Due diligence on RAFI Pure Small Value

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
caklim00
Posts: 2420
Joined: Mon May 26, 2008 10:09 am

Re: PXSV Fee Waiver extended to Aug 31, 2015

Post by caklim00 »

SnowSkier wrote:FYI, one thing to add on PXSV (PowerShares Fundamental Pure Small Value)

The Invesco PowerShares PXSV website says that the current fee waiver extends through "at least August 31, 2014"
( https://www.invesco.com/portal/site/us/ ... uctId=PXSV )

I emailed them and asked if it would be extended. Their response is yes:

"While the prospectus has not yet been updated, we did extend this fee waiver through at least Aug 31st, 2015"
That's good news. I really can't imagine a company deciding to increase the ER by .3 or more given the competition that is out there.
User avatar
Topic Author
Robert T
Posts: 2803
Joined: Tue Feb 27, 2007 8:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Due diligence on RAFI Pure Small Value

Post by Robert T »

.
There have been several other threads showing a higher value load for pxsv than other value indexes (since it started tracking the RAFI pure value series), but higher current P/B. How do we reconcile the two? The analyses have also shown less negative (zero) alpha on pxsv relative to some of the other indexes. What the source of this effect?

Here's an attempt to answer these two questions:

As I understand, the methodology for the RAFI series has an in-built ‘dynamic’ relative weighting to growth and value – in essence giving greater weight to higher book-to-market stocks when the spread in book-to-market between value and growth stocks is larger, and giving smaller weights when the spreads are small. If valuation spreads influence subsequent performance then there should be a ‘relative valuation based rebalancing’ return to the RAFI funds (i.e. overweighting value when spreads are wide, underweighting when spreads are small), if not, then there won’t be value-add from this approach.

By way of example (at least my understand). Lets take a weighting by book value only which is one of the RAFI metrics. As B = M*B/M, RAFI essentially weights stocks by B/M. Suppose the average B/M for value stocks is 1.5 and for growth stocks is 0.5 then value stocks will have 3 times the weight as growth stocks (1.5/0.5). Suppose the spread narrows and the average B/M for value stocks is 1.25 and for growth stocks is 0.75 then the weighting to value stocks will decline to 1.67 times the weight of growth stocks (1.25/0.75). This seems to imply that the P/B for a RAFI linked fund may be more variable over time than a 'pure value fund', reflective of this ‘relative valuation based rebalancing'. There is some evidence of this in the Powershares RAFI US 1000 (PRF) and RAFI US 1500 small-mid (PRFZ) – according to M*, since inception the standard deviation in ‘prospective’ P/B has been larger for PRFZ than for DFA SV, Russell 2000 Value (IWN), S&P 600 Value (IJS), or S&P 600 Pure Value (RZV), and has been larger for PRF than DFA LV, and Russell 1000 value.

My understanding of the RAFI Pure Value (pxsv) sorts is that after the book value sorts into size (which as indicated above is equivalent to M*B/M), they sort by B/M [plus other metrics]. So in effect it seems – something along the lines of a M*(B/M) squared. This would increase the ‘relative valuation based rebalancing’ effect. Following the above example, of a B/M of 1.5 for value stocks and 0.5 for growth, instead of value stocks having three times the weight (1.5/0.5), they would have nine times the weight (1.5*1.5/0.5*0.5= 2.25/0.25), and when variations narrow to 1.25 and 0.75, value stocks will get almost 3 times the weight as growth stocks (1.25*1.25/0.75*0.75 = 1.56/0.54 = 2.9), compared to 1.67 times in the previous example.

My sense is that the 'relative valuation based rebalancing' is what can lead to: (i) high longer-term value loads (on average, a value tilt over the long term), but shorter term P/B variation; and (ii) the observable non-negative (zero) alphas on the RAFI series, which will likely only continue if relative valuation differences continue to influence subsequent return differences.

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

Re: Due diligence on RAFI Pure Small Value

Post by grap0013 »

^ That's such an intelligent post you may not get any responses, but I will try. :-)

I think basically Robert T is saying that the more prices are "misaligned" from historic norms, the more that PXSV will load up on what's cheap relative to other investment options. Buy low, sell high at its finest.

Neyseyers want a factor model to explain all of the outperformance, but just because it cannot be explained (yet) doesn't mean it isn't there.
There are no guarantees, only probabilities.
caklim00
Posts: 2420
Joined: Mon May 26, 2008 10:09 am

Re: Due diligence on RAFI Pure Small Value

Post by caklim00 »

grap0013 wrote:^ That's such an intelligent post you may not get any responses, but I will try. :-)

I think basically Robert T is saying that the more prices are "misaligned" from historic norms, the more that PXSV will load up on what's cheap relative to other investment options. Buy low, sell high at its finest.

Neyseyers want a factor model to explain all of the outperformance, but just because it cannot be explained (yet) doesn't mean it isn't there.
Klaatu verata nik...PXSV
User avatar
grap0013
Posts: 1892
Joined: Thu Mar 18, 2010 1:24 pm

Re: Due diligence on RAFI Pure Small Value

Post by grap0013 »

^ Ha ha! Thanks for the chuckle!
There are no guarantees, only probabilities.
User avatar
wjo
Posts: 509
Joined: Fri Jan 22, 2010 11:17 am

Re: Due diligence on RAFI Pure Small Value

Post by wjo »

Robert T wrote: My sense is that the 'relative valuation based rebalancing' is what can lead to: (i) high longer-term value loads (on average, a value tilt over the long term), but shorter term P/B variation; and (ii) the observable non-negative (zero) alphas on the RAFI series, which will likely only continue if relative valuation differences continue to influence subsequent return differences.

