Due diligence on RAFI Pure Small Value

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
User avatar
Topic Author
Robert T
Posts: 2593
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Due diligence on RAFI Pure Small Value

Post by Robert T » Fri Nov 22, 2013 9:37 am

.
RAFI recently came out with a US fundamental pure value series http://www.researchaffiliates.com/Our%2 ... _Style.pdf

This is now tracked by the Powershares Fundamental Pure Value Series (PXSV) starting June 16, 2011 (note that PXSV tracked a different index before June 2011, so any historical analysis of the fund before this date is a bit meaningless. We need to look at the underlying RAFI index to get a sense of the longer term characteristics).

Interesting that the value sorts use by the RAFI pure value series, like their size sorts, use 5 year averages, rather than one year observations. This seems to have kept portfolio turnover in check, and has maintained a fairly high value and size load of the series (at least in the back test).
Using the earlier data series generously share by Rick, here is the FF size and value loads

Monthly data – Jan. 1976 – Sept 2013.

Alpha…….0.08
Market….1.03
Size……….0.89
Value….…0.73

Here are the long-term annual returns compared to DFA Targeted Value, DFA Small Value (Research), and FF Small Value xU (which I understand is the benchmark for the DFA Small Value Fund, its live annual returns were used from 1994).

Image

Source: Earlier RAFI data shared by Rick, DFA matrix book 2013 (for DFA Targeted Value and FF Small Value research), and Small Value ex. U shared by Jeff Troutner (now many years ago). http://www.ifa.com/pdf/DFA/matrix_book_us_2013.pdf

Historical returns track fairly closely the DFA series.

And RAFI has made a reasonable start at tracking the index - 1 year index return = 35.9, fund NAV return = 35.52, a difference of 0.38, close to the expense ratio of 0.39.

A few caveats:
  • - The expense ratio is 0.39, as “The Advisor has contractually agreed to waive fees and/or pay certain fund expenses through at least August 31, 2014”. Without the fee waiver, operating expenses were 0.70, its not certain what will happen after August 31, 2014.

    - AMU = 58.6m. If it remains low there is a risk that the fund could close or change its benchmark (as in the past with some of the powershares funds – however its seems that RAFI has stuck with the other RAFI funds with relatively low assets).
Overall looks fairly good IMO. Other views?

Rob
.

garlandwhizzer
Posts: 2474
Joined: Fri Aug 06, 2010 3:42 pm

Re: Due diligence on RAFI Pure Small Value

Post by garlandwhizzer » Fri Nov 22, 2013 11:15 am

Looks fairly good also in my opinion for SCV exposure.

Garland Whizzer

User avatar
Rick Ferri
Posts: 8923
Joined: Mon Feb 26, 2007 11:40 am
Location: Georgetown, TX. Twitter: @Rick_Ferri
Contact:

Re: Due diligence on RAFI Pure Small Value

Post by Rick Ferri » Fri Nov 22, 2013 11:26 am

I think PXSV it is a good fund under RAFI methdology. My biggest concern is the low asset base $58.6mm. InvescoPowershares has closed ETFs that do not attract $100 million and this fund is having difficulty attracting assets.

Rick Ferri
The Education of an Index Investor: born in darkness, finds indexing enlightenment, overcomplicates everything, embraces simplicity.

User avatar
Ketawa
Posts: 2189
Joined: Mon Aug 22, 2011 1:11 am
Location: DC

Re: Due diligence on RAFI Pure Small Value

Post by Ketawa » Fri Nov 22, 2013 11:33 am

Very interesting results. 756 holdings is a good number for that small of a fund, about the same amount as VBR.

I don't think I have the RAFI data. Would you be able to run a regression from July 2001 to the present so we could compare it against the CRSP Small Value index? Would you be able to run a regression against momentum and/or profitability? I'm curious to know if the higher expense ratio would be worthwhile for the factor loads.

I could see myself adding it to my portfolio, maybe a 50/50 split with VBR.

berntson
Posts: 1366
Joined: Mon Oct 29, 2012 12:10 pm

Re: Due diligence on RAFI Pure Small Value

Post by berntson » Fri Nov 22, 2013 2:08 pm

Ketawa wrote: I don't think I have the RAFI data.
I've posted the RAFI small value index returns that I have (through 9/2011) here.

caklim00
Posts: 1914
Joined: Mon May 26, 2008 10:09 am

Re: Due diligence on RAFI Pure Small Value

Post by caklim00 » Fri Nov 22, 2013 2:16 pm

Hmmm, just randomly checked bid/ask spread and it currently is 23.97 x 100, 23.99 x 300

I can definitely live with that kind of spread... Not sure I'm willing to sell out of VBR and buy PXSV to do a round trip though, but maybe willing to put new money to PXSV. ER is .29 more than VBR

FWIW, I'm in with RAFI on the International side (although I know this is like RAFI on steroids since its already cut into the Value before the RAFI sort). I still hold a small RZV position that I have not added to in quite some time.

How low is the turnover on this index? Not sure looking at 1 year is going to be very helpful.

User avatar
Topic Author
Robert T
Posts: 2593
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Due diligence on RAFI Pure Small Value

Post by Robert T » Fri Nov 22, 2013 2:23 pm

Ketawa wrote:Would you be able to run a regression from July 2001 to the present so we could compare it against the CRSP Small Value index?
This is what I get for the factor loads using monthly data from July 2001 to September 2013:

RAFI Fundamental Small Value
  • Market=1.01
    Size=0.99
    Value=0.75
CRSP Small Value
  • Market=0.96
    Size=0.57
    Value=0.48
Annualized return over this period
  • RAFI Fundamental Small Value = 13.9%
    CRSP Small Value = 10.7%
I am much less certain that the (Novy-Marx) profitability loads give meaningful results (particularly for smallcap stocks). For example the profitability ("company quality") load on S&P 600 pure value is about +0.30. This makes little sense to me. The RAFI fund also has a relatively large profitability load over the full 1976 to 2013 period, and a large negative momentum load (both much larger than CRSP Small Value). Since 07/2001, the momentum load on RAFI SV was -0.19 vs. CRSP Small Value = -0.05. So it seems RAFI is picking up much more negative momentum than most small value funds (including DFA), but the overall performance (all things considered) seems relatively good.

