"Andrew Koenig" <ark[at]acm.org> wrote in message
news:6TuSh.29158$VU4.18626[at]bgtnsc05-news.ops.worldnet.att.net...
- quote -

Many modern *indexes* have hysteresis built in - they call the mechanism
"buffers". If a stock crosses a boundary, it is not immediately removed
from (or added to) the index. It is in a buffer zone, from which it may
return to meet the index criteria, or may subsequently be removed. This is
a key reason why Vanguard switched to MSCI indexes.

Here's a brief description of how the Morningstar index buffers work - by
overlapping regions (what you described, except that the overlap is built
into the index, not into the tracking fund):
http://indexes.morningstar.com/Index...gy.asp#Method6

Here's Vanguard's description of theMSCI buffers:
https://institutional.vanguard.com/V...MSCIBuffer.jsp

The implicit idea that index funds literally hold exactly what is in the
index, when it is in the index, is a bit simplistic. First of all, sampling
is used by funds tracking a significant number of indexes; only certain
indexes have a sufficiently small number of components, all of which are
sufficiently liquid, to make full replication practical. Once you go to
sampling, all bets are off as to when (or even whether) particular
securities are added or removed.

Even with full replication, timing comes into play - you suggested this
(e.g. not selling when the price is depressed), but (I think) from a
different perspective. A fund may add a new index component a few days
prior to its actual inclusion into its benchmark index, knowing that the
price will rise on the inclusion date (due to other funds being forced to
buy on that particular day). Options can be used to keep the performance
tracking better (no, I haven't thought this sentence through carefully; just
repeating market-speak here).

Vanguard uses a variety of such techniques, which is why I credit them with
a few more basis points than I would another index fund company. (I'm
excluding DFA here, since their funds have limited access.)

In another post, in talking about DFA differences, you wrote:
- quote -

That's not as unusual as DFA might lead you to believe. (That's a problem
with referencing market-speak; it's prone to hyperbole, aka "puffing").
According to S&P's methodology:

"Treatment of IPOs. Initial public offerings should be seasoned for 6 to 12
months before being considered for addition to an index."
http://www2.standardandpoors.com/spf...dology_Web.pdf

There are various techniques that can enhance performance. Dimensional
implicitly acknowedges its extensive use of such techniques by not even
calling its funds index funds (except for one, if memory serves correctly).

Mark Freeland
BnetOnewsX[at]sbcglobal.net

On Apr 10, 2:09 am, dsmo...[at]stat.purdue.edu (David Moore) wrote:
- quote -

I think Vanguard tries to mimick or copy Dimensional Funds' value
funds but they haven't been peformed as well as the DFA funds. It's
interesting how Vanguard can't replicate what DFA does even though
Fama/French have put out a paper on the value premium works many
years. Obviously not all indices are built the same way. Some are
better built than others.

pago

<pago_boss[at]yahoo.com.au> wrote in message
news:1176177658.995350.201510[at]n59g2000hsh.googlegroups.com...

- quote -

The phrase "track the market" is too vague, because there is no single index
that represents "the market" (unless maybe you consider some composite of
all the stocks available worldwide). So in general, an index fund doesn't
track "the market;" it tracks a particular index.

My understanding of how DFA works is that each of their funds tracks an
index. However, the index they track isn't usually one of the widely
published ones. In other words, they don't decide on the basis of
individual research whether to include a company; but rather they have rules
that they apply mechanically, just like any other index fund.

Here are two differences I've seen on their website between their indices
and some of the others. Of course I may be misinterpreting their
statements, or their policies might have changed since I read about them,
but this is what I remember:

1) They don't include newly public companies until one year post-IPO.
They say that their experience is that there's too much volatility in the
first year without an attendant performance gain, on average.

2) They exclude companies for which there are too few market makers, as
it is difficult to buy and sell those companies quickly.

In addition to having slightly unconventional indices, they allow their
funds to deviate slightly from the index in the interest of minimizing
transaction costs. For example, their funds have some hysteresis about
companies leaving or entering the index, so that they rarely have to buy a
stock and then sell it again quickly.

I personally consider these strategies, as described, to be very far from
what I would ordinarily expect from an actively managed fund.

On Apr 10, 12:58 am, Michael Siemon <mlsie...[at]sonic.net> wrote:
- quote -

so Dimensional Fund managers are not passive managers like Vanguard,
ie they do not track the market and they manually intervene to achieve
their outperformance by being using smart buying and selling
strategies.

pago


======================================= MODERATOR'S COMMENT:
Please trim the post to which you are responding. "Trim" means that except for a FEW lines to add context, the previous post is deleted.