Robert
Good analysis and squares with what Research Affiliates says about it - basically they are capturing the value effect more effectively through mean reversion harvesting.

Is essence, they continue to claim that the FI approach builds a better mousetrap to capturing returns. As long as mean reversion works(i.e. volatility works as in the past and we are not in a secular bull/bear market) and they don't have much market impact (i.e.., too much of what they are doing will over-mine the premia), then FI should relatively out perform simple sorts such as P/B.

Downsides:

(1) tends to result in portfolios of smaller numbers of stocks (or maybe they just start with a smaller universe for the pure style series??)
(2) in its own way concentrates holdings, depsite protests to the contrary
(3) is not style consistent if you are focusing capturing size and value - Vanguard has a white paper about this that suggested that the best way to consistently harvest value and size factors is with cap weighted approaches.

Positives -

(1) does appear to capture value effectively, provided you believe the research that says value is not best expressed through a simple sort but rather though mean reverting processes
(2) does intuitively appeal as a logical approach to security selection. I actually like that FI approaches do attempt to determine securities pricing as a opposed to just screening on a size/value sort. John Cochrane would also appreciate this view as well; in Cochrane's 2010 address ("Discount Rates") to the finance society, he asked the question why have we let price be on the right side of the equations (used as an input instead of an output to our models)?

I suppose philosophically, if one believes in the FI approach to index construction, why not go whole hog and just use FI for all your portfolio? If instead one wants factor selection via Fama French, why believe in an FI derived approach and not just get the simplest/cheapest slice of value/size?
jaab
Posts: 112
Joined: Thu Aug 07, 2014 12:52 pm

Re: Due diligence on RAFI Pure Small Value

Post by jaab »

So, in other words the style drift is a feature, not a bug, and Rob Arnott DID invent something himself (regarding retail client products)? Ha, take that, Asness! :D
User avatar
Topic Author
Robert T
Posts: 2803
Joined: Tue Feb 27, 2007 8:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Due diligence on RAFI Pure Small Value

Post by Robert T »

.
The 'closer to zero alpha' on the RAFI series relative to traditional value indexes, IMO is mainly from a rebalancing effect between value and growth stocks. There is an in-built rebalancing effect in the value premium (HML) itself which is comprised of large growth, large value, small growth, and small value stocks. If you take a portfolio of 63% FF LV and 37% FF LG you can replicate the RAFI US 1000 index fairly (very) closely as I recall - on factor loads, annualized returns and standard deviation. This 63:37 LV:LG (if I am remembering the shares correctly) has a rebalancing return - selling value when it gets relatively expense, buying growth when it gets relatively cheap and vice-versa - so rebalancing is based on changes in relative valuation between value and growth - which is what the RAFI methodology seems to do (together with HML itself). Traditional value indexes only focus on one side of HML - the LV and SV side - so the common 'negative alpha' may simply be a reflection of the abscence of the rebalancing return. In this respect the DFA vector and core funds, which hold some growth stock, will perhaps also have less 'negative alpha' than traditional value funds.

Robert
.
User avatar
wjo
Posts: 509
Joined: Fri Jan 22, 2010 11:17 am

Re: Due diligence on RAFI Pure Small Value

Post by wjo »

Robert T wrote: The 'closer to zero alpha' on the RAFI series relative to traditional value indexes, IMO is mainly from a rebalancing effect between value and growth stocks. There is an in-built rebalancing effect in the value premium (HML) itself which is comprised of large growth, large value, small growth, and small value stocks. If you take a portfolio of 63% FF LV and 37% FF LG you can replicate the RAFI US 1000 index fairly (very) closely as I recall - on factor loads, annualized returns and standard deviation. This 63:37 LV:LG (if I am remembering the shares correctly) has a rebalancing return - selling value when it gets relatively expense, buying growth when it gets relatively cheap and vice-versa - so rebalancing is based on changes in relative valuation between value and growth - which is what the RAFI methodology seems to do (together with HML itself). Traditional value indexes only focus on one side of HML - the LV and SV side - so the common 'negative alpha' may simply be a reflection of the abscence of the rebalancing return. In this respect the DFA vector and core funds, which hold some growth stock, will perhaps also have less 'negative alpha' than traditional value funds.

Robert
.
Robert - unclear about the DFA core funds. A recent paper by Ed Tower suggests that the Core funds have under performed RAFI and Wisdom Tree. (See: http://papers.ssrn.com/sol3/papers.cfm? ... id=2179188) I was trying to understand the differences between FI and DFA Core construction. It appears that DFA doesn't capture much rebalancing from the growth side of its holdings - still a good bit of negative alpha. However, I haven't looked at the data to compare negative alphas of say, Vector Core versus Targeted Value.

In any case, it does seem that the RAFI indices are fundamentally a different mousetrap. DFA Core seems to mostly lower transactional costs from trading out of value bands.
Last edited by wjo on Wed Aug 27, 2014 3:20 pm, edited 1 time in total.
User avatar
Topic Author
Robert T
Posts: 2803
Joined: Tue Feb 27, 2007 8:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Due diligence on RAFI Pure Small Value

Post by Robert T »

.
Just did a quick check on the FF3F analysis on RAFI vs FF LV:FF LG – its closer to a 2/3 FF Large Value:1/3 FF Large Growth. Data are from Jason Hsu’s website (for RAFI) and Ken French website (for FF LG and FF LV).

