Australian Super member direct

Great responses as usual @dunno. I'm in Sunsuper choosing my allocations to pooled index options. Is there any evidence that the highest profile active option gets the best tax outcome and by how much? It sounds dodgy that they can arbitrarily choose the allocations, not that these companies aren't dodgy.

Sounds like Vanguard will be the go once they pull their finger out. I've always been wary of SMSF because of costs and administrative overheads, plus I have no compulsion to buy property, paintings etc in super - all I need is a world index and bonds.

 

Superannuation is heavily regulated by APRA who even provide guidelines for funds to follow RG 94 Unit pricing: Guide to good practice | ASIC - Australian Securities and Investments Commission

Funds are held in trust so must be separated from company operations so I think it’s pretty safe to say savings from wiped capital gains are returned to members. One would hope they are returned to the same buckets of funds where they were incurred as per the APRA guidelines rather than boosting returns of flagship offerings.

I wonder if these retirement transfer bonuses are actually working against members. That is, wouldn’t you be better off receiving the compounding benefits of wiped deferred capital gains from other investors immediately as they retire than as a lump sum when you retire.

A benefit of pooled investments is being able to change asset allocations without incurring capital gains. If you are in direct options, this is not possible. Eg inflows may be insufficient to glide up the defensive allocation as you approach retirement.

I’d love to find a way to quantify this deferred capital gains drag by comparing the returns of Sunsuper Australian Shares - Index to say VAS within a super structure after 15% tax on income, real and deferred capital gains and franking credit refunds are applied. You can get the sunsuper return easily enough, I just haven’t worked out a way to do an apples with apples comparison to VAS over say 10 years. Any ideas?

 

There seems to be a gap in the superannuation market for a passive option (Aussie domiciled version of VT ihttps://investor.vanguard.com/etf/profile/VT is fine or even something with a moderate home country bias) with low fees and internally attributed taxation. Let's hope Vanguard can do that. I'd even be happy if the options were extreme, high, moderate, low risk with respective equity allocations of 100, 80, 60, 40 and the rest going to bonds / fixed interest.

Why can't an Australian just say, give me the market return with no fees?

Any entrepreneurs out there interested and capable in starting such a fund? It would have to be not for profit like Vanguard in the US.

 

First attempt comparing SunSuper Australian Shares - Index to VAS investing $100000 on Nov 30 2010 (1675 units @ $59.70) and modelling a 10y return using ShareSight for VAS

SunSuper Pension = $240,454.61 =FV(9.17%, 10, 0, -100000)
VAS Pension = $230,528.18 or 8.71% =RATE(10,0,-100000,230528.18)

SunSuper Super = $220,330.84 =FV(8.22%, 10, 0, -100000)
VAS Super = $218,571.33 or 8.13% p.a. =RATE(10,0,-100000,218,571.33)

VAS Pension value is
$211,194.12 share value (2532 units x $83.41) on Nov 30 2020 with DRP turned on
+ 19,334.06 franking credits refund

VAS Super value is
VAS Pension value - $11,956.85 tax on gross dividends (15%)

Ignoring $410.34 cost base adjustments and $50.08 tax withheld in VAS over the period.

On face value, it doesn't look like direct investment is getting a better return, however the big omission on the VAS side that would likely tilt the result towards ETFs is factoring in compounding return on the franking credit refund each year, although it would a lower impact during accumulation due to the 15% tax Eg in super there is only a $7k refund of franking credits over 10 years.

When do these franking credits get refunded to the investor in DIO or SMSF options? Is it quarterly or annually? Knowing that we could model a purchase of shares each period in ShareSight.

 

OK, second attempt, moving to a June - June 10y timeframe so we can model financial years reinvesting refunded franking credits from SMSF tax return.

TLDR, direct investment might be worth 9 basis points before costs. You would need a sizeable balance for it to be worthwhile

Invest $100,000 in VAS on June 30, 2010 (1788 units @ $55.93) with DRP enabled and $100,000 in SunSuper Australian Shares - Index in accumulation phase

VAS assumptions

• Holdings are in a SMSF where the audit and tax return is lodged promptly and any net tax refund of franking credits are reinvested on Oct 31.
• When reinvesting, round to the nearest number of whole shares
• Zero brokerage as purchases are added on to inflows
• Ignore SMSF admin and audit costs
• Ignore cost base adjustments and tiny amounts of tax withheld

On June 20, 2020 we have

SunSuper = $214,497.27 or 7.93% =FV(7.93%, 10, 0, -100000)
VAS = $216,294.64 or 8.02% =RATE(10,0,-100000,216294.64)

VAS value calculated as

• VAS value on June 30, 2020 = $214,439.05 (2855 units @ $75.11)
• Pending dividend on July 16, 2020 = $588.20
• Pending tax refund on Oct 31, 2020 = $1267.39

Now this is on a very tax advantaged franking credits system. I’d expect the savings to be less for international shares as you cannot get a refund on withholding tax

Hopefully someone can check my work

 

Actually on second thought, I shouldn't include the July 16, 2020 dividend as that is part of the 2020/21 FY so on June 20,2020 we have a 6 basis point p.a. benefit to VAS before costs. I'm surprised the benefit is so small.

