Which S&S ISA provider - Low cost trackers

Edit: The following was the second part of an attempt to use the stats provided on fund factsheets to analyse the performance of funds over time. Please note that if you read the following without reading the rest of the thread, you may get a misleading idea of the usefulness of these stats. The first part of the attempt was Post #22. Posts #47 and #49 on this thread describe the outcome.

gadgetmind wrote: »

So what's the summary? Vanguard generally better, but only if you expect to hold for several years without changing provider/platform/funds/etc.?

My summary is that I'm a bit out of my depth. But here goes...

On the basis of very limited data (and expertise!), over the last five years the HSBC All-Share Index appears to have a material tracking difference (I’m going to stop using the term ‘tracking error’, because it appears to be a technical term with a very specific meaning, as explained here).

gadgetmind wrote: »

1 yr, HSBC=6.7%, FTSE=7.3%
3yrs, HSBC=5.9%, FTSE=8.5%
5yrs, HSBC=6%, FTSE=10.7%

Five-year tracking difference: -4.7%

Compounded tracking difference due to TER: (((((100 x 0.9936) x 0.9936) x 0.9936) x 0.9973) x 0.9973) – 100 = -2.44%

Compounded tracking difference not due to TER: -4.7% minus -2.44% = -2.26%

(NB: The fund TER changed to 0.27% on 1/9/2009 and according to Rollinghome's post from the other thread, the TER prior to the change was 0.64%. I don’t know that it was 0.64% for the whole period 2006-2009, but that’s the only source I have. If the TER was higher than 0.64% at any point after 1/9/2006, then the compounded tracking difference due to TER above is understated and the compounded tracking difference not due to TER is overstated.)

Of course, past performance is no guide to the future. But if this tracking difference were to persist, and if the Vanguard All-Share fund were to continue (as it appears) to lose more or less its TER against the index, and assuming other charges were to remain constant, then to compare the HSBC and Vanguard funds purely on the basis of TER, initial charges and platform/wrapper/dealing fees would be to slightly overstate the performance of the HSBC fund relative to the Vanguard fund. Whether this slight overstatement of performance would be enough to justify the extra fees for the Vanguard fund (initial dilution levy+platform/wrapper/dealing fees) would depend, I think, on the level of investment, the term of the investment, and the precise platform/wrapper/dealing fees.

This is my hunch, anyway. I’m reluctant to call it more than a hunch, because I’m not confident that I have the knowledge and experience to interpret the data in a meaningful way. Also, I’m only too aware of the difficulty of making meaningful comparisons between trackers from different fund managers, as argued by Snowman in that thread linked to by JamesU.

If you think the above figures are useful and would like to calculate similar figures for the other HSBC funds, here is the performance data:

HSBC European
1 year, HSBC -13.0, FTSE -13.6
3 years, HSBC -0.6, FTSE 2.3
5 years, HSBC -7.0, FTSE -1.4

HSBC American
1 year HSBC 3.6 S&P 1.7
3 years HSBC 22.4 S&P 16.3
5 years HSBC 8.0 S&P 9.2

HSBC Pacific
1 year HSBC -8.9 FTSE -8.6
3 years HSBC 44.1 FTSE 49.5
5 years HSBC 42.8 FTSE 52.8

HSBC Japan
1 year HSBC 2.2 FTSE 1.9
3 years HSBC 11.5 FTSE 14.3
5 years HSBC -10.4 FTSE -6.6

Source: HSBC fund factsheets dated 30 September 2011

You can get the TERs from that Rollinghome post. Note that these are 30/9/2011 figures, not 31/8/2011, so when calculating the TER, it’s not 3 years on the old and 2 years on the new but 3 years minus one month and 2 years plus one month. Having the 31/8/2011 data would make it simpler!

The most obvious feature of these figures is the smaller tracking difference of the American index fund (again, if you don't mind your tracker outperforming the index!). Perhaps this is because the US has no SDRT equivalent. (As an aside, the current situation with HSBC and Vanguard trackers seems somewhat analogous to the Investment Trust/Unit Trust debate. Broadly speaking, it seems that less dealing for ITs means lower dealing costs and therefore less drag on performance, but this is balanced for the most part by higher fees to buy into the fund.) I haven’t done the detailed calculations, but another feature of these figures is that the tracking differences on the other funds seem less obvious than that on the FTSE All-Share fund. For example, the European fund has a cumulative tracking difference of -5.6% over five years, but if you consider that the fund’s TER prior to 1/9/2009 was 1.2% (!), that doesn’t look too bad. It’s a similar story for the Japan fund.

It’s difficult to go further in comparing the HSBC and Vanguard trackers because only the UK and European funds track the same index. In the case of the US, Pacific, and Japan trackers, the HSBC and Vanguard funds track different indexes, and the variations in the performance of the indexes might play as big a part as the variations in performance between the funds. Similarly, the HSBC and Vanguard UK gilt trackers also track different indexes and, although I haven’t looked closely at the figures, at first sight the variations in performance between the indexes look quite marked, more so than with the equity funds. But analysing the components of the bond indexes and whether it would be preferable to track one or the other feels like a step too far for me.

You've probably seen this already, but other people (if they've read this far!) might be interested in this:
http://monevator.com/2011/01/18/tracking-error-%E2%80%93-a-hidden-cost/

Edit: Oops - original post had a reference to 'HSBC UK index-linked gilt tracker', which doesn't currently exist. FYI, Vanguard does do a UK index-linked gilt tracker.

JamesU wrote: »

Nufc has been lost on the thread...anybody using iii know why there are difficulties with L+G EM regular investment here?

Perhaps we never will, iii is switching to its new platform this weekend.