The factor loads are almost identical, although RAFI still has a slight edge (-0.01 vs. -0.04 monthly alpha). Compare this with the alpha on the FF LV series over the same period of -0.12. With the addition of growth, the negative alpha was significantly reduced.

Here are the respective factor loads, annualized returns, and standard deviations for the two portfolios.

01/1964-06/2011
  • RAFI US
    Alpha = -0.01
    Beta = 1.03
    Size = -0.05
    Value = 0.39

    Annualized return = 11.7%
    Standard deviation = 15.4

    0.34% FF LG:66% FF LV
    Alpha = -0.04
    Beta = 1.03
    Size = -0.05
    Value = 0.39

    Annualized return = 11.4%
    Standard deviation = 15.5
I think this matters more to those who have constructed their portfolio to target a specific size and value load (matters to me). Having a close to zero alpha provides more assurance of realizing the factor load implied premium shares, and avoiding the haircut from a large negative alpha. i.e. a more accurate factor matching of a portfolio to ‘needed’/‘expected returns’.

Robert
.
caklim00
Posts: 2420
Joined: Mon May 26, 2008 10:09 am

Re: Due diligence on RAFI Pure Small Value

Post by caklim00 »

Thanks Robert. I think a key thing though with PXSV is that it is a Value fund where a RAFI sort is then performed. So, its more valuey to begin with. I wish the same option was available in international where recently I've been adding to RAFI Small. RAFI Small Value just seems better...

And this is interesting, for RAFI Small Value has a lower concentration of Financials than RAFI Small Growth: http://www.researchaffiliates.com/Our%2 ... _Style.pdf

Financials
GROWTH 27.70%
CORE 23.28%
VALUE 25.39%
JohnnyFive
Posts: 181
Joined: Mon Sep 08, 2014 1:14 pm

Re: Due diligence on RAFI Pure Small Value

Post by JohnnyFive »

Is this ETF offered commission free at any of the online brokerages? I don't have any brokerage accounts at this point. My entire portfolio is held within Vanguard's funds. I have not yet bought ETF's, however, I'm looking to buy into PXSV and given they don't have a fund equivalent, will have to open up a brokerage account somewhere. Any help would be greatly appreciated. If this particular ETF is NOT offered commission free anywhere, are there any very inexpensive online brokerages out there any of you guys would recommend?
vesalius
Posts: 794
Joined: Tue Jul 13, 2010 7:00 pm
Location: Texas

Re: Due diligence on RAFI Pure Small Value

Post by vesalius »

JohnnyFive wrote:Is this ETF offered commission free at any of the online brokerages? I don't have any brokerage accounts at this point. My entire portfolio is held within Vanguard's funds. I have not yet bought ETF's, however, I'm looking to buy into PXSV and given they don't have a fund equivalent, will have to open up a brokerage account somewhere. Any help would be greatly appreciated. If this particular ETF is NOT offered commission free anywhere, are there any very inexpensive online brokerages out there any of you guys would recommend?
Generally speaking no, I do not know of any brokerage that offer PXSV as part of a free trade group of ETF's. Although I trade it free at WellsTrade through a program that no longer is available to new customers. The cost to trade this or almost any other ETF is going to be between $7-$9 at almost any of the reputable online brokers. If you are only doing a few pooled or quarterly buys for year the trading cost are relatively minor.

EDIT: A few smaller brokerages offer cheaper trades. OptionsHouse and TradeKing offer etfs and stock trades for $4.75 and $4.95 respectively.
Last edited by vesalius on Fri Sep 12, 2014 12:45 pm, edited 2 times in total.
SnowSkier
Posts: 268
Joined: Fri Mar 16, 2012 12:21 pm
Location: Austin

Re: Due diligence on RAFI Pure Small Value

Post by SnowSkier »

Johnny5, You can open a Vanguard Brokerage acct and trade ETFs, but the Fundamental Index ETFs would not be commission free.

If you want to do "Fundamental Indexing" and go commission-free, one option would be a Schwab account where I believe the following are commission free:
FNDX, US Large Fundamental, similar to PRF
FNDB, US Broad Fundamental
FNDF, Developed Large Fundamental, similar to PXF
FNDC, Developed Small Fundamental, similar to PDN
FNDE, Emerging Fundamental, similar to PXH

I'm not sure if FNDA is commission free?
FNDA, US Small Fundamental, similar to PRFZ
User avatar
Rick Ferri
Posts: 9707
Joined: Mon Feb 26, 2007 10:40 am
Location: Georgetown, TX. Twitter: @Rick_Ferri
Contact:

Re: Due diligence on RAFI Pure Small Value

Post by Rick Ferri »

SnowSkier,

The issue with this approach is that you're paying a lot of money for beta exposure.

FNDB has an expense ratio of 0.32%. You could engineer a lower cost portfolio with the same factor exposures by investing in VTI and adding a small cap value fund.

