PXSV has changed indexes

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Topic Author
RJM
Posts: 9
Joined: Wed Aug 01, 2007 5:36 pm

PXSV has changed indexes

Post by RJM »

I just read that PXSV, a favorite small cap value fund of many on the forum, has changed from the RAFI small value index to Russell 2000 pure small value index. Anybody have any backtests and factor loads on this new index? Robert T.? Switch is effective May 22,2015.
User avatar
backpacker
Posts: 1620
Joined: Mon Sep 22, 2014 2:17 pm

Re: PXSV has changed indexes

Post by backpacker »

Ugh. I was afraid of this sort of thing from PowerShares. More here.
exeunt
Posts: 907
Joined: Tue Jan 08, 2008 5:54 pm

Re: PXSV has changed indexes

Post by exeunt »

Gross. The Russell 2000 is a terrible, flawed index. It loses 1% or more a year to arbitrageurs front-running the trades of funds tracking it.
User avatar
backpacker
Posts: 1620
Joined: Mon Sep 22, 2014 2:17 pm

Re: PXSV has changed indexes

Post by backpacker »

An article on the front-running woes of the Russell 2000.
RNJ
Posts: 863
Joined: Mon Apr 08, 2013 9:06 am

Re: PXSV has changed indexes

Post by RNJ »

Lest we forget, companies such as Invesco exist to make money from investors, not for them. As such, they are marketing companies masquerading as investment companies. Most of the funds in question trade by appointment, and all failed to attract enough assets to satisfy the good people at Invesco, Ltd. (I believe they all have <$100M AUM).

RAFI may be a wonderful indexing strategy with respect to obtaining factor loads. But Russell is a better, far more familiar brand for obtaining investors' money.



(FULL DISCLOSURE: I own PXSV).
User avatar
in_reality
Posts: 4529
Joined: Fri Jul 12, 2013 6:13 am

Re: PXSV has changed indexes

Post by in_reality »

RNJ wrote: RAFI may be a wonderful indexing strategy with respect to obtaining factor loads. But Russell is a better, far more familiar brand for obtaining investor's money.
RAFI and Russell are not mutually exclusive. Schwab's fundamental indexes track a Russell index that is created through the RAFI methodology.

RAFI actually has no "pure" style. Powershares states RAFI defines what is "pure value" for PXSV which sorta struck me as odd because the notion itself is not in the RAFI literature. RAFI rather trumps up the fact that their methodology is a value tilt that doesn't make sector bets because it selects even from sectors that are not very valuey. As such, it's always seemed a little inconsistent to me that these "pure" style funds existed.

Anyway, my Russell fundamental small cap fund has a healthy dose in mid-cap growth which is how RAFI seems to think it should be. So I wonder if RAFI itself didn't have something to say about the decision. From a factor analysis point of view of course, the "pure style" may be desirable, and I am not saying it isn't in any way.

I really appreciate RobertT's analysis on PXSV as it greatly influenced by current holdings in the Russell fundamental index [which is likely not small enough or valuey enough for those true PXSV fans]
Last edited by in_reality on Sat Mar 14, 2015 10:57 am, edited 1 time in total.
JohnnyFive
Posts: 181
Joined: Mon Sep 08, 2014 1:14 pm

Re: PXSV has changed indexes

Post by JohnnyFive »

This is a bummer, as it was the best option available to us as investors, in the domestic small cap value space, without having to pay advisors for access to DFA's options.

Would love Robert T to chime in here.
RNJ
Posts: 863
Joined: Mon Apr 08, 2013 9:06 am

Re: PXSV has changed indexes

Post by RNJ »

lazyday
Posts: 3849
Joined: Wed Mar 14, 2007 10:27 pm

Re: PXSV has changed indexes

Post by lazyday »

Just from a very quick search, there's probably more:

Dec 19 2012 http://www.invescopowershares.com/news/ ... Family.pdf
As part of its ongoing commitment to putting investors’ interests first, Invesco PowerShares announced that it plans to
close 13 ETFs. The affected funds represent less than 1% of Invesco PowerShares’ total assets. .... The final day of trading
on The NASDAQ Stock Market LLC (“NASDAQ”) and NYSE Arca, Inc. (“NYSE Arca”) will be Feb. 26, 2013.
December 18, 2013 http://www.invescopowershares.com/news/ ... E-2013.pdf
Reflective of its ongoing commitment to place investors’ interests first, Invesco PowerShares announced
that it plans to change the underlying indexes and names of ten ETFs. In addition, it also announced
plans to close four ETFs, which together represent less than 1% of Invesco PowerShares’ total assets. .... These changes are scheduled to take
effect on February 19, 2014
March 13, 2015 http://www.prnewswire.com/news-releases ... 50487.html
As part of the firm's commitment to aligning its fund offerings to meet an evolving investment landscape, Invesco PowerShares announced it will change the underlying indexes and index providers of ten ETFs. In addition, four ETFs will see a reduction in management fees* and four ETFs will receive new product tickers. .... These changes are anticipated to go into effect at the close of markets on May 22, 2015.
Looks like they give about 2 months notice. Easy for some of us to miss, or do they send a letter to all shareholders?

Anyone seen a comparison of SSGA vs iShares vs Powershares for % of funds (not % assets) closed, merged, or changed per year?
garlandwhizzer
Posts: 3565
Joined: Fri Aug 06, 2010 3:42 pm

Re: PXSV has changed indexes

Post by garlandwhizzer »

Question: Why is it that many of those who trumpet the merits of factor investing, of the small cap value premium in particular, seem to be constantly changing their fund and etf indexes designed to capture it? Isn't such change an admission that the old SCV index they used didn't achieve its desired result? It seems a bit like a never ending fox hunt where the fox always seems too quick and too agile for the dogs and hunters.

