Fund Selection for first timer

dunstonh

When i was referring to ramping up,i was referring to the kind of literature,often confusing,put out by the actual fund providers rather than IFAs and similar.

Very true. Providers are marketing a product and want you to see the upsides, not the downsides. Even IFAs that retail direct to public (and not advice) do this as well. Hargreaves Lansdown is a good example of that in the way they promote certain funds.

You have to take a lot of the marketing with a pinch of salt.

Of course Fund providers and IFAs will talk about risk but in reality,everyone has an expectation that they will make money. Thats the very reason why they sign up !

you wouldnt do it otherwise.

As i say,just a bystander..not an expert.

thank you for your reasoned response.:)

xyy123

Aegis wrote: »

Claiming that IFAs are useless, which is what you seem to be implying, is just plain silly.

Wasn't what I was claiming at all. But clearly there is more attached value to financial advisors recommending actively managed products since it is something that a lay investor has less knowledge of and create a market for a service that, arguably, does not need to exist. Financial advisors admitting picking actively managed funds as an exercise in futility would be like McDonalds coming out and saying its customers should not eat Big Macs as they were bad for their health.

You really aren't getting this pretty simple issue, are you? In the US, index trackers are taxed MORE FAVOURABLY than managed funds. It's not because they trade less, it's because they actually are not taxed on capital gains made within the fund when they trade.

Comparing trackers in the US to managed mutuals there and claiming that the same automatically applies here is like claiming that the net returns on a corporate bond fund here will be the same for a taxpayer whether they invest in the fund within an ISA or through direct holding. It's a ridiculous suggestion.

Cannot find any evidence to back this up, maybe I'm looking in the wrong places. All I can see is that both funds are taxed and pass capital gains taxes onto the shareholders but because index trackers only trade a few times per year these will obviously be less.

But, regardless of this fact, I found a couple of interesting studies for you to read:

1. www.investmentfunds.org.uk/press/2002/20021014-01.pdf

and the FSA followup:

2. http://www.fsa.gov.uk/pubs/other/pastperf_mutalfunds.pdf

The first states:

"Clearly the treatment of the 434 funds, almost half the sample, that die during the period of the study is crucial."

So a survivorship of just over 50%...

"...funds in the All-Companies sector in the bottom decile (bottom 10%) there is a 31% chance of dying within the next year. In contrast for the top five deciles (top 50%) there is an average probability of only 5%."

So mainly the poor performing funds dying.

*Note that this was a study commissioned by the IMA, not designed to but a corrollary of which was to show how past performance can be used as a tool for future indication*

#2 Says amongst other things:

"6.7. The CRA2 study should be commended for using a large sample of funds including dead funds and accounting for the effect of survivorship bias. Survivorship bias (the effect on reported returns of excluding funds that have died some time during the sample period) can lead to important distortions in the statistical analysis. Our earlier work (Lunde, Timmermann and Blake (1999)) found that underperforming unit trusts do eventually merge with more successful unit trusts, but that on average it takes some time for this to occur. Overall, our study found that about 40% of funds over the period 1972-1995 were eventually wound up or merged. The most frequently occurring (modal) duration for a fund was 4.25 years (51 months), but the average duration was about 16 years. Across the whole unit trust industry, the average return on funds that survived the whole period was 13.7% per annum, while the average return on funds that

32 were wound up or merged during the period was 11.3% per annum. This implies that a typical unit trust investor might find himself locked into an underperforming trust that is eventually wound up or merged into a more successful trust, experiencing an underperformance of 2.4 percentage point per annum over a 16 year period. This translates into a fund value that is 19% lower after 16 years than a fund that is not wound up or merged."

So fund survivorship of 60%? Hmmm... worse than I thought. And it drags the annualised performance of funds down 2.5%. The FSA also concluded it would be more had they included things like initial charges, even moreso when compared to index trackers which have lower AMC, no initial, and less often initials since they rarely close down and invest in another fund, resulting in further initial fees to investors.

All interesting and justified data which no doubt you will disagree with. But the facts, as they say, do not lie.

fg22

Aegis wrote: »

The accepted form of the efficient market hypothesis (i.e. the weak form) rules out the possibility of using past performance to guarantee future performance, but it in no way rules out the requirement for good asset allocation, investments into funds which have stock picking based on strong fundamentals (something not ruled out by the EMH unless you accept the strong form in the face of all evidence)

It's actually the semi-strong version which is the most debated and for which there is inconclusive evidence either way. If believed this does rule out fundamental analysis as a way of producing consistently higher, risk adjusted returns.

