I think I would also be concerned whether the index can be timed on the inclusions and exclusions a la Russell.

Do you happen to know?

Petey

raddr wrote:
Interestingly, they are using the MSCI® US Small Cap Value Index instead of Russell or S&P for small cap. I wonder how that will play out and how the selection process compares to the Russell or S&P.



This table is hard to read but I did a Fama/French 3 factor analysis on the MSCI and S&P600 value indices (monthly data since 12/93) and it would appear that value loading (HML) is about the same but small cap loading (SMB) is slightly greater for the S&P index (0.642 vs. 0.413 - statistically significant). Interestingly, "market loading" is also a little greater for the S&P index though I doubt that it means anything. :?

Thus it would appear that you might capture the small cap effect with a bit more efficiency with the S&P600 index but I don't know if that will make much difference down the road. Returns were almost the same for the two indices over this time period (R-squared = 0.90).


Code: Select all

MSCI ScV                         
                         
Regression Statistics                         
Multiple R   0.939271901                     
R Square   0.882231703                     
Adjusted R Square   0.879185971                     
Standard Error   1.443985015                     
Observations   120                     
                         
ANOVA                         
   df   SS   MS   F   Significance F         
Regression   3   1811.914198   603.9713993   289.6616501   1.07125E-53         
Residual   116   241.8707561   2.085092725               
Total   119   2053.784954                   
                         
   Coefficients   Standard Error   t Stat   P-value   Lower 95%   Upper 95%   Lower 95.0%   Upper 95.0%
Intercept   -0.123102648   0.136368956   -0.902717534   0.368546521   -0.393198385   0.146993089   -0.393198385   0.146993089
Mkt-RF   0.980559664   0.034248328   28.63087688   2.0366E-54   0.912726576   1.048392752   0.912726576   1.048392752
SMB   0.413283522   0.036306224   11.38326924   1.29377E-20   0.341374515   0.48519253   0.341374515   0.48519253
HML   0.80421496   0.04649734   17.29593462   6.62564E-34   0.712121174   0.896308745   0.712121174   0.896308745
                         
S&P600v                         
                         
Regression Statistics                         
Multiple R   0.943096601                     
R Square   0.889431199                     
Adjusted R Square   0.886571661                     
Standard Error   1.706792539                     
Observations   120                     
                         
ANOVA                         
   df   SS   MS   F   Significance F         
Regression   3   2718.311472   906.1038239   311.0401778   2.77118E-55         
Residual   116   337.9243296   2.913140773               
Total   119   3056.235801                   
                         
   Coefficients   Standard Error   t Stat   P-value   Lower 95%   Upper 95%   Lower 95.0%   Upper 95.0%
Intercept   -0.240148835   0.161188319   -1.489865001   0.138973751   -0.559402417   0.079104747   -0.559402417   0.079104747
Mkt-RF   1.121170519   0.040481577   27.69582111   5.88809E-53   1.040991703   1.201349335   1.040991703   1.201349335
SMB   0.6416535   0.042914013   14.95207396   7.9931E-29   0.556656941   0.726650059   0.556656941   0.726650059
HML   0.793522552   0.054959929   14.43820195   1.13215E-27   0.684667554   0.90237755   0.684667554   0.90237755

Kramer
* Rookie
Posts: 18
Joined: Mon Aug 25, 2003 3:53 pm

Post by Kramer » Wed Feb 04, 2004 10:49 pm

I am pretty sure that the MSCI value small cap index includes REITs, the S&P 600 value does not. This accounts for the huge dividend yield difference.

As far as methodology, I believe that the MSCI is superior. Instead of having hard boundaries like S&P 600 value has on value/growth division within the index which can cause high turnover, MSCI has a range around the value/growth boundary so that stocks will not keep changing from growth to value and vice versa. This will be the way that most indexes are constructed in the future, IMO.

I personally hold lots of IJS, ishares S&P 600 small cap value, because my holdings are in a taxable account. Because of the REITs, the MSCI is not really appropriate for a taxable account unless you are in the 15% bracket (both because of the high dividends, and because REIT dividends are not favored under the new tax law). If I were to hold small cap value in a tax deferred account, I would choose the MSCI option.

Kramer

raddr
*** Veteran
Posts: 265
Joined: Mon Nov 25, 2002 3:25 am
Contact:

Post by raddr » Thu Feb 05, 2004 3:41 am

peteyperson wrote:]I think I would also be concerned whether the index can be timed on the inclusions and exclusions a la Russell.

Petey



Supposedly the MSCI and S&P indexes are harder to time than the Russell but I haven't really looked into it.

raddr
*** Veteran
Posts: 265
Joined: Mon Nov 25, 2002 3:25 am
Contact:

Post by raddr » Thu Feb 05, 2004 3:46 am

Kramer wrote:I am pretty sure that the MSCI value small cap index includes REITs, the S&P 600 value does not. This accounts for the huge dividend yield difference.

Because of the REITs, the MSCI is not really appropriate for a taxable account unless you are in the 15% bracket


Hi Kramer,

What is the yield difference between IJS and VBR? I can't imagine that it would be so big as to seriously undermine the tax-adjusted performance of the MSCI product.

Kramer
* Rookie
Posts: 18
Joined: Mon Aug 25, 2003 3:53 pm

Post by Kramer » Thu Feb 05, 2004 11:36 pm

Howdy raddr,

The yield difference is pretty high:

VAnguard MSCI small cap value (visvx): 1.90%
S&P 600 Value (ijs): 0.76%

Next year the yield difference should be larger, since VISVX was S&P600 value until the spring, so the 1.90% is weighted down by lower dividends earlier in the year. On the other hand, a small bit of the VISVX dividend is probably untaxed return of capital from the REITs (depreciation), so these two probably balance out.

Also, the REIT component of the VISVX dividend, which provides a large portion of the dividends, is non-qualified, so those part of the dividends are taxed at your marginal rate, not the new lower qualified dividend tax rate.

Kramer

User avatar
ataloss
**** Heavy Hitter
Posts: 559
Joined: Mon Nov 25, 2002 3:00 am

Post by ataloss » Fri Feb 06, 2004 4:59 am

Also, the REIT component of the VISVX dividend, which provides a large portion of the dividends, is non-qualified, so those part of the dividends are taxed at your marginal rate, not the new lower qualified dividend tax rate.


except for the part characterized as return of capital, right (I am not sure if that is much)
Have fun.

Ataloss

peteyperson
**** Heavy Hitter
Posts: 525
Joined: Tue Nov 26, 2002 6:46 am

Post by peteyperson » Fri Feb 06, 2004 7:41 pm

True. But then you to adjust accordingly for that which takes a bit more work.

Petey

ataloss wrote:
Also, the REIT component of the VISVX dividend, which provides a large portion of the dividends, is non-qualified, so those part of the dividends are taxed at your marginal rate, not the new lower qualified dividend tax rate.


except for the part characterized as return of capital, right (I am not sure if that is much)

Post Reply