Robert
.

User avatar
grap0013
Posts: 1884
Joined: Thu Mar 18, 2010 1:24 pm

Re: Due diligence on RAFI Pure Small Value

Post by grap0013 » Fri Nov 22, 2013 4:32 pm

I'm glad somebody smart finally posted about this ETF. :)

Here's what I have to say: http://www.bogleheads.org/forum/viewtop ... 10&t=91257

I think a few richer BHs need to dump their VBR for PXSV to keep this thing afloat. Vanguard will survive without your dollars I promise. You can even buy it easily through their brokerage link. I'm only half kidding about any of this.
There are no guarantees, only probabilities.

User avatar
grap0013
Posts: 1884
Joined: Thu Mar 18, 2010 1:24 pm

Re: Due diligence on RAFI Pure Small Value

Post by grap0013 » Fri Nov 22, 2013 5:20 pm

caklim00 wrote: How low is the turnover on this index? Not sure looking at 1 year is going to be very helpful.
Fairly low, but it doesn't matter much. It's trailing its' index by roughly its' expense ratio.
There are no guarantees, only probabilities.

User avatar
Ketawa
Posts: 2189
Joined: Mon Aug 22, 2011 1:11 am
Location: DC

Re: Due diligence on RAFI Pure Small Value

Post by Ketawa » Fri Nov 22, 2013 9:22 pm

berntson wrote:
Ketawa wrote: I don't think I have the RAFI data.
I've posted the RAFI small value index returns that I have (through 9/2011) here.
Ha, just realized that grap posted the same data in this thread: Comparison of small-cap value ETFs.. The data was in my spreadsheet at home, just didn't have access to it at work.

Code: Select all

RAFI Small Value    1976-01 to 2011-09  index entire history
                          HML       SMB    Mkt-RF       Mom     alpha yearly alpha
loadings                 0.678     0.878     0.996    -0.184     0.243       2.953%
standard error           0.028     0.027     0.019     0.018     0.082
r2                       0.923     1.620   #N/A      #N/A      #N/A
F                     1266.582   424.000   #N/A      #N/A      #N/A
ssreg                13303.566  1113.373   #N/A      #N/A      #N/A

t-stat                  24.048    32.847    53.402   -10.363     2.978
My factor regressions are a little different from Robert T's. The alpha is a lot higher. Is this because I haven't bothered to incorporate profitability into my spreadsheet?

In the other thread, we dumped on RZV because of the high negative momentum loading and the small number of holdings. PXSV is similarly hurt by a large negative momentum loading, but has about as many holdings as VBR. If it also has the high profitability load that Robert T mentions, I think it might be worth some more consideration. I could see myself shifting half my VBR allocation to PXSV with some convincing...

User avatar
grap0013
Posts: 1884
Joined: Thu Mar 18, 2010 1:24 pm

Re: Due diligence on RAFI Pure Small Value

Post by grap0013 » Fri Nov 22, 2013 9:37 pm

A few other thoughts I had.

Although I agreed that Fama/French should be the benchmark against which other SCV is compared, should we really be that surprised to find a superior alternative? Although there is definitely a risk story to SCV, it really is in a way a great backtest of that asset class. Therefore, it should not be too hard to imagine someone developing a formulaic strategy with SCV components that beats Fama/French. I think Arnott tested the BeJesus out of this strategy before he found the "magic" formula. It does not matter either that his strategy cannot be explained soley by FF factors because they are not perfect themselves. Hence, we have 36 years of data showing superiority and early live data demonstrating the same. This reminds me of what Larry has recently stated about profitability premiums and I'm paraphrasing, "you want to get on board early".

Also of interest: We have 7 negative years here. PXSV has lower draw down that FF SCV ex util in 6 of those periods. Higher returns with less risk?

Finally, we often talk about tracking error in the context of stocks vs. bonds. Or TSM vs. SCV. Look at 2000 and 2001 between those 4 indices. Some very large spreads in annualized returns. Therefore, in order to really make money on SCV, you have to endure not only tracking error in relation to other asset classes, but within the SCV class itself. It takes true grit and discipline to follow through. You'd better believe in what you are doing.
There are no guarantees, only probabilities.

User avatar
Ketawa
Posts: 2189
Joined: Mon Aug 22, 2011 1:11 am
Location: DC

Re: Due diligence on RAFI Pure Small Value

Post by Ketawa » Fri Nov 22, 2013 9:46 pm

I agree with everything you said, just curious why my alpha is so high but not Robert T's. Presumably he is also using F/F data for small, value, and momentum, so I'm guessing it's my lack of profitability.

Arnott seems well-respected by Larry too. Seriously considering putting some into PXSV.

User avatar
grap0013
Posts: 1884
Joined: Thu Mar 18, 2010 1:24 pm

Re: Due diligence on RAFI Pure Small Value

Post by grap0013 » Fri Nov 22, 2013 9:49 pm

Ketawa wrote:I agree with everything you said, just curious why my alpha is so high but not Robert T's. Presumably he is also using F/F data for small, value, and momentum, so I'm guessing it's my lack of profitability.