Garland Whizzer
User avatar
backpacker
Posts: 1620
Joined: Mon Sep 22, 2014 2:17 pm

Re: PXSV has changed indexes

Post by backpacker »

garlandwhizzer wrote:Question: Why is it that many of those who trumpet the merits of factor investing, of the small cap value premium in particular, seem to be constantly changing their fund and etf indexes designed to capture it? Isn't such change an admission that the old SCV index they used didn't achieve its desired result? It seems a bit like a never ending fox hunt where the fox always seems too quick and too agile for the dogs and hunters.
PowerShares doesn't care whether its funds are good for investing in small value stocks. They care whether their funds are good at making money. The switch is good for PowerShares, so why should they care if the result is a worse fund for investors?
randomguy
Posts: 11295
Joined: Wed Sep 17, 2014 9:00 am

Re: PXSV has changed indexes

Post by randomguy »

garlandwhizzer wrote:Question: Why is it that many of those who trumpet the merits of factor investing, of the small cap value premium in particular, seem to be constantly changing their fund and etf indexes designed to capture it? Isn't such change an admission that the old SCV index they used didn't achieve its desired result? It seems a bit like a never ending fox hunt where the fox always seems too quick and too agile for the dogs and hunters.

Garland Whizzer
Why do market cap weighted people keep switching their indexes around? TSM for example has gone from Wilshire 5k, MSCI, and then to CRSP. Was that a never ending fox hunt where the fox seems too agile?

That being said I think when you switch up indexes, you need to provide a reason why and some information about the new index. As far as I can tell the only mention of these indexes on the web are from the Invesco press release. Does someone want to call that phone number monday morning and let us know what they say?
User avatar
Ketawa
Posts: 2521
Joined: Mon Aug 22, 2011 1:11 am
Location: DC

Re: PXSV has changed indexes

Post by Ketawa »

All the more reason I'm happy to be switching to QSMLX.
JohnnyFive
Posts: 181
Joined: Mon Sep 08, 2014 1:14 pm

Re: PXSV has changed indexes

Post by JohnnyFive »

garlandwhizzer wrote:Question: Why is it that many of those who trumpet the merits of factor investing, of the small cap value premium in particular, seem to be constantly changing their fund and etf indexes designed to capture it? Isn't such change an admission that the old SCV index they used didn't achieve its desired result? It seems a bit like a never ending fox hunt where the fox always seems too quick and too agile for the dogs and hunters.

Garland Whizzer
I respect you greatly as someone who knows what he's talking about, and, I've learned ALOT from your posts in the past, but, this is, truly, one of the silliest things you have ever posted. Those who trumpet the merits of factor investing didn't change the index. The ETF provider did. There is a HUGE difference. Hence the reason that many of us who DO ACTUALLY trumpet the merits of factor investing are so upset...because this was, without question, the best available SCV fund available to retail investors.
boglety
Posts: 60
Joined: Sun Jun 15, 2008 1:44 am

Re: PXSV has changed indexes

Post by boglety »

Is the index now changed to similar what Ishares and Vanguard 2000 offers? Russell 2000 small value? It seems like a marketing gimmick to add that pure Russell 2000 small cap value if it's the same index as what Ishares and Vanguard offers. I for one will be looking to transition my future small cap value allocation elsewhere as Russel 2000 small value has significant reconstitution issues
pauliec84
Posts: 414
Joined: Wed Feb 09, 2011 10:29 am

Re: PXSV has changed indexes

Post by pauliec84 »

I did a look for data on the Russell 2000 Pure Value Index but cannot find any. If anyone finds the data I am happy to run the appropriate factor regressions. My guess is that is that it will be similar to the RZV pure factor loadings. Will be interesting to see if it hit with the negative alpha as the result of the Russell 2000 value.

Guess this settles my angst about the extra 0.30 bps of fees I was paying to invest in PXSV vs. VBR.
garlandwhizzer
Posts: 3565
Joined: Fri Aug 06, 2010 3:42 pm

Re: PXSV has changed indexes

Post by garlandwhizzer »

I think this is the second time that PXSV, a favorite SCV fund, has changed its index. Perhaps its performance has something to do with that. It has underperformed VB, Vanguard's SC blend fund, YTD, 1yr, 5 yr, 10 yr, and since PXSV's inception more than 10 yr ago. PXSV dismissed its early underperformance, saying they were tracking the wrong index and therefore they changed indexes years ago. Since that change despite a slight outperformance relative to VB at 3 yr, PXSV has once again fallen behind the plain old vanilla small cap blend fund. Now they plan to change the index approach again, more than a coincidence, I suspect.

Perhaps this is a case of the SCV premium disappearing for a time which historically it has done for periods of up to 15 yrs. Perhaps the market is more driven by professionals now, all of whom know about factors and arbitrage away much of the premiums. Perhaps it is just harder to exploit the SCV premium in real world funds with real expenses than it is to write academic papers based on cost-free indexes researched retrospectively over long periods of time in much of which many market participants were unaware of the existence of these premiums. Perhaps with the growing popularity of DFA, RAFI, fundamental indexing, etc., the available alpha is overgrazed.

