From: "GB" 
Newsgroups: uk.finance
Subject: Re: Starting a pension - advice please!
Date: Wed, 20 Dec 2006 02:19:45 -0000


"Ronald Raygun"  wrote in message 
news:JzWhh.18832$k74.3512@text.news.blueyonder.co.uk...

>>
>> Well, I agree with that. But to do that, you have to maximise the 
>> *future*
>> performance net of charges. Given that past and future performance are
>> uncorrelated, how do you do that?
>
> Because, although past performance is no guarantee that future performance
> will be similar, to say they are uncorrelated is simply rubbish.

Oh dear. In two words 'simply rubbish' you give us your apparently 
considered verdict on dozens of academic research papers plus the advice of 
the FSA to private investors.


> Good fund
> managers achieve, on average, good performance year after year.
>

And how do you know they are good managers? Because they have a string of 
good years behind them, I guess. And next year they have a 50/50 chance of 
having another good year. And those that do, you say 'see, he's a good 
manager'. And those that don't, you forget about, or you find some excuse, 
or you fall back on the 'on average' part above.

Really, it's the same type of fallacy that gamblers have, remembering the 
good wins and forgetting the losses.

> Their acumen which brings this about is worth paying for

Yes, of course it would be, if it's repeatable in the future and not just a 
string of luck.

Last time I looked, there were several thousand unitised funds listed. If 
you divide each year's performance into above average and below average for 
its sector, then each fund has a 50% chance of being average in any one 
year, even if the process is completely random.

So, on average, out of 1024 funds, one of them will be above average for 10 
straight years in a row. 45 of them will have 8 out of 10 good years, and so 
on. This is all on a completely random basis.

You can surely see that this random process could make people think that 
certain managers were outstandingly gifted.

The investment industry knows this, and any consistently underperforming 
funds just get quietly lost, ie closed, merged with other funds, or 
restructured.

> , and thus funds
> which return consistently good performance tend to charge more (but not
> so much more as would cancel the benefit of using them).

If it's random, as nearly all the academic research suggests, then this 
would be a con, of course.