An IFA who knows Monkey with a Pin

Linton

grey_gym_sock wrote: »

1) we can predict that some segments will do better than others, but can we predict which 1s? if it were obvious, wouldn't the market adjust the prices immediately, so that the expected difference in future return disappeared?

2) this is only relevant if an ethical, or income, or defensive, investment remit is actually likely to give below-average returns. which is far from obvious. if defensive includes sitting on a lot of cash, it's probably true for it, but that doesn't help with beating (e.g.) an equity index.

You can check by looking at the historical data from trustnet. For ethical - I found 1 ethical fund in the top quartile and 7 in the bottom quartile for 5 year growth for funds in the UK AllShare sector.

Equity Income sector averages are pretty much the same as the Allshare.

Another very clear effect is the midcap funds within the Allshare sector. Over 10 years 7 out of the top 20 funds had "midcap" or (FTSE)250 in their name. I didnt find any below that. Why should midcap funds across the piece all be in the top 13% of the league table if performance is random?

These examples I believe demonstrate that the model that fund managers effectively pick shares at random is not a good one.

One general effect I have noticed with the successful niche sectors is that they are in general significantly more volatile than the main indexes. They also tend to do relatively better in the good years compared with the broad sector index than they do poorly in the bad years. As we get more good years than bad ones, after all we wouldnt be investing if that wasnt the case, the niche sectors tend to outperform the broad indexes over time.

Why doesnt the perfect market cause the outperformance of niche sectors to disappear in a short timeframe? One possible answer is that they are seen as "risky" and specialist and so they are avoided by many individual investors and perhaps more importantly by restricted environments such as employer pensions. Another reason could be that trackers and perhaps many general funds are heavily weighted towards larger companies.

A third potential reason is that sector outperformance is a long term effect. Many fund managers would appear to be more concerned with taking advantage of short term opportunities. For that type of investing there are disadvantages in gambling with smaller less well researched shares and no obvious advantages.

jamesd

gadgetmind wrote: »

Please can I have links to these studies as they seem to have all come up with very different conclusions to the large body of academic research that I have encountered to date?

All have been covered in past discussions here, so looking back at them should find them.

jamesd

grey_gym_sock wrote: »

if trackers hold a constant 90% of XYZ plc, and active managers the other 10%, then any relative gains 1 active manager makes by going over- (or under-)weight in XYZ plc in a period when it out- (or under-)performs, are matched by relative losses for other active managers.

It's the tracker holders who are the losers in this situation, because the active funds buying or selling first either increase or decrease the price the trackers have to pay or receive for their own transactions.

grey_gym_sock wrote: »

there is also the possibility for active managers as a group to outperform by anticipating when trackers as a group will have to buy or sell due to changes in demographics, savings rates, etc. i.e. trackers may need to invest another £1bn, or take out £1bn, thus increasing their holding in XYZ plc to 90.01%, or decreasing it to 89.99% (and the same percentages for every other company on the market). active managers may be able to ensure that prices are higher when trackers are buying (so they get fewer shares for their money), and lower when they're selling (so they have to sell more shares to raise the required money).

The opportunity I was thinking of was changes in index composition. When a new share enters an index the trackers are forced to buy or accept an increased tracking error.

grey_gym_sock wrote: »

this suggests to me that, if you're trying not to be too clever, being "buy and hold" may be at least as important as being passive.

Buy and hold has other problems, though. But I assume you don't mean to be a purist and hold a predictably declining stock.

jamesd

gadgetmind wrote: »

BTW, here is how active funds are quietly side-lined and then closed when they underperform.

http://www.investmentweek.co.uk/investment-week/news/1651965/axa-liquidate-uk-small-cap-alpha-fund

Q: Where is this one in the data?
A: Conveniently removed from it.

Here's how that starts:
" Axa Rosenberg is to shut down its UK Small Cap Alpha fund as its assets have dwindled to £3m. The fund, which is managed by Gideon Smith, halved in size in the 12 months to March 2009" It was an unpopular fund that investors had rejected, with very little money invested in it, so as a portion of the total market it was insignificant. But it'd get the same
weighting as every other fund in the studies using averages. For context, Old Mutual's UK Select Smaller Companies fund currently has £556.6 million under management. Note that the amount invested in the closed fund had halved: investors bailed out instead of sitting still and accepting the under-performance.