I don't know. What I do know is that the performance of real SCV funds in recent years, with portfolios designed by acknowledged experts in factor investing, people who are very smart and very well versed in the literature, is not convincingly impressive relative to simple market cap weighting of comparable cap weighted portfolios. The proponents of strong factor tilts tend to quote statistics that look impressive if they are very long term (before factors were widely appreciated by the market) and they must include the LCG bubble burst of 2000 - 2003. That single event made SCV look compelling but as the years pass it looks less and less impressive in hindsight. After the unwinding of that massive bubble about 10 years ago and especially with the rising dominance of knowledgeable professional investors in the market and all the dollars chasing after a limited supply of factor alpha, the case for exploiting factors in real world funds has gotten much weaker in my opinion.

In the initial SCV literature all you had to do was to select low P/B and low market cap and the magic happened if you stayed patient. The fact that there is such a proliferation of factor strategies now says something. In medicine, something I know a bit about being a retired MD, whenever you find many treatments advocated for the same disease that usually means that none of them work reliably. I'm wondering if it's the same story in the factor zoo now.


Garland Whizzer
pauliec84
Posts: 414
Joined: Wed Feb 09, 2011 10:29 am

Re: PXSV has changed indexes

Post by pauliec84 »

I found some information on the underlying index from the Powershares SEC Filing about the new funds http://www.sec.gov/Archives/edgar/data/ ... oc888553_6.

It sounds like an Russell 2000 version of the S&P 600 Pure Value, although I am a bit confused by the Composite Value score giving weight to "growth characteristics."
Hopefully Russell publishes the historic index returns, but this looks like a stinker...

For those that held PXSV, what are your plans with the holding?
Principal Investment Strategies

The Fund generally will invest at least 90% of its total assets in the component securities that comprise the Underlying Index. The Underlying Index is composed of securities selected from the Russell 2000 Index, which is composed of the smallest 2,000 securities of the Russell 3000® Index, an index designed to measure the performance of the largest 3,000 U.S. companies.

Frank Russell Company (the “Index Provider”) first identifies securities in the Russell 2000 Index with lower price-to-book ratios and lower forecasting growth values. In order to further concentrate the Underlying Index, the constituents are further narrowed to those with the highest Composite Value Score (“CVS”), which is calculated based on three characteristics (book to price (a value characteristic), sales per share growth (a growth characteristic) and medium-term growth forecast (a growth characteristic)).

The constituent securities are then weighted in proportion to their CVS compared to the CVS of securities excluded from the Underlying Index due to being deemed value securities. The effect of this weighting methodology is to give greatest weight to the securities demonstrating the most growth characteristics, instead of using market capitalization to determine component weights.

Additionally, the Index Provider caps the sector weights within the Underlying Index at 10% above the respective sector weight in the Russell 2000 Value Index, which utilizes the same constituent selection criteria as the Underlying Index, but applies a traditional, market capitalization weighting.

The Underlying Index is completely reconstituted annually at the close of the last Friday in June.

Concentration Policy. The Fund will concentrate its investments (i.e., invest 25% or more of the value of its total assets) in securities of issuers in any one industry or sector only to the extent that the Underlying Index reflects a concentration in that industry or sector. The Fund will not otherwise concentrate its investments in securities of issuers in any one industry or sector.
User avatar
grap0013
Posts: 1892
Joined: Thu Mar 18, 2010 1:24 pm

Re: PXSV has changed indexes

Post by grap0013 »

pauliec84 wrote:I found some information on the underlying index from the Powershares SEC Filing about the new funds http://www.sec.gov/Archives/edgar/data/ ... oc888553_6.

It sounds like an Russell 2000 version of the S&P 600 Pure Value, although I am a bit confused by the Composite Value score giving weight to "growth characteristics."
Hopefully Russell publishes the historic index returns, but this looks like a stinker...

For those that held PXSV, what are your plans with the holding?
Principal Investment Strategies

The Fund generally will invest at least 90% of its total assets in the component securities that comprise the Underlying Index. The Underlying Index is composed of securities selected from the Russell 2000 Index, which is composed of the smallest 2,000 securities of the Russell 3000® Index, an index designed to measure the performance of the largest 3,000 U.S. companies.

Frank Russell Company (the “Index Provider”) first identifies securities in the Russell 2000 Index with lower price-to-book ratios and lower forecasting growth values. In order to further concentrate the Underlying Index, the constituents are further narrowed to those with the highest Composite Value Score (“CVS”), which is calculated based on three characteristics (book to price (a value characteristic), sales per share growth (a growth characteristic) and medium-term growth forecast (a growth characteristic)).

The constituent securities are then weighted in proportion to their CVS compared to the CVS of securities excluded from the Underlying Index due to being deemed value securities. The effect of this weighting methodology is to give greatest weight to the securities demonstrating the most growth characteristics, instead of using market capitalization to determine component weights.

Additionally, the Index Provider caps the sector weights within the Underlying Index at 10% above the respective sector weight in the Russell 2000 Value Index, which utilizes the same constituent selection criteria as the Underlying Index, but applies a traditional, market capitalization weighting.

The Underlying Index is completely reconstituted annually at the close of the last Friday in June.

Concentration Policy. The Fund will concentrate its investments (i.e., invest 25% or more of the value of its total assets) in securities of issuers in any one industry or sector only to the extent that the Underlying Index reflects a concentration in that industry or sector. The Fund will not otherwise concentrate its investments in securities of issuers in any one industry or sector.
This is mildly annoying. It is a first world problem after all. :) I'm becoming more acutely aware of the "intangible risks" of SCV. Both garlandwhizzer and nisiprius have both made some pretty good points about this in this thread and other tilting threads. Vanguard SCV changing indexes multiple times. My 403B plan eliminates my DFA value fund and replaces it with a core fund after 5 years of value underperformance, and now my absolute favorite domestic SCV fund will not stay the course either. Make no mistake, I will always SCV tilt, but it is frustrating and likely hurting premiums. At least if you sell PXSV one is selling high as it has been about the best SCV fund since its last index change on 6/11/11.

