DFA Core-- Why bother with bottom 1000 names?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
countmein
Posts: 653
Joined: Fri Dec 06, 2013 8:10 pm

DFA Core-- Why bother with bottom 1000 names?

Post by countmein »

Was just persuing the holdings in DFVEX: DFA US Vector Equity. It has about 2800 names. About 97% of the fund is in the top 1800 stocks. Why bother having to deal with the bottom 1000 stocks when it only makes up 3% of your fund? Seems like the logistical expense would dwarf any benefit.
jginseattle
Posts: 1125
Joined: Fri Jul 01, 2011 7:33 pm

Re: DFA Core-- Why bother with bottom 1000 names?

Post by jginseattle »

countmein wrote:Was just persuing the holdings in DFVEX: DFA US Vector Equity. It has about 2800 names. About 97% of the fund is in the top 1800 stocks. Why bother having to deal with the bottom 1000 stocks when it only makes up 3% of your fund? Seems like the logistical expense would dwarf any benefit.
My guess is that one of the "bottom stocks" could become the next Microsoft.
Topic Author
countmein
Posts: 653
Joined: Fri Dec 06, 2013 8:10 pm

Re: DFA Core-- Why bother with bottom 1000 names?

Post by countmein »

Sure but they own so little of it that it wouldn't matter.
freddie
Posts: 920
Joined: Sat Feb 08, 2014 10:06 pm

Re: DFA Core-- Why bother with bottom 1000 names?

Post by freddie »

Isn't this true of any broad market index? What percentage of Vanguard total market is in the top 1000 stocks.
countmein wrote:Sure but they own so little of it that it wouldn't matter.
User avatar
Blue
Posts: 1178
Joined: Sat Jul 12, 2008 10:18 pm

Re: DFA Core-- Why bother with bottom 1000 names?

Post by Blue »

Migration by Fama paper - it's a lottery effect. A few will have outsized gains as they migrate from small value to large growth.
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: DFA Core-- Why bother with bottom 1000 names?

Post by larryswedroe »

Just add this, if you are non indexer willing to live with tracking error risk (random) then you can be seller of liquidity in asset class where liquidity is expensive

A fund like this can buy significant percentage of these stocks at discounts below last trade, maybe averaging 4% or so, and maybe 20% of trades.

And the smallest stocks have the largest premiums and they also carry the largest securities lending fees

Larry
steve_14
Posts: 1507
Joined: Wed Jun 20, 2012 12:05 am

Re: DFA Core-- Why bother with bottom 1000 names?

Post by steve_14 »

As we have now almost 10 years of data, we can see they aren't making any difference. This is a mid cap fund, as M* shows, and will track other such funds:

Image

Notes: This compares to Vanguard's higher fee investor share class, and does not include advisor fees required for the DFA fund.
Calm Man
Posts: 2917
Joined: Wed Sep 19, 2012 9:35 am

Re: DFA Core-- Why bother with bottom 1000 names?

Post by Calm Man »

steve_14 wrote:As we have now almost 10 years of data, we can see they aren't making any difference. This is a mid cap fund, as M* shows, and will track other such funds:

Image

Notes: This compares to Vanguard's higher fee investor share class, and does not include advisor fees required for the DFA fund.
This. Just like Index 500 vs TSM at Vanguard. Its academically and esthetically pure but has no meaning. OP, I concur with your instincts. But how could DFA make money if they admitted that their fund mirrors the Index 500 except they charge more and you have to pay an adviser to get it.
Topic Author
countmein
Posts: 653
Joined: Fri Dec 06, 2013 8:10 pm

Re: DFA Core-- Why bother with bottom 1000 names?

Post by countmein »

larryswedroe wrote:Just add this, if you are non indexer willing to live with tracking error risk (random) then you can be seller of liquidity in asset class where liquidity is expensive

A fund like this can buy significant percentage of these stocks at discounts below last trade, maybe averaging 4% or so, and maybe 20% of trades.

And the smallest stocks have the largest premiums and they also carry the largest securities lending fees

Larry

Larry,

I don't doubt any of this, but I don't think it addresses my question. Why pack 1000 stocks into 3% of a fund? How can this possibly make sense? All of the advantages you cite will amount to...squat.
afan
Posts: 8191
Joined: Sun Jul 25, 2010 4:01 pm

Re: DFA Core-- Why bother with bottom 1000 names?

Post by afan »

These observations will apply to any cap weighted portfolio. The largest stocks are so much larger than the others that their returns dominate results. If you have the patience you can do your own portfoio of perhaps 30-50 of the largest stocks, cap weighted, and get returns that closely track the S&P 500, which closely tracks the total stock market. The smaller the number of stocks you hold, the larger the potential tracking error, but you save your 4 to 6 basis points of expense ratio.