That's a fine example of one of the massive systematic flaws in studies that promote passive funds: the use of average performance. Add in the failures that closed because nobody was interested due to their poor performance and you make that flaw even worse.

This sort of thing should be obvious. There are at least three Vanguard managed FTSE index tracker funds, with fees of around 1% for two of them (via Standard Life, one Vanguard brand, one an SL fund it's managing under contract) and around 0.3% from some other places. The studies and ETF fans then go on to argue this way:

The average price for Vanguard managed FTSE tracking is 0.77%. That's way too high, you can get 0.3% in ETFs, so it's obvious that buying even the tracker funds of a company that claims to be cheap is a rip-off and everyone should buy the cheaper ETFs instead and get the extra 0.47% performance that they give compared to the average performance of Vanguard funds.

Just substitute active for tracker and tracker for ETF and you have one of the flaws in a nutshell: the use of averages distorts the results because those with a free choice would pick the around 0.3% Vanguard fund, not the 1% versions. Same for active funds: the consistently poor ones don't get so much money invested in them.

If you're looking at fund performance one of the things you need to do if you're trying to provide an accurate study is be sure that you're using performance weighted by amount invested. If you're after performance for investors paying attention you should also be looking at the free market funds, not those that people who are locked in end up stuck with. The performance of that fund with 556.6 million in it matters much more to investors than the one with three million in it because the big one is where the money is and hence is a big part of the average performance that investors end up seeing, a very different thing from the average fund performance.

There's a nice piece by Merryn Somerset Webb in this Saturday's Financial Times, "A new take on the active vs passive debate" which says in part:

"Take the equity income sector for example. At the beginning of 2012 there were 92 funds in it. But Neil Woodford, via just three Invesco Perpetual funds, was responsible for running 40 per cent of all the money invested in it; his funds together were bigger than 84 of the remaining 89 funds combined. Yet if you simply added up the performance of the funds and divided by 92, as any conventional analysis would, his funds would account for only 3.3 per cent of the sector performance.

The point is that if you want to see if the average investor (rather than the average fund) is beating the market you need to calculate a weighted average of fund performance. This is, after all, how most stock indices are run – they are weighted by market capitalisation. Redefine average like this, says Evan-Cook, and you find that over the past five years, “in almost all cases the average investor is doing better than the average fund”."

But it's harder work to do properly weighted studies that show what investors really do and it's unlikely that tracker fund managers would want to pay for the research because it's less likely to show what they want shown.

gadgetmind

jamesd wrote: »

It was an unpopular fund that investors had rejected, with very little money invested in it, so as a portion of the total market it was insignificant.

By the time it closed it certainly was insignificant, but what about earlier when it was underperforming?

Note that the amount invested in the closed fund had halved: investors bailed out instead of sitting still and accepting the under-performance.

Yes, all that hot money charged off somewhere else trying to find outperformance, and it will have gone into some active fund that had done well recently. And then along comes reversion to mean.

Same for active funds: the consistently poor ones don't get so much money invested in them.

Such consistency of either under or over-performance is not something that active funds show in any way that's useful to an investor.

Neil Woodford, via just three Invesco Perpetual funds, was responsible for running 40 per cent of all the money invested in it; his funds together were bigger than 84 of the remaining 89 funds combined.

Yes, hot money does tend to flock together in that way, and this performance-chasing surge is part of what changes the fortunes of funds.

There's a very good section in "Smarter Investing" on exactly this, and if I can get my copy back next week I'll pull out some quotes and references to the key studies.

BTW, various of those IP income funds have been taken off the well-known "white list" for exactly this reason.

it's unlikely that tracker fund managers would want to pay for the research because it's less likely to show what they want shown.