Rick Ferri
The Education of an Index Investor: born in darkness, finds indexing enlightenment, overcomplicates everything, embraces simplicity.
berntson
Posts: 1366
Joined: Mon Oct 29, 2012 12:10 pm

Re: Due diligence on RAFI Pure Small Value

Post by berntson »

^ Agree with Rick. Fundamental funds are nice for international markets because there are not widely available small value or midcap value funds. For domestic stocks, the right combination of VTI and VBR (or VOE) will get the same result with lower expense.
berntson
Posts: 1366
Joined: Mon Oct 29, 2012 12:10 pm

Re: Due diligence on RAFI Pure Small Value

Post by berntson »

vesalius wrote: A few smaller brokerages offer cheaper trades. OptionsHouse and TradeKing offer etfs and stock trades for $4.75 and $4.95 respectively.
Interactive Brokers has $1 commissions. They charge a $10 monthly minimum, but the minimum is waved if you have more than $100,000 invested. Given that I only make a few trades a year, my yearly commissions would be less than the cost of a large pizza. (I have my investments at Vanguard but am considering moving them to IB to reduce my fees.)
User avatar
in_reality
Posts: 4529
Joined: Fri Jul 12, 2013 6:13 am

Re: Due diligence on RAFI Pure Small Value

Post by in_reality »

Rick Ferri wrote:SnowSkier,

The issue with this approach is that you're paying a lot of money for beta exposure.

FNDB has an expense ratio of 0.32%. You could engineer a lower cost portfolio with the same factor exposures by investing in VTI and adding a small cap value fund.

Rick Ferri
I don't believe that Russell fundamental indexes should be viewed as adding the same factor exposure as traditional value style funds. Yes, they load on the value factor but in a different way.

If they were the same, why is the alpha on these fundamental indexes higher (about zero or maybe positive but not statistically significantly so) than value style funds (which tend to be statistically significantly negative)?

A traditional cap-weighted value will certainly maintain a constant value exposure at a lower cost if that is what you want.

As for me, I plan to add FNDB as rolling 36-month Fama-French regressions shows fundamental type funds have a dynamic range of exposures to HML (value premium). This is as the weighting scheme will have it buy value after a period of under performance and sell after a period of over performance (buy low sell high). We certainly know the value premium is time varying and I believe it has a tendency toward mean (or average) reversion.

Also, the weighting scheme of the Russell funds means it will also select companies that are relatively valuey in relatively non-valuey sectors. You can see this in the 9 box, so if you are looking for straight value funds, the Schwab fundamental indexes hardly seem to fit the bill. FNDA has larger allocation to mid-cap growth than you would get in a traditional cap-weighted mid-cap or small-cap fund. I don't mind having exposure to more companies in sectors that are not typically valuey.

To suggest you are getting the same exposure from a traditional cap-weighted value, I think is inaccurate.

The RAFI funds seem to be doing fine in terms of overcoming their ER. The question perhaps on the Russell series is since it doesn't tilt to such deep value, will it also be worth holding. I don't pretend to have a definitive answer on this, but I think it will.

I do agree that if FNDA offers the same exposure, then the higher ER is a waste. Is it really the same???

http://www.russell.com/documents/indexe ... esting.pdf
http://www.rallc.com/Production%20conte ... tegies.pdf
http://www.researchaffiliates.com/Produ ... Series.pdf
berntson
Posts: 1366
Joined: Mon Oct 29, 2012 12:10 pm

Re: Due diligence on RAFI Pure Small Value

Post by berntson »

^ Thanks in_reality. I just compared PowerShare RAFI 1000 (PRF) to Vanguard small, mid, and large value and DFA large value. Anyone who is interested can find the results here. One of the things that surprised me is how strong the value loading is for PRF. It has a value loading that is significantly higher than for the mid cap and large cap value funds from Vanguard. That helps to make the fund more worthwhile.

None of the funds have alphas statistically different than zero. For instance, looking at the p-values, there is about a 45% chance that the positive alpha for PRF since 2006 is just the result of chance. So I wouldn't put too much weight on the positive alphas.

FWIW I prefer PowerShare's RAFI funds to the funds from Schwab.
in_reality wrote: To suggest you are getting the same exposure from a traditional cap-weighted value, I think is inaccurate.
I at least partially agree. For the same price, I prefer RAFI fundamental funds to similar cap weighted value funds from Vanguard. The problem is that PRF is 30 basis points more expensive than the value funds from Vanguard. Better fund construction may make some of that back. Higher value loadings also help. But I'm not at all sure that PRF can make back all of those extra 30 basis points.
User avatar
in_reality
Posts: 4529
Joined: Fri Jul 12, 2013 6:13 am

Re: Due diligence on RAFI Pure Small Value

Post by in_reality »

berntson wrote:^ Thanks in_reality.
All thanks to Robert T.
berntson wrote: I just compared PowerShare RAFI 1000 (PRF) to Vanguard small, mid, and large value and DFA large value. Anyone who is interested can find the results here. One of the things that surprised me is how strong the value loading is for PRF. It has a value loading that is significantly higher than for the mid cap and large cap value funds from Vanguard. That helps to make the fund more worthwhile.
Thanks for the link.

If you look at the four factor model though, the PRF value loading falls in line with the others (droping from .41 to .30).

Similarly, if you look at PRFZ (PowerShares RAFI 1500 Mid Small) the loading in three factor models goes from .31 to .19. FNDA (Schawb Mid Small)goes the opposite from .25 to .34 Not sure what to make of this.
berntson wrote: FWIW I prefer PowerShare's RAFI funds to the funds from Schwab.

I at least partially agree. For the same price, I prefer RAFI fundamental funds to similar cap weighted value funds from Vanguard. The problem is that PRF is 30 basis points more expensive than the value funds from Vanguard. Better fund construction may make some of that back. Higher value loadings also help. But I'm not at all sure that PRF can make back all of those extra 30 basis points.
Well, looking at comparing Schwab Fundamental Indexes to Vanguard values funds, it seems the value loading is about the same (at least under four factor model). RAFI Large cap is the same, and not sure why the loading on RAFI small drop. Anway, the RAFI and Schwab funds have much more in the Growth style (9 box). So to me it's about diversification. I'll guess I'll take a chance on paying the ER to weight by fundamentals as that gives me exposure to relatively valuey companies in less valuey industries. Not sure I'd ever go 100% fundamental indexes or even more than 50% though.

