We'd like to remind Forumites to please avoid political debate on the Forum. This is to keep it a safe and useful space for MoneySaving discussions. Threads that are - or become - political in nature may be removed in line with the Forum’s rules. Thank you for your understanding.

Low costs are good predictor of performance

I recently came across some research that I thought was interesting:
http://www.tcfinvestment.com/sites/default/files/pdfs/Low_cost_funds_beat.pdf

It shows that in a given investment sector funds with the lowest costs outperform the funds with the highest costs. What I found interesting is that although this seems similar to the figures that are often used by fans of passive funds, this particular research was not focusing on passive funds. It is looking at the three managed sectors (active, balanced and cautious), which I don't think normally have any passive funds. So, it seems to be showing that low-cost active funds perform better than high cost active funds. (Or, at least, high cost active funds don't deliver enough gross outperformance to compensate for their higher costs.)

Of course, this does not mean that all low-cost funds deliver better net performance than high-cost funds, but choosing a low-cost fund significantly increases your chance of better than average performance.
koru
«1

Comments

  • Linton
    Linton Posts: 17,733 Forumite
    Part of the Furniture 10,000 Posts Name Dropper Hung up my suit!
    Presented as a sales document it may look pretty convincing but lets see how significant the numbers actually are:

    Looking at annual performance of the Active Managed sector over 5 years using the performance data from Trustnet we get

    1st Quartile: 1.1% to 7.4%
    2nd Quartile: 0.2% to 1%
    3rd Quartile: -1.0 % to 0.0%
    4th Quartile: -4% to -1.1%

    So it would seem that charges actually are not a very good predictor of performance as the variation between top and bottom quartiles based on charges is about 1% annually whereas that actually experienced is nearer 6% annually.

    Note that my figures are less than those presented in the advert I guess because their calculation was done a few months ago when share prices were higher.

    A second point is that the advert may imply they have carried out a full calculation. There are of course some 33 IMA sectors, they have chosen to only look at 3 of them.
  • koru
    koru Posts: 1,529 Forumite
    Part of the Furniture 1,000 Posts Name Dropper Combo Breaker
    They looked at another four categories shortly after, with the same conclusion:

    http://www.fundweb.co.uk/home/news/tcf-investment-repeats-cost-as-performance-predictor-research/1038478.article

    I note also that Morningstar did a similar analysis recently, for every asset class, and found that “in every asset class over every time period, the cheapest quintile produced higher total returns than the most expensive quintile":

    http://www.advisorone.com/2010/08/09/high-fees-are-bad-for-investment-performance

    http://news.morningstar.com/articlenet/article.aspx?id=347327

    In fact, Morningstar found that cost was a better predictor of performance than its own star ratings.

    I agree that TCF's “research" is more than a bit self-serving, and therefore should not necessarily be taken at face value. But I don't see why Morningstar would have any bias towards low-cost funds.
    koru
  • koru
    koru Posts: 1,529 Forumite
    Part of the Furniture 1,000 Posts Name Dropper Combo Breaker
    Linton wrote: »
    Presented as a sales document it may look pretty convincing but lets see how significant the numbers actually are:

    Looking at annual performance of the Active Managed sector over 5 years using the performance data from Trustnet we get

    1st Quartile: 1.1% to 7.4%
    2nd Quartile: 0.2% to 1%
    3rd Quartile: -1.0 % to 0.0%
    4th Quartile: -4% to -1.1%

    So it would seem that charges actually are not a very good predictor of performance as the variation between top and bottom quartiles based on charges is about 1% annually whereas that actually experienced is nearer 6% annually.

    Note that my figures are less than those presented in the advert I guess because their calculation was done a few months ago when share prices were higher.

    A second point is that the advert may imply they have carried out a full calculation. There are of course some 33 IMA sectors, they have chosen to only look at 3 of them.
    How does one sort the Trustnet data in order of charges? I can't see this as a data field in the comparative tables here: http://www.trustnet.com/Investments/Perf.aspx?ctr=QS&univ=O

    Are you using some other part of Trustnet?
    koru
  • Linton
    Linton Posts: 17,733 Forumite
    Part of the Furniture 10,000 Posts Name Dropper Hung up my suit!
    koru wrote: »
    How does one sort the Trustnet data in order of charges? I can't see this as a data field in the comparative tables here: http://www.trustnet.com/Investments/Perf.aspx?ctr=QS&univ=O

    Are you using some other part of Trustnet?

    I dont think one can.

    To look at charges, I am afraid you will have to do each fund in turn, and quite a few dont supply the data to trustnet.

    A non-scientific random check suggested that most in the Active Managed sector had an AMC of 1.5% with a few at 1.25%. I dont see how a difference of 0.25% in charges can directly cause a 1% difference in total return. Perhaps the higher charged ones are both inefficient and incompetent?!?! Or perhaps the data is actually very noisy.
  • koru
    koru Posts: 1,529 Forumite
    Part of the Furniture 1,000 Posts Name Dropper Combo Breaker
    Linton wrote: »
    I dont think one can.

    To look at charges, I am afraid you will have to do each fund in turn, and quite a few dont supply the data to trustnet.