Arnott seems well-respected by Larry too. Seriously considering putting some into PXSV.
Anyone want to run the 7 negative years returns between the 4 indices on annualized basis? I think this could tip you over the edge. :)
There are no guarantees, only probabilities.

User avatar
Topic Author
Robert T
Posts: 2593
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Due diligence on RAFI Pure Small Value

Post by Robert T » Sat Nov 23, 2013 5:04 am

Ketawa wrote:I agree with everything you said, just curious why my alpha is so high but not Robert T's. Presumably he is also using F/F data for small, value, and momentum, so I'm guessing it's my lack of profitability.
Ketawa,

I get the same results on the alpha when momentum is added. Run the regression again without momentum added and you should get and alpha of 0.08 (and not statistically significant). So why it the momentum load so large and negative, and why does the alpha increase from 0.08 to 0.24 when you add momentum? Here are some possible explanations (just my views):
  • On the magnitude of the momentum load: In general value indexes have negative momentum loads. Over this period the correlation coefficient between HmL and Mom was -0.17. RAFI’s size screen already adds a value weight, e.g. if you sort by book value, as book value = market * (book-to-market), it automatically gives more weight to higher book-to-market stocks than does the overall market (a value tilt). This value weight is further amplified by the subsequent price-to-book sorts on the pure value series. The HmL tilt leads to a negative moment load – and perhaps it is larger than other cap weighted small value funds because of the value tilt already implicit in RAFI’s size sorts.

    On the alpha differences: On possible explanation is that RAFI’s ‘contra-trading’ approach (and perhaps also, in part, the way it sorts and weights stock holdings) offsets the negative moment effects. i.e. rather than trying to screen out negative momentum (which may act to reduce the value load), it uses its ‘contra-trading’ (rebalancing) and ‘economic footprint’ sorts to offset this negative effect (which may reduce the sampling error risks in momentum screens). Using the regression results: Alpha = 0.24%. Over this period the momentum ‘premium’ was 0.71% per month, and the negative momentum load on the RAFI fund was -0.18% - so exposure to negative momentum reduced the funds return by -0.13% per month (-0.18*0.71=-0.13) which leaves a residual ‘alpha’ of 0.11 (0.24-0.13), which is close to the alpha with no momentum added.
This is slightly different from the S&P600 Pure Value series where the alpha in the 3 factor model is -0.22 (compared to the positive alpha on the RAFI pure value series). The negative alpha seems to be explained by negative momentum (i.e. the S&P size sorts and rebalancing effects do not offset the negative momentum effects). For example when adding momentum to the model, the alpha on the S&P600 pure value series declined to zero (the momentum load = -0.31).

The ‘profitability’ loads on both the RAFI pure value series and the S&P600 pure value series are positive and significant, and is larger on the S&P600 pure value series (+0.36), than the RAFI series (+0.2) [this compares to about a +0.06 ‘profitability’ load on the DFA small value fund). This does not make much sense to me. The correlation coefficient between the RAFI series and the ‘profitability premium’ is zero, and between HmL and the ‘profitability premium’ its -0.27. So if anything one would expect a negative coefficient.

IMO, just using the three factor model gets you most of the way, is simple, and the relative alphas likely embody the effects of omitted variables. A close to zero alpha seems to suggest no omitted variables or simply that they cancel each other out.

Robert

PS:
Here is a paper of Jason Hsu (of RAFI) + others, on style indexes - http://www.jasonhsu.org/uploads/1/0/0/7 ... ndices.pdf
And if interested, he has some Fundamental Index series available on his website (US, Global-ex US, EM)
http://www.jasonhsu.org/research-data.html. And usefully, he also has Non-US and EM factor data (mkt, size, value, and momentum), which may be helpful in calculating factor loads on Non-US and EM funds (they are calculated using the same methodology as Fama-French US factors).

.

User avatar
grap0013
Posts: 1884
Joined: Thu Mar 18, 2010 1:24 pm

Re: Due diligence on RAFI Pure Small Value

Post by grap0013 » Sat Nov 23, 2013 9:47 am

Average negative year drop:

PXSV
-12.2%

DFA small value research
-13%

DFA targeted value
-13.2%

FF SCV ex. util
-14.7%

I'd also be curious about 5 and 10 year rolling returns between the 4 indices. My guess is PXSV has slightly more consistent long term returns.
There are no guarantees, only probabilities.

User avatar
grap0013
Posts: 1884
Joined: Thu Mar 18, 2010 1:24 pm

Re: Due diligence on RAFI Pure Small Value

Post by grap0013 » Sat Nov 23, 2013 10:02 am

Robert T wrote:
On the alpha differences: On possible explanation is that RAFI’s ‘contra-trading’ approach (and perhaps also, in part, the way it sorts and weights stock holdings) offsets the negative moment effects. i.e. rather than trying to screen out negative momentum (which may act to reduce the value load), it uses its ‘contra-trading’ (rebalancing) and ‘economic footprint’ sorts to offset this negative effect (which may reduce the sampling error risks in momentum screens). Using the regression results: Alpha = 0.24%. Over this period the momentum ‘premium’ was 0.71% per month, and the negative momentum load on the RAFI fund was -0.18% - so exposure to negative momentum reduced the funds return by -0.13% per month (-0.18*0.71=-0.13) which leaves a residual ‘alpha’ of 0.11 (0.24-0.13), which is close to the alpha with no momentum added.[/list]
This makes sense to me. When the S**t hits the fan, cheap/value companies that are bad (poor fundamentals, pardon the pun) go down more (ie. negative momentum). However, cheap/value high quality (RAFI) companies should not have the same negative momentum. And why should they? They are profitable, generate strong earnings, cash flow, etc...

