Small Cap Tilting

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.

Do you tilt towards Small Growth, Value, or Blend?

Small Growth
6
4%
Small Blend
32
22%
Small Value
100
68%
No SC tilt, but Mid Cap tilt
10
7%
 
Total votes: 148

FillorKill
Posts: 1007
Joined: Sat Aug 06, 2011 7:01 am

Re: Small Cap Tilting

Post by FillorKill »

The OP is considering a size tilt. Maybe I’m having a middle aged moment but I didn’t notice a single mention of Vanguard’s Tax Managed Small Cap fund VTMSX in this thread. Especially for a taxable account, a VG investor that has 10K, wants a blend fund (but not an ETF) to tilt to SmB and wants a considerably greater SmB loading than VSMAX, this fund is worth consideration. The superior tax efficiency is a nice feature as well when comparing options.

Not really the point but as some issue has been made of the relative performance of VISGX, NAESX and VISVX, note that since inception in 1999 VTMSX it outperforms the other three choices easily.
pingo
Posts: 2638
Joined: Sat Sep 19, 2009 8:24 pm

Re: Small Cap Tilting

Post by pingo »

I don't have a dog in the Small Cap tilt vs. not debate, nor in the Small Growth vs. Small Value vs. Small Blend debate, but it occurred to me to simply go to Morningstar.com, pull up a chart and play around with the scrubber bar/time adjustment bar thingy below to see how each performed over varying, random periods. Here's some of what I saw:

Image

Image

Image

Image

Image

Image

Image

Image

Image

Image

The only thing I can see with me eyeball is something I seem to recall reading in Common Sense on Mutual Funds: that small cap value appears to have offered better risk-adjusted returns and with small blend appears to have done so as well. My memory may be incorrect (I don't have a copy of the book on hand) and I certainly don't know how to look at data to judge, but someone who made a lump sum investment in Small Value in 1974 would still be waiting 40 years later for Small Value to outperform Small Growth, etc. It looks like a reversal of fortunes happened circa 2005, so an investor starting in 1994 wouldn't have been as disappointed. But who knows the right time to invest in an asset class so that they can see a benefit in their lifetime? Small Blend seems to have been the best bet on a total return and risk-adjusted basis, if only Vanguard's index fund had been able to track it!

I suppose what might not be viewable to the eyeball is whether some rebalancing bonus might have existed to bring a portfolio with Small Value tilting ahead of…well what do we compare with it? But that's rarely something I see addressed, except in Larry Swedroe's articles about portfolio construction where including Lg, LV, Sm, SV with bonds beat the S&P 500 with bonds, but I don't know if it applies.

I don't know how long the VG Small Blend has been around (probably since circa 1974), but I think it makes a case against being an early adopter of new funds/ETFs that propose to give investors access to a theoretically important asset class.
Last edited by pingo on Thu Mar 06, 2014 1:46 pm, edited 1 time in total.
Topic Author
Paden
Posts: 19
Joined: Tue Feb 11, 2014 11:00 am

Re: Small Cap Tilting

Post by Paden »

BBL wrote:The OP is considering a size tilt. Maybe I’m having a middle aged moment but I didn’t notice a single mention of Vanguard’s Tax Managed Small Cap fund VTMSX in this thread. Especially for a taxable account, a VG investor that has 10K, wants a blend fund (but not an ETF) to tilt to SmB and wants a considerably greater SmB loading than VSMAX, this fund is worth consideration. The superior tax efficiency is a nice feature as well when comparing options.

Not really the point but as some issue has been made of the relative performance of VISGX, NAESX and VISVX, note that since inception in 1999 VTMSX it outperforms the other three choices easily.

Wow, I did not consider checking the M* box on VTMSX. I had no idea it would have a much stronger tilt than VSMAX. I feel like the tax managed funds don't get talked about much around here. Is there a reason for that? Is it the higher ER? What exactly makes them more tax efficient than the non tax-managed funds?
FillorKill
Posts: 1007
Joined: Sat Aug 06, 2011 7:01 am

Re: Small Cap Tilting

Post by FillorKill »

Paden wrote:
Wow, I did not consider checking the M* box on VTMSX. I had no idea it would have a much stronger tilt than VSMAX. I feel like the tax managed funds don't get talked about much around here. Is there a reason for that? Is it the higher ER? What exactly makes them more tax efficient than the non tax-managed funds?
Yes the size loading is much better because of the index this fund targets (S&P 600) which differs from the other three VG small index funds.

The tax efficiency is a product of fund management. In Vanguards words: "The advisor uses portfolio optimization techniques to select a sample of stocks that, in aggregate, reflect the characteristics of the benchmark index. Vanguard uses proprietary software to implement trading decisions that accommodate cash flow and maintain close correlation with index characteristics. In addition, a disciplined sell process minimizes the realization of net capital gains and may include the realization of losses to offset unavoidable gains. The experience and stability of Vanguard’s Equity Investment Group have permitted continuous refinement of indexing techniques designed to minimize tracking error and provide tax-efficient returns".

Why are they often overlooked? I have no idea...
cb474
Posts: 905
Joined: Tue Jan 19, 2010 5:32 am

Re: Small Cap Tilting

Post by cb474 »

BBL wrote:The OP is considering a size tilt. Maybe I’m having a middle aged moment but I didn’t notice a single mention of Vanguard’s Tax Managed Small Cap fund VTMSX in this thread. Especially for a taxable account, a VG investor that has 10K, wants a blend fund (but not an ETF) to tilt to SmB and wants a considerably greater SmB loading than VSMAX, this fund is worth consideration. The superior tax efficiency is a nice feature as well when comparing options.

Not really the point but as some issue has been made of the relative performance of VISGX, NAESX and VISVX, note that since inception in 1999 VTMSX it outperforms the other three choices easily.
That is an interesting suggestion, if one is looking only for a small (and not value) tilt and wants to use a mutual fund, rather than an ETF. I suspect the reason VTMSX doesn't come up much is because most tilters (perhaps especially the vocal ones in the forum) are both size and value tilters. So if they want a fund based on the S&P 600, they choose iShares IJS (and more recently Vanguard's VIOV), the S&P 600 value fund. IJS does come up a lot and is probably even the most preferred small fund, other than VISVX/VSIAX/VBR.

People may be surprised how much smaller VTMSX is than NAESX/VSMAX/VB, but that very same surprise at the relative smallness of the S&P 600 comes up all the time in discussions of IJS versus VBR.

What's interesting, is to look at the factor loads of VTMSX in comparison not just to VSMAX, but also to VBR. It looks like VTMSX has a slight value tilt, which varies depending on the time period, and is about half the value tilt of IJS. Although if you look at it since inception, it looks like it has almost as much value tilt as the current CRSP version of VBR (VTMSX = .38 HmL, VBR = .48 HmL). I don't know, this is where I get confused about how to compare funds based on factor loads, given that they vary over time and different funds have different inception dates (plus the added complication that VBR's index has changed from MSCI to CRSP--and once was Russell, I think). Perhaps I'll start a separate thread to ask this question about these funds. [Edit: Started the thread here: http://www.bogleheads.org/forum/viewtop ... 0&t=134612. Hopefully some of the factor load proficient bogleheads will respond.] Still, it's a good example of how you really can't compare funds well using the style box charts.

