From: Norman Wells 
Newsgroups: misc.invest.stocks uk.finance misc.invest.mutual-funds
Subject: Re: Why Managed Funds are Bad for your Wealth
Date: Sun, 1 Jul 2007 11:19:43 +0100
Bytes: 12425

In message <5eo454F38t61hU1@mid.individual.net>, Andy Pandy 
 writes
>
>"Norman Wells"  wrote in message
>news:LkYoqt2Z$rhGFw5A@myard.demon.co.uk...
>> >> That's a worse than
>> >> random result.  Why, if they have expertise?
>> >
>> >What?? Worse than random?
>>
>> In 2005, yes.  Random selection of shares that are included in the
>index
>> should give a result approximating to the index.  How can you end up
>> with a performance that is below random if the fund managers have
>> particular expertise in selecting shares, even if it is for only one
>> year?
>
>You really are getting desperate now. So if a football team won the
>league 14 years out of 15 you'd be asking "why, if they have
>expertise, did they fail to win that one year?"

No.  You really haven't got to grips with this analogy, have you?  It's 
not about failing to win the league that one year, it's about how they 
failed in that year to beat even random selection, ie finishing below 
half way in a league of teams that I have selected from the electoral 
roll using a pin. That's really quite serious, I would have thought, and 
worth questioning.

In that year their fund would not have beaten me with my pin, nor a five 
year old with a pin.  How good is that?  And how do you explain it, 
given all the expertise they claim to possess?
>
>> I can envisage a scenario in which a number of obvious basket case
>> economies are included in the index, but no shares from those
>countries
>> are actually included in the fund.  That would of course give an
>> artificially low value for the index, and a correspondingly inflated
>> overperformance when the fund is compared with it.
>
>Whereas an index tracker, which you seem so fond of, wouldn't avoid
>the basket cases....

No it wouldn't, but I've already said that any indexes that cover more 
than one country are highly suspect.  It would be like investing in 'the 
world economy', something which no-one does.  So I wouldn't be at all 
fond of any fund that did that.
>
>> > So the manager has about a 15 year track
>record
>> >(at least) of consistent excellent results.
>>
>> No, not consistently.  See 2005 as discussed above, which you need
>to
>> address.
>
>No I don't. I'm absolutely delighted with his performance. You stick
>to index trackers, which will fail to beat the index every single year
>as they have charges too.

They will, but their charges are lower, which means that if non-tracker 
fund managers are investing at random, as I say the evidence supports, 
the tracker will be the better investment on average.

>> Without knowing all other indices, how can I possibly say?  However,
>> judging by the performance of the Chinese stock market over the last
>ten
>> years or so, I'd say it's quite likely you'd have been better off in
>a
>> China-only tracker fund.
>
>One of my current funds is Invesco HK & China, which I've held since
>2001. That has beaten the relevant index too.

By how much each year?
>
>> and it is that the best 30 domestic mutual funds in each five year
>> period in Dimensional Fund Advisors and S&P/Micropal's ongoing
>study,
>> as a whole underperformed the S&P index going forward, sometimes by
>a
>> large margin.
>>
>> Do you actually dispute that finding?  If so, what evidence do you
>have
>> that is not purely anecdotal?
>
>Anwer my questions about it then, which I see you've snipped yet
>again.

There's no inconsistency at all in what I've said.  So, do you dispute 
the above finding, or are you hiding your head in the sand wishing it 
would go away because it doesn't suit your position?
>>
>> Oh dear, you're really scraping the bottom of the ideas barrel now.
>
>Oh really? So you don't think a sticky coin could possibly influence
>the result of a toss?
>
>I might have had a jam sandwich and inadvertantly got one side of the
>coin sticky.

Again, you're hiding from the truth because you don't like it.
>
>> >> If you get results from tossing it that appear to be otherwise,
>> >they're
>> >> merely as a result of statistical variance.
>> >
>> >Anyone who understands statistics or epidemiology would not simply
>> >dismiss a result as I've described as statistical variance.
>>
>> A statistician certainly would.
>
>Bullshit. It's well outside the normally accepted realms of
>statistical variance. A statistician would assume the result was *not*
>random and would look for evidence of bias.

Not so.  If a statistician ever assumed a result was anything in 
advance, he would be biased in any further investigation he might make. 
He'd be looking for evidence that supports his assumption, and not be 
approaching the matter with an open mind.  Just as you are actually.
>
>> But epidemiology?  What have epidemics to do with it?
>
>Look up epidemiology and the methodology used to determine causal
>relationships in disease. It's not dissimilar to what we've been
>discussing. Consider the biased coin to be "ill".