Given the apparent inefficiencies I see in the markets and the likes of Warren Buffet, I guess in the end I come down on the side of not believing in EMH.

However, I still own index trackers (as well as active funds). Why? Because markets although not quite efficient appear to be pretty close to it. In some markets I don't believe it's possible to pick good retail fund managers who have demonstrated superior knowledge/ability and hence justify their higher fees (remembering that even monkeys are likely to outperform the market at some point!).

I'd be quite interesting to seeing an analysis of how the top 10 fund managers recommended in 2000 by advisors performed in subsequent years.

Aegis

xyy123 wrote: »

Wasn't what I was claiming at all. But clearly there is more attached value to financial advisors recommending actively managed products since it is something that a lay investor has less knowledge of and create a market for a service that, arguably, does not need to exist. Financial advisors admitting picking actively managed funds as an exercise in futility would be like McDonalds coming out and saying its customers should not eat Big Macs as they were bad for their health.

There's nothing to admit. Good managers can and do beat the index, the performance tables support that (and so does one of the sources you posted, for that matter), certainly for the UK All Companies sector.

Cannot find any evidence to back this up, maybe I'm looking in the wrong places. All I can see is that both funds are taxed and pass capital gains taxes onto the shareholders but because index trackers only trade a few times per year these will obviously be less.

Indeed, that's where the difference comes from. The fact that the US system encourages the fund managers to buy and hold rather than trade puts a hugely unfair disadvantage on mutual fund managers trying to beat the index. From wikipedia:

According to a study conducted by John Bogle over a sixteen-year period, investors get to keep only 47% of the cumulative return of the average actively managed mutual fund, but they keep 87% in a market index fund. This means $10,000 invested in the index fund grew to $90,000 vs. $49,000 in the average actively managed stock mutual fund.

So effectively this is saying that tracker funds beat managed funds only when managed funds have to give up half of their returns.

If the total returns are taxed twice as much for managed funds as for trackers in the US but not for the UK, the comparison is utterly invalid.

But, regardless of this fact, I found a couple of interesting studies for you to read:

1. www.investmentfunds.org.uk/press/2002/20021014-01.pdf

and the FSA followup:

2. http://www.fsa.gov.uk/pubs/other/pastperf_mutalfunds.pdf

The first states:

"Clearly the treatment of the 434 funds, almost half the sample, that die during the period of the study is crucial."

So a survivorship of just over 50%...

"...funds in the All-Companies sector in the bottom decile (bottom 10%) there is a 31% chance of dying within the next year. In contrast for the top five deciles (top 50%) there is an average probability of only 5%."

So mainly the poor performing funds dying.

*Note that this was a study commissioned by the IMA, not designed to but a corrollary of which was to show how past performance can be used as a tool for future indication*

This study's ultimate conclusion is that persistence is actually quite an important feature, i.e. that picking funds based on past performance will beat picking random funds because if you pick something from the top quartile it has a much better than even chance of being in the top two quartiles for the coming years too. Since the top quartile is almost invariable filled with managed funds rather than trackers, this would imply that the study actually concludes that managed funds really do offer good returns.

#2 Says amongst other things:

"6.7. The CRA2 study should be commended for using a large sample of funds including dead funds and accounting for the effect of survivorship bias. Survivorship bias (the effect on reported returns of excluding funds that have died some time during the sample period) can lead to important distortions in the statistical analysis. Our earlier work (Lunde, Timmermann and Blake (1999)) found that underperforming unit trusts do eventually merge with more successful unit trusts, but that on average it takes some time for this to occur. Overall, our study found that about 40% of funds over the period 1972-1995 were eventually wound up or merged. The most frequently occurring (modal) duration for a fund was 4.25 years (51 months), but the average duration was about 16 years. Across the whole unit trust industry, the average return on funds that survived the whole period was 13.7% per annum, while the average return on funds that

32 were wound up or merged during the period was 11.3% per annum. This implies that a typical unit trust investor might find himself locked into an underperforming trust that is eventually wound up or merged into a more successful trust, experiencing an underperformance of 2.4 percentage point per annum over a 16 year period. This translates into a fund value that is 19% lower after 16 years than a fund that is not wound up or merged."

So fund survivorship of 40%? Hmmm... worse than I thought. And it drags the annualised performance of funds down 2.5%. The FSA also concluded it would be more had they included things like initial charges, even moreso when compared to index trackers which have lower AMC, no initial, and less often initials since they rarely close down and invest in another fund, resulting in further initial fees to investors.