It's almost like a GMO high quality + SCV strategy rolled into 1 fund.
There are no guarantees, only probabilities.

RNJ
Posts: 812
Joined: Mon Apr 08, 2013 9:06 am

Re: Due diligence on RAFI Pure Small Value

Post by RNJ » Sat Nov 23, 2013 2:42 pm

This has been an interesting and very helpful thread to follow. Thank you all for your work and generosity.

I've seen that Morningstar's tax efficiency ratings have received mixed reviews on the board. Other than the not-insubstantial risks that iShares will change the index or liquidate the fund due to low AUM, how might the ETF look from a tax perspective (I.e., is it appropriate for a taxable account)?

Thanks again.

EDIT: The Powershares site reports no cap gains distributions (short or long) going back to at least 2005.
Last edited by RNJ on Sat Nov 23, 2013 4:19 pm, edited 2 times in total.

User avatar
grap0013
Posts: 1884
Joined: Thu Mar 18, 2010 1:24 pm

Re: Due diligence on RAFI Pure Small Value

Post by grap0013 » Sat Nov 23, 2013 3:56 pm

RNJ wrote:This has been an interesting and very helpful thread to follow. Thank you all for your work and generosity.

I've seen that Morningstar's tax efficiency ratings have received mixed reviews on the board. Other than the not-insubstantial risks that iShares will change the index or liquidate the fund due to low AUM, how might the ETF look from a tax perspective (I.e., is it appropriate for a taxable account)?

Thanks again.
I believe PXSV had 100% qualified dividends last year. Current yield is about 1.5%. Very tax efficient. Only fear is Powershares closing the fund due to lack of AUM and/or expense ratio climbing due to the waiver drop. I previously had stated that I wouldn't pay 0.70% for this fund. However, now I'm not so sure. Heck, people pay 0.54% for the DFA fund (sometimes adviser fees too!) and I believe this ETF to be better.

Rick said wait until it gets to 100 AUM before taxable and that's probably smart. However, I guess it depends on how much we are talking about investing in it. For example: 50K or 5K. The former I'd be hesitant until AUM>100 million and the latter I'd have no problem.

Full disclosure: I liquidated my taxable account a couple years ago and haven't added to it since. Therefore, no PXSV in taxable yet, but am planning on it in the near future.
There are no guarantees, only probabilities.

vencat
Posts: 231
Joined: Thu Sep 10, 2009 6:30 pm

Re: Due diligence on RAFI Pure Small Value

Post by vencat » Sun Nov 24, 2013 12:32 pm

I'm not a quant but I had a couple of comments:
1) Despite criticism of RZV, its' total AUM now are 135 million and currently it's the top performing SV ETF. (Altruist Haas must finally feel vindicated :wink:) It may have some role added on to another SV fund such as Vanguard Small cap value.

2) Has anyone else noted the odd behavior of Vanguard Tax managed small cap fund. It has actually beaten DFSVX YTD, 3 and 10 years and IJS over all time periods! Is this a fluke or some mysterious momentum/profitability/low cost factor(s) at work?
I vaguely remember Robert T mentioning this fund favorably when there were less SV options.

Venkat

User avatar
BlueEars
Posts: 3779
Joined: Sat Mar 10, 2007 12:15 am
Location: West Coast

Re: Due diligence on RAFI Pure Small Value

Post by BlueEars » Sun Nov 24, 2013 5:35 pm

When I look at an M* chart of PXSV versus VBR, it doesn't jump out that I should consider a switch. From Robert's post the relative PXSV loadings look attractive but the 8 years of real data seem to show something I would not have thought. Both etf's went down together in the 2008 sell off. VBR recovered better it appears.

I've got a open mind on this, am I missing something here? Do we have to wait another decade to get the full story?

Image

User avatar
Ketawa
Posts: 2189
Joined: Mon Aug 22, 2011 1:11 am
Location: DC

Re: Due diligence on RAFI Pure Small Value

Post by Ketawa » Sun Nov 24, 2013 5:41 pm

PXSV changed indices a couple years ago.

User avatar
BlueEars
Posts: 3779
Joined: Sat Mar 10, 2007 12:15 am
Location: West Coast

Re: Due diligence on RAFI Pure Small Value

Post by BlueEars » Sun Nov 24, 2013 5:53 pm

Ketawa wrote:PXSV changed indices a couple years ago.
OK thanks, I did not read Robert's post carefully enough. I guess this means that real time data comparison is somewhat meaningless given only 2+ years of data.

exeunt
Posts: 907
Joined: Tue Jan 08, 2008 6:54 pm

Re: Due diligence on RAFI Pure Small Value

Post by exeunt » Sun Nov 24, 2013 11:37 pm

I wouldn't touch this fund. It can own really tiny, illiquid, micro-cap stocks, potentially creating adverse price impact during reconstitutions.

User avatar
Topic Author
Robert T
Posts: 2593
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Due diligence on RAFI Pure Small Value

Post by Robert T » Mon Nov 25, 2013 2:28 am

exeunt wrote:I wouldn't touch this fund. It can own really tiny, illiquid, micro-cap stocks, potentially creating adverse price impact during reconstitutions.
Yes, market impact is a concern with all smaller-cap funds. It is arguably a larger risk with S&P600 pure value (RZV) which weights stocks by value metric (within the small size bucket) leading to a smaller average market cap than RAFI fundamental pure value (PXSV) which weights stocks by fundamental size but sorted into 'value-growth' buckets (as I understand). This is reflected in the lower average market cap and higher size load of RZV than PXSV. You would expect some of the market impact effects to show up index tracking error, although some of it will be also reflected in the actual index returns, depending of the preciseness of the reconstitution timing of the fund. In any event the returns of RZV since inception have essentially been the index returns minus the expense ratio (6.32% vs. 6.76%, current expense ratio = 0.38%). In addition, the RAFI fundamental pure value index uses a liquidity and capacity screen. From the 'rule book':
Liquidity and Capacity Screens

RA assumes a certain notional amount is invested in each index and computes an implied holding for each security. The fundamental weight of each security is reduced, if necessary, to limit the implied amount of average daily trading volume free float.
Robert
.