On the other hand, if you look at the Morningstar's growth chart for VTMSX and compare it to VISVX and IJS (starting from the inception date of IJS, since it has not been around as long as VTMSX and VISVX), what you see is that IJS and VISVX (the small cap value funds) perfrom very similarly to each other (despite the different size loads) and do outperform VTMSX by varying margins depending on the time period, contrary to what BBL says above. A good example of how you can make the charts say anything, depending on the start date one chooses, and are not a great way to compare funds either.

I'd say, if all one wants is a size tilt, there's an argument for VTMSX over NAESX/VSMAX/VB. But I wouln't assume that there is no potential benefit to Vanguard's small cap value fund (VBR/VISVX/VSIAX) or to iShares IJS. Styleboxes and back testing returns with Morningstar charts are not the way to judge these funds.

Also, let's not forget that everyone gets obsessed with returns, but part of the whole point of small and value (if you buy the arguments for them) is that they correlate less to large and blend funds. This gives one potentially better risk adjusted returns over the long haul. So in forgoing the value tilt, one may forgo benefits other than the potential extra return.
cb474
Posts: 905
Joined: Tue Jan 19, 2010 5:32 am

Re: Small Cap Tilting

Post by cb474 »

Found this thread comparins VTMSX and VSMAX/NAESX:

http://www.bogleheads.org/forum/viewtop ... 1&t=114788

A couple interesting points made in posts by nisiprius. First, VTMSX seems to have had a burst of outperformance soon after it's inception, but in subsequent years the two have tracked very closely (more evidence of how what Morningstar charts appear to show all depends on the point in time at which one starts and ends them). Second, although VTMSX in the last decade has reduced the amount of taxes one pays, the overall after tax returns for NAESX were slightly better (despite even the fact that this is the investor shares version with a higher expense ratio). Differences may just be noise and expecting to see much of a difference in the performance of the two funds, going forward, may be unfounded.

Also note that Vanguard now has an ETF that tracks the S&P 600 (VIOO) and of course there's an iShares version too. So one can track that index, without the tax managment strategies. And there are supposedly potential tax benefits to an ETF over a mutual fund making the likelihood of any differences pretty minimal.

The last post of the thread also points out that NAESX has had five different incarnations, starting as an active fund in 1960, becoming a Vanguard active fund in 1989, later tracking the Russell 2000, then the MSCI, now the CRSP. So, as I've been saying, the history of its past returns in a bit of an apples and oranges phenomenon.
FillorKill
Posts: 1007
Joined: Sat Aug 06, 2011 7:01 am

Re: Small Cap Tilting

Post by FillorKill »

cb474 wrote:
BBL wrote:
Not really the point but as some issue has been made of the relative performance of VISGX, NAESX and VISVX, note that since inception in 1999 VTMSX it outperforms the other three choices easily.
On the other hand, if you look at the Morningstar's growth chart for VTMSX and compare it to VISVX and IJS (starting from the inception date of IJS, since it has not been around as long as VTMSX and VISVX), what you see is that IJS and VISVX (the small cap value funds) perfrom very similarly to each other (despite the different size loads) and do outperform VTMSX by varying margins depending on the time period, contrary to what BBL says above. A good example of how you can make the charts say anything, depending on the start date one chooses, and are not a great way to compare funds either.
cb474 - good post, I mostly agree with what you have to say, however, I was making a very specific point about the comparison of the two blend funds and was leaving the value/growth element out of it. That is why I said the comparison including VISGX & VISVX wasn't the point. However, what I wrote and what you suggest I wrote are not the same thing.


What I said was that since inception of VTMSX, it outperforms the other 3. That is not in question (03/25/1999 – 03/06/2014):

VTMSX: 50,495.25
VISVX: 47,933.89
VISGX: 46,986.91
NAESX: 42,498.66

Comparing a small blend fund (VTMSX) to anything other than a small blend fund VSMAX (NAESX) isn’t a good idea due to the differing strategies. I can certainly find periods where either VISVX or VISGX (or NAESX is outperforming VTMSX) are outperforming the rest of the group but that isn’t a relevant comparison. If you only want to tilt to SmB then why would you use a fund that has an additional undesired tilt? VTMSX gives you much more SmB bang for your buck than VSMAX/NAESX, so even if VSMAX/NAESX happened to outperform over some period of time that is nothing but a classic strategy/outcome issue.
FillorKill
Posts: 1007
Joined: Sat Aug 06, 2011 7:01 am

Re: Small Cap Tilting

Post by FillorKill »

cb474 wrote: The last post of the thread also points out that NAESX has had five different incarnations, starting as an active fund in 1960, becoming a Vanguard active fund in 1989, later tracking the Russell 2000, then the MSCI, now the CRSP. So, as I've been saying, the history of its past returns in a bit of an apples and oranges phenomenon.
As much as I love Vanguard and I take them as they are, this point you're making here is, to me, a red flag against NAESX/VSMAX. When the fund company keeps changing targets/strategies the buy and hold investor gets the results of the ongoing strategy changes either way - good or bad. So the comparision including those strategy changes is fully relevant - that is the reality for the investor. The fact that the headline VG index funds keep changing targets is an additional unnecessary, unknowable, uncontrollable variable added into the equation. What makes anyone think this is over? I thought it was after the MSCI change but I was wrong. I consider this an unwelcome additional risk.
The Wizard
Posts: 13356
Joined: Tue Mar 23, 2010 1:45 pm
Location: Reading, MA

Re: Small Cap Tilting

Post by The Wizard »

BBL wrote:
As much as I love Vanguard and I take them as they are, this point you're making here is, to me, a red flag against NAESX/VSMAX. When the fund company keeps changing targets/strategies the buy and hold investor gets the results of the ongoing strategy changes either way - good or bad. So the comparision including those strategy changes is fully relevant - that is the reality for the investor. The fact that the headline VG index funds keep changing targets is an additional unnecessary, unknowable, uncontrollable variable added into the equation. What makes anyone think this is over? I thought it was after the MSCI change but I was wrong. I consider this an unwelcome additional risk.
I'm not sure why a tracking-index change would be an additional risk. VG probably puts out a statement giving the rationale, so you'd have to review the facts to see if it matters to you.
There's probably a transition period where they are selling out existing company stocks and buying new ones that are needed for the new index, so you could have higher than usual CGDs, I suppose...
Attempted new signature...
FillorKill
Posts: 1007
Joined: Sat Aug 06, 2011 7:01 am

Re: Small Cap Tilting

Post by FillorKill »

The Wizard wrote: I'm not sure why a tracking-index change would be an additional risk
Different targets have different factor exposures and a mid course change from one to the other will result in factor exposure changes for the index investor and that adds an element of luck into the equation from the transitional period forward. It also requires an investor that targets specific exposures to make tactical adjustments to maintain their strategy and doing this within your existing constraints isn't always so simple.
cb474
Posts: 905
Joined: Tue Jan 19, 2010 5:32 am

Re: Small Cap Tilting

Post by cb474 »

BBL wrote:cb474 - good post, I mostly agree with what you have to say, however, I was making a very specific point about the comparison of the two blend funds and was leaving the value/growth element out of it. That is why I said the comparison including VISGX & VISVX wasn't the point. However, what I wrote and what you suggest I wrote are not the same thing.