40% in 23 years means fewer than 2% of funds shut a year. That's really not all that many, especially when you consider how much investment rules, taxation, etc, changed over the period of study.

There are a couple of comments where they claim that the worst performing funds are the most likely to close, however this is going to have less of an effect than you might hope for, especially since they come predominantly from the bottom of the performance table, while the study maintains that it is possible to pick funds that persistently outperform and are therefore unlikely to close.

Rebalance your portfolio every year and you've got a very good chance of having most of your funds in the top two quartiles, i.e. beating trackers.

You will notice that I and others give people some introductory guidance on spotting the overly expensive underperformers most likely to be in the bottom half of the performance tables, and therefore most likely to close.

All interesting and justified data which no doubt you will disagree with. But the facts, as they say, do not lie.

I don't think the data says what you think it says. After all, both studies conclude that persistence happens, i.e. that the funds in the top quartile are likely to continue to outperform.

Aegis

fg22 wrote: »

It's actually the semi-strong version which is the most debated and for which there is inconclusive evidence either way. If believed this does rule out fundamental analysis as a way of producing consistently higher, risk adjusted returns.

Given the apparent inefficiencies I see in the markets and the likes of Warren Buffet, I guess in the end I come down on the side of not believing in EMH.

However, I still own index trackers (as well as active funds). Why? Because markets although not quite efficient appear to be pretty close to it. In some markets I don't believe it's possible to pick good retail fund managers who have demonstrated superior knowledge/ability and hence justify their higher fees (remembering that even monkeys are likely to outperform the market at some point!).

I'd be quite interesting to seeing an analysis of how the top 10 fund managers recommended in 2000 by advisors performed in subsequent years.

I'll agree, that would be a great study to see, and certainly much better than other data available.

However, I think a better study would be one that looks at the funds recommended in any one year followed by the relative performance over the next 5 or so. Beyond that most clients in a servicing relationship would have had their fund choices reviewed a couple of times, and if a fund had become unpopular it's more than likely that a switch would have occurred.

xyy123

Aegis wrote: »

There's nothing to admit. Good managers can and do beat the index, the performance tables support that (and so does one of the sources you posted, for that matter), certainly for the UK All Companies sector.

If you'd cared to read the studies, as I did, you'd have seen how the first study was basically flawed, documented by the FSA's response. I used it merely to highlight the level of survivorship.

Indeed, that's where the difference comes from. The fact that the US system encourages the fund managers to buy and hold rather than trade puts a hugely unfair disadvantage on mutual fund managers trying to beat the index. From wikipedia:

Double standards. When I used a wikipedia source you rubbished it.

Apart from that... maybe... just maybe... over-trading is a key factor in failing to beat a fund's benchmark? Which would tie in with either the strong or the weak version of the EMH...??

This study's ultimate conclusion is that persistence is actually quite an important feature, i.e. that picking funds based on past performance will beat picking random funds because if you pick something from the top quartile it has a much better than even chance of being in the top two quartiles for the coming years too. Since the top quartile is almost invariable filled with managed funds rather than trackers, this would imply that the study actually concludes that managed funds really do offer good returns.

Again, either you haven't bothered to read either or both or misunderstood. This study 'concluded' there was a performance persistence. The FSA demonstrated how the study was invalid. The FSA did concede there was some correlation between bottom quartile funds, i.e. being somewhat indicative of remaining in the bottom quartile, but not any other quartile.

Further, the study shows an approximately 25-30% chance of the top quartile funds remaining top quartile. In other words, there was more chance of a fund not being in the top quartile the following year.

40% in 23 years means fewer than 2% of funds shut a year. That's really not all that many, especially when you consider how much investment rules, taxation, etc, changed over the period of study.

Er... what?! It means over that time period 40% of funds closed... if 2% of funds closed on average per year then over a 23 year period then it would still be 2%, not 40%. Mental.

Rebalance your portfolio every year and you've got a very good chance of having most of your funds in the top two quartiles, i.e. beating trackers.

Wrong. I guess you ignored the Morningstar reference I gave you with iShares FTSE 100 being 'above average' in its category of UK equity large cap. And the fact that only 37% of the funds on the site managed to beat the becnhmark. (Which doesn't include those that closed because they're no longer registered.)

I don't think the data says what you think it says. After all, both studies conclude that persistence happens, i.e. that the funds in the top quartile are likely to continue to outperform.

Again, no it doesn't:

"Studies in the empirical literature on persistence in mutual fund performance provide fairly strong evidence of persistence in negative performance, but much weaker evidence of persistence in positive performance."

