Page 1 of 1
Vanguard VIPER ETF launch
Posted: Tue Feb 03, 2004 8:31 pm
by peteyperson
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
Posted: Wed Feb 04, 2004 1:31 am
by aedelswil2429
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
Posted: Wed Feb 04, 2004 7:22 am
by ElSupremo
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
Posted: Wed Feb 04, 2004 7:56 am
by raddr
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
Posted: Wed Feb 04, 2004 4:26 pm
by peteyperson
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
Posted: Wed Feb 04, 2004 10:49 pm
by Kramer
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
Posted: Thu Feb 05, 2004 3:41 am
by raddr
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.
Posted: Thu Feb 05, 2004 3:46 am
by raddr
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.
Posted: Thu Feb 05, 2004 11:36 pm
by Kramer
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
Posted: Fri Feb 06, 2004 4:59 am
by ataloss
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)
Posted: Fri Feb 06, 2004 7:41 pm
by peteyperson
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)