Trackers are rather light on managers and marketing. Perhaps the active funds, which invest heavily in both of these, might like to fund some studies that investigate what active investors manage to achieve by hopping around from fund to fund in the hope of getting it right next time?

doughnutmachine

jamesd wrote: »

But it's harder work to do properly weighted studies that show what investors really do and it's unlikely that tracker fund managers would want to pay for the research because it's less likely to show what they want shown.

??? would it not be easy to look at a sample of unit trust investors portfolios and see how they compare against a tracker?

ohhh blast someone has beat me to it!!

"In a study of US mutual fund investor behaviour from 1991 to 2004 Friesen and Sapp (2007) find that the typical investor loses 1.56% per year due to the timing of their investment decisions. And since they use disaggregated data they are able to identify, for example, that this performance loss averages 0.13% per month for actively managed funds but only 0.05% per month for passive funds. They also find that there is a greater degree of poor investment timing associated with better performing funds."

and in case someone says US investors are less smart than us british ones....

"Using data collated by the Investment Management Association (IMA)3 that spanned the period from 1992 to the end of 2009 Clare and Motson (2010) examined market level data on the net investment into broad categories of UK mutual funds (known in the UK as unit trusts). Their results, the first of their kind on non-US mutual funds, were broadly in keeping with those achieved using US mutual fund data. As such their work suggests that the performance gap phenomenon is probably global rather than being confined to any one set of investors from a particular market. Clare and Motson find that on average UK retail investors have lost performance equivalent to just under 1.2% per year over the eighteen year period of their study. Although 1.2% may not sound very high, compounded over 18 years it represents a cumulated underperformance of 20%, compared with a simple buy and hold strategy. By distinguishing between retail and institutional flows into these funds, Clare and Motson also show that the UK’s performance gap phenomenon is a feature that is almost entirely due to retail investor behaviour, since the performance gap for institutional flows is virtually 0.00%pa."

http://www.cass.city.ac.uk/__data/assets/pdf_file/0005/69926/comparing-the-performance-of-retail.pdf

nice idea about retail investors being able to outperform the market, but the evidence doesn't back your thesis

Linton

doughnutmachine wrote: »

??? would it not be easy to look at a sample of unit trust investors portfolios and see how they compare against a tracker?

ohhh blast someone has beat me to it!!

"In a study of US mutual fund investor behaviour from 1991 to 2004 Friesen and Sapp (2007) find that the typical investor loses 1.56% per year due to the timing of their investment decisions. And since they use disaggregated data they are able to identify, for example, that this performance loss averages 0.13% per month for actively managed funds but only 0.05% per month for passive funds. They also find that there is a greater degree of poor investment timing associated with better performing funds."

and in case someone says US investors are less smart than us british ones....

"Using data collated by the Investment Management Association (IMA)3 that spanned the period from 1992 to the end of 2009 Clare and Motson (2010) examined market level data on the net investment into broad categories of UK mutual funds (known in the UK as unit trusts). Their results, the first of their kind on non-US mutual funds, were broadly in keeping with those achieved using US mutual fund data. As such their work suggests that the performance gap phenomenon is probably global rather than being confined to any one set of investors from a particular market. Clare and Motson find that on average UK retail investors have lost performance equivalent to just under 1.2% per year over the eighteen year period of their study. Although 1.2% may not sound very high, compounded over 18 years it represents a cumulated underperformance of 20%, compared with a simple buy and hold strategy. By distinguishing between retail and institutional flows into these funds, Clare and Motson also show that the UK’s performance gap phenomenon is a feature that is almost entirely due to retail investor behaviour, since the performance gap for institutional flows is virtually 0.00%pa."

http://www.cass.city.ac.uk/__data/assets/pdf_file/0005/69926/comparing-the-performance-of-retail.pdf

nice idea about retail investors being able to outperform the market, but the evidence doesn't back your thesis

The article doesnt seem to me to say anything about trackers vs active. What I read it as saying is that private investors on average lose money by the timing of their buy and sell decisions and that a buy and hold strategy would be on average better for them. There is nothing to say that it makes any difference whether these funds are passive or managed. Indeed, their worked example compares something that follows the FTSE100 being bought and sold as the average investor does with a hypothetical fixed duration derivative based product that looks similar to ones you can get from the banks but rather more generous.