Anyway, there is a nice write up on the Russell series at Schwab - better metrics to measure value and better rebalancing http://news.morningstar.com/articlenet/ ... ?id=640658
SnowSkier
Posts: 268
Joined: Fri Mar 16, 2012 12:21 pm
Location: Austin

Re: Due diligence on RAFI Pure Small Value

Post by SnowSkier »

in_reality wrote:FNDA (Schawb Mid Small)goes the opposite from .25 to .34 Not sure what to make of this.
If analyzing any of the FND_ funds, which have only been around for a little over a year, you may want to use the currently-equivalent Schwab mutual funds (SF__X), which have been around much longer.

FNDX ~ SFLNX (Schwab Fundamental US Large Co), similar to PRF
FNDA ~ SFSNX (Schwab Fundamental US Small Co), similar to PRFZ
FNDF ~ SFNNX (Schwab Fundamental Intl Large Co), similar to PXF
FNDC ~ SFILX (Schwab Fundamental Intl Small Co), similar to PDN
FNDE ~ SFENX (Schwab Fundamental Emerging Mkts Large Co), similar to PXH

At one point, the Schwab mutual funds changed from FTSE RAFI Fundamental Indexes to the similar Russell Fundamental Indexes, but all the Schwab ETFs and Funds above are now using Russell Fundamental indexes. The PowerShares ones use FTSE RAFI Fundamenal indexes.
in_reality wrote:Anyway, there is a nice write up on the Russell series at Schwab - better metrics to measure value and better rebalancing http://news.morningstar.com/articlenet/ ... ?id=640658
Agreed, interesting & helpful article from Samuel Lee.

One thing to add,
In William Bernstein's recent book "Rational Expectations", he lists the following Fundamental Index funds in his short lists of recommended funds:

PRF, for US Large Cap Value
PXSV, for US Small Cap Value
PXF, for Dev Large Cap Value
PDN, for Dev Small Cap Value
PXH, for EM Large Cap Value

When William Bernstein speaks, it definitely catches my attention. :happy
So, I'm doing more investigation, and currently reading "The Fundamental Index" book by Rob Arnott et. al.
vesalius
Posts: 794
Joined: Tue Jul 13, 2010 7:00 pm
Location: Texas

Re: Due diligence on RAFI Pure Small Value

Post by vesalius »

berntson wrote:
vesalius wrote: A few smaller brokerages offer cheaper trades. OptionsHouse and TradeKing offer etfs and stock trades for $4.75 and $4.95 respectively.
Interactive Brokers has $1 commissions. They charge a $10 monthly minimum, but the minimum is waved if you have more than $100,000 invested. Given that I only make a few trades a year, my yearly commissions would be less than the cost of a large pizza. (I have my investments at Vanguard but am considering moving them to IB to reduce my fees.)
Won't be long before all trades are free at several smaller brokerages, at least for larger accounts. Thanks for adding in more info.
berntson
Posts: 1366
Joined: Mon Oct 29, 2012 12:10 pm

Re: Due diligence on RAFI Pure Small Value

Post by berntson »

SnowSkier wrote: One thing to add,
In William Bernstein's recent book "Rational Expectations", he lists the following Fundamental Index funds in his short lists of recommended funds:

PRF, for US Large Cap Value
PXSV, for US Small Cap Value
PXF, for Dev Large Cap Value
PDN, for Dev Small Cap Value
PXH, for EM Large Cap Value

When William Bernstein speaks, it definitely catches my attention. :happy
So, I'm doing more investigation, and currently reading "The Fundamental Index" book by Rob Arnott et. al.
I had forgotten about this, so went back to look it up. Bernstein's recommended funds are all from three companies: Vanguard, DFA, and PowerShares (including all of the above listed funds). This makes me all the more happy about my recent decision to use PXF for my large developed international fund. I already have PXSV for domestic small value and may switch to PXH for emerging markets. I seem to be moving in the direction of an all PowerShares portfolio.

You should definitely read the fundamental index book. Be warned that, while it's one of my favorite books on investing, it's also essentially an extended advertisement for their funds. Also, as Robert T has pointed out, something is wacky with their emerging markets data in the book.
JohnnyFive
Posts: 181
Joined: Mon Sep 08, 2014 1:14 pm

Re: Due diligence on RAFI Pure Small Value

Post by JohnnyFive »

What fundamental index book are you folks referring to?
jaab
Posts: 112
Joined: Thu Aug 07, 2014 12:52 pm

Re: Due diligence on RAFI Pure Small Value

Post by jaab »

The book is called exactly that: The Fundamental Index :D
JohnnyFive
Posts: 181
Joined: Mon Sep 08, 2014 1:14 pm

Re: Due diligence on RAFI Pure Small Value

Post by JohnnyFive »

Gotcha, thanks. Will download it now for my ereader.
User avatar
Blue
Posts: 1178
Joined: Sat Jul 12, 2008 10:18 pm

Re: Due diligence on RAFI Pure Small Value

Post by Blue »

Does PowerShares give back 100% of securities lending revenue?
JohnnyFive
Posts: 181
Joined: Mon Sep 08, 2014 1:14 pm

Re: Due diligence on RAFI Pure Small Value

Post by JohnnyFive »