"Andrew Koenig" <ark[at]acm.org> writes:

- quote -

As I recall, DFA claims to have *negative* expense ratios in
some of their microcap funds, because as a major market-maker
in the space, they are capturing spreads instead of paying them.

--
Rich Carreiro rlcarr[at]animato.arlington.ma.us

The question is, can DFA index funds outperform "the market."

If by "the market" you mean the entire US equity market (represented
by e.g. Vanguard Total Market Index Fund to allow for transaction
costs etc), the answer is YES.

DFA funds represent distinct asset classes within the overall market.
Academic research strongly suggests that "small" and "value" stocks
do better than the overall market. (Like all statements about market
returns, this is true only on the average in the long run.) So
if you choose, for example DFA US Small Cap Value Fund, you can indeed
expect to do better (on the average in the long run).

In fact (I only have 5-yr data, from Morningstar):

DFA US Small Cap Value 17.56%/yr std dev 14.08%
Vanguard Total Market 8.30%/yr std dev 7.73%

Note that the usual holds true: higher return accompanies higher
risk (that is, higher volatility).

David

On Apr 9, 6:28 pm, pago_b...[at]yahoo.com.au wrote:
- quote -

They outperform the market -i.e., they obtain a higher rate of return
than the market average- because the type of stocks they typically
invest in (small cap value stocks) are riskier than the market
average. That's actually what the 3-factor Fama and French model tells
you: in the long run, the only way to obtain higher returns is if you
are willing to accept higher risks (that's a consequence of the
efficient market hypothesis).

In another group -http://groups.google.com/group/small-microcap-value?
hl=en - I posted the 1926-2005 data used by Fama and French in their
research work (Fama and French.xls file). For small cap value stocks,
the average return for that period has been 15.48% per year, while
the broad market -proxied by large blend stocks- had an average annual
return of 10.55%. The riskiness of small cap value stocks -standard
deviation of these returns- is 28.6%, clearly higher than the
riskiness of the broad market - standard deviation of 21%-.

<pago_boss[at]yahoo.com.au> wrote in message
news:1176132716.635910.166840[at]o5g2000hsb.googlegroups.com...

- quote -

They can't. What they can do is select their markets more carefully than
the published indices. So, for example, their small-cap funds stick to
smaller compmanies than most others.

As another example, just try to find an international small-cap value index
fund that's open to new investors.

The other thing they claim to do is to deviate slightly from the index when
it would be more expensive to stick to it. For example, suppose a company
in a small-cap fund grows to be a little too big for the fund. Then if the
fund is trying to stick to the index, it has to sell that stock immediately.
Then the company shrinks a bit, so it fits in the index again, and so the
index has to buy it back. Sell high, buy low.

DFA claims to solve this problem through hysteresis. When a company crosses
the boundary of their index, they don't sell (or buy) it immediately.
Instead, they wait until it has gone past the boundary by some amount.

Another thing they claim to do is pay attention to transaction costs in a
way that index funds generally don't. For example, when they sell stock,
they look for buyers who are willing to pay a premium for fast execution.

There are a few other tweaks that they claim to make. However, I think that
the biggest advantage they offer is access to market segments that aren't
well covered in general.

In article <1176132716.635910.166840[at]o5g2000hsb.googlegroups.com> ,
pago_boss[at]yahoo.com.au wrote:

- quote -

An index adds/drops stocks as they meet, or depart from, the defining
parameters. A mutual fund tracking an index necessarily has some
transaction costs at these points. DFA "models" the index parameters
but does not necessarily buy or sell stocks on the same schedule as
the index itself, spreading large transactions. In the case of their
micro-cap fund (their first), they are a large player in the market
and have developed some clout/strategies that enable them to do a
particularly good job of managing spreads in their transactions.

Similarly, their modeling of an asset class is nuanced (e.g. with
respect to value, their criteria are a bit more sophisticated than
just book-to-market).

You can see their general comments on their approach on their
web page, http://www.dfaus.com/

Whether they "beat" an index (on a regular basis) is arguable,
but in general they often do better than most other fund providers
in their target asses class funds.

Dimensional Funds claim to offer superior returns with lower risks.
The key to their success is using the 3 factor Fama/French to select a
portfolio of value stocks. It claims to have all the benefits of
indexing while still outperforming the market.

My question is, how can an index manager outperform the market? by
definition, an indexer cannot outperform the market.

Thanks,

Pago