    A non-scientific random check suggested that most in the Active Managed sector had an AMC of 1.5% with a few at 1.25%. I dont see how a difference of 0.25% in charges can directly cause a 1% difference in total return. Perhaps the higher charged ones are both inefficient and incompetent?!?! Or perhaps the data is actually very noisy.
    Ah, then I think I must have misunderstood your first post. I thought your data showed the relative performance of funds ranked in order of cost. Were you simply saying that the top quartile of performers outperformed the bottom quartile by about an annual average of 6%, which is a much bigger difference than the difference in TER?
    koru
  • Linton
    Linton Posts: 17,733 Forumite
    Part of the Furniture 10,000 Posts Name Dropper Hung up my suit!
    koru wrote: »
    Ah, then I think I must have misunderstood your first post. I thought your data showed the relative performance of funds ranked in order of cost. Were you simply saying that the top quartile of performers outperformed the bottom quartile by about an annual average of 6%, which is a much bigger difference than the difference in TER?


    What I am suggesting is that both high cost and low cost active managed funds include a broad range of good and bad performers. The average performance of the low cost ones may be slightly higher than the average performance of the high cost ones, but if the difference is small compared with the variation within each class then its not a very helpful predictor. Your chance of chosing at random and picking a dog may be much the same for both high and low cost funds.

    Unfortunately there doesnt seem to be the data available to prove things either way.

    Most people contributing to this often repeated discussion tend to have a line to sell or an axe to grind. My line is that looking at what a fund invests in is far more important than focussing on costs which I see as a secondary matter.
  • koru
    koru Posts: 1,529 Forumite
    Part of the Furniture 1,000 Posts Name Dropper Combo Breaker
    Linton wrote: »
    What I am suggesting is that both high cost and low cost active managed funds include a broad range of good and bad performers. The average performance of the low cost ones may be slightly higher than the average performance of the high cost ones, but if the difference is small compared with the variation within each class then its not a very helpful predictor. Your chance of chosing at random and picking a dog may be much the same for both high and low cost funds.
    Well, I agree that choosing a low-cost active managed fund is certainly no guarantee that you will get out performance. But the TCF figures seem to indicate that this would make a significant difference in the likely performance of your fund. The average low-cost active managed fund has provided 1% better total return on average per year over the last five years than the average high cost active managed fund. There are similar differences of 1% in several other investment categories, such as balanced managed, global growth and UK equity income. You might feel this is a small difference, but it seems pretty big to me.

    Clearly, this would not be so impressive if there were another way of reliably knowing in advance which funds will be in the top quartile. You appear to be suggesting that a better way of improving your odds is to examine what each fund is investing in. Personally, I wouldn't know how to judge in advance which fund managers are following a strategy that is likely to outperform. That's why I'm interested to see that there is a factor that does appear to be correlated with higher net total returns and that can be easily and objectively identified.

    I'm not saying I will never invest in high cost funds, but I am saying that unless I can see some other clear reason to prefer a high cost fund, I will make my choices from amongst the low-cost funds.
    koru
  • koru
    koru Posts: 1,529 Forumite
    Part of the Furniture 1,000 Posts Name Dropper Combo Breaker
    Linton wrote: »
    A non-scientific random check suggested that most in the Active Managed sector had an AMC of 1.5% with a few at 1.25%. I dont see how a difference of 0.25% in charges can directly cause a 1% difference in total return. Perhaps the higher charged ones are both inefficient and incompetent?!?! Or perhaps the data is actually very noisy.
    I found that Hargreaves Lansdown will give you tables of all the funds in a sector (do a Fund Search by sector) and if you select the Discounts, Savings & Charges tab, you can see the AMC for each fund. You can click to sort in order of AMC. By eyeballing the AMCs for funds in the Active Managed sector I would say that the 25% of funds with the lowest AMC probably average about 1.2% AMC whilst the 25% of funds with the highest AMC probably have an average AMC of about 1.6%. Obviously, the differences in the TER might not be exactly the same as the differences in AMC, but it seems unlikely that the low-cost funds have a cost advantage of more than 0.5%.

    And yet, on average they outperform by about 1%. Which seems to suggest that the high cost funds are, on average, performing worse than the low-cost funds even before costs are taken into account. Half of the average outperformance of low-cost funds is explained by the lower costs themselves and the other half of the performance seems to be from generating better gross total returns than the guys who are charging a higher TER.

    That's puzzling. It seems to mean that those fund managers who work for funds with below average costs are actually better fund managers than the guys who work for the funds with high costs. Perhaps, as Linton says, this is just random noise. It would have been interesting to see if TCF would have got the same results over different time periods (that is, starting in different years). Then again, the Morningstar figures seem consistent with TCF, although the Morningstar figures are for US mutual funds, not UK unit trusts.
    koru
  • I thought things like trading costs were additional overheads over and above the TER. I think I heard that active managers tend to churn their funds quite a lot. So maybe higher TER also implies additional costs due to higher churn. But this is pure speculation.
  • darkpool
    darkpool Posts: 1,671 Forumite
    Linton wrote: »
    Unfortunately there doesnt seem to be the data available to prove things either way.

    yeah, hardly evidence at all.

    "Expense ratios are strong predictors of performance. In every asset class over every time period, the cheapest quintile produced higher total returns than the most expensive quintile."
This discussion has been closed.
Meet your Ambassadors

Categories

  • All Categories
  • 346.7K Banking & Borrowing
  • 251.4K Reduce Debt & Boost Income
  • 451.4K Spending & Discounts
  • 238.9K Work, Benefits & Business
  • 614.3K Mortgages, Homes & Bills
  • 174.8K Life & Family
  • 252.1K Travel & Transport
  • 1.5M Hobbies & Leisure
  • 16K Discuss & Feedback
  • 15.1K Coronavirus Support Boards

Is this how you want to be seen?

We see you are using a default avatar. It takes only a few seconds to pick a picture.