Looks like PXSV is now offered commission free on Schwab as of VERY recently. Great news.
vesalius
Posts: 794
Joined: Tue Jul 13, 2010 7:00 pm
Location: Texas

Re: Due diligence on RAFI Pure Small Value

Post by vesalius »

JohnnyFive wrote:Looks like PXSV is now offered commission free on Schwab as of VERY recently. Great news.
Glad you brought that here. Schwab just expanded their lineup of free to trade ETFs 2 days ago I think. There is your free trades and no fee account Johnny.
stlutz
Posts: 5585
Joined: Fri Jan 02, 2009 12:08 am

Re: Due diligence on RAFI Pure Small Value

Post by stlutz »

Does PowerShares give back 100% of securities lending revenue?
Yes.
berntson
Posts: 1366
Joined: Mon Oct 29, 2012 12:10 pm

Re: Due diligence on RAFI Pure Small Value

Post by berntson »

JohnnyFive wrote:Looks like PXSV is now offered commission free on Schwab as of VERY recently. Great news.
This is good news. Unfortunately, the other PowerShares funds I use aren't offered by Schwab (PXF and PXH). This is probably because they have their own funds that directly compete with PXF and PXH.
Last edited by berntson on Sat Sep 20, 2014 10:38 am, edited 2 times in total.
stlutz
Posts: 5585
Joined: Fri Jan 02, 2009 12:08 am

Re: Due diligence on RAFI Pure Small Value

Post by stlutz »

Also, Schwab's ETF One Source program has relatively high fees for the ETF provider. As a result, the platform is full of ETFs that are looking for assets. If the ETF has a $1b AUM, the product is likely profitable for the provider and as such, there is no reason for the provider to give a lot of that profit to Schwab.

http://www.schwab.com/public/schwab/nn/ ... .html#etfs
Schwab receives payments from the third-party ETF sponsors or their affiliates participating in ETF OneSource for recordkeeping, shareholder services and other administrative services that Schwab provides to participating ETFs. .. ETF sponsors or their affiliates pay a fixed program fee to Schwab each year for each ETF participating in ETF OneSource. The program fees vary, but can range up to $250,000 per year for each participating ETF. ETF sponsors or their affiliates also pay Schwab an asset-based fee based on a percentage of total ETF assets purchased by Schwab customers after the ETF was added to ETF OneSource. The amount of the asset-based fee can range up to 0.15% annually.
berntson
Posts: 1366
Joined: Mon Oct 29, 2012 12:10 pm

Re: Due diligence on RAFI Pure Small Value

Post by berntson »

^ Maybe this means that PowerShares can finally get the PSXV AUM over the magical 100 million threshold. :)
User avatar
grap0013
Posts: 1892
Joined: Thu Mar 18, 2010 1:24 pm

Re: Due diligence on RAFI Pure Small Value

Post by grap0013 »

berntson wrote:^ Maybe this means that PowerShares can finally get the PSXV AUM over the magical 100 million threshold. :)

Yeah! Then we can all buy in taxable!

I'm going to do my part and still pay my once per year $7.95 commission thru VG brokerage link to buy PXSV in my Roth until then. These fees are outrageous! :P
There are no guarantees, only probabilities.
JohnnyFive
Posts: 181
Joined: Mon Sep 08, 2014 1:14 pm

Re: Due diligence on RAFI Pure Small Value

Post by JohnnyFive »

Grap and Robert, what are your thoughts on JKL (Ishares Morningstar Small Value fund). Compares to PXSV, seems to have fewer holdings (258 vs 799), higher turnover, but a lower ER (0.3% vs 0.39%, for now, and might rise). It has significantly more AUM (almost $400mil vs less than $70mil). It also seems to have a greater value load, at least according to the morningstar style boxes. PXSV seems to be slightly smaller, but relatively less valuey. I'm curious as to your thoughts.
ralph124cf
Posts: 2985
Joined: Tue Apr 01, 2014 11:41 am

Re: Due diligence on RAFI Pure Small Value

Post by ralph124cf »

grap0013 wrote:
berntson wrote:^ Maybe this means that PowerShares can finally get the PSXV AUM over the magical 100 million threshold. :)

Yeah! Then we can all buy in taxable!

I'm going to do my part and still pay my once per year $7.95 commission thru VG brokerage link to buy PXSV in my Roth until then. These fees are outrageous! :P
Does anybody know how often an ETF updates its AUM? I know that a number of people on this forum have bought shares, but that does not necessarily equate to new creation units. As we know we buy ETF shares from others, not from the ETF itself, so buying the shares does not directly change the AUM.

Ralph
User avatar
Vegomatic
Posts: 385
Joined: Tue Feb 27, 2007 10:16 pm

Re: Due diligence on RAFI Pure Small Value

Post by Vegomatic »

JohnnyFive wrote:

  • ...It [JKL] also seems to have a greater value load [than PXSV], at least according to the morningstar style boxes. PXSV seems to be slightly smaller, but relatively less valuey. I'm curious as to your thoughts.
If you look at both (JKL,PXSV) over the post-June 2011 period in which both have been around in their current form, PXSV would seem to have higher market, size and value factor loadings (coefficients) than JKL.