What I said was that since inception of VTMSX, it outperforms the other 3. That is not in question (03/25/1999 – 03/06/2014):

VTMSX: 50,495.25
VISVX: 47,933.89
VISGX: 46,986.91
NAESX: 42,498.66

Comparing a small blend fund (VTMSX) to anything other than a small blend fund VSMAX (NAESX) isn’t a good idea due to the differing strategies. I can certainly find periods where either VISVX or VISGX (or NAESX is outperforming VTMSX) are outperforming the rest of the group but that isn’t a relevant comparison. If you only want to tilt to SmB then why would you use a fund that has an additional undesired tilt? VTMSX gives you much more SmB bang for your buck than VSMAX/NAESX, so even if VSMAX/NAESX happened to outperform over some period of time that is nothing but a classic strategy/outcome issue.
Yes, I think I understood your point, that VTMSX tracks the S&P 600 and gives one more of a small tilt than NAESX, so it's an interesting option that's not often mentioned in the forum.

I made the point about VISVX and VIGSX outperforming VTMSX during different time periods, because you had actually comapred the returns of these funds in the post I was responding to and you indicated that since the inception of VTMSX it outperformed VISVX and VIGSX, in addition to NAESX. (You also said it outperforms these funds "easily," which is only true of NAESX during the entire time period and only barely true of VISVX starting in about 2011.) So my point was that choosing the inception date of VTMSX as the point from which to make comparisons (and implicitly the present as the relevant end point) is entirely arbitrary and doesn't mean anything. Yes, your statement about outperformance was mostly true, but I was questioning what the value of the information is. In fact, I would argue the same thing even in comparison to NAESX, which as you note is the other small blend fund; start in 2003 and NAESX outperforms. I guess, unless one plans to go back in a time machine and invest in VTMSX on the date of it's inception and then withdrawn ones investements today, I don't see how that information is meaningful.

As you say, the real point is that VTMSX has more SmB than the other funds in question and if all one wants to do is tilt to SmB, as was the OP's initial question, then VTMSX is an interesting choice. So I agree with you there. (I'm less convinced that the tax management strategies of VTMSX are effective or worthwhile.)

But I was also trying to say, when one looks at factor loads, NAESX may not even be the relevant comparison to VTMSX. If you look at the factor loads for VTMSX, starting from its inception date, it has a HmL of .38. That makes it seem a lot more like a value fund like VISVX. So one can't rely on the name of the fund (or perhaps even its stated strategy) to decide what to compare it to. In fact, if you look at the factor loads for NAESX from the same date, it has more SmB than VTMSX. And the fund that looks like the most truly blend fund in this period, with the smallest HmL of .11 (thought still a value tilt!) is VISGX (it also has a slightly higher but basically equivalent SmB as VTMSX over this period). So perhaps VTMSX's outperformance was partly due to its value tilt? And VTMSX is also not obviously the fund with the most SmB (only VISVX has less in this time period).

Of course, at this point in time the factor loads for these funds don't work out in the same way. And as I said, this is where I get a little bit confused about how to use factor loads to compare funds, given that they vary over time. But my general point was, it is not so obvious what fund is the relevant point of comparison for VTMSX.

Then one has to throw on top of that the tax management strategies of VTMSX and whether they were effective. Is it relevant to compare a tax managed fund to one that isn't? Perhaps the better comparison is VIOO, the Vanguard's S&P 600 ETF.

So, I entirely agree with you that VTMSX is an interesting choice that seems to get overlooked. And I agree that right now it has more SmB than NAESX, so perhaps it's a better choice for small tilt. But I also think there are a lot of apples and oranges and kumquats here and how to make comparisons is complicated. What should we expect going foward? Will VTMSX over extended future time periods continue to have a better SmB than NAESX (or VISVX or VISGX)? Will it continue to be significantly different from VISVX in terms of HmL, even though one is supposed to be a blend fund and the other a value fund? I'm not trying to quibble, I honestly don't know and it seems messy and hard to say.

But given all of that I do really think looking at Morningstar charts over any particular time period (unless one has thirty of forty years of returns) is probably pretty arbitrary. Especially with extremely similar funds. I hope that makes more sense of where I was coming from.

*

All that aside, I do agree with you that it's annoying that Vanguard keeps changing indices and that this throws in an additional element of unpredictability. Maybe it wouldn't be practical, but I prefer it if Vanguard contructed funds around specific indices (in the way that VIOV/IJS are designed specifically to track the S&P 600 Value). If they want to introduce a new fund that tracks the CRSP, fine. Then let customers decide if they want to switch to that fund.
User avatar
nedsaid
Posts: 19275
Joined: Fri Nov 23, 2012 11:33 am

Re: Small Cap Tilting

Post by nedsaid »

I recommend small cap tilting. It might be as simple as pairing the US Total Stock Market Index with the US Extended Market Index. The reason is different than what most would think. Yes, I think Small Cap Tilting will enhance your performance over long periods of time. I tilt towards the mid-cap/small cap stocks. The main reason is that the top 100 stocks by market capitalization make up just over 50% of the US Total Stock Market Index. A small-cap tilt dilutes the effect of those 100 stocks. The US Total Market Index is pretty much a Large Cap Index.

Tilt but don't overdo it. I shoot for 60% large caps vs. 40% mid-cap/small-caps. I gently tilt towards value. The Total Market is about 70% large caps vs. 30% mid-cap/small-caps.
A fool and his money are good for business.
FillorKill
Posts: 1007
Joined: Sat Aug 06, 2011 7:01 am

Re: Small Cap Tilting

Post by FillorKill »

The investor that is targeting particular factor loads for their porfolio, within practical and individual constraints, would be wise to consider VTMSX as an instrument that can give them exposure to the smaller deciles that VSMAX doesn't. The investor may even use a combination of the two depending on what they consider appropriate exposure.

The two primary reasons that VTMSX remains attractive is that many VG investors simply do not want to deal with ETFs and the tax efficiency is attractive. The investor that doesn't want to deal with ETFs and requires meaningful exposure to smaller deciles that they don't get with the portfolio composition of VSMAX should consider the fund. The investor that will use ETFs obviously has many more options available within VG.
cb474 wrote:So my point was that choosing the inception date of VTMSX as the point from which to make comparisons (and implicitly the present as the relevant end point) is entirely arbitrary and doesn't mean anything.
I've repeatedly said that the return comparison does not matter. Decompose the returns and you simply see the results of the factor exposures, most of which will be obvious, and likely a little influence from the patient trading that tax management requires. However, in any comparative analysis, relevant or not, the appropriate amount of data to use is the entire universe of relevant data. In this case that would be from inception of VTMSX forward. Calling that data set arbitrary and meaningless is a pretty strong statement. Are you sure that is an appropriate characterization?
YDNAL
Posts: 13774
Joined: Tue Apr 10, 2007 4:04 pm
Location: Biscayne Bay

Re: Small Cap Tilting

Post by YDNAL »

BBL wrote:
cb474 wrote: wrote:So my point was that choosing the inception date of VTMSX as the point from which to make comparisons (and implicitly the present as the relevant end point) is entirely arbitrary and doesn't mean anything.

I've repeatedly said that the return comparison does not matter. Decompose the returns and you simply see the results of the factor exposures, most of which will be obvious, and likely a little influence from the patient trading that tax management requires.
Returns matter since we don't invest - at least no one I know - just to practice how to run factor loadings. I also don't see a glaring difference in return. VTMSX has taken less of a beating in declines, but lower increase in a rebound (2002/03, 2008/09, 2011/12).