I got in touch with HSBC and asked them to send me the factsheets I needed to complete the stats in my previous posts. I've just received them and my mind is completely blown.

HSBC FTSE All-Share Index Fund

On the 31 August 2010 factsheet, the performance figure for 31/8/2009-31/8/2010 is 5.4%. On the 31 August 2011 factsheet, the performance figure for the exact same period (31/8/2009-31/8/2010) is 8.3%.

On the 31 August 2011 factsheet, the 5-year fund/index comparison is fund 6.0% vs index 10.7%. On the 30 September 2011 factsheet, the 5-year comparison is fund 0.5% vs index 3.6%. In the space of one month, the 5-year tracking difference has fallen from 4.7% to 3.1%.

HSBC European Index Fund

On the 31 August 2010 factsheet, the performance figure for 31/8/2009-31/8/2010 is -6.0%. On the 31 August 2011 factsheet, the performance figure for the exact same period (31/8/2009-31/8/2010) is -3.8%.

On the 31 August 2011 factsheet, the 5-year fund/index comparison is fund 3.8% vs index 11.6%. On the 30 September 2011 factsheet, the 5-year comparison is fund -1.4% vs index -7.0%. In the space of one month, the 5-year tracking difference has fallen from 7.8% to 5.6%.

I won’t bother you with the rest of the stats. Suffice to say, it’s exactly the same story for all the regional trackers.

Why would the performance figures for the same period change from one year to the next? I’ve emailed the helpful HSBC adviser who sent me the factsheets, and I’ll report back what he says, but I’m not holding my breath. I already felt out of my depth when I thought I had a grip, however tenuous, on what I was looking at. Now, I fully expect any explanation to complicate, not clarify things. Certainly, whatever the explanation, I don’t expect to be able to continue the analysis.

As for the tracking difference – what could explain these sudden swings? Even allowing for the fact that one month at the much higher pre-1/9/2009 TERs has fallen out of the data, this suggests a ridiculously high degree of volatility. Maybe its the volatility of the markets? Whatever. What seems clear to me is that there's no point trying to calculate tracking difference with these figures, even over a 5-year period, because everything depends on the data point. To get some sort of accuracy, you would have to use many, many data points - monthly, say, or even better daily - and then average the results, or something like that. Quite impossible.

I’m going to put a note on my earlier posts to alert any reader to this post, and then I’m going to leave it at that (unless I get an answer from HSBC, which I will post for the sake of completeness). My first attempt to grapple with the stats has been a very sobering experience.

...and it’s as underwhelming as I expected. Apparently the performance figures for 31/8/09-31/8/10 differ on the 31/8/10 and 31/8/11 factsheets because HSBC moved from quoting ‘capital only’ to quoting ‘total return’. Oh, well – at least it makes sense.

So, for what it’s worth, here are those figures in full:

HSBC

Performance: 31/8/09–31/8/10 (CAPITAL ONLY); 31/8/10–31/8/11 (TOTAL RETURN)

HSBC FTSE All-Share: 5.4%; 6.7%
FTSE All-Share Index: 7.0%; 7.3%

HSBC European: -6.0%; 3.3%
FTSE World Europe ex-UK Index: -3.5%; 3.4%

HSBC American: 7.1%; 12.2%
S&P 500 Composite: 9.0%; 11.2%

HSBC Pacific: 15.8%; 10.7%
FTSE World Asia Pacific ex-Japan Index: 16.9%; 11.4%

HSBC Japan: -2.6%; 0.9%
FTSE Japan Index: -2.4%; 0.7%

Vanguard

Performance: 31/8/09–31/8/10; 31/8/10–31/8/11 (ALL TOTAL RETURN)

Vanguard UK Equity: 10.4%; 7.14%
FTSE All-Share Index: 10.57%; 7.26%

Vanguard Developed Europe ex-UK Equity: -1.98%; 2.73%
FTSE All-World Developed Europe ex-UK Index: -0.78%; 2.96%

Vanguard US Equity: 11.0%; 12.11%
S&P Total Market Index: 11.88%; 12.3%

Vanguard Pacific ex-Japan: 17.32%; 11.45%
MSCI Pacific ex-Japan Index: 17.5%; 11.53%

Vanguard Japan: -0.3%; -0.03%
MSCI Japan Index: -0.23%; 0.27%

Judging by the index stats, the only performance figures that are directly comparable between HSBC and Vanguard are the 31/8/10-31/8/11 figures for the FTSE All-Share trackers. I thought that the European tracker figures for the same period ought to be comparable, because I thought that the two funds were following the same index, but the index figures differ, so I don’t think that comparison can be made. Is it possible that there are two different FTSE Developed Europe ex-UK indexes? I don’t believe it, but I don’t think it’s worth the hassle to find out, because it’s only one figure for one year. Sorry.

As previously established, the volatility of these point-to-point figures makes calculating tracking difference a very complicated matter. I would suggest that it makes comparing performance very difficult too, since one fund might be lagging the index at a particular date, but then shoot ahead the next day, owing to the way the tracking works. I still think that the HSBC funds seem to lag the index by more than the Vanguard funds, which is logical when you consider the dilution levy that Vanguard takes, but interestingly the extra Vanguard stats I obtained for 31/8/09-31/8/10 do show that Vanguard funds are not purer-than-pure in this respect either (see Europe, US and Japan above).

Can I stop now?

JamesU wrote: »

correlations look pretty good really:

I also resorted to the "squint at graphs with Mk 1 eyeball" approach in the end! Over the longer term, the higher TER and "dilution in the tracking error" of the HSBC trackers does cause them to lag, but it's chicken feed compared to the lag of averaged managed funds.

Of course, you could pick a managed fund that's above average and hope it stays that way, but that's a high risk strategy.