"Will Trice" <wwtrice[at]paragondynamics.com> wrote in message
news:4412F4AE.2050404[at]paragondynamics.com...
- quote -

It is an issue of small cap vs large cap combined with a capitalization
weighted index. Once everyone is aware of a certain strategy generating
excess returns, the strategy will stop working.

Elle wrote:

- quote -

All of this is true. I don't know if this helps your perspective or
not, but Jason Zweig (you mentioned him in another thread) wrote an
article that accepts the data (but rejects the conclusions) of the
article linked above in this month's Money magazine (Zweig's article is
the original source for this thread).

-Will

catalpa wrote:

- quote -

Casting aside the rebalancing problem this presents, does this hold up
for longer periods of time?

Tad Borek wrote:
- quote -

I'm don't know what data you're looking at, but could it suffer from the
equal-weighted funds skew that the author claims?

- quote -

They claim that their results are nearly invariant with time, but they
only check two time periods and then show a chart where it seems they
checked many time periods where the results differ greatly. The
implication of my first post was that maybe managers are getting better.
But then again, maybe they've just had a good run.

- quote -

Could this be just because the results of larger companies swamp that of
the smaller companies?

- quote -

No, she's right. They included expenses, but not loads on the managed
funds.

-Will

"Will Trice" <wwtrice[at]paragondynamics.com> wrote in message
news:4411D4A7.1070604[at]paragondynamics.com...
- quote -

Which index or how the index is composed makes all the difference when
comparing a managed fund against an index . On a pure price basis in the 7
years from 12/31/1998 to 12/31/2005 the S&P 500 was unchanged and the
Russell 2000 was up over 60%.

You can beat the S&P 500 with the S&P 500 stocks just by equal weighting
them. From multiple sites: "The equal-weight S&P 500 has held its lead over
the past decade, gaining 11.9% annualized versus a 9.1% average yearly rise
for the cap-weighted index, according to S&P."

Will Trice wrote:
- quote -

Will, thanks for the link. Didn't read closely but just a couple
thoughts on quick read...one possible source of differences could be
their "rolling 12 month returns" methodology. Just intuitively...you
look at the "fund comparison vs. S&P 500" kinds of charts and you'll see
much higher percentage of "winning" numbers on the one-year data than on
the three year, five year or ten year. Because of that basic problem of
"performance not repeating." Now if they run the data right it should
account for this but I'd need to look at exactly what they're doing. I
wonder if this study is basically compounding forward all those one-year
figures, while ignoring the fact that a given fund subsequently
underperformed. It seems it depends on how they do their asset
weighting. If a fund lost 50% do they asset-weight that (-50%) return
based on the initial assets, or do they underweight it based on the
lower, end-of-period asset levels? Do they look at all at fund
inflows/outflows to see how many dollars experienced posted returns?
There's this tendency towards buying into gains with actively managed
funds. Devil's in the details!

I'm also curious about their choice of 1/75 as a start date - just after
some very big losses. I don't know the effect of that but as the article
itself notes (so did Buffett's latest letter to shareholders) the choice
of a start and end date can lead to dramatically different results.

Also - they take as a given that CRSP 1-10 and S&P 500 are
interchangable, citing "discussions with academics." Essentially they
seem to be using the one-factor model rather than the three-factor model
to adjust for risk. Over the period they cited I ran these two against
each other and the standard deviations and geometric returns are close.
But I don't consider these interchangable. There were substantial
discrepancies along the way and again, they're picking a certain
start-end date where the two happened to line up over the long haul. The
assumption tacitly implies that there's no small-company risk, and
Fama's Nobel suggests otherwise! =)

I saw also the language Elle quoted but a sentence a bit later seems to
contradict that - it's unclear to me exactly what investment costs, they
incorporated into the analysis. Maybe if it's rainy this wknd I'll give
it a closer look...

-Tad

"Will Trice" <wwtrice[at]paragondynamics.com> wrote
- quote -

Did you notice the following statement from the article
linked above?

"Neither the Lipper data nor the Barra data include loads,
commissions, or investment management fees."

Other statements in this article reflect that it's being
published in a trade journal, not an academic one; that its
publisher is an association whose mission is to assist those
who are financial planners //for a living//; that the author
is the President of an asset management company and chair of
a mutual fund company (Bullfinch; two of its funds that I
located quickly have expense ratios of 1.6% and 2%). There
are enormous conflicts of interest here.

Those responding to my comments: I think it would be helpful
if you disclosed whether your livelihood derives, in part or
in full, from giving advice on financial planning.

Elle
Private investor of 20+ years; never employed in the
finances yada industry.

Tad Borek wrote:
- quote -

I don't have the data, but here is the article:
http://www.fpanet.org/journal/articl...TOKEN=57741705

- quote -

The author claims no survivorship bias.

- quote -

By computing the standard deviation of the S&P 500 index vs. the
standard deviation of the average return of managed funds. I only
skimmed the article before, but a quick read-through shows that the
author did not actually adjust the returns, but concluded that the
standard deviations were nearly the same, so the risk was comparable.

-Will

"Will Trice" <wwtrice[at]paragondynamics.com> wrote
- quote -

That's the salient point to me, given that allocation is
darn near everything. For example, I presume there are some
high grade bond funds that beat the S&P 500 for a few years.
That doesn't mean one should put all one's money in high
grade bonds.

I thought the following was interesting, and will take its
basic argument to be true (based on other reading) until
further notice:

In a recent study of 147 large investment funds, researchers
studied the
contribution of stock picking, market timing, and asset
allocation to
long-term returns. What percentage of returns was the result
of asset
allocation, as measured by a portfolio of index funds?
A. more than 95%
B. 70%
C. 50%
D. 30%
E. less than 5%

Correct answer is A, or so claims
http://www.indexfunds.com/ (click on "Risk Capacity
Survey").

Will Trice wrote:
- quote -

Will-
A good starting point would be seeing whether the numbers on which you
base the premise are accurate, they don't look right to me (though I
don't keep 7-year data at my fingertips!). Where are those from? Do they
correct for survivorship bias, or just show the returns of the funds
that still exist now? How do they "adjust for risk"? My 5 and 10 year
numbers show the usual - active funds lagging once you adjust for their
asset-class risks - even without factoring in survivorship bias.

-Tad

OK, so the subject was a bit dramatic, but the average managed U.S.
equity mutual fund has now beat "the market" for seven years straight,
and that's after adjusting for risk. Of course, "the market" here is
defined as the S&P 500 which is not really "the market". The Wilshire
5000 may be a better benchmark, but the average managed fund beat that,
too last year (Average managed fund: 6.7%, W5000: 6.4%, SP500: 4.9%).

It may not be fair to compare all managed funds against the S&P (or
Wilshire). One could argue (correctly, I think) that managed funds
should be compared against the benchmark for their investing category.
If you do this, you'll see that managed funds on average almost always
get killed (I'm looking at 1, 3, and 10 year data). Aha! Indexing
isn't dead!

I myself have been an advocate of indexing, and in particular indexing
with an S&P 500 fund. But perhaps a better strategy is to build a
market weighted portfolio of managed funds, that apparently would beat
the S&P 500. Is this practical? How would one go about this? Or is
seven years too little data to be formulating this kind of strategy?
Maybe the strategy would be even better by buying index funds across
investment categories with the same weights as managed mutual funds (e.g
if 3% of the total money in managed U.S. equity funds is in microcap
funds, 3% of my portfolio should be in a microcap index).

Just tossing around some ideas, what do you think?

Thanks in advance,
-Will