I think this makes a compelling case to mildly diversify amongst SCV funds. You wouldn't want a SCV fund to tilt to value a certain way only to watch it underperform for a decade and then change strategies. P/B, P/S, P/CF, and P/D can have quite variable returns. Best to include some value funds with different value metrics IMO. For the record I own VSIAX, QSMLX, and PXSV.

PXSV is about 6% of my portfolio. I'm not going to make any hasty decisions as I do not think that is the right approach. I want to know more about the new index first. P/B has the lowest turnover of the value sorts, P/S gives a nice profitability load, but what is a medium term growth forecast sort? Momentum? Sounds like net effect is value + profitability + momentum which is exactly what AQR does. If not, I may go back to PRFZ/SFSNX possibly: https://www.portfoliovisualizer.com/fac ... F02%2F2006

PRFZ avoids negative alpha through sales/profitability tilts. I think owners of PXSV should employ a "wait and see" approach.
There are no guarantees, only probabilities.
User avatar
grap0013
Posts: 1892
Joined: Thu Mar 18, 2010 1:24 pm

Re: PXSV has changed indexes

Post by grap0013 »

One other note: I want to know if the new index is fundamental or not. The backtests between RAFI fundamental indexes and Russell fundamental indexes only differ in about 10 basis points annualized over the length of the series (from 1976?) per Arnott. I think historically RAFI has had a slightly larger value tilt while Russell has had a slightly higher profitability tilt. 10 basis points is noise and profitability is more like small cap blend or "grow" which also reduces tracking error.

Wait and see folks.
There are no guarantees, only probabilities.
Fryxell
Posts: 183
Joined: Sat Mar 02, 2013 9:08 pm
Location: Dallas

Re: PXSV has changed indexes

Post by Fryxell »

pauliec84 wrote:
Principal Investment Strategies

The Fund generally will invest at least 90% of its total assets in the component securities that comprise the Underlying Index. The Underlying Index is composed of securities selected from the Russell 2000 Index, which is composed of the smallest 2,000 securities of the Russell 3000® Index, an index designed to measure the performance of the largest 3,000 U.S. companies.

Frank Russell Company (the “Index Provider”) first identifies securities in the Russell 2000 Index with lower price-to-book ratios and lower forecasting growth values. In order to further concentrate the Underlying Index, the constituents are further narrowed to those with the highest Composite Value Score (“CVS”), which is calculated based on three characteristics (book to price (a value characteristic), sales per share growth (a growth characteristic) and medium-term growth forecast (a growth characteristic)).

The constituent securities are then weighted in proportion to their CVS compared to the CVS of securities excluded from the Underlying Index due to being deemed value securities. The effect of this weighting methodology is to give greatest weight to the securities demonstrating the most growth characteristics, instead of using market capitalization to determine component weights.
That's odd. It looks like the description of the strategies for the small value and small growth indexes are identical. Perhaps it was an oversight?
User avatar
in_reality
Posts: 4529
Joined: Fri Jul 12, 2013 6:13 am

Re: PXSV has changed indexes

Post by in_reality »

Fryxell wrote:
pauliec84 wrote:
Principal Investment Strategies

The Fund generally will invest at least 90% of its total assets in the component securities that comprise the Underlying Index. The Underlying Index is composed of securities selected from the Russell 2000 Index, which is composed of the smallest 2,000 securities of the Russell 3000® Index, an index designed to measure the performance of the largest 3,000 U.S. companies.

Frank Russell Company (the “Index Provider”) first identifies securities in the Russell 2000 Index with lower price-to-book ratios and lower forecasting growth values. In order to further concentrate the Underlying Index, the constituents are further narrowed to those with the highest Composite Value Score (“CVS”), which is calculated based on three characteristics (book to price (a value characteristic), sales per share growth (a growth characteristic) and medium-term growth forecast (a growth characteristic)).

The constituent securities are then weighted in proportion to their CVS compared to the CVS of securities excluded from the Underlying Index due to being deemed value securities. The effect of this weighting methodology is to give greatest weight to the securities demonstrating the most growth characteristics, instead of using market capitalization to determine component weights.
That's odd. It looks like the description of the strategies for the small value and small growth indexes are identical. Perhaps it was an oversight?
It is almost certainly a mistake. Compare it to the Russell Top 200 Pure Value Portfolio which states:
The Underlying Index’s constituent securities are then weighted in proportion to their Composite Value Score (“CVS”), which is calculated based on three characteristics (book to price (a value characteristic), sales per share growth (a growth characteristic) and medium-term growth forecast (a growth characteristic)), compared to the CVS of securities excluded from the Underlying Index due to being deemed growth securities. The effect of this weighting methodology is to give greatest weight to the securities demonstrating the most value characteristics, instead of using market capitalization to determine component weights.
pauliec84
Posts: 414
Joined: Wed Feb 09, 2011 10:29 am

Re: PXSV has changed indexes

Post by pauliec84 »