Code: Select all

Total Returns		
YEAR  	VTMSX	  VSMAX
2013	  41.00%	 37.81%
2012	  16.03%	 18.24%
2011	   1.22%	 -2.69%
2010	  25.99%	 27.89%
2009	  25.59%	 36.33%
2008	 -30.82%	-36.00%
2007	   0.51%	  1.24%
2006	  14.15%	 15.77%
2005	   7.74%	  7.49%
2004	  22.84%	 20.02%
2003	  38.51%	 45.76%
2002	 -14.44%	-19.95%
2001	   5.44%	  3.17%
2000**	13.44%	  1.75%
1999**	26.28%	

** VSMAX inception 11/13/2000		
VTMSX inception 3/25/1999
Landy | Be yourself, everyone else is already taken -- Oscar Wilde
cb474
Posts: 905
Joined: Tue Jan 19, 2010 5:32 am

Re: Small Cap Tilting

Post by cb474 »

BBL wrote:
cb474 wrote:So my point was that choosing the inception date of VTMSX as the point from which to make comparisons (and implicitly the present as the relevant end point) is entirely arbitrary and doesn't mean anything.
I've repeatedly said that the return comparison does not matter. Decompose the returns and you simply see the results of the factor exposures, most of which will be obvious, and likely a little influence from the patient trading that tax management requires. However, in any comparative analysis, relevant or not, the appropriate amount of data to use is the entire universe of relevant data. In this case that would be from inception of VTMSX forward. Calling that data set arbitrary and meaningless is a pretty strong statement. Are you sure that is an appropriate characterization?
I do honestly think the inception date is an arbitrary point of measurement. As I said, I'm not trying to quibble. Yes, you're right that it gives one the most possible information about past returns for VTMSX. But there's nothing special about that date, vis-a-vis the S&P 600. It's just the point in time one company, Vanguard, happened to start a fund based on that index. Further, as a starting point to compare to other funds based on other indices, it's just luck of the draw (NAESX, after all, if that's what we're comparing it to, has a much longer history that is ignored by selecting the inception date of VTMSX as a point of comparison). I think it's pretty commonly pointed out in this forum that the starting point, from which one measures returns, hugely effects the outcome one perceives.

A better reason, if there is one, to expect VTMSX to outperform NAESX is because of the greater factor load of SmB (as you point out), not because of the outperformance of VTMSX over what is essentially an arbitrary period of time. That being said, I'm not convinced that fifteen years is enough time over which to expect to see the consequences of differing factor loads show up. There are lots of long periods of underperformance. So factor loads may not explain what we see since the inception of VTMSX. If we had thirty, forty, fifty years for VTMSX, then I think one could start to more confidently say it's the factor loads.

Indeed, as I pointed out above, but to which you did not respond, if we take the inception date of VTMSX as a starting point and check the factor loads over that time period to the present (I'm using the PortfolioVisualizer website) NAESX has greater SmB over that time period. NAESX = .74 SmB, VTMSX = .68 SmB. So the supposed greater smallness of VTMSX does not seem to explain the outperformance over that period. On the other hand, VTMSX has a higher HmL over that period. VTMSX = .38 HmL, NAESX = .25 HmL. That shows VTMSX was capturing a fair amout of the small value premium over that period, a little more than NAESX, and if one is going to point to factor loads it seems one might conclude that it's the value permium that accounts for VTMSX's outperformance.

So I think the story about these funds is not as simple as one wants them to be. It's not even obvious which one will capture more of the small premium over time, given that factor loads vary and are not a static, set in stone, thing.

I do agree with you, however, that if one wants an S&P 600 fund and one doesn't want to fuss with ETFs, then VTMSX is an often overlooked option at Vanguard to get into the S&P 600 using a mutual fund. As I said initially, I imagine the reason that VTMSX is overlooked is because tilters who want the S&P 600 usually also want the value premium, don't mind an ETF, and go for IJS/VIOV instead. It also seems, from the other thread that I pointed to above, that the benefit of the tax management strategies with VTMSX does not clearly show up in the after tax returns in a signficant way (as compared to NAESX). So perhaps that's another reason it does not attract the more sophisticated investors, who are even thinking about things like the S&P 600 to begin with and who view ETFs as having their own tax advantages.
MapleHermit
Posts: 68
Joined: Sat Oct 26, 2013 2:37 am

Re: Small Cap Tilting

Post by MapleHermit »

When you see everyone tilting to small cap value, you know it is over valued. The main reason value stocks have done well in the past were because their price was cheaper than their real worth. However, the price of small cap value stocks is being driven up by those making that "historical bet". Warren Buffet made his fortune buying cheap stocks but today, his strategy isn't really working as well. You can visually see that the market is slowly becoming more and more efficient as an inefficiency becomes more well known.

I'd personally recommend mid-caps because they were able to stay on par with small cap value's historical results and they are extremely unpopular compared to small caps. Beating the market means going against the market and finding a secret, rather than following the general consensus. Can we not agree that most people think small cap value will outperform? Therefore, the chances of it outperforming are greatly reduced.
animule
Posts: 151
Joined: Sun Jan 06, 2008 11:29 am

Re: Small Cap Tilting

Post by animule »

VTMSX is different in one important regard.

It is based on the S&P 600 which requires a company to have at least 3 years of profitable results.

Because of this the fund did not load up on dot com garbage during the tech boom.

This offered some downside protection vs other small cap indexes.
User avatar
grabiner
Advisory Board
Posts: 35307
Joined: Tue Feb 20, 2007 10:58 pm
Location: Columbia, MD

Re: Small Cap Tilting

Post by grabiner »

BBL wrote:
Paden wrote:
Wow, I did not consider checking the M* box on VTMSX. I had no idea it would have a much stronger tilt than VSMAX. I feel like the tax managed funds don't get talked about much around here. Is there a reason for that? Is it the higher ER? What exactly makes them more tax efficient than the non tax-managed funds?
Yes the size loading is much better because of the index this fund targets (S&P 600) which differs from the other three VG small index funds.

The tax efficiency is a product of fund management. In Vanguards words: "The advisor uses portfolio optimization techniques to select a sample of stocks that, in aggregate, reflect the characteristics of the benchmark index. Vanguard uses proprietary software to implement trading decisions that accommodate cash flow and maintain close correlation with index characteristics. In addition, a disciplined sell process minimizes the realization of net capital gains and may include the realization of losses to offset unavoidable gains. The experience and stability of Vanguard’s Equity Investment Group have permitted continuous refinement of indexing techniques designed to minimize tracking error and provide tax-efficient returns".
With the development of ETFs, this is less important than it used to be, but TM Small-Cap still has the advantage of 100% qualified dividends, which none of Vanguard's other small-cap index funds and ETFs have. The Vanguard S&P 600 ETF, which tracks the same index, has about 80% qualified dividends.

Thus, if you want to hold a separate small-cap blend fund in taxable, TM Small-Cap is the natural choice.
Wiki David Grabiner
thisismyusername123
Posts: 95
Joined: Tue Nov 05, 2013 7:52 am

Re: Small Cap Tilting

Post by thisismyusername123 »

I am reallocating and have decided to allocate domestic equities (60% of portfolio) as follows: 2/3 total market, with rest split between FSSVX (small cap index), VTMSX (tax managed small cap), VBR (small and mid value) and RZV (microcap value). Not crazy about the number of funds but it cones down to available space in retirement accounts and taxable. Wish me luck :)
User avatar
baw703916
Posts: 6681
Joined: Sun Apr 01, 2007 1:10 pm
Location: Seattle

Re: Small Cap Tilting

Post by baw703916 »

MapleHermit wrote:Can we not agree that most people think small cap value will outperform? Therefore, the chances of it outperforming are greatly reduced.
People on this board seem to think that, but apparently not everybody does.