Roy
Posts: 970
Joined: Wed Sep 10, 2008 9:34 am

Re: Due diligence on RAFI Pure Small Value

Post by Roy » Mon Nov 25, 2013 8:13 am

vencat wrote: 2) Has anyone else noted the odd behavior of Vanguard Tax managed small cap fund. It has actually beaten DFSVX YTD, 3 and 10 years and IJS over all time periods! Is this a fluke or some mysterious momentum/profitability/low cost factor(s) at work?
I vaguely remember Robert T mentioning this fund favorably when there were less SV options.
Venkat
Thanks, Vencat. I use Vanguard for SV exposure and don't have a compelling reason to switch, but I had not noticed this about Vanguard Tax managed small cap (VTMSX) and would like to examine it more. It's a blend fund, and is smaller and growthier than the SCV Vanguard funds (another "black hole" success story since inception?) and just a bit more expensive than VBR and with a higher minimum than VISVX (Inv. class but the same minimum as Admiral, VSIAX). But VTMSX has been here since 1999 only, so that's a consideration too.

RNJ
Posts: 812
Joined: Mon Apr 08, 2013 9:06 am

Re: Due diligence on RAFI Pure Small Value

Post by RNJ » Mon Nov 25, 2013 8:56 am

Roy wrote:
vencat wrote: 2) Has anyone else noted the odd behavior of Vanguard Tax managed small cap fund. It has actually beaten DFSVX YTD, 3 and 10 years and IJS over all time periods! Is this a fluke or some mysterious momentum/profitability/low cost factor(s) at work?
I vaguely remember Robert T mentioning this fund favorably when there were less SV options.
Venkat
Thanks, Vencat. I use Vanguard for SV exposure and don't have a compelling reason to switch, but I had not noticed this about Vanguard Tax managed small cap (VTMSX) and would like to examine it more. It's a blend fund, and is smaller and growthier than the SCV Vanguard funds (another "black hole" success story since inception?) and just a bit more expensive than VBR and with a higher minimum than VISVX (Inv. class but the same minimum as Admiral, VSIAX). But VTMSX has been here since 1999 only, so that's a consideration too.
According to M*, VTMSX has outperformed both IJS and VBR, and done so with a lower standard deviation. But choose your time parameters wisely.

Roy
Posts: 970
Joined: Wed Sep 10, 2008 9:34 am

Re: Due diligence on RAFI Pure Small Value

Post by Roy » Mon Nov 25, 2013 9:08 am

RNJ wrote: According to M*, VTMSX has outperformed both IJS and VBR, and done so with a lower standard deviation. But choose your time parameters wisely.
Yep, saw that too. One concern is the short time period for any of these. But I'll be thinking about it.

RNJ
Posts: 812
Joined: Mon Apr 08, 2013 9:06 am

Re: Due diligence on RAFI Pure Small Value

Post by RNJ » Mon Nov 25, 2013 9:36 am

Roy wrote:
RNJ wrote: According to M*, VTMSX has outperformed both IJS and VBR, and done so with a lower standard deviation. But choose your time parameters wisely.
Yep, saw that too. One concern is the short time period for any of these. But I'll be thinking about it.
For the sake of clarity, what I meant to imply was that winners and losers here are determined by where the starting and finish lines are placed. Look at the funds before 2009, since 2009, etc.

User avatar
Ketawa
Posts: 2189
Joined: Mon Aug 22, 2011 1:11 am
Location: DC

Re: Due diligence on RAFI Pure Small Value

Post by Ketawa » Mon Nov 25, 2013 11:13 am

I moved half my VBR allocation to PXSV. My domestic stock allocation is now split about equally between VBR, PXSV, and the S Fund.

The spread is about 2 or 3 cents and there were only 100 shares available at the ask. TD Ameritrade Level II quotes showed the next available bid was about 4 cents higher. I was buying more than 100 shares, so I put in a limit order at the ask. The entire order filled practically instantaneously in several transactions, so I suspect that whoever was offering 100 shares had their computers programmed to keep offering that many. I'm guessing if I used a market order, I would have ended up purchasing the majority of the shares at a higher price. Use a limit order!

User avatar
grap0013
Posts: 1884
Joined: Thu Mar 18, 2010 1:24 pm

Re: Due diligence on RAFI Pure Small Value

Post by grap0013 » Mon Nov 25, 2013 11:31 am

exeunt wrote:I wouldn't touch this fund. It can own really tiny, illiquid, micro-cap stocks, potentially creating adverse price impact during reconstitutions.
This actually could be advantageous. Just google Larry Swedroe and "liquidity premium".

Naysayers are always good for business. :wink: I will continue to make a lot of money with this fund inside my Roth. Just don't say I didn't tell you 20 years from now.
There are no guarantees, only probabilities.

caklim00
Posts: 1914
Joined: Mon May 26, 2008 10:09 am

Re: Due diligence on RAFI Pure Small Value

Post by caklim00 » Mon Nov 25, 2013 12:00 pm

I'm trying to get there...

But, when I go to the Powershares page for PXSV: http://www.invescopowershares.com/produ ... icker=PXSV
From 6/30/11 to 9/30/13 I still see better performance for S&P600V vs Fundamental Small Pure Value. Am I missing something here. At least the past 2 years it seems you would have been better off in S&P600V.