I think this makes a compelling case to mildly diversify amongst SCV funds. You wouldn't want a SCV fund to tilt to value a certain way only to watch it underperform for a decade and then change strategies. P/B, P/S, P/CF, and P/D can have quite variable returns. Best to include some value funds with different value metrics IMO. For the record I own VSIAX, QSMLX, and PXSV.
I agree with this, both as a hedge against one strategy being superior to the other, and as a hedge against strategy change as in PXSV. I was lucky to have my PXSV holding in tax protected space due to powershares distrust. So will not be too costly to change (although trading big sums for powershare etfs is no fun...).
PXSV is about 6% of my portfolio. I'm not going to make any hasty decisions as I do not think that is the right approach. I want to know more about the new index first. P/B has the lowest turnover of the value sorts, P/S gives a nice profitability load, but what is a medium term growth forecast sort? Momentum? Sounds like net effect is value + profitability + momentum which is exactly what AQR does. If not, I may go back to PRFZ/SFSNX possibly: https://www.portfoliovisualizer.com/fac ... F02%2F2006
It is almost certainly a mistake. Compare it to the Russell Top 200 Pure Value Portfolio which states:
One other note: I want to know if the new index is fundamental or not. The backtests between RAFI fundamental indexes and Russell fundamental indexes only differ in about 10 basis points annualized over the length of the series (from 1976?) per Arnott. I think historically RAFI has had a slightly larger value tilt while Russell has had a slightly higher profitability tilt. 10 basis points is noise and profitability is more like small cap blend or "grow" which also reduces tracking error.

Wait and see folks.
I hope that is equivalent to a pure value approach to the Russell fundamental index, however the CVS scores in the press release are consistent with the wording in the Russell 2000 value index, which I interpret as being more equivalent to relationship between S&P 600 pure value and S&P 600. With that said my biggest issue with the S&P 600 pure value is the ~ -0.50 negative momentum load. If Russell 2000 pure value is able to avoid this, and also avoid the negative alpha associated Russell 2000 then sure it is worth considering keeping the holding in PXSV. These are two big ifs though. I do agree with your wait and see approach, but if I am going to trade out of it, would like to do so before it switches indexes so as not to incure the implicit cost of all of the turnover from the index change.
User avatar
in_reality
Posts: 4529
Joined: Fri Jul 12, 2013 6:13 am

Re: PXSV has changed indexes

Post by in_reality »

pauliec84 wrote:
I hope that is equivalent to a pure value approach to the Russell fundamental index, however the CVS scores in the press release are consistent with the wording in the Russell 2000 value index
It's almost certainly not a pure value version of the Russell Fundamental:

1) it rebalances yearly...the RF does it in quarterly tranches
2) the sectors weightings conflict .. It says no more than 10% away from the Russell 2000 value which the RF series couldn't fit into without trimming.
3) the ER reductions given for some of the other funds seem too high. RAFI isn't that cheap.
User avatar
nedsaid
Posts: 19275
Joined: Fri Nov 23, 2012 11:33 am

Re: PXSV has changed indexes

Post by nedsaid »

This is what is frustrating to an investor. If I remember right, Vanguard changed its small cap value index too. So much for staying the course.
A fool and his money are good for business.
acompton5
Posts: 16
Joined: Fri Jun 07, 2013 4:23 pm

Re: PXSV has changed indexes

Post by acompton5 »

I voted with my feet and sold. There was a reasonable bid/ask spread, so I took it. PXSV looks like it is going to be the same as IWN Russell 2000 Value with a higher ER and lower volume. I replaced it with 80% SLYV (SPDR S&P 600 Small Value) and 20% RZV (Guggenheim S&P Small Cap 600 Pure Value). My ER is lower, P/B is lower, and market cap is lower. I've added a bit to the volatility though. All three were commission free at at Schwab, but the bid/ask spread got me for 0.1% on the buy and the sell.
alex_686
Posts: 13320
Joined: Mon Feb 09, 2015 1:39 pm

Re: PXSV has changed indexes

Post by alex_686 »

backpacker wrote:An article on the front-running woes of the Russell 2000.
Do we know if this still holds true? The article you cite is from 2009 and Russell said they were going to make changes to their methodology to prevent this.
Former brokerage operations & mutual fund accountant. I hate risk, which is why I study and embrace it.
garlandwhizzer
Posts: 3565
Joined: Fri Aug 06, 2010 3:42 pm

Re: PXSV has changed indexes

Post by garlandwhizzer »

Confession: having argued on this forum that I am skeptical of real fund's ability to capture the SCV and other factor premiums going forward, I actually continue hold some SCV myself, changing from one SCV fund to another from time to time as frustration mounts. I do this in spite of the fact that I really have no certainty whatsoever that any of them will outperform Vanguard's Small Cap Index Fund or Extended Market Index Fund. I do like having more small cap and mid cap exposure than TSM alone allows for so I need some increased SC and MC exposure. I think the valuations of SCG at present are overstretched. Such is my current rationalization for holding VBR, Vanguard's SCV fund which during its lifetime (17 years) has slightly underperformed Vanguard's SCG option. Waiting for outperformance seems more and more like Waiting for Godot as time passes.