Since these figures are averages, it could be that the majority of investors, whether in trackers or managed funds are naive in their investing buying when the going looks good and selling when its bad and so losing potential return. From the results there is nothing to say that there arent a minority of sophisticated investors gaining return with a different buy/sell approach.

Interestingly it also says that this average poor investing isnt found amongst the professionals.

Have I missed something?

gadgetmind

Linton wrote: »

What I read it as saying is that private investors on average lose money by the timing of their buy and sell decisions and that a buy and hold strategy would be on average better for them.

Yes, but how do we square this with your view that the average investor is good at picking which active funds to use, and furthermore is able to sense when they've gone off the boil and switch to one that's going to perform better?

I'm not sure what the solution is for the average investor as they seem to be their own worst enemies.

What then of the more clued-up investor? Can they identify the active funds to use? Well, the evidence for sustained outperformance (and underperformance) is pretty thin, and "hot hands" more of a marketing myth than an investing reality.

If some think they can do this, and at the same time ensure the fund's holdings when aggregated across their entitre portfolio achieves the asset mix and risk level they desire, then hats off to them.

While I'm not going to say it's impossible, I certainly don't think I could do it well, and I have objective evidence to hand that suggests I'm smarter than the average bear.

I prefer to use trackers to gain exposure to the major markets (tilted slightly towards value, small/mid caps and more growthy areas), add some bonds, REITs and infrastructure to the mix, season *lightly* with some active investments (mainly ITs) in the more specialist areas, rebalance every year or so for a few decades and then serve.

Simple, effective, mostly hands-off, shown to give top-notch results, and avoids having to keep monitoring whether Red Leicester is still managing the "First Mutual Global Strategic Unconstrained Absolute Return Enhanced Accumulation Fund" or not.

Linton

gadgetmind wrote: »

Yes, but how do we square this with your view that the average investor is good at picking which active funds to use, and furthermore is able to sense when they've gone off the boil and switch to one that's going to perform better?

Investors can be good at choosing sectors that outperform in the long term. My intention is to hold for the long term. Usually the complete dogs are pretty obvious, as are the small funds, and the funds from people you have never heard of. The only buy and sell decisions are in the eventuality that you do happen to chose a dog. As long as the fund isnt significantly worse than average over a few years I would stay with it. In any case switching from one fund to another in the same sector isnt quite the buy and sell decisions that the paper is about which I would think are more to do with switch to cash for a while.

wrote:

I'm not sure what the solution is for the average investor as they seem to be their own worst enemies.

Many are, I am sure.

wrote:

What then of the more clued-up investor? Can they identify the active funds to use? Well, the evidence for sustained outperformance (and underperformance) is pretty thin, and "hot hands" more of a marketing myth than an investing reality.

In my view choosing one fund rather than another is a secondary concern. The important thing is the sector asset allocation. I agree that hot hands is largely a myth, though not completely. Perhaps it can be the case that a fund manager's natural style is appropriate for a particular market for a particular possibly lengthy time period.

.......

wrote:

I prefer to use trackers to gain exposure to the major markets (tilted slightly towards value, small/mid caps and more growthy areas), add some bonds, REITs and infrastructure to the mix, season *lightly* with some active investments (mainly ITs) in the more specialist areas, rebalance every year or so for a few decades and then serve.

I started with a core/satellite approach, but was finding it difficult to see what overall good the core was doing for me apart from some diversification into less volatile areas. But over time I havent seen volatility as a problem, even during 2008. This is mainly because I hold a fair amount of cash to pay the next few years living expenses.

Simple, effective, mostly hands-off, shown to give top-notch results, and avoids having to keep monitoring whether Red Leicester is still managing the "First Mutual Global Strategic Unconstrained Absolute Return Enhanced Accumulation Fund" or not.

Agreed that detailed consideration of one fund against another in the same sector is not worth the effort. Just the basics to ensure you havent got something appalling or something which doesnt do what its sector allocation implies it should.

Linton

As to what the less confident investor should do, if you believe the paper they should go for 5 year guaranteed capital bonds which stop them tweaking.