AUM
Vanguard Small Value Index (all share classes): $12.1 billion
Vanguard Small Growth Index (all share classes): $15.2 billion
Most of my posts assume no behavioral errors.
FillorKill
Posts: 1007
Joined: Sat Aug 06, 2011 7:01 am

Re: Small Cap Tilting

Post by FillorKill »

YDNAL wrote: Returns matter since we don't invest - at least no one I know - just to practice how to run factor loadings. I also don't see a glaring difference in return. VTMSX has taken less of a beating in declines, but lower increase in a rebound (2002/03, 2008/09, 2011/12).
** VSMAX inception 11/13/2000
VTMSX inception 3/25/1999[/code]
Ignoring the factor loadings, I'm curious what information you think you've gleaned from that return comparison when one of the funds (VSMAX) has had three different targets along the way (and the hard to define results of the two several-month-long transitional periods); other than the fact that the headline VG index funds seem to have commitment issues with index providers? The current incarnation is not even close to two years old and the previous incarnations don't exist. See how that's a problem?

FWIW - many investors do target specific exposures as a matter of course in their investing - I do. It seems a little more precise than choosing a fund merely by name/asset class and assuming that what's under the hood will do the job without further inquiry/analysis.
cb474 wrote: Indeed, as I pointed out above, but to which you did not respond, if we take the inception date of VTMSX as a starting point and check the factor loads over that time period to the present (I'm using the PortfolioVisualizer website) NAESX has greater SmB over that time period. NAESX = .74 SmB, VTMSX = .68 SmB. So the supposed greater smallness of VTMSX does not seem to explain the outperformance over that period. On the other hand, VTMSX has a higher HmL over that period. VTMSX = .38 HmL, NAESX = .25 HmL. That shows VTMSX was capturing a fair amout of the small value premium over that period, a little more than NAESX, and if one is going to point to factor loads it seems one might conclude that it's the value permium that accounts for VTMSX's outperformance.
See the question I posed to Landy above. How can I respond? What difference do the comparative factor loadings make when VSMAX was attempting to replicate the MSCI index, or R2K prior to that? That has nothing to do with the product available today. That's why analyzing the historic factor loadings for VSMAX is, IMO, not a good use of time.

One insight an investor might draw from this thread is that VTMSX hasn't changed strategy one single time and that has some value beyond the smaller deciles or quality tax management approach an investor may be targeting with the fund. VSMAX has had three incarnations (and two transitional periods) over the same period. So what difference does the performance comparison or anything else prior to the full integration of the current strategy for VSMAX make? IMO - none.
cb474
Posts: 905
Joined: Tue Jan 19, 2010 5:32 am

Re: Small Cap Tilting

Post by cb474 »

BBL wrote:See the question I posed to Landy above. How can I respond? What difference do the comparative factor loadings make when VSMAX was attempting to replicate the MSCI index, or R2K prior to that? That has nothing to do with the product available today. That's why analyzing the historic factor loadings for VSMAX is, IMO, not a good use of time.
To be fair, VTMSX has had as much historical variability in its factor loads as NAESX/VSMAX, over time, despite tracking a single index. Perhaps the differences over time have more to do with shifts in the market in general. That being said (and this is an honest question to which I feel perplexed about the right approach) why should we consider current factor loads any better of a predictor of future performance? Given the past history of factor loads for VTMSX, shouldn't I expect it to vary as much in the future? Perhaps there's some reason over time that the S&P 600, despite have a lower cut off for capitalization, tends to capture no more of SmB over time than VSMAX? It just seems very unclear to me given the current evidence. (I also feel like this is usually the point at which forum member Rodc jumps in to say over any of our given lifetimes the variability in these differences between mostly similar funds are more likely to be due to statistical noise than anything else.)
BBL wrote:One insight an investor might draw from this thread is that VTMSX hasn't changed strategy one single time and that has some value beyond the smaller deciles or quality tax management approach an investor may be targeting with the fund. VSMAX has had three incarnations (and two transitional periods) over the same period. So what difference does the performance comparison or anything else prior to the full integration of the current strategy for VSMAX make? IMO - none.
I agree with you that it is a big complication that NAESX has changed indices several times. I wouldn't say that it's changed strategy, exactly. In fact, if one wanted to make the argument in favor of this, it would be that it has honed its strategy with indices ever better constructed to realize it; it has allowed for the continual reduction of costs; and it has helped avoid problems like front running. So I don't think one can say that the indices changes are entirely without potential value. Still, I too find it annoying to have the index change mid course.

Further, given the indices changes, I agree with your statement: "What difference does the performance comparison or anything else prior to the full integration of the current strategy for VSMAX make?" But that to me only compounds my point that your comparison of historical returns between these funds, based on the inception date of VTMSX, is pretty arbitrary. Both for the reason you state and the ones I've stated above. That being said, one can, I think, look at CRSP returns back to the inception date of VTMSX. Although I'm not sure how to get that data. That would respond to your criticism about comparing historical returns of the funds as the currently stand, though not mine.
FillorKill
Posts: 1007
Joined: Sat Aug 06, 2011 7:01 am

Re: Small Cap Tilting

Post by FillorKill »

cb474 wrote: To be fair, VTMSX has had as much historical variability in its factor loads as NAESX/VSMAX, over time, despite tracking a single index. Perhaps the differences over time have more to do with shifts in the market in general. That being said (and this is an honest question to which I feel perplexed about the right approach) why should we consider current factor loads any better of a predictor of future performance? Given the past history of factor loads for VTMSX, shouldn't I expect it to vary as much in the future? Perhaps there's some reason over time that the S&P 600, despite have a lower cut off for capitalization, tends to capture no more of SmB over time than VSMAX? It just seems very unclear to me given the current evidence. (I also feel like this is usually the point at which forum member Rodc jumps in to say over any of our given lifetimes the variability in these differences between mostly similar funds are more likely to be due to statistical noise than anything else.).
IMO the shifts you see in VTMSX are a combination of the market shifts in general with the added variable of the tax management trading screens. That’s a unique variable for this type of fund. I think you get some momentum effects because of it and probably some additional noise that is hard to define. An investor today would likely be best served to consider the current form of VSMAX and anticipate that it will remain (relatively) similar.

VSMAX in current form doesn’t have much track of a track record. Although there are small differences in the numbers for each fund as reported by VG versus Morningstar (both as of 01/31/2014) the size difference is pronounced. I would anticipate that despite inevitable fluctuation the relative size difference will continue. Of course we know that S&P can change methodology as can CRSP and either fund can change strategy – but guessing at that isn’t worthwhile. As it stands:

VG reported median market cap: 3.0B/1.5B. MS reported average market cap: 2.7B/1.5B. Using Morningstar’s size breakdown (skipping giant) VSMAX/VTMSX:

Large: 0.3/1.06
Med: 45.1/7.84
Small: 46.37/64.78
Micro: 8.24/26.33

Using VG’s (skipping large):

Medium/Large: 0.4/0.0
Medium: 2.1/0.2
Medium/small: 51.0/12.2
Small: 46.5/87.6

What I see is a mid cap/small cap fund (VSMAX) and a small cap fund.