Garland Whizzer
pauliec84
Posts: 414
Joined: Wed Feb 09, 2011 10:29 am

Re: PXSV has changed indexes

Post by pauliec84 »

I'd say it is still significant. Since 2009 negative alphas on the value and regular index are negative at a statistic and economically significant level. https://www.portfoliovisualizer.com/fac ... F01%2F2009

Edit: Link Fixed
Last edited by pauliec84 on Mon Mar 16, 2015 12:38 pm, edited 2 times in total.
User avatar
backpacker
Posts: 1620
Joined: Mon Sep 22, 2014 2:17 pm

Re: PXSV has changed indexes

Post by backpacker »

pauliec84 wrote:I'd say it is still significant. Since 2009 negative alphas on the value and regular index are negative at a statistic and economically significant level. https://www.portfoliovisualizer.com/fac ... sisResults
I think you need to repost the link. This one just goes to an empty PV page.
Last edited by backpacker on Mon Mar 16, 2015 12:50 pm, edited 1 time in total.
pauliec84
Posts: 414
Joined: Wed Feb 09, 2011 10:29 am

Re: PXSV has changed indexes

Post by pauliec84 »

Sorry bout that. Link should be fixed here (and in original post):

https://www.portfoliovisualizer.com/fac ... F01%2F2009
Last edited by pauliec84 on Mon Mar 16, 2015 12:38 pm, edited 1 time in total.
User avatar
Robert T
Posts: 2803
Joined: Tue Feb 27, 2007 8:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: PXSV has changed indexes

Post by Robert T »

.
It seems the decision on the RAFI fundamental style indexes was RAFIs, rather than Powershares. I wrote to both:

The response from Powershares .... "Unfortunately the decision to discontinue the use of the RAFI Fundamental Small Value Index underlying PXSV was not ours. Research Affiliates who is the index provider had decided to terminate the calculation of the 9 RAFI Fundamental style indexes (including Fundamental Small value Index) as of 5/31/2015."

The response from RAFI ..."Unfortunately, as a firm we decided to discontinue the calculation of the RAFI Fundamental US Style Index Series due to lack of asset growth and cost of maintaining the index series."

Seems a bit shortsighted from RAFI's perspective IMO. Nevertheless, I am sure they have their reasons. And it was not the assets under any of the individual 9 style indexes (e.g. SV) that mattered but the aggregate across them all.
.
User avatar
Doc
Posts: 10607
Joined: Sat Feb 24, 2007 12:10 pm
Location: Two left turns from Larry

Re: PXSV has changed indexes

Post by Doc »

nedsaid wrote:This is what is frustrating to an investor. If I remember right, Vanguard changed its small cap value index too. So much for staying the course.
:D
Vanguard' wrote:S&P SmallCap 600 Value Index (formerly known as the S&P SmallCap 600/Barra Value Index) through May 16, 2003; MSCI US Small Cap Value Index through April 16, 2013; CRSP US Small Cap Value Index thereafter.
And before the S&P SmallCap 600 I think it was the Russell 2000 value.

It may have been more but I don't want to take my shoes off to count.
A scientist looks for THE answer to a problem, an engineer looks for AN answer and lawyers ONLY have opinions. Investing is not a science.
User avatar
backpacker
Posts: 1620
Joined: Mon Sep 22, 2014 2:17 pm

Re: PXSV has changed indexes

Post by backpacker »

Robert T wrote:. The response from RAFI ..."Unfortunately, as a firm we decided to discontinue the calculation of the RAFI Fundamental US Style Index Series due to lack of asset growth and cost of maintaining the index series."
What "cost" could they possibly be referring to? They're already generating a fundamental weighting for every stock. Once you have those weightings, chopping up the universe into smaller "pure" indexes is trivial. It would take a competent finance grad student an hour. Tops. And that's including a 45 minute beer break. :beer
e5116
Posts: 867
Joined: Mon Oct 05, 2009 11:22 am

Re: PXSV has changed indexes

Post by e5116 »

So, what are you guys going to? I have held PXSV at Schwab in a Roth once it went onto the no transaction fee ETF list. Thinking of going to FNDA or SFILX instead. I'd rather have a MF, but the extra 0.16% ER doesn't seem worth it. Hold most of my SCV at Vanguard, though, so my Schwab holding isn't that important...

Sounds like SLYV, RZV, or FNDA are the way to go...Opinions between these?
Last edited by e5116 on Mon Mar 16, 2015 1:49 pm, edited 1 time in total.
User avatar
Taylor Larimore
Posts: 32842
Joined: Tue Feb 27, 2007 7:09 pm
Location: Miami FL

Re: PXSV vs Total Stock Market

Post by Taylor Larimore »

Bogleheads:

These are returns for PXSV (PowerShares Fundamental Pure Small Value ETF ) and VTSMX (Total Stock Market Index Fund):

FUND----1 Year----3 Years----5 Years----10 Years
PXSV------5.49%----17.41%----14.85%-----7.20%
VTSMX---12.41%----16.15%----14.74%-----8.17%

It is difficult to beat the total stock market.

Best wishes.
Taylor
"Simplicity is the master key to financial success." -- Jack Bogle
JohnnyFive
Posts: 181
Joined: Mon Sep 08, 2014 1:14 pm

Re: PXSV has changed indexes

Post by JohnnyFive »

Robert, you beat me to the punch, as I got the EXACT same boilerplate response from RAFI (I emailed them directly but did not email Powershares).

Anyway, I would be curious as to what you, Robert, will be doing for your domestic SCV exposure now that PXSV is off the table.
randomguy
Posts: 11295
Joined: Wed Sep 17, 2014 9:00 am

Re: PXSV vs Total Stock Market

Post by randomguy »

Taylor Larimore wrote:Bogleheads:

These are returns for PXSV (PowerShares Fundamental Pure Small Value ETF ) and VTSMX (Total Stock Market Index Fund):

FUND----1 Year----3 Years----5 Years----10 Years
PXSV------5.49%----17.41%----14.85%-----7.20%
VTSMX---12.41%----16.15%----14.74%-----8.17%

It is difficult to beat the total stock market.