"This is not because of the traditional argument over whether superior performance might or might not persist, which we regard as inconclusive, but rather because of the evidence that inferior performance seems to persist."

"It is far better for the FSA to publish past performance data on a risk-adjusted basis for all funds and to give a ‘health warning’ along the lines that the mass of existing empirical research shows that ‘losers generally repeat, while winners do not necessarily repeat’."

Need I go on?

The point I am making here, which has been lost, is that trackers will be beaten by some funds over any time period you wish to measure. Of course they will. Its pretty much impossible that they won't be. However, as the FSA reference shows, performance persistence in top performing funds is inconclusive. By law of averages, some funds will remain in the top quartile the following year - about 1/3 - 1/4. Then the following year they have about the same chance of remaining.

Overall, unless you're very lucky (or extremely skilful), then you won't have:

1) Only one fund in your portfolio that remains in the top quartile for a sustained period of time.
2) Even more unlikely, a number of funds in your portfolio that remain in the top quartile over a sustained period of time, since the odds of each of those funds remaining top quartile the subsequent year, all things being equal, is about 25-30%.

Once you take this into consideration, with the costs associated with switching funds around, initial fees, advisory fees when you want to rebalance, over the long haul you are simply more likely to be better off in staying with a boring old tracker.

Aegis

xyy123 wrote: »

If you'd cared to read the studies, as I did, you'd have seen how the first study was basically flawed, documented by the FSA's response. I used it merely to highlight the level of survivorship.


Double standards. When I used a wikipedia source you rubbished it.

I rubbished it because it was the wrong tax system, not because it was wikipedia.

Apart from that... maybe... just maybe... over-trading is a key factor in failing to beat a fund's benchmark? Which would tie in with either the strong or the weak version of the EMH...??

So you're drawing false conclusions and claiming to be correct?

The simple fact is that you cannot take a study done in a system which offers an advantage to non-managed funds and claim that it's due to overtrading where internal tax means that almost twice as much of the gross return is lost in a managed fund than in a passive one.

How on earth can you draw any conclusions other than "the tax has a very large effect" from that?

Again, either you haven't bothered to read either or both or misunderstood. This study 'concluded' there was a performance persistence. The FSA demonstrated how the study was invalid. The FSA did concede there was some correlation between bottom quartile funds, i.e. being somewhat indicative of remaining in the bottom quartile, but not any other quartile.

Even if you accept the FSA's , if funds in the bottom quarter are more likely to stay in the bottom quarter, that would imply that funds selected from the upper three quartiles will tend to remain in the upper three quartiles. If trackers stay more or less mid table, which they do, this still means that picking a fund at random from the top quartile should still, on average, beat the trackers.

Of course, picking at random would be silly, which is why I think that these studies are generally pretty meaningless.

Further, the study shows an approximately 25-30% chance of the top quartile funds remaining top quartile. In other words, there was more chance of a fund not being in the top quartile the following year.

Yes, but importantly, there was a much higher chance of a top quartile fund ending up in the top two quartiles than the bottom two quartiles.

Er... what?! It means over that time period 40% of funds closed... if 2% of funds closed on average per year then over a 23 year period then it would still be 2%, not 40%. Mental.

So you're telling me that if you have a sample size of 100 and 2 close each year, then 2 have closed in 23 years?


Anyway, I'm bored of this discussion now. My stance is that managed funds can beat index trackers, and that decent managers have a good chance of doing so regularly. I don't accept your stance, based largely on the actual performance tables over the last 5 and 10 years for the majority of sectors.

End of story really.

jamesd

You might find this story in the New York Times of interest when it comes to taxes on US funds:

"Expenses were the culprit. For both the actively managed fund and the hedge fund, those expenses more than ate up the large amounts — 3.5 and 9 percentage points a year, respectively — by which they beat the index fund before expenses."

'What if you’re investing in a tax-sheltered account, like a 401(k) or an I.R.A.? In that case, Mr. Kritzman conceded, the odds are relatively more favorable for active management, because, in his simulations, taxes accounted for about two-thirds of the expenses of the actively managed mutual fund and nearly half of the hedge fund’s. But he emphasized the word "relatively." "Even in a tax-sheltered account,” he said, “the odds of beating the index fund are still quite poor."'

"he assumed that long-term capital gains were subject to a 15 percent federal tax and a 6.85 percent state tax; short-term capital gains and dividends were taxed at a combined federal and state rate of nearly 42 percent. The index fund’s average after-expense return was 8.5 percent a year, versus 8 percent for the actively managed fund and 7.7 percent for the hedge fund."