Considering factors as a predictor of future returns really depends (and this is where you end up in the ‘higher risk adjusted returns’ v ‘higher returns’ debates and ‘risk’ story v ‘behavioral’ story v both/neither ) on your personal investment views/philosophy. I happen to believe that factors have explanatory value. That is not the singular worldview by any means and I don’t think anyone else should use my conclusions as their own without having conducted their own research. Whatever decisions informed investors conclude are correct for themselves are the right decisions.
cb474 wrote: I wouldn't say that it's changed strategy, exactly. In fact, if one wanted to make the argument in favor of this, it would be that it has honed its strategy with indices ever better constructed to realize it; it has allowed for the continual reduction of costs; and it has helped avoid problems like front running. So I don't think one can say that the indices changes are entirely without potential value. Still, I too find it annoying to have the index change mid course.
IMO for an index fund the only strategy is cost effective/tax efficient replication of the performance of the index. If you change the index you are attempting to replicate, you changed strategy, maybe this is semantic (or rigidity on my part). Let’s hope CRSP is the final chapter.
cb474 wrote:Further, given the indices changes, I agree with your statement: "What difference does the performance comparison or anything else prior to the full integration of the current strategy for VSMAX make?" But that to me only compounds my point that your comparison of historical returns between these funds, based on the inception date of VTMSX, is pretty arbitrary. Both for the reason you state and the ones I've stated above. That being said, one can, I think, look at CRSP returns back to the inception date of VTMSX. Although I'm not sure how to get that data. That would respond to your criticism about comparing historical returns of the funds as the currently stand, though not mine.
Maybe. CRSP has data sets that go way back but (and I could be wrong) AFAIK they didn’t develop the index construction/methodology prior to 2011 which they then backtested against a data set. If the index construction/methodology didn’t exist in real time over that period I’m not sure you can effectively reverse engineer the combination of an index construction methodology that didn’t exist (and was developed after another decade of real time knowledge/experience/innovation) and overlay a guesstimate of what the VG index approach circa 2000 would have been and come out with anything useable. Or maybe it can be effectively done and we’ve just eclipsed my ability to conceptualize it.
cb474
Posts: 905
Joined: Tue Jan 19, 2010 5:32 am

Re: Small Cap Tilting

Post by cb474 »

Thanks for your thoughts on the makeup of the two funds. A couple remarks:
BBL wrote:IMO the shifts you see in VTMSX are a combination of the market shifts in general with the added variable of the tax management trading screens. That’s a unique variable for this type of fund. I think you get some momentum effects because of it and probably some additional noise that is hard to define.
You see the same variation with IJS, so I don't think the tax management strategy has much to do with it. It doesn't seem to effect composition or performance hardly at all.
BBL wrote:Using Morningstar’s size breakdown (skipping giant) VSMAX/VTMSX:

Large: 0.3/1.06
Med: 45.1/7.84
Small: 46.37/64.78
Micro: 8.24/26.33

Using VG’s (skipping large):

Medium/Large: 0.4/0.0
Medium: 2.1/0.2
Medium/small: 51.0/12.2
Small: 46.5/87.6

What I see is a mid cap/small cap fund (VSMAX) and a small cap fund.
I don't really put much faith in the significance, if any, of the Morningstar style charts. The factor loads are more meaningful to me, despite their limitations. I think the S&P 600 and CRSP (and even MSCI before it) are much more similar in their small cap makeup than the style charts suggeset. To the extent (or not) that they can be viewed as part mid cap and part small cap, I think that's true of both S&P 600 small cap and CRSP. Indeed, some people argue that most of the small cap premium is in the micro-caps, so these funds are equally missing out of that.

These wiki entries show the different cut off points that the different indices use and Rick Ferri provides a useful translation of the CRSP % cutoff point method into number of companies:

http://www.bogleheads.org/wiki/Stock_market_indexing
http://www.bogleheads.org/wiki/CRSP_equity_indexes
http://www.rickferri.com/blog/investmen ... p-shuffle/

The S&P 600 includes the 901 - 1500 largest firms. CRSP small includes 646 to 2,359 (but in terms of percentage of total market cap that's still the bottom 85% - 98% in terms of size). So CRSP has a higher cutoff point, but also goes much smaller. It's a bit apples and oranges and, once again, hard to make direct comparisons. But this is why, to me, the factor loads are more useful and the style charts are pretty misleading (if not useless).
BBL wrote:Considering factors as a predictor of future returns really depends (and this is where you end up in the ‘higher risk adjusted returns’ v ‘higher returns’ debates and ‘risk’ story v ‘behavioral’ story v both/neither ) on your personal investment views/philosophy. I happen to believe that factors have explanatory value. That is not the singular worldview by any means and I don’t think anyone else should use my conclusions as their own without having conducted their own research. Whatever decisions informed investors conclude are correct for themselves are the right decisions.
Yes, I don't need to redebate whether or not the small and value premiums exist or what they are. That's been done enough in this forum. I also except that the factor loads are meaningful (though not always so simple to apply in practice).
BBL wrote:IMO for an index fund the only strategy is cost effective/tax efficient replication of the performance of the index. If you change the index you are attempting to replicate, you changed strategy, maybe this is semantic (or rigidity on my part). Let’s hope CRSP is the final chapter.
I agree with you, I was just trying to present the other way of looking at Vanguard's index shifts. I'm not convinced its entirely a negative.
BBL wrote:Maybe. CRSP has data sets that go way back but (and I could be wrong) AFAIK they didn’t develop the index construction/methodology prior to 2011 which they then backtested against a data set. If the index construction/methodology didn’t exist in real time over that period I’m not sure you can effectively reverse engineer the combination of an index construction methodology that didn’t exist (and was developed after another decade of real time knowledge/experience/innovation) and overlay a guesstimate of what the VG index approach circa 2000 would have been and come out with anything useable. Or maybe it can be effectively done and we’ve just eclipsed my ability to conceptualize it.
If one develops a methodology for indexing in the present, but has all the data from the past necessary to see what would have happened had one applied that methodology, I'm not sure I see why one can't go back and see what the returns would have been. In any case, I've seen people in this forum cite factor loads for the CRSP going back to 2001. And I also found a wiki entry that gives returns for small cap CRSP going back to 2002. They compare favorably to the S&P 600. See:

http://www.bogleheads.org/forum/viewtop ... 0#p1511454
http://www.bogleheads.org/wiki/Small_Cap_Index_Returns
FillorKill
Posts: 1007
Joined: Sat Aug 06, 2011 7:01 am

Re: Small Cap Tilting

Post by FillorKill »