Best wishes.
Taylor

Seems pretty easy to me: 10 year (10k at start versus end)
Vanguard total stock market: 8.10 (21,790)
Vanguard. mid cap: 9.47 (24,714)
Vanguard Small cap: 9.13 (23,958 )

What about if we go back to the max of VIMAX (mid cap): You would have 52.4k versus 27.5k. Small cap is 34.6k versus 20.6k. Those are pretty big differences in my book. Granted if those funds had another 2 or 3 years of data, total market would look better.

Now you can ask why buy PXSV versus those funds. Well none of the funds listed are on the same index (PXSV switched to RAFI in 2011, vanguard to CRSP ) so you would have to travel back in time and figure out which one did what you wanted back then. I picked the vanguard midcap because that was what my 401(k) offered.
caklim00
Posts: 2420
Joined: Mon May 26, 2008 10:09 am

Re: PXSV has changed indexes

Post by caklim00 »

Anyone have back tested results for Russell Fundamental SCV similar to the analysis that was done for RAFI SCV? I'm upset about this move, but am not going to act hastily. And, I have about 4 months before I can move my money back to my free trade brokerage (took part of a transfer bonus), so would have to incur trade fees in addition to bid/ask spreads if I wanted to do anything right now.
Hank Moody
Posts: 279
Joined: Fri Oct 17, 2014 9:41 pm

Re: PXSV has changed indexes

Post by Hank Moody »

True. Annualized return for Russell 2000 is actually LESS than Russell 1000 from 12/1979 to 1/2015. Front running the index couldn't explain that much of a lousy outcome. I wonder if there isn't more to it.
exeunt wrote:Gross. The Russell 2000 is a terrible, flawed index. It loses 1% or more a year to arbitrageurs front-running the trades of funds tracking it.
-HM
User avatar
Robert T
Posts: 2803
Joined: Tue Feb 27, 2007 8:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: PXSV has changed indexes

Post by Robert T »

.
For those that don't have a 100% small value tilt, there are many other ways to achieve a specific factor load target. Not sure about the Russell 2000 Pure Value, but don't have high expectations given performance of Russell 2000 Value.

Here is just one example:

30% MTUM
30% VOE
40% PXSV

Simulated factor loads: 8/1996 - 12/2014
Alpha = +0.07
Mkt = +1.05
Size = +0.38
Value = +0.49
Momentum = -0.05
Quality =+0.20
R^2 = 0.96

Using MSCI US Momentum: Russell MidCap Value/CRSP MidCap Value: RAFI Fundamental Small Value to derive the factor load estimates

The factor loads can fairly closely be approximated by:

20% MTUM
30% VOE
50% FNDA

Simulated factor loads: 8/1996 - 12/2014
Alpha = +0.01
Mkt = +1.07
Size = +0.35
Value = +048
Momentum = -0.04
Quality =+0.24
R^2 = 0.96

Using MSCI USA Momentum:Russell MidCap Value/CRSP MidCap Value:Russell Fundamental Small Cap to derive the factor load estimates. 8/1996 is earliest date of available data for Russell Fundamental Small Cap.

Robert
.
User avatar
grap0013
Posts: 1892
Joined: Thu Mar 18, 2010 1:24 pm

Re: PXSV has changed indexes

Post by grap0013 »

^ That's informative.

I think FNDA and PRFZ/SFSNX are all pretty interchangeable.

For us 100% SCV tilters I'm guessing something like:

70% FNDA
30% MTUM

looks pretty for factor loads?
There are no guarantees, only probabilities.
JohnnyFive
Posts: 181
Joined: Mon Sep 08, 2014 1:14 pm

Re: PXSV has changed indexes

Post by JohnnyFive »

I'm really torn as to what to do here. Like Grap and Robert, I'm considering just simplifying and going to something like VISVX (Vanguard Small Value), which is really more of a mid cap value fund, cause it isn't that small (and for that matter, isn't that valuey), or possibly going with something like RZV, which is probably the smallest and most valuey of all the domestic small cap value funds out there, but, has a hard time with momentum. It's available to trade for free on Schwab at about the same ER as PXSV. I'm just not really sure.
Beat The Street
Posts: 336
Joined: Wed Sep 26, 2012 3:09 pm

Re: PXSV has changed indexes

Post by Beat The Street »

Here is a cheap portfolio for U.S. allocation with good factor loadings....https://www.portfoliovisualizer.com/fac ... F01%2F2011
“Never ask anyone for their opinion, forecast, or recommendation. Just ask them what they have—or don’t have—in their portfolio.” -Taleb
User avatar
matjen
Posts: 2189
Joined: Sun Nov 20, 2011 10:30 pm

Re: PXSV has changed indexes

Post by matjen »

What happened to good old iShares S&P Small-Cap 600 Value IJS?

The elephant in the room is that this is another potential reason for DFA it seems to me. They don't mess about on the indexes.
A man is rich in proportion to the number of things he can afford to let alone.
Wade Garrett
Posts: 235
Joined: Wed Jul 10, 2013 2:41 pm
Location: Jasper, MO/Double Deuce

Re: PXSV has changed indexes

Post by Wade Garrett »

JohnnyFive wrote:I'm really torn as to what to do here. Like Grap and Robert, I'm considering just simplifying and going to something like VISVX (Vanguard Small Value), which is really more of a mid cap value fund, cause it isn't that small (and for that matter, isn't that valuey), or possibly going with something like RZV, which is probably the smallest and most valuey of all the domestic small cap value funds out there, but, has a hard time with momentum. It's available to trade for free on Schwab at about the same ER as PXSV. I'm just not really sure.
Have a look at JKL. It rarely gets discussed. JKL isn't very small but is very valuey. I also own RZV and PXSV and will also have to make a decision on what to do about PXSV before late May.
"I'm not an inventor. I'm an improver. I see things that are wrong, and I improve them." - Larry David, Curb Your Enthusiasm
Beat The Street
Posts: 336
Joined: Wed Sep 26, 2012 3:09 pm

Re: PXSV has changed indexes

Post by Beat The Street »

matjen wrote:What happened to good old iShares S&P Small-Cap 600 Value IJS?