See link/result: http://tinyurl.com/pxsv-jkl-ff3f

Code: Select all

Ticker  	                        Bmkt  Size  Value   Alpha   Annual-Alpha   R2
JKL                                1.05   0.48  0.39  -0.05%   -0.57% 	   0.98
PXSV                               1.09   0.82  0.52   0.02%    0.30% 	   0.98
Explanation/DFA-marketing info on factor loads available here: http://tinyurl.com/m-f-i-fama
User avatar
grap0013
Posts: 1892
Joined: Thu Mar 18, 2010 1:24 pm

Re: Due diligence on RAFI Pure Small Value

Post by grap0013 »

ralph124cf wrote:
Does anybody know how often an ETF updates its AUM?

Ralph
Daily I believe.
There are no guarantees, only probabilities.
JohnnyFive
Posts: 181
Joined: Mon Sep 08, 2014 1:14 pm

Re: Due diligence on RAFI Pure Small Value

Post by JohnnyFive »

How reliable is PortfolioVisualizer? I ask because it's reporting, for example, a value load for RZV for that exact same time period, of 0.19, versus 0.52 for PXSV, which doesn't seem possible, to be honest.
User avatar
Ketawa
Posts: 2521
Joined: Mon Aug 22, 2011 1:11 am
Location: DC

Re: Due diligence on RAFI Pure Small Value

Post by Ketawa »

It is accurate, and that is possible.
JohnnyFive
Posts: 181
Joined: Mon Sep 08, 2014 1:14 pm

Re: Due diligence on RAFI Pure Small Value

Post by JohnnyFive »

Not sure how that is possible. By almost all accounts, RZV is probably the most "valuey" fund out there, and, the big knock on it has always been it's negative momentum load, which, by the way, is a product of how "valuey" it is. By almost any measureable, RSV is much more value tilted than PXSV, and, yet, porfolio visualizer indicates that PXSV has, essentially, 3 times the value load? Would love someone with more knowledge to explain that one to me. PE, PB, PS, PCF as per morningstar all significantly in favor of RZV, so, how is it that portfolio visualizer deems PXSV has 3 times the value factor load?
pradador
Posts: 156
Joined: Thu Jun 14, 2012 9:20 pm
Contact:

Re: Due diligence on RAFI Pure Small Value

Post by pradador »

JohnnyFive wrote:Not sure how that is possible. By almost all accounts, RZV is probably the most "valuey" fund out there, and, the big knock on it has always been it's negative momentum load, which, by the way, is a product of how "valuey" it is. By almost any measureable, RSV is much more value tilted than PXSV, and, yet, porfolio visualizer indicates that PXSV has, essentially, 3 times the value load? Would love someone with more knowledge to explain that one to me. PE, PB, PS, PCF as per morningstar all significantly in favor of RZV, so, how is it that portfolio visualizer deems PXSV has 3 times the value factor load?
There's some good articles on the wiki (http://www.bogleheads.org/wiki/Fama-Fre ... l_analysis) and on the net (http://www.efficientfrontier.com/ef/101/roll101.htm) on how a site like Portfolio Visualizer runs it's factor regression analysis.
User avatar
Ketawa
Posts: 2521
Joined: Mon Aug 22, 2011 1:11 am
Location: DC

Re: Due diligence on RAFI Pure Small Value

Post by Ketawa »

A 3 year period isn't enough time to draw meaningful conclusions. For 1995-07 to 2011-09, the longest period I have data for both the RAFI Pure Small Value Index and S&P 600 Pure Value Index, I get:

Code: Select all

                          HML     SMB   Mkt-RF    Mom     PMU      α   yearly α
S&P 600 Pure Value       1.15    0.95    1.05   -0.30    0.36   -0.19   -2.3%
RAFI Pure Small Value    0.80    0.84    0.95   -0.24    0.20    0.26    3.2%
All those t-stats are >2 or <-2 except for alpha on S&P 600PV, where it is -0.9, still pretty big.

For fun, the same time period in a 4-factor model. Portfolio Visualizer doesn't incorporate profitability (PMU):

Code: Select all

                          HML     SMB  Mkt-RF     Mom      α   yearly α
S&P 600 Pure Value       1.10    0.92    1.01   -0.31   -0.01   -0.1%
RAFI Pure Small Value    0.77    0.83    0.93   -0.24    0.37    4.5%
Almost exactly the same except both have higher alpha since the model doesn't account for the positive loading on profitability.

I have been keeping my eye on QSMLX as a potential holding. I'm still undecided, but it looks pretty good in Portfolio Visualizer. There isn't much data on it yet and no index data available to evaluate it.
JohnnyFive
Posts: 181
Joined: Mon Sep 08, 2014 1:14 pm

Re: Due diligence on RAFI Pure Small Value

Post by JohnnyFive »

Ketawa, which of the two would be your choice in the small value space if you had to pick one? PXSV or RZV?
User avatar
grap0013
Posts: 1892
Joined: Thu Mar 18, 2010 1:24 pm

Re: Due diligence on RAFI Pure Small Value

Post by grap0013 »

JohnnyFive wrote:Not sure how that is possible. By almost all accounts, RZV is probably the most "valuey" fund out there, and, the big knock on it has always been it's negative momentum load, which, by the way, is a product of how "valuey" it is. By almost any measureable, RSV is much more value tilted than PXSV, and, yet, porfolio visualizer indicates that PXSV has, essentially, 3 times the value load? Would love someone with more knowledge to explain that one to me. PE, PB, PS, PCF as per morningstar all significantly in favor of RZV, so, how is it that portfolio visualizer deems PXSV has 3 times the value factor load?
Don't get too hung up on factor loads. This stuff isn't hard science. RZV has an index that goes back to the mid nineties. If you back test RZV as far back as you can go and then compare PXSV's underlying index during that same time frame you find that PXSV's index had higher returns and less volatility. I call it "grap's backtest factor loading".
There are no guarantees, only probabilities.
JohnnyFive
Posts: 181
Joined: Mon Sep 08, 2014 1:14 pm