Holding those small allocations of smaller stocks does expose you to the big gains, and huge losses, that are more typical of the small companies. No matter how well they do, Apple and Google are not going to see 100 fold increases in their market caps.
We don't know how to beat the market on a risk-adjusted basis, and we don't know anyone that does know either | --Swedroe | We assume that markets are efficient, that prices are right | --Fama
Roy
Posts: 970
Joined: Wed Sep 10, 2008 9:34 am

Re: DFA Core-- Why bother with bottom 1000 names?

Post by Roy »

Calm Man wrote:
steve_14 wrote:As we have now almost 10 years of data, we can see they aren't making any difference. This is a mid cap fund, as M* shows, and will track other such funds:

Image

Notes: This compares to Vanguard's higher fee investor share class, and does not include advisor fees required for the DFA fund.
This. Just like Index 500 vs TSM at Vanguard. Its academically and esthetically pure but has no meaning. OP, I concur with your instincts. But how could DFA make money if they admitted that their fund mirrors the Index 500 except they charge more and you have to pay an adviser to get it.
Hi, guys: I agree with some of these points but I am not sure the comparison to a Midcap fund is appropriate (any more than comparing the DFA Small Value fund to the Vanguard equivalent is appropriate).

Here, the DFA fund spreads the allocation out much more among asset classes (where the midcap is not surprisingly concentrated in the middle), and the DFA fund therefore has a significantly larger average market cap, and differing tilts.

Seems you'd need several funds to replicate the DFA fund. DFA also seems to have several types of all-in-one funds that tilt to varying degrees. In each case, it likely would require several other funds even to try and replicate the loadings.
User avatar
JoMoney
Posts: 16260
Joined: Tue Jul 23, 2013 5:31 am

Re: DFA Core-- Why bother with bottom 1000 names?

Post by JoMoney »

It would seem that someone could create a similar "style box weighting" tilt using Morningstar's Instant X-Ray tool, still covered a broader range of stocks using various index funds, and still done the same job. I played around creating a similar weighted portfolio using IVV (S&P500), IJH (S&P400 Mid-Cap), IJJ (S&P Mid-Cap Value), and IJR (S&P600 Small-Cap). The similar weighting would have outperformed DFVEX, but really all of those funds individually beat DFVEX except for the S&P500 (link to chart). What's interesting (to me anyway) is that the S&P indexes are essentially just sampling (not trying to own everything in that size/style section of the market) and they explicitly avoid stocks with poor liquidity.
How is a fund investor supposed to realize a liquidity premium? What if the poor liquidity is still being discounted into the price when the time comes and they want to sell? Does DFA sell the more liquid stocks first and leave the share holders left behind concentrated in the illiquid stocks?
"To achieve satisfactory investment results is easier than most people realize; to achieve superior results is harder than it looks." - Benjamin Graham
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: DFA Core-- Why bother with bottom 1000 names?

Post by larryswedroe »

JoMoney
The way it works is that you can be seller if you have cash to invest and you are patient trader, price maker not taker.
So say in 2008 when almost all funds were experiencing outflows DFA was experiencing inflows. So say a small stock is trading 10 bid 10.2 asked and a fund must sell to raise cash or it wants to sell for whatever reason. In that illiquid market if the fund needed/wanted to sell a large block it might be willing to sell say at 9.5 or 9.3 to move it all at once, either because it was forced to or thought it was good move.

Same thing happens on daily basis though liquidity is less expensive when markets aren't crashing. But still a significant number of buys can be made below market prices if you can take tracking error risk

One other issue to consider in looking at DFA value (or vector) funds vs index funds---index funds will own REITs and regulated utilities that DFA won't. So you can at times have that lead to significant positive or negative impact on returns. (Note an investor using DFA funds who wants REIT exposures would own them separately). REITS would have helped relative to SV for the 5/10 year period, though subtracted over 15 years
Utilities would have helped some in last 10 years and been about draw in last 5.

Larry
steve_14
Posts: 1507
Joined: Wed Jun 20, 2012 12:05 am

Re: DFA Core-- Why bother with bottom 1000 names?

Post by steve_14 »

Roy wrote:Here, the DFA fund spreads the allocation out much more among asset classes (where the midcap is not surprisingly concentrated in the middle), and the DFA fund therefore has a significantly larger average market cap, and differing tilts.