User avatar
grap0013
Posts: 1884
Joined: Thu Mar 18, 2010 1:24 pm

Re: Due diligence on RAFI Pure Small Value

Post by grap0013 » Mon Nov 25, 2013 12:09 pm

caklim00 wrote:I'm trying to get there...

But, when I go to the Powershares page for PXSV: http://www.invescopowershares.com/produ ... icker=PXSV
From 6/30/11 to 9/30/13 I still see better performance for S&P600V vs Fundamental Small Pure Value. Am I missing something here. At least the past 2 years it seems you would have been better off in S&P600V.
You are correct. From 6/16/11 - current:

VTI +50%
VBR +50.4%
DFSVX +55.7%
PXSV +59.3%
IJS +60%
RZV +62.7%

Not best in class....yet. However, VBR is not looking too good comparitively. Should be fun to see how it all unfolds.
There are no guarantees, only probabilities.

caklim00
Posts: 1914
Joined: Mon May 26, 2008 10:09 am

Re: Due diligence on RAFI Pure Small Value

Post by caklim00 » Mon Nov 25, 2013 1:49 pm

grap0013 wrote:
caklim00 wrote:I'm trying to get there...

But, when I go to the Powershares page for PXSV: http://www.invescopowershares.com/produ ... icker=PXSV
From 6/30/11 to 9/30/13 I still see better performance for S&P600V vs Fundamental Small Pure Value. Am I missing something here. At least the past 2 years it seems you would have been better off in S&P600V.
You are correct. From 6/16/11 - current:

VTI +50%
VBR +50.4%
DFSVX +55.7%
PXSV +59.3%
IJS +60%
RZV +62.7%

Not best in class....yet. However, VBR is not looking too good comparitively. Should be fun to see how it all unfolds.
Thanks. My biggest concern moving forward is with Powershares.

My US SCV is basically split between DFFVX (DFA US Targeted Value) and VBR right now (very small allocation to VIOV and RZV). DFFVX and PXSV have roughly the same ER. PXSV theoretically has slightly higher expected returns. But, DFA has a very long consistent track record. Plugging my no DFFVX allocation (old 401k can no longer contribute) to Powershares instead of Vanguard is slightly concerning... As you know I'm in with Schwab for FNDC/SFILX (RAFI Developed Small) so I suppose I've already made this decision on the Intl side.

User avatar
grap0013
Posts: 1884
Joined: Thu Mar 18, 2010 1:24 pm

Re: Due diligence on RAFI Pure Small Value

Post by grap0013 » Mon Nov 25, 2013 5:20 pm

caklim00 wrote: My US SCV is basically split between DFFVX (DFA US Targeted Value) and VBR right now (very small allocation to VIOV and RZV). DFFVX and PXSV have roughly the same ER. PXSV theoretically has slightly higher expected returns. But, DFA has a very long consistent track record. Plugging my no DFFVX allocation (old 401k can no longer contribute) to Powershares instead of Vanguard is slightly concerning... As you know I'm in with Schwab for FNDC/SFILX (RAFI Developed Small) so I suppose I've already made this decision on the Intl side.
Decisions...decisions...if tax deferred I think it is a pretty easy one. Taxable is another animal.

I'm not crazy about Powershares either, but I am confident they have to follow the law and cannot murder you in your sleep.
There are no guarantees, only probabilities.

User avatar
Topic Author
Robert T
Posts: 2593
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Due diligence on RAFI Pure Small Value

Post by Robert T » Tue Nov 26, 2013 3:33 am

caklim00 wrote:How low is the turnover on this index? Not sure looking at 1 year is going to be very helpful.
I don't know what the longer term turnover (trading costs) of the fund will be. But two interesting findings from RAFI research.

1. Using a 5 year average of fundamentals to weight stocks both reduced turnover and increased returns relative to shorter periods. See linked chart - moving from 1 year to 5 year averages of fundamentals reduced turnover, and raised returns. RAFI indexes use 5 year averages. They don't present the results for value metric averages.

Image

2. Less frequent rebalancing (annually vs. less) also helped to reduce turnover and raise returns (slightly). On the return side, this is a similar result I found with the FF data, annually rebalanced portfolios had higher returns than quarterly rebalanced portfolios. RAFI indexes are rebalanced annually.

Image

While we only have 1 year of live turnover data for PXSV of 31% for the year ending April 2013, this compares to 30% for the Powershares FTSE US 1500 fund (PRFZ). From 2009 to 2013 average turnover for PRFZ was 19%. I would expect PXSV turnover to be a bit higher than PRFZ.

No sure that this provides an exact answers, but provides some indicative information.

On Powershares vs. Schwab for US small cap (value).

From 2009-2011 (before Schwab changed over to the Russell indexes), Powershares and Schwab tracked the same index FTSE RAFI US Mid Small 1500 Index. Over this period, annualized (NAV) returns were:

Powershares (PRFZ) = 3.88%
Schwab (SFSNX) = 3.30%
Index = 4.00%

Robert
.

User avatar
grap0013
Posts: 1884
Joined: Thu Mar 18, 2010 1:24 pm

Re: Due diligence on RAFI Pure Small Value

Post by grap0013 » Tue Nov 26, 2013 8:11 am

Robert T wrote:
From 2009-2011 (before Schwab changed over to the Russell indexes), Powershares and Schwab tracked the same index FTSE RAFI US Mid Small 1500 Index. Over this period, annualized (NAV) returns were:

Powershares (PRFZ) = 3.88%
Schwab (SFSNX) = 3.30%
Index = 4.00%

Robert
Powershares may have better implementation? Securities lending?