Re: Due diligence on RAFI Pure Small Value

Post by JohnnyFive »

grap0013 wrote:
JohnnyFive wrote:Not sure how that is possible. By almost all accounts, RZV is probably the most "valuey" fund out there, and, the big knock on it has always been it's negative momentum load, which, by the way, is a product of how "valuey" it is. By almost any measureable, RSV is much more value tilted than PXSV, and, yet, porfolio visualizer indicates that PXSV has, essentially, 3 times the value load? Would love someone with more knowledge to explain that one to me. PE, PB, PS, PCF as per morningstar all significantly in favor of RZV, so, how is it that portfolio visualizer deems PXSV has 3 times the value factor load?
Don't get too hung up on factor loads. This stuff isn't hard science. RZV has an index that goes back to the mid nineties. If you back test RZV as far back as you can go and then compare PXSV's underlying index during that same time frame you find that PXSV's index had higher returns and less volatility. I call it "grap's backtest factor loading".
Great, thanks Grap, where can I get returns for the two underlying indices going back further than than fund inception?
User avatar
Ketawa
Posts: 2521
Joined: Mon Aug 22, 2011 1:11 am
Location: DC

Re: Due diligence on RAFI Pure Small Value

Post by Ketawa »

JohnnyFive wrote:Ketawa, which of the two would be your choice in the small value space if you had to pick one? PXSV or RZV?
PXSV since it is more diversified and has 4-5% higher alpha than RZV. It is currently my largest equity holding. I am skeptical of its ability to maintain a large positive alpha, so I am considering QSMLX for the AQR implementation and explicit positive momentum and profitability targeting.

Search the forum and you can find a post where Rick Ferri linked the RAFI index data, I think. I don't think S&P index data is publicly available anymore.
jaab
Posts: 112
Joined: Thu Aug 07, 2014 12:52 pm

Re: Due diligence on RAFI Pure Small Value

Post by jaab »

Ketawa wrote:Portfolio Visualizer doesn't incorporate profitability (PMU):
Nowadays you can select the Frazzini/Pedersen data set, not just F/F research and benchmark. Then you get two additional options; QMJ (Asness) and BAB (low beta).

QMJ seems to be a better idea (multiple metrics), for similar reasons as for value, I think: vague concept in general (not like size), every single metrics has made or may make problems, every metric is just a second degree proxy in itself and may carry idiosyncratic noise or whatever. also there does not seem to be anything special about gross profitability: http://www.alphaarchitect.com/blog/2014 ... asurement/

Here is a paper with some reasons to include a separate beta factor (e.g. BAB). Unfortunately I myself still don't fully understand the differences between a market constant + a beta factor, a market factor + a beta factor or a market factor + Frazzini's leveraged BAB/low beta factor. Anyway, same paper, four links. First three are summaries, last one is the link to the actual paper:
http://www.cfapubs.org/doi/abs/10.2469/faj.v70.n5.3
http://www.q-group.org/wp-content/uploa ... L-5514.pdf (pages 15-17)
http://www.q-group.org/wp-content/uploa ... tation.pdf
http://papers.ssrn.com/sol3/papers.cfm? ... id=2321434

PS: This Q Group seems to be some kind of "financial old boy network". :D Nevertheless they provide several other interesting summaries of their meetings on their website. Often presentations of academic research papers that show up in the usual journals later (or have done so sooner). see http://www.q-group.org/seminar-reviews/ For full Powerpoint slides and playable audio of presentations see http://www.q-group.org/search-archives/
User avatar
grap0013
Posts: 1892
Joined: Thu Mar 18, 2010 1:24 pm

Re: Due diligence on RAFI Pure Small Value

Post by grap0013 »

Ketawa wrote: Search the forum and you can find a post where Rick Ferri linked the RAFI index data, I think. I don't think S&P index data is publicly available anymore.
JohnnyFive,

I agree with Ketawa about the S&P index data. However, you can find it in an old Robert T thread somewhere. I'd link it, but I don't have the thread handy and I can't remember which one it is in. I ran a bunch of backtested returns with the RZV data and then I discarded it once I had decided I did not like RZV. Looks like you've got some homework. :-)
There are no guarantees, only probabilities.
User avatar
grap0013
Posts: 1892
Joined: Thu Mar 18, 2010 1:24 pm

Re: Due diligence on RAFI Pure Small Value

Post by grap0013 »

Ketawa wrote: I have been keeping my eye on QSMLX as a potential holding. I'm still undecided, but it looks pretty good in Portfolio Visualizer. There isn't much data on it yet and no index data available to evaluate it.
Not that you asked me, but the biggest bugaboo with QSMLX is the lack of underlying index. If this fund were to trail PXSV for 10 years you'd have no idea if it were due to luck of the draw in differences of indexes or if AQR had poor implementation and "frictional costs" were dragging 2% per year or something to that nature. You'd be in the dark. At that point do you switch to PXSV or stay the course for QSMLX hoping to turn it around? On the other hand, if PXSV trails QSMLX for 10 years I have no problem with it as long as PXSV keeps its returns close to its underlying index. I will not confuse strategy with outcome.

Therefore, PXSV is much more transparent than QSMLX in my view and I've read the Fundamental Index and I like their methodology.
There are no guarantees, only probabilities.
Post Reply