Although we have probably provided some interesting information (or at least our takes on it) I doubt that continuing this is all that productive or valuable so I’ll give it a rest after this post.
cb474 wrote: It doesn't seem to effect composition or performance hardly at all.
They're good at it. If they weren't it would be more visible but a small difference isn't no difference.
cb474 wrote: The S&P 600 includes the 901 - 1500 largest firms. CRSP small includes 646 to 2,359 (but in terms of percentage of total market cap that's still the bottom 85% - 98% in terms of size). So CRSP has a higher cutoff point, but also goes much smaller. It's a bit apples and oranges and, once again, hard to make direct comparisons. But this is why, to me, the factor loads are more useful and the style charts are pretty misleading (if not useless).
You don't need to run a regression to know that if you have two cap weighted funds and one has a higher cutoff there is almost no possible methodology that would cause the fund with higher cutoff to have a greater SmB loading. Look at the deciles of the US market. Take a fund that cap weights the entire market and a fund that cap weights deciles 2-10 and what do you get? The variation in the difference will wander over time but when and how would the 1-10 fund ever have a higher SmB loading than the 2-10 fund? You can debate the significance or the time varying nature of the difference but not the inherent existence.
cb474 wrote:I was just trying to present the other way of looking at Vanguard's index shifts. I'm not convinced its entirely a negative.
Agreed.
cb474 wrote: If one develops a methodology for indexing in the present, but has all the data from the past necessary to see what would have happened had one applied that methodology, I'm not sure I see why one can't go back and see what the returns would have been. In any case, I've seen people in this forum cite factor loads for the CRSP going back to 2001.
Have at it. Factor loads for a CRSP data set are one thing but that’s just a piece of the puzzle. I don’t see this as much different than retrospectively constructing something that didn’t exists and contemplating realities that never existed and letting that influence current thinking. Maybe you will enjoy the experience and you will find it worthwhile but there is little chance that applying the 2011 CRSP developed index methodology (you don’t know what this would have been if developed in 1999/2000 and you can’t dismiss the unknown difference(s) or how they would have impacted the result). You also don’t have the circa 1999/2000 Vanguard proprietary indexing software/technology/process that would have been used at implementation. You also can’t apply the changes across time in either CRSPs index methodology that didn’t exist but would have been likely different from 1999/2000 to the only know version of 2011. Just like you can’t apply refinements to Vanguards software/technology/process across that time. Because of these limitations you will never produce a set of data whereby you could confidently state that the CRSP small index based VG small cap fund would have this or that exact size loading in 2001 or 2004 or 2007, etc. I don’t see how you can get that level of precision with a reverse engineered fund that requires many assumptions of unknown and unknowable information.

Although I think you could approximate reasonably well, this just isn’t that far away from the theoretical arguments like: ‘if I could have invested in a US SV index from 1932-1954 then…’; ‘when I construct the EM index from 1971-1987 then…’

IMO - backtesting, simulating and recreating have obvious limits and create traps. That’s why I think an investor should accept the differences in these funds as they are constructed today and make their decisions based on this information.
cb474 wrote: And I also found a wiki entry that gives returns for small cap CRSP going back to 2002. They compare favorably to the S&P 600.
You’re using one of your own disliked apples to oranges comparisons here. We agree that returns are a product of factor loading so a decomposition of returns will clarify what happened. Sometimes size is rewarded more in decile 5 or 6 than 7 or 8 (or not at all). But you’re comparing strategy A to strategy B without decomposing the returns and then comparing the outcomes. You target specific exposures within your constrains, whatever they may be, and then you get the results that you get. A relative comparison of outcomes of strategies you did pursue to strategies you didn’t pursue isn’t really that useful.
YDNAL
Posts: 13774
Joined: Tue Apr 10, 2007 4:04 pm
Location: Biscayne Bay

Re: Small Cap Tilting

Post by YDNAL »

BBL wrote:
YDNAL wrote: Returns matter since we don't invest - at least no one I know - just to practice how to run factor loadings. I also don't see a glaring difference in return. VTMSX has taken less of a beating in declines, but lower increase in a rebound (2002/03, 2008/09, 2011/12).
** VSMAX inception 11/13/2000
VTMSX inception 3/25/1999[/code]
Ignoring the factor loadings, I'm curious what information you think you've gleaned from that return comparison when one of the funds (VSMAX) has had three different targets along the way (and the hard to define results of the two several-month-long transitional periods); other than the fact that the headline VG index funds seem to have commitment issues with index providers? The current incarnation is not even close to two years old and the previous incarnations don't exist. See how that's a problem?
1. I responded to your claim/post that return comparison doesn't matter.
BBL » Sat Mar 08, 2014 11:02 am wrote:
cb474 wrote:So my point was that choosing the inception date of VTMSX as the point from which to make comparisons (and implicitly the present as the relevant end point) is entirely arbitrary and doesn't mean anything.
I've repeatedly said that the return comparison does not matter. <snip>
2. No, I don't see "a problem" and see the same information anyone else can see. S&P 600 Index compensated Small Cap investors comparatively similar with "no glaring difference in return" against Russell 2000 Index through May 16, 2003; MSCI US Small Cap 1750 Index through January 30, 2013; CRSP US Small Cap Index thereafter.
Landy | Be yourself, everyone else is already taken -- Oscar Wilde
User avatar
Garco
Posts: 1078
Joined: Wed Jan 23, 2013 1:04 am
Location: U.S.A.

Re: Small Cap Tilting

Post by Garco »

All this effort to find the "G-spot" of asset allocation is fraught with problems for average investors, especially those like me who have a highly constrained set of choices in their 401k plan. If the funds that they have available aren't truly small or truly value, or if the premium really isn't there for the year or two when they try out some investment, they may learn very little or earn nothing for the (small) effort they've put into searching for the G-spot.

Although I have only limited options within my plan, I do have an option to buy the haystack (total market index) and to overweight large or small caps -- as defined by how they are named or by what one can see via an X-Ray and the style box. But I can't go in search of even better funds -- more truly value or truly small -- if they truly exist. I'm stuck with my plan. That said, I would use the word "satisfice" to describe my approach: do the best I can with what I've got, and add some extended market index funds to my total market index. Is it small enough, valuey enough? I don't care. I know what it's NOT. It's not the S&P500. And because I can buy VEMSX/VEXMX in my 401k, I buy some. I satisfice. Will it always give a higher return than SP500 index or the total market index? No. It has for some years, however, and the money is real, so I take it. I satisfice.
YDNAL
Posts: 13774
Joined: Tue Apr 10, 2007 4:04 pm
Location: Biscayne Bay

Re: Small Cap Tilting

Post by YDNAL »

Garco wrote:Although I have only limited options within my plan, I do have an option to buy the haystack (total market index) and to overweight large or small caps -- as defined by how they are named or by what one can see via an X-Ray and the style box. But I can't go in search of even better funds -- more truly value or truly small -- if they truly exist. I'm stuck with my plan.
:thumbsup

The amount of time spent thinking of investing as "Science" (NOT), and searching for things like needles in a haystack that can stick you in the rear end, could be more productive if the time is devoted to search for ways to maximize our earning-power - thus maximizing our savings rate.
Landy | Be yourself, everyone else is already taken -- Oscar Wilde
User avatar
Taylor Larimore
Posts: 32842
Joined: Tue Feb 27, 2007 7:09 pm
Location: Miami FL

"Buy the haystack" -- Jack Bogle quote

Post by Taylor Larimore »

The amount of time spent thinking of investing as "Science" (NOT), and searching for things like needles in a haystack that can stick you in the rear end, could be more productive if the time is devoted to search for ways to maximize our earning-power - thus maximizing our savings rate.
YDNAL is right (again):
Forget the needle, buy the haystack. -- Jack Bogle
Best wishes.
Taylor
"Simplicity is the master key to financial success." -- Jack Bogle
User avatar
abuss368
Posts: 27850
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!
Contact:

Re: "Buy the haystack" -- Jack Bogle quote

Post by abuss368 »

Taylor Larimore wrote:
Forget the needle, buy the haystack. -- Jack Bogle
Best wishes.
Taylor
When we left the evil stocking picking world, a big factor in our decision was this exact quote. Jack Bogle's books on this made so much sense.
John C. Bogle: “Simplicity is the master key to financial success."
cb474
Posts: 905
Joined: Tue Jan 19, 2010 5:32 am

Re: "Buy the haystack" -- Jack Bogle quote

Post by cb474 »

Taylor Larimore wrote:
The amount of time spent thinking of investing as "Science" (NOT), and searching for things like needles in a haystack that can stick you in the rear end, could be more productive if the time is devoted to search for ways to maximize our earning-power - thus maximizing our savings rate.
YDNAL is right (again):
Well, except that I would say this is a caricature of what people who look at things like factor loads, etc., are doing. I think that the more sophisticated investors who get into factor loads and regressions, etc., might be wasting time like YDNAL says, but often they are much more aware, than the so called lumpers, of just how inexact investing is and just how wide the dispersion of possible outcomes is, given a particular strategy. They are often the people I see in this forum who understand best the statistical and probabilistic variables at play.