I'll assume that you know that in the UK short term capital gains and dividends within a fund are not taxed at 42% compared to 21.85% for longer term gains. In the US the effect of this tax difference is to impose a significant tax penalty on funds that are more active. Enough to get rid of the outperformance that their active management generates.

That's why the tax difference is a key driver for the popularity of ETFs and simple trackers in the US: the US tax system strongly favors them. For the same reason it favors simple buy and long term hold strategies for shares.

Applying methods that are based on the quirks of the US tax system isn't prudent for a UK based investor, except in areas where the UK system has similar tax quirks.

xyy123

Even if you accept the FSA's , if funds in the bottom quarter are more likely to stay in the bottom quarter, that would imply that funds selected from the upper three quartiles will tend to remain in the upper three quartiles.

Until those in the bottom quartiles die and the dataset is lost, improving the average performance of active funds overall.

If trackers stay more or less mid table, which they do

iShares FTSE 100 = Above Average. i.e. not mid-table.

Yes, but importantly, there was a much higher chance of a top quartile fund ending up in the top two quartiles than the bottom two quartiles.

I never saw this mentioned in the study, actually. Simply because a fund is in the top quartile in year 1 does not mean it cannot be in the bottom quartile in year 2.

Anyway, I'm bored of this discussion now. My stance is that managed funds can beat index trackers, and that decent managers have a good chance of doing so regularly. I don't accept your stance, based largely on the actual performance tables over the last 5 and 10 years for the majority of sectors.

End of story really.

Not really. I checked out your so-called evidence on Morningstar.

http://www.morningstar.co.uk/uk/fundscreener/results.aspx?lang=en-GB&Category=EUCA000550&Universe=FOGBR%24%24ALL&InvestorType=-1&PEAStatus=-1&EquityStyle=0%7c1%7c0%7c0%7c0%7c0%7c0%7c0%7c0&Return=NULL%7cNULL%7cNULL%7cNULL%7cNULL%7c

Of the top 100 performing over 5 years out of 354 funds in UK large cap blend (FTSE 100), 19 were index funds.

Of those 100, 42, almost half, were 'Above Average Risk' or 'High Risk' relative to sector. I guess that's one way to outperform the benchmark - just increase the level of risk in the fund and over time some will do better. This is what the FSA study also alluded to - funds in the top quartile will invariably be higher risk since over time higher risk funds have a better chance of success and those that don't work out, get closed. Only 14 were 'Below Average Risk' or 'Low Risk' relative to sector, which is what we all want really - less than average risk with above average returns, including:

• Majedie Asset UK Equity B Acc
• Newton Income Ex 2 Net Acc
• BlackRock Armed Forces Common In...
• Morgan Stanley UK Equity Alpha I...

Not exactly household name funds we're talking about here.

I can understand why you think the way you do, I honestly can. IFAs and those that work with them have little choice to think any differently. Apart from the opinion coming from a private investor with a background on the buy-side arguing the case for passive investing, standing to gain nothing, against someone who's defending his job and industry, the weight of evidence for passive vs active investing is overwhelming. I never suggested trackers will (or ever) beat actively managed funds in any particular year, but its simply a question of odds.

What are the odds of me choosing, or being advised, only funds which outperform their designated benchmark for the majority of the time I hold them? The answer is - low. The risk is therefore - not worth it. Slow and steady wins the race.

Trackers > Managed funds.

fg22

In a totally unscientific study I've looked at how 6 IFA's UK choices have performed over the last few years. Selections taken from http://www.thisismoney.co.uk/ifafunds as at March 2006.
I've then put a + or - by them if I think they've out or underperformed a FTSE all-share tracker, however this might be wrong as I've just looked at rough graphs of total returns.

UK Growth - Rensburg UK Select Growth -
UK Income - Invesco Perpetual Income +

UK Growth - Artemis Capital -
UK Income - Rathbone Income -

UK Growth - Standard Life Growth & Income (think this is now uk equity high alpha)+
UK Income - Invesco Pertepual Income +

UK Growth - New Star UK Select Growth -
UK Income - Invesco Perpetual UK Higher Income +

UK Growth - Liontrust First Growth -
UK Income - Jupiter Income -

UK Growth - M&G UK Select +
UK Income - Legal & General Fixed Interest - err this is a corporate bond fund

So it looks like only 5/11 beat the tracker - not the best of results. Both of H-L's Mark Dampier's choices seem to have done pretty poorly.