Robert, do you have the numbers for SFILX/PDN?

Thanks!

grap
There are no guarantees, only probabilities.

User avatar
Topic Author
Robert T
Posts: 2593
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Due diligence on RAFI Pure Small Value

Post by Robert T » Tue Nov 26, 2013 10:21 am

grap0013 wrote:Robert, do you have the numbers for SFILX/PDN?

We only have three full calendar years (2009-2011) of implementation in which both funds tracked the same index - the FTSE RAFI Developed ex-US Mid Small 1500. This is a fairly short time period - but here are the results.

2009-2011

Powershares (PDN) = 17.06
Schwab (SFILX) = 15.29
Index = 17.79

The difference in PDN returns with the index was 0.74 percent over this period, and if I am not mistaken the expense ration was 0.75 (before being reduced to 0.49 in 2012).

Both were way (way) off the RAFI Emerging Market index, but I think this was more a problem of the actual index (not accounting for 'free-float') than Powershares/Schwab. I understand RAFI have been working on the issue of free-float in EM, introducing a new way to account for this is the index weights in 2011 or 2012. So there may be less tracking error with the EM index going forward, but time will tell.

Robert
.

User avatar
camontgo
Posts: 324
Joined: Tue Oct 27, 2009 1:46 pm
Location: Chicago, IL
Contact:

Re: Due diligence on RAFI Pure Small Value

Post by camontgo » Tue Nov 26, 2013 12:19 pm

The factor loadings and positive alpha look attractive, but there has been a lot written about other RAFI funds benefiting from the fortuitous timing of their index reconstitution in 2009.

Based on the table in the original post, 2009 looks extreme for this index as well. It would be interesting to look at time series of the FF3F residuals. How much is 2009 affecting the results? Just curious to know if the relative outperformance of this index can be attributed to a single home run or if it is the result of consistent base hits.
"Essentially, all models are wrong, but some are useful." - George E. P Box

caklim00
Posts: 1914
Joined: Mon May 26, 2008 10:09 am

Re: Due diligence on RAFI Pure Small Value

Post by caklim00 » Tue Nov 26, 2013 1:00 pm

Robert T wrote:
grap0013 wrote:Robert, do you have the numbers for SFILX/PDN?

We only have three full calendar years (2009-2011) of implementation in which both funds tracked the same index - the FTSE RAFI Developed ex-US Mid Small 1500. This is a fairly short time period - but here are the results.

2009-2011

Powershares (PDN) = 17.06
Schwab (SFILX) = 15.29
Index = 17.79

The difference in PDN returns with the index was 0.74 percent over this period, and if I am not mistaken the expense ration was 0.75 (before being reduced to 0.49 in 2012).

Both were way (way) off the RAFI Emerging Market index, but I think this was more a problem of the actual index (not accounting for 'free-float') than Powershares/Schwab. I understand RAFI have been working on the issue of free-float in EM, introducing a new way to account for this is the index weights in 2011 or 2012. So there may be less tracking error with the EM index going forward, but time will tell.

Robert
.
Delete out the first year and you should see different results. I bought into SFILX, but only after checking that Schwab had gotten their tracking error under control. It was really bad when they first opened the fund. I wonder if it has to do with running a mutual fund vs an ETF. Anyway, I didn't get the exact numbers but PDN and SFILX seemed to track each other extremely closely up to the Dividend Distribution time in late December.

User avatar
Topic Author
Robert T
Posts: 2593
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Due diligence on RAFI Pure Small Value

Post by Robert T » Tue Nov 26, 2013 1:11 pm

.
The factor loadings and positive alpha look attractive, but there has been a lot written about other RAFI funds benefiting from the fortuitous timing of their index reconstitution in 2009.

Based on the table in the original post, 2009 looks extreme for this index as well. It would be interesting to look at time series of the FF3F residuals. How much is 2009 affecting the results? Just curious to know if the relative outperformance of this index can be attributed to a single home run or if it is the result of consistent base hits.
I also looked at this earlier (on similar concerns). There seem to be two 'outlier' years 2001 (41% vs. 22% of FF/DFA series) and 2009 (73% vs. 35-45% for FF/DFA series). If we simply take out those two years - the annualized returns for the 1976-2012 period (ex. 2001 and 2009) becomes 16.7%, more in-line with the DFA series in the OP. Although without these years the DFA targeted value series returns also declines to 15.7 - so its not just outliers driving return differences.

Similar if we take out these two years from the factor analysis, the factor loads don't change much, although the alpha declined from 0.08 to -0.02 (but still close to zero).

1976-2012 (excluding 2001 and 2009).

Alpha=-0.02
Market = 1.01
Size = 0.81
Value = 0.72

Robert
.

User avatar
camontgo
Posts: 324
Joined: Tue Oct 27, 2009 1:46 pm
Location: Chicago, IL
Contact:

Re: Due diligence on RAFI Pure Small Value

Post by camontgo » Tue Nov 26, 2013 1:56 pm

Robert T,

Thanks for the additional analysis! Looks like the fund is worth watching as a potential SCV option.
"Essentially, all models are wrong, but some are useful." - George E. P Box

caklim00
Posts: 1914
Joined: Mon May 26, 2008 10:09 am

Re: Due diligence on RAFI Pure Small Value

Post by caklim00 » Tue Nov 26, 2013 3:16 pm

I've been watching the bid/ask spread for the past couple days since PXSV has piqued my interest.

At 2:36pm today
Bid: 24.22 x 1800
Ask: 24.31 x 1200

That's roughly a .40% cost on the spread. Not sure I'm ready to go round trip on this yet. That seems like a lot of eat on the front-end. Not sure what it is about Powershares, but their spreads always seem awful (especially PDN)...