I think many of these people aren't looking for a needle in the haystack at all. They simply accept the arguments for the small and value premiums (which one need not accept) and they are looking for the best way to apply this strategy. This need not at all mean that they don't understand the huge potential variability in outcomes. And it hardly means that they believe they have some sort of science for investing. They just have a more sophisticated way of understanding the odds and are trying to play the odds as best they can.

In addition, it really doesn't take that much time and if people find it interesting and worthwhile, I don't know why other people need to leap in and tell people what to do with their time. I really don't see the tilters constantly dumping on the lumpers and telling them how ridiculous their efforts are. But many people seem to feel compelled to do the same in every thread that starts about small and value tilting. Even those who have seen the debate many times and must (or should) know that it's never going to play out differently. There will always be both sides of the argument. Why not just let the people who find the effort worthwhile have their conversations, without the snarky interjections and condescension?
cb474
Posts: 905
Joined: Tue Jan 19, 2010 5:32 am

Re: Small Cap Tilting

Post by cb474 »

BBL wrote:Although we have probably provided some interesting information (or at least our takes on it) I doubt that continuing this is all that productive or valuable so I’ll give it a rest after this post.
Yes, we probably have gone over our take on the questions at hand enough times. I think I mostly agree with what you've said and I certainly see the logic of your way of looking at things. For me there are just some comparisons that I find less clear value in than you seem to and I guess I'm a little more skeptical about historical data over relativley short periods of time (less than at least twenty or thirty years), whether applied to real funds or retrospectively contructed hypothetical ones. I'm also trying to understand better how to use and how meaningful factor loads are. So I appreciate the conversation. Thanks.
YDNAL
Posts: 13774
Joined: Tue Apr 10, 2007 4:04 pm
Location: Biscayne Bay

Re: Small Cap Tilting

Post by YDNAL »

cb474 wrote:I think that the more sophisticated investors who get into factor loads and regressions, etc., might be wasting time like YDNAL says, but often they are much more aware, than the so called lumpers, of just how inexact investing is and just how wide the dispersion of possible outcomes is, given a particular strategy. They are often the people I see in this forum who understand best the statistical and probabilistic variables at play.
I challenge you to provide a quote that says "wasting time." That said, each one of us is free to use our time as we see fit.

Now, who told you that be a "sophisticated investor one must get into factor loads and regressions, etc."? Or, that would make the individual "much more aware"? Or,....?
  • 1. The highest level of awareness and, yes, sophistication is to diversify - while knowing to accept the performance (+/-) of diversified investments - at reduced cost (including taxes).
    2. Reviewing historical information is what it is; and making bets on HmL or SmB can actually be considered bets *against* diversification.
BBL » Wed Mar 12, 2014 9:02 am wrote:Although we have probably provided some interesting information (or at least our takes on it) I doubt that continuing this is all that productive or valuable so I’ll give it a rest after this post.
I agree, and many side conversations have been getting further and further away from the original post.
Paden [OP] » Tue Mar 04, 2014 2:33 pm wrote:Do you tilt towards Small Growth, Value, or Blend? <snip>

My question, is if you have a tilt in your portfolio toward small cap, do you tilt toward SG, SV, or both (small blend)? I would also like to here hear a reason or justification you use for that tilting. Would using half SG and half SV be essentially equivalent to the Small Blend? Using Morning* X-ray it looks like it is slightly different.
Landy | Be yourself, everyone else is already taken -- Oscar Wilde
gc527
Posts: 16
Joined: Sat Feb 04, 2012 10:35 am

Re: "Buy the haystack" -- Jack Bogle quote

Post by gc527 »

cb474 wrote: Well, except that I would say this is a caricature of what people who look at things like factor loads, etc., are doing. I think that the more sophisticated investors who get into factor loads and regressions, etc., might be wasting time like YDNAL says, but often they are much more aware, than the so called lumpers, of just how inexact investing is and just how wide the dispersion of possible outcomes is, given a particular strategy. They are often the people I see in this forum who understand best the statistical and probabilistic variables at play.

I think many of these people aren't looking for a needle in the haystack at all. They simply accept the arguments for the small and value premiums (which one need not accept) and they are looking for the best way to apply this strategy. This need not at all mean that they don't understand the huge potential variability in outcomes. And it hardly means that they believe they have some sort of science for investing. They just have a more sophisticated way of understanding the odds and are trying to play the odds as best they can.
I agree. You need some sort of framework to "stay the course". It could be "3-fund" portfolio or it could be "tilt to small & value". Personally, my progression in the investing world has been from individual stock selection to the 3-fund portfolio and moved on to tilt to small & value camp after studying and reading a lot (especially in this forum). I have also come to the conclusion that future is unknown and you need to play the probabilities. My way of playing the probabilities is 3FF.

But, I do want to add that for folks who are brand new to managing invesments or don't have the interest or inclination to deal with it, a simpler framework (that doesn't have too much tracking error) might be of benefit. When people hear value/small premium and jump into it without fully comprehending what they are getting into, this could be hugely detrimental to their portfolio. So, the dissenting (but civilized :happy) comments can help to show the other side of the story.

It is all about taking risk that you can understand. And, that includes 3-fund portfolio!

GC
cb474
Posts: 905
Joined: Tue Jan 19, 2010 5:32 am

Re: "Buy the haystack" -- Jack Bogle quote

Post by cb474 »

gc527 wrote:But, I do want to add that for folks who are brand new to managing invesments or don't have the interest or inclination to deal with it, a simpler framework (that doesn't have too much tracking error) might be of benefit. When people hear value/small premium and jump into it without fully comprehending what they are getting into, this could be hugely detrimental to their portfolio. So, the dissenting (but civilized :happy) comments can help to show the other side of the story.
Yes, the three fund portfolio is great for a lot of people, could easily end up beating tilters over any number of different time frames (long and short), and is a great starting (as well as ending) place. Tilting to small and value is not for the feint of heart and I think people need to do a lot of reading for themselves to develop their own understanding of the arguments for it, before constructing such a portfolio. As you say, "It is all about taking risk that you can understand"; that is very well put.

Still, it is striking to me how much I see some "lumpers" jumping into threads about small and value tilting and telling people how much of fools they are and how they don't understand anything, while making arguments that have already been made a bazillions times before and which often really portray a lack of understanding of what they're criticizing. That's not to say that there are not good criticisms of the small and value premiums, but I rarely see them made. Maybe I'm blind to it, but I don't see so much snarkiness and condescension from the tilters towards the lumpers, especially interjected into threads that aren't about tilting but rather purely about things like the three fund portfolio.
Post Reply