Now you're being absurd.  It's a matter of statistics, pure and simple.

>> I think we need to look at which indexes are being compared to which
>> funds to determine whether they are in fact the relevant indexes,
>for
>> the reasons I gave above.
>
>Oh right. What about the other quotes, the 73% in the OP, the 90%
>quoted by Motley Fool, the "great majority" by someone or other, the
>"30 funds" ambiguous rubbish. Don't you feel compelled to check they
>were using the relevant indexes etc? Nah - just blindly accept any
>evidence that suits your case and assume any evidence which doesn't
>must be flawed.

Whatever the actual figures, it's universally accepted, even by you, 
that most managed funds don't beat the relevant index after charges.  In 
those circumstances it is wholly unnecessary to go back and do the 
fundamental research all over again.  If you want to query it, you do 
it.
>
>> Not so.  Mutual fund managers probably make up only a small
>proportion
>> of the market in any one particular index.  Company pension funds
>and
>> other independents are huge investors.
>
>Yeah - and those funds are managed, there's no reason why a manager of
>a company pension fund should be any less skilled than a manager of a
>unit trust fund.

No, they're all investing randomly.  But they're investing differently 
too, which means that the mutual fund managers are certainly not 'the 
market' as you previously claimed.
>
>> In any case, I've shown that I can construct my own portfolio at
>random
>> from within any specific index with a 50% chance of beating it.  If
>a
>> fund manager can't do any better than that, it's a piss-poor show.
>
>Yeah - I've successfully avoided such piss poor managers.

Not in 2005 you didn't.  How do you account for that?
>
>> >Now - look at it another way. If funds really are investing at
>random,
>> >then only a very tiny minority would ever beat the index over the
>long
>> >term after charges.
>>
>> I agree.  And it's borne out by the figures that I constantly quote
>that
>> 73% of managed funds, after charges, do not beat their relevant
>index in
>> any one year.  It gets worse every succeeding year too.
>
>So what are the odds for 14 years out of 15, or thereabouts? Of 6
>funds beating the index over 5 years on average?

As I've said, it depends somewhat on whether they are comparing 
themselves with a relevant index.  I've already pointed out the reasons 
why any multi-national index is highly suspect, for example.
>
>> >But suffice to say, with random investment results, if only say 25%
>> >beat the index in any one year, then less than 25% would beat the
>> >index over 3 years, and considerably less than that would beat it
>over
>> >5 etc.
>>
>> Again, that is so.
>
>What are the odds for 5 years? 15 years?

Before charges, if they're investing at random, it's 50% over either 
period, and that's all that can be said with certainty.  After charges, 
it will be less depending on how high those charges are.  And it will be 
less after 15 years than after 5 because there are 10 further years of 
those charges to take into account.
>
>> A randomly selected portfolio of shares contained in an index must
>have
>> a 50% chance of beating the index over any period you choose.  It's
>the
>> charges that drag the average fund below the index level.
>>
>> But I can create a randomly selected portfolio with a pin that has
>just
>> the same 50% chance of beating the index, with no charges.
>
>Well done. You keep droning on and on about that. So what? It's
>irrelavent to the question of whether some fund managers can produce
>consistently good results.

Any fund manager investing at random can produce consistently good 
results in a rising market.  You can't analyse how good he is, however, 
or even if he is in fact investing other than at random, without knowing 
what a random investment would return, and how much better he has done 
than the index.  And that depends on using a relevant and applicable 
index in each case, and on knowing what the standard deviation from that 
index is.  I've questioned whether the indexes with which some of your 
funds are compared are relevant and appropriate, and we have no 
information whatsoever about standard deviations.  It follows that we 
can't determine whether your fund manager is any good or not.  We can't 
say whether his success is due to skill or luck.  Both are possible.
>
>Yawn. Banging on about that again, the average fund manager being crap
>doesn't mean there aren't good ones.

But that's exactly what we're trying to determine, whether they are in 
fact good or just plain lucky.  I think the evidence points one way, you 
obviously think it points the other.  But if you consort with knaves 
it's hardly surprising if you're taken as one yourself.

>> >A much more
>> >relevant study than the above one would be what happens if an
>investor
>> >chose the top long-term consistently performing fund (therefore
>> >excluding high volatility ones) in each sector.
>>
>> Point me in its direction then.  Otherwise, we have to go with what
>> we've got.
>
>I already have. You carry on believing ambiguous bullshit on any old
>website, and that "most fail, therefore all are crap" bollox.
>
>I'll believe my bank balance.

And you'll probably deny it when it goes down, just as any gambler 
denies their losers.

-- 
Norman Wells
NG