Variance in index fund prices

aroominyork

Fidelity Index World rose today by 0.20% while HSBC FTSE All World Index fell by 0.51%. Both have valuation points of midday and the only material difference, so far as I am aware, is that the Fidelity fund excludes emerging markets. That would not explain the difference in fund price movement so, while I expect such variances iron out over the days, it makes me assume that fund prices on any given day are not an accurate reflection of the underlying share prices. Could someone explain?

Thrugelmir

If the individual holdings differ then so will the performance. 

SonOf

The asset mix is different. 
On the 12th Fidelity went up and HSBC down. On the 13th Fidelity went down. HSBC went up.  On the 14th it was the other way around again.

aroominyork

Thrugelmir said:

If the individual holdings differ then so will the performance. 

But the holdings only vary by HSBC including EMs and Fidelity being slightly more large cap focused. For the most part the underlying holdings are very similar, and the differences surely are not material enough to explain the fund prices moving in different directions.

SonOf said:

On the 12th Fidelity went up and HSBC down. On the 13th Fidelity went down. HSBC went up.  On the 14th it was the other way around again.

Exactly. They move in different directions but over time closely track each other (if you strip out EMs, hence HSBC has drifted a little lower than Fidelity over the last two years). That is why it seems they are not recording the underlying companies' share prices in the same way.

nothing_to_fear

Could it be that either (or both) HSBC and Fidelity are using "floating" unit pricing? I.e. they may be "floating" the daily price to fractionally above the NAV on days when there are more units being created than redeemed, and/or "floating" it to fractionally below the NAV on days when there are more units being redeemed than created?

The purpose of this approach to unit pricing is to make the investors buying and selling units each day (instead of the long-term holders of the fund) pay the costs of actually buying or selling the underlying investments, when the latter is necessary (i.e. in so far as units creations and redemption don't neatly match each other). An alternative way to do this would be to charge a fixed (percentage) dilution levy on unit creation and/or redemption; as Vanguard UK used to do on some funds, before they switched to floating pricing. IMHO, a dilution levy is more transparent, but presumably most investors don't agree (or don't understand it properly ).

This could explain the differences in daily price moves, since the flows of funds into and out of the 2 funds will be different. And it is also possible that only 1 of them uses floating prices, or that 1 of them uses it more aggressively (i.e. with bigger swings in prices) than the other.

AnotherJoe

Please show the maths you've used to show that the difference must be due to a conspiracy or cockup at one or both companies rather than due to different mixes of assets in different currencies making such trivial differences between them on a day to day basis, 

Argument from personal incredulity doesn't wash as an explanation which is all you've got so far. 

Linton

The difference between the two indexes is larger than a few EMs. HSBC FTSE All World is 54% US and 9% Far East, Fidelity Index World is 63% US and 2% Far East.  So its hardly surprising that the funds dont always move in step.  For example consider the reaction of the different markets to the Coronavirus.

aroominyork

AnotherJoe said:

Please show the maths you've used to show that the difference must be due to a conspiracy or cockup at one or both companies rather than due to different mixes of assets in different currencies making such trivial differences between them on a day to day basis, 

Argument from personal incredulity doesn't wash as an explanation which is all you've got so far. 

I have no maths to show. All I have is the pattern of them consistently diverging from and then converging back towards each other. eg here are the last few weeks - only a snapshot, I know, but illustrating an ongoing pattern.

Linton said:

The difference between the two indexes is larger than a few EMs. HSBC FTSE All World is 54% US and 9% Far East, Fidelity Index World is 63% US and 2% Far East.  So its hardly surprising that the funds dont always move in step.  For example consider the reaction of the different markets to the Coronavirus.

Most emerging markets are in the far east so that is where you would Fidelity to have lower exposure than HSBC. And you would also expect a goodly amount of that 'spare' 7% to be invested in to US. (Small point - I think US allocations have risen in the last few months and the 54%/63% might be drawn from different months). Again, it's the consistent diverge then convergence that makes me think (or, as Linton would say, be personally incredulous) that it is to do with a difference in how the funds are priced.  

Linton

You are worrying about short term noise which is irrelevent to investing.  For example the different markets are open at differrent times and data is published at different times so an event which causes a blip may affect one day's markets in the UK or US and the next day's markets in the far east.  That would explain convergence as would a blip just affecting one area.

aroominyork

Linton said:

You are worrying about short term noise which is irrelevent to investing.  For example the different markets are open at differrent times and data is published at different times so an event which causes a blip may affect one day's markets in the UK or US and the next day's markets in the far east.  That would explain convergence as would a blip just affecting one area.

I'm not worrying about anything. It's just out of interest. I'll do a bit of work to try to understand nothing_to_fear's post (it's a shame posts are no longer numbered).

bowlhead99

aroominyork said:
I'll do a bit of work to try to understand nothing_to_fear's post

Each only has one published price per day and it's true that they can both use swing pricing - effectively a dilution adjustment to ensure that existing investors are not impacted by particularly large inflows or outflows on a given day exceeding a given threshold. HSBC's prospectus notes their dilution adjustment rate estimates of 0.17-0.20% for their All-World fund, as of about 6 months ago. 

So it's feasible that HSBC All-World had large inflows on Thursday causing them to hike the price, which means HSBC's fund has further to fall on Friday (especially if there are big outflows that day and they swing the price to the downside by a similar 0.2%, because that gives them a net 0.4% drop between the 'bid and offer' swing). If Fidelity's developed world fund happened to be swinging to the opposite directions at the same time (e.g. loading the price by 0.1-0.2% in the opposite directions) the two funds' published prices could diverge by 0.6-0.8 % opposite directions even if they had the same underlying assets.

As they don't have the same underlying assets the prices will diverge and also the investor demand factors will be different - so perhaps a really substantial investor piled into HSBC Allworld on Thursday believing it was now safe to invest in an index that included Emerging markets after some recent market turmoil, while some other substantial investors decided to redeem out of the fund on Friday having seen the news that Coronavirus cases in China just shot up (due to the new reporting methodology change within Coronavirus accounting in China).

If after a few years each fund has delivered within an acceptable range the result of the index it's tracking, investors will generally conclude that tracking error is fine for each, so both funds can be investible even though there are quirks of pricing along the way.

An alternative is that somehow the datafeed is off and the price published by one fund on a given day is really the previous day's price, published with a rolling delay. You could buy a unit of one of the funds to check your contract note against what was published for that day.

aroominyork said:
I have no maths to show. All I have is the pattern of them consistently diverging from and then converging back towards each other. eg here are the last few weeks - only a snapshot, I know, but illustrating an ongoing pattern.

You would expect them to differ each day by some fraction of a percent while generally getting to the same sort of place in the end - because they invest in similar stuff rather than investing in the same stuff, while the core driver of their returns is common. Looking at your graph there were only two days (28 Jan and 13 Feb) where the divergence was enough that one went down while the other went up, which could be swing pricing effects or something else, but probably not enough to lose sleep over if it is only happening once every couple of weeks.

 (it's a shame posts are no longer numbered).

As a side note if you want to direct link someone else's recent post you can click to their profile, look at their Replies and grab the hyperlink to their reply from there, like this, though it would be handy if you could just say Post #5 above without needing to count them for yourself first