Finding and evaluating tracking error

puk999

Hi,

The core of my investment is currently in the Fidelity Index World Fund P-Acc which is supposed to closely match the performance of the MSCI World Index. I remember reading something a while back suggesting Fidelity's trackers are pretty bad for following the indices.

The Performance & Risk tab of the page I linked to above seems to have some pretty wild fluctuations against the index. I'm not sure I trust the figures and they I would prefer something a bit more detailed/accurate. Trustnet's multi-charting tool knows about the fund but doesn't seem to know about the MSCI World Index.

Would like help to find the tracking error in this fund and evaluate whether it's acceptable.

bowlhead99

Not sure if it's just that I'm looking at it on a phone, but the fluctuations didn't seem "pretty wild" against the orange line on the chart on that link.

Obviously there would be big differences against the lower red line on the chart, "Morningstar category" which is just global large cap blend equity. Because that red line is a sector average that nobody is trying to track, and will always have a different risk profile to the specific fund in which you're invested.

If you go to Morningstar.co.uk you can pull up the fund info and examine the chart in more detail with the chart options - zoom in a bit.

Obviously the chart will flatter the fund's likely performance net to an investor, for most people, because the 0.2% management fees are included in reported fund performance but an investor might be paying 0.35% on top to Fidelity for platform access and tax wrapper (or 0.2%-0.5% for that access with other providers)

guymo

The "returns vs index" data on that page show that the fund returns around 0.8 or 0.9% less than the index. That's the tracking error. It seems high to me but I'm no expert.

InvestInPoker

Since December 2012 that tracking fund has returned £1328 whilst the index would have returned £1320 if I read the graphs correctly? For an initial investment of £1000

Tracking errors can work in both directions with various types of replication I guess, although charges will always be a drag downwards which causes a "tracking error". A small deviation like that graph shows some error but it seems fine to me. Certainly not at the bad end of the scale.

Ryan_Futuristics

I'd have thought any error was just as likely to be an upside as a downside (which is why they don't drift drastically over time)

There is no 'real' global index of course

Funnily enough, I read a bit of research the other day that claimed if you weighted the companies in an index by how many board members wore bow ties, it would still tend to beat the regular index ... Goes along with this idea that an index is really just about averaging out risk, and that any arbitrary set of rules can build a functional index - the reason you may get outperformance is because you'd not be buying all the same big companies as everyone else, and therefore picking up some better value

InvestInPoker

Ryan_Futuristics wrote: »

I'd have thought any error was just as likely to be an upside as a downside (which is why they don't drift drastically over time)

Good point and I would say this is true with the replication errors but fees can also be considered part of the "tracking error" imo, especially over a long period of time and that is something the buyer can control.

bowlhead99

InvestInPoker wrote: »

Since December 2012 that tracking fund has returned £1328 whilst the index would have returned £1320 if I read the graphs correctly? For an initial investment of £1000

Tracking errors can work in both directions with various types of replication I guess, although charges will always be a drag downwards which causes a "tracking error". A small deviation like that graph shows some error but it seems fine to me. Certainly not at the bad end of the scale.

That was my thought, generally seems ok when you look life to date, same ball park. However reading the stats, I agree to lag the index by a almost a whole percent in a year seems veryexcessive... but if it is still ahead of the index according to the life-to-do date chart, that's just reversing / normalizing the opening position - i.e. it must have started further ahead of the index than it ended but still ended up a reasonable place in the end.

Probably something shown in the annual or one month stats is the effect of changes in the index constituents at quarter ends and Fidelity either being slow to change or pre-empting them (jumping the gun). If they are sometimes a bit up and sometimes a bit down, no harm done. If they continually lag the index performance by more than their AMC, it points to a problem.

Some of the snapshot data looks a little wrong compared to the graph or vice versa. Table says the fund is up 8.5% in the last month while saying that's 12.5% better than the index (implied down 4%). I can see on the chart the fund is marginally up while index is down, but not by those amounts: 8.5% is >£100 up for the fund, and 4% is £50 down for the index, which is not what the graph (allegedly also to 21 October) is showing. Perhaps the table is not apples to apples, e.g. last reported calendar month end for one while last actual month (21st to 21st) on the other?

puk999

bowlhead99 wrote: »

Not sure if it's just that I'm looking at it on a phone, but the fluctuations didn't seem "pretty wild" against the orange line on the chart on that link.

On the RHS of the Performance and Risk tab trailing returns of the fund compared to the index are shown:



Is it a coincidence that 12.53% appears in the returns for this fund from Oct 2013 to Sept 2014 and in the Trailing Returns table? Also, I'm reading from this table that the fund did 12.53% better than its benchmark index in the last month. That's what I mean about "wild fluctuations" and skepticism on what I'm seeing. The other figures there seem pretty high too (almost a full percentage different than the index for last quarter/YTD).

I agree that the blue and orange lines on the chart look close.

puk999

bowlhead99 wrote: »

for most people, because the 0.2% management fees are included in reported fund performance but...

Just want to make sure I understand. Are you saying that the reported fund performances figures are after fees have been paid to the manager, etc? I understand the bit about platform fees being completely separate.

Is it always the case that performance figures are shown net (i.e. part of regulation)?

I guess this tallies with your latter point about being worried if the performance continually lags the index by more than the AMC (which in this case is 18bp, so 1% does seem high).

puk999

bowlhead99 wrote: »

Some of the snapshot data looks a little wrong compared to the graph or vice versa. Table says the fund is up 8.5% in the last month while saying that's 12.5% better than the index (implied down 4%). I can see on the chart the fund is marginally up while index is down, but not by those amounts: 8.5% is >£100 up for the fund, and 4% is £50 down for the index, which is not what the graph (allegedly also to 21 October) is showing. Perhaps the table is not apples to apples, e.g. last reported calendar month end for one while last actual month (21st to 21st) on the other?

My post about this anomaly got crossed with yours. Looking on morningstar.co.uk, the figure shows as -0.78, not 12.53, so yes, probably some error in Fidelity's IT (or data feed from Morningstar to them).

InvestInPoker

puk999 wrote: »

so yes, probably some error in Fidelity's IT (or data feed from Morningstar to them).

I agree, if you send a note to their CS they will have this clarified for you/fixed I am sure.

This fund is not lagging or suffering from any bad error at all. The fact it has returned £1328 instead of the index's £1320 over 2 years suggests there is a small error - even though it has resulted in a small upside here, it could equally have been a small downside. However no fund will be absolutely perfect and I think this tracker is a suitable choice for anyone wishing to follow the MSCI World index.