For comparison, 3:14pm VBR is
Bid: 96.28 x 100
Ask: 96.30 x 500

User avatar
BlueEars
Posts: 3779
Joined: Sat Mar 10, 2007 12:15 am
Location: West Coast

Re: Due diligence on RAFI Pure Small Value

Post by BlueEars » Tue Nov 26, 2013 5:32 pm

You can check the 30 day average of bid-ask spreads for VG ETF's here: https://advisors.vanguard.com/VGApp/iip ... daskspread
It's 0.05 for VBR.

User avatar
Ketawa
Posts: 2189
Joined: Mon Aug 22, 2011 1:11 am
Location: DC

Re: Due diligence on RAFI Pure Small Value

Post by Ketawa » Wed Nov 27, 2013 3:10 am

caklim00 wrote:I've been watching the bid/ask spread for the past couple days since PXSV has piqued my interest.

At 2:36pm today
Bid: 24.22 x 1800
Ask: 24.31 x 1200

That's roughly a .40% cost on the spread. Not sure I'm ready to go round trip on this yet. That seems like a lot of eat on the front-end. Not sure what it is about Powershares, but their spreads always seem awful (especially PDN)...

For comparison, 3:14pm VBR is
Bid: 96.28 x 100
Ask: 96.30 x 500
I bought PXSV on Monday, and the spread was only 2-3 cents from 10:30 to 11:00 EST. However, there was a lot less available at the bid and ask. It was typically 200x100 or 100x200.

User avatar
grap0013
Posts: 1884
Joined: Thu Mar 18, 2010 1:24 pm

Re: Due diligence on RAFI Pure Small Value

Post by grap0013 » Wed Nov 27, 2013 8:33 am

Ketawa wrote:
caklim00 wrote:I've been watching the bid/ask spread for the past couple days since PXSV has piqued my interest.

At 2:36pm today
Bid: 24.22 x 1800
Ask: 24.31 x 1200

That's roughly a .40% cost on the spread. Not sure I'm ready to go round trip on this yet. That seems like a lot of eat on the front-end. Not sure what it is about Powershares, but their spreads always seem awful (especially PDN)...

For comparison, 3:14pm VBR is
Bid: 96.28 x 100
Ask: 96.30 x 500
I bought PXSV on Monday, and the spread was only 2-3 cents from 10:30 to 11:00 EST. However, there was a lot less available at the bid and ask. It was typically 200x100 or 100x200.
Nice work! That's why AUM has increased 0.76 = 760K since the OP. :D
There are no guarantees, only probabilities.

caklim00
Posts: 1914
Joined: Mon May 26, 2008 10:09 am

Re: Due diligence on RAFI Pure Small Value

Post by caklim00 » Wed Nov 27, 2013 11:22 am

Ketawa wrote:
caklim00 wrote:I've been watching the bid/ask spread for the past couple days since PXSV has piqued my interest.

At 2:36pm today
Bid: 24.22 x 1800
Ask: 24.31 x 1200

That's roughly a .40% cost on the spread. Not sure I'm ready to go round trip on this yet. That seems like a lot of eat on the front-end. Not sure what it is about Powershares, but their spreads always seem awful (especially PDN)...

For comparison, 3:14pm VBR is
Bid: 96.28 x 100
Ask: 96.30 x 500
I bought PXSV on Monday, and the spread was only 2-3 cents from 10:30 to 11:00 EST. However, there was a lot less available at the bid and ask. It was typically 200x100 or 100x200.
Hmmm, wonder if it has anything to do with Thanksgiving coming up, but the lowest I've seen it the past 2 days is 7 cents.

User avatar
abuss368
Posts: 15639
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!

Re: Due diligence on RAFI Pure Small Value

Post by abuss368 » Wed Nov 27, 2013 11:45 am

I have no clue. This is why I prefer the "Total" markets appraoch.
John C. Bogle: "Simplicity is the master key to financial success."

User avatar
grap0013
Posts: 1884
Joined: Thu Mar 18, 2010 1:24 pm

Re: Due diligence on RAFI Pure Small Value

Post by grap0013 » Wed Nov 27, 2013 11:55 am

abuss368 wrote:I have no clue. This is why I prefer the "Total" markets appraoch.
Just had to get that in there didn't ya! :D

I do have a clue that's why I tilt heavily. Happy Thanksgiving abuss!
There are no guarantees, only probabilities.

User avatar
abuss368
Posts: 15639
Joined: Mon Aug 03, 2009 2:33 pm
Location: Where the water is warm, the drinks are cold, and I don't know the names of the players!

Re: Due diligence on RAFI Pure Small Value

Post by abuss368 » Wed Nov 27, 2013 12:12 pm

grap0013 wrote:
abuss368 wrote:I have no clue. This is why I prefer the "Total" markets appraoch.
Just had to get that in there didn't ya! :D

I do have a clue that's why I tilt heavily. Happy Thanksgiving abuss!
That was good!

Happy Thanksgiving!
John C. Bogle: "Simplicity is the master key to financial success."

michaelsieg
Posts: 570
Joined: Mon Jan 07, 2013 11:02 am

Re: Due diligence on RAFI Pure Small Value

Post by michaelsieg » Wed Nov 27, 2013 5:02 pm

Happy Thanksgivings to everyone, what a great thread!
Still trying desperately to find a good SV option for my 403b TD Ameritrade brokerage that does not allow me to buy ETFs, only mutual funds. The Schwab option just does not look that good. TDA lists the DFA funds, I might just try that route, but I am pretty sure they will not execute my order as I have no FA....
Thanks everyone for posting.
Happy Thanksgivings
Michael

Post Reply