Seems you'd need several funds to replicate the DFA fund. DFA also seems to have several types of all-in-one funds that tilt to varying degrees. In each case, it likely would require several other funds even to try and replicate the loadings.
These funds hold the same asset class - stocks. They even hold stocks of the same country. While they do own different groups of companies, as my graph shows this doesn't make much difference in their performance. The funds are both broadly diversified, with a similar average market cap, so are going to track each other closely.
Roy
Posts: 970
Joined: Wed Sep 10, 2008 9:34 am

Re: DFA Core-- Why bother with bottom 1000 names?

Post by Roy »

steve_14 wrote:
These funds hold the same asset class - stocks. They even hold stocks of the same country. While they do own different groups of companies, as my graph shows this doesn't make much difference in their performance. The funds are both broadly diversified, with a similar average market cap, so are going to track each other closely.
Steve: The market cap of the DFA fund is about 50% greater, which may well matter over time. And their stocks are spread over more asset classes (as expressed on a 9-box grid, for example). So, I'm not sure the comparison is apt even if the 10-year returns happen to be similar. (I don't own any DFA funds, just Vanguard.) As to the OP's question, I've no idea why DFA allocates as it does to the 3%, seems odd, but they seem to have been doing a good job as a fund family.
steve_14
Posts: 1507
Joined: Wed Jun 20, 2012 12:05 am

Re: DFA Core-- Why bother with bottom 1000 names?

Post by steve_14 »

Roy wrote:
steve_14 wrote:
These funds hold the same asset class - stocks. They even hold stocks of the same country. While they do own different groups of companies, as my graph shows this doesn't make much difference in their performance. The funds are both broadly diversified, with a similar average market cap, so are going to track each other closely.
Steve: The market cap of the DFA fund is about 50% greater, which may well matter over time. And their stocks are spread over more asset classes (as expressed on a 9-box grid, for example). So, I'm not sure the comparison is apt even if the 10-year returns happen to be similar. (I don't own any DFA funds, just Vanguard.) As to the OP's question, I've no idea why DFA allocates as it does to the 3%, seems odd, but they seem to have been doing a good job as a fund family.
Roy, the definition of an asset class is here: http://www.bogleheads.org/wiki/Asset_class
Roy
Posts: 970
Joined: Wed Sep 10, 2008 9:34 am

Re: DFA Core-- Why bother with bottom 1000 names?

Post by Roy »

steve_14 wrote: Roy, the definition of an asset class is here: http://www.bogleheads.org/wiki/Asset_class
As explained, when you look under the hood, your comparison was inappropriate. Significantly different stock distribution and different market cap weighting.
Topic Author
countmein
Posts: 653
Joined: Fri Dec 06, 2013 8:10 pm

Re: DFA Core-- Why bother with bottom 1000 names?

Post by countmein »

Another thing that comes to mind is that this "bottom 1000" might not be the smallest cap stocks in the fund. Perhaps they are low-weighted for a combination of reasons (e.g. they're growthy). In which case I would even more emphatically say "why bother"?
jdilla1107
Posts: 849
Joined: Sun Jun 24, 2012 8:31 pm

Re: DFA Core-- Why bother with bottom 1000 names?

Post by jdilla1107 »

larryswedroe wrote:JoMoney
The way it works is that you can be seller if you have cash to invest and you are patient trader, price maker not taker.
So say in 2008 when almost all funds were experiencing outflows DFA was experiencing inflows. So say a small stock is trading 10 bid 10.2 asked and a fund must sell to raise cash or it wants to sell for whatever reason. In that illiquid market if the fund needed/wanted to sell a large block it might be willing to sell say at 9.5 or 9.3 to move it all at once, either because it was forced to or thought it was good move.

Larry
First, why would you believe that dfa is better at this than the banks and prop trading firms whose business is doing this? I'll bet the banks had a field day off dfas trades too.

Second, what happens if next time if dfa experiences outflows when everyone else does? This "advantage" seemed like a one time thing.
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: DFA Core-- Why bother with bottom 1000 names?

Post by larryswedroe »

jdilla
DFA isn't better, never said they were. No idea where you got the idea that I did. DFA tends to generate more securities lending revenue because they tend to own smaller stocks which get the higher lending fee. In this rate environment securities lending revenue is of course way down.

As to cash flows, of course we cannot know what will happen in future, but there is only one reason why DFA was only equity fund family that received inflows and that's because the advisors did a better job of keeping clients disciplined, even rebalancing (adding equities) than investors did on their own or likely with advisors who are active. DFA has experienced cash inflows throughout the bear markets we have had.

Larry
Post Reply