The elephant in the room is that this is another potential reason for DFA it seems to me. They don't mess about on the indexes.


IJS and VIOV (same index) are great substitutes for PXSV, not quite as good on value but almost as close as you can get.
“Never ask anyone for their opinion, forecast, or recommendation. Just ask them what they have—or don’t have—in their portfolio.” -Taleb
countmein
Posts: 653
Joined: Fri Dec 06, 2013 8:10 pm

Re: PXSV has changed indexes

Post by countmein »

grap0013 wrote: For us 100% SCV tilters I'm guessing something like:

70% FNDA
30% MTUM

looks pretty for factor loads?
Not really. See: https://www.portfoliovisualizer.com/fac ... ctor=false

FNDA has had an almost non-detectable value load since inception, but hard to say as it's not stat-significant.

MTUM has a bit of negative value and some ugly negative alpha, -3.5% pa and it's almost stat-significant. I have real doubts about ishares' ability to implement this strategy cheaply/effectively.

More and more I'm tempted towards the simplicity of 3-fund + QSPIX (or QMNIX). Paying up a little for the stewardship, knowing what you're getting, and the mentally-easy compartmentalization of exposures. It's like having 3 redwood trees and a rosebush instead of a motley collection of dying houseplant ETFs.
User avatar
matjen
Posts: 2189
Joined: Sun Nov 20, 2011 10:30 pm

Re: PXSV has changed indexes

Post by matjen »

countmein wrote:
grap0013 wrote: For us 100% SCV tilters I'm guessing something like:

70% FNDA
30% MTUM

looks pretty for factor loads?
Not really. See: https://www.portfoliovisualizer.com/fac ... ctor=false

FNDA has had an almost non-detectable value load since inception, but hard to say as it's not stat-significant.

MTUM has a bit of negative value and some ugly negative alpha, -3.5% pa and it's almost stat-significant. I have real doubts about ishares' ability to implement this strategy cheaply/effectively.

More and more I'm tempted towards the simplicity of 3-fund + QSPIX (or QMNIX). Paying up a little for the stewardship, knowing what you're getting, and the mentally-easy compartmentalization of exposures. It's like having 3 redwood trees and a rosebush instead of a motley collection of dying houseplant ETFs.
Love the metaphor! This has been my central point when we get into these 3-fund vs. slice and dice spats. If I had a fresh lump sum or were a young high income investor I would take the middle ground moving forward and utilize a 3-fund or perhaps 4-fund portfolio that had all the factor loads/tilts taken care of for you by DFA. Throw in a cheap & solid intermediate term bond fund and call it a day. Something like:

DFA US Vector Equity I DFVEX for US

DFA World ex US Targeted Val Instl DWUSX for International and Emerging Markets

Vanguard Intermediate-Term Govt Bd ETF VGIT for fixed income
A man is rich in proportion to the number of things he can afford to let alone.
pauliec84
Posts: 414
Joined: Wed Feb 09, 2011 10:29 am

Re: PXSV has changed indexes

Post by pauliec84 »

Not really. See: https://www.portfoliovisualizer.com/fac ... ctor=false
FNDA has had an almost non-detectable value load since inception, but hard to say as it's not stat-significant.
MTUM has a bit of negative value and some ugly negative alpha, -3.5% pa and it's almost stat-significant. I have real doubts about ishares' ability to implement this strategy cheaply/effectively.
3 years is not enough time to have a resalable sense estimate of the factor loads, especially with a fund like FNDA which shifts its value exposures over time. If you look at loads since the index inception you get more realiable estimates which is what Robert T used in his earlier posting.

More and more I'm tempted towards the simplicity of 3-fund + QSPIX (or QMNIX). Paying up a little for the stewardship, knowing what you're getting, and the mentally-easy compartmentalization of exposures. It's like having 3 redwood trees and a rosebush instead of a motley collection of dying houseplant ETFs.
Can't disagree with you here. AQR is going to give you some nice risk premium exposures, although you mention at a little steeper price.
Have a look at JKL. It rarely gets discussed. JKL isn't very small but is very valuey. I also own RZV and PXSV and will also have to make a decision on what to do about PXSV before late May.
JKL has done pretty well since compared to the other main small value funds providing more value exposure than most and positive alpha. It would be nice to see what the morningstar small value index factor loads look like over a longer timeframe, however I could not find out how to downlaod the index data from morningstars website. $MSVLT is the index code. If anyone can figure out how to get it I am happy to run the regressions on it.
With that said, I am happy to pay the extra 30bp of ER to access PDN compared to VSS to get a small value tilt, and was happy to pay an extra 30bp of ER to get really deep value in PXSV compared to VBR. I am not sure I will bring myself to pay the extra 20-30bp in fees to access of these funds compared to VBR. Per Grap0013's advice, will wait.

https://www.portfoliovisualizer.com/fac ... F01%2F2007
Post Reply