Maybe the net deferred capital gains are not as high as we expected with capital gains deferred vs wiped as members convert to pension roughly canceling each other out

SunSuper = $214,497.27 or 7.93% =FV(7.93%, 10, 0, -100000)
VAS = $215,706.44 or 7.99% =RATE(10,0,-100000,215706.44)

VAS value calculated as

• VAS value on June 30, 2020 = $214,439.05 (2855 units @ $75.11)
• Pending tax refund on Oct 31, 2020 = $1267.39

On second thought, maybe the impact on Australian shares is less due to the high proportion of income vs capital gains. With international shares, the deferred capital gains would be much higher. Next I'll try a SunSuper International Shares Index vs VGS model

 

Modelling International Shares is a little trickier as

• SunSuper follows the MSCI World ex Australia IMI index of which there isn’t a single ETF in Australia that follows it.
• IWLD isn’t ex Australia but does follow MSCI World IMI so might be close enough given Australia is only about 2% of the index
• Another limitation is IWLD has only been around for 4 years.
• VGS/VISM is another option, but then you have the complication of rebalancing and unfortunately VISM has only been around for less than 2 years.

Lets go with IWLD for 3 years buying $100,000 of each on June 30, 2017 (IWLD 3411 units @ 29.32)

SunSuper International Shares - Index (unhedged)
= $131,509.18 or 9.56% =FV(9.56%, 3, 0, -100000)

IWLD on June 30, 2020 after DRP = 3647 units @ $35.71
= $130,234.37 or 9.20% =RATE(3,0,-100000,130234.37)

So IWLD is 36 basis points worse than the SunSuper pooled fund.

On top of that, gross dividends over the time period is $11,231.88 with 15% tax payable of $1684.78, except IWLD withholding tax was only $1299.14 so the SMSF would need to fund the $385.64 tax shortfall.

Scratching my head here. Perhaps this is just the shortcomings of IWLD not exactly tracking the index we want and the inherent tracking error in this ETF or the time period us just too short

In the absence of further insights, pooled index funds are looking pretty good with negligible real world impact of deferred gains.

What am I missing?

 

@Hockey Monkey unless I've misunderstood you're trying to do something I also attempted earlier in this thread (starting around Nov 2019 - with a couple of updates over the following months). I attached my spreadsheet to each of those posts. When I have some free time I'll look at your posts in more detail against my work in this space. In the meantime really interested in your thoughts on what I'd previously put together and if applicable/equivilent how it relates to your findings so far.

 

Would it be that 10 years might be too small a number of compounding periods for the provision of unrealised gains to take affect??

 

Hey @ChrisP73, when I started looking at this issue, I also started with the Vanguard wholesale performance data but given they are also a pooled fund, I wasn’t sure how their returns compare to ETFs. This is what led me to use actual data from ShareSight to model returns, including reinvesting refunded franking credits.

I guess a SMSF using the wholesale funds could probably do the same although with lower tax efficiency than the ETF structure. I also wasn’t confident how accurate the vanguard wholesale tax accounting is on their website as it is all a little opaque.

I don’t think VGS is a good comparison to Sunsuper international as one is just large/mid caps and the other is total market. Large/mid have outperformed in the last decade.

Looking at your latest 10y figures and using the pension result as the baseline, we have Vanguard 5 basis points better for Australian Shares, roughly inline with my calculated 6 basis points

And for international, using pension as a baseline your numbers also indicate a worse outcome for Vanguard during accumulation similar to mine, although I'm much less confident about my IWLD numbers given the short 3y timeframe.

 

Not really. If my sharesight model is accurate it would be 6 basis points ongoing, eg 1.0006^30 = 1.8% total after 30 years

That said, it is entirely possible I’ve made a mistake in the modelling

 

Interesting alternative to consider while we wait for VPI
Fintech Superhero to launch super fund

 

Australia's largest superannuation fund, AustralianSuper this week became the country's first nest-egg manager to surpass $200 billion in assets

 

@ChrisP73 a couple more thoughts about your analysis.

1) Wholesale funds are pooled so pass any capital gains onto current investors so would have a similar drag to deferred capital gains in super. My modeling uses ETF returns so should avoid this issue.
2) The MER of SunSuper Australian Index is 0.10% vs 0.16% for Vanguard Wholesale. This would explain some of the difference in the Pension analysis. Further differences might be from the different indexes ASX200 vs ASX300

 

I completed some further 10y Australian Shares accumulation phase modelling today, this time using STW which tracks the ASX 200 (VAS tracks the ASX 300)

SunSuper = $214,497.27 or 7.93% =FV(7.93%, 10, 0, -100000)
VAS = $215,706.44 or 7.99% =RATE(10,0,-100000,215706.44)
STW = $214,197.21 or 7.91% =RATE(10,0,-100000,214197.21)

So STW is 2 basis points worse than SunSuper. This makes sense given the MER of STW is 3 basis points higher than SunSuper

Still searching for a real world advantage of direct investment over pooled index funds. Not finding evidence of a deferred capital gains drag

 

Thanks @Hockey Monkey this is really helpful and shows that although this is an arbitrary period, the difference is not as much as I thought. Compounding over 30 years may increase the margin, but given I'm approaching old fart phase, it won't make a significant difference, especially given my priority is laziness. That being said, I may consider a Vanguard option when released.

 

I'd love to find a way to model a 5-10 year period for International shares using the index followed by SunSuper (MSCI World ex-Australia IMI) capturing 99% of 22/23 developed countries but am struggling to find something suitable.

VGS is large and mid cap only
VISM is only 18 months old
IWLD is 4 years old and missing US mid caps
IVV/IVE is large and mid cap only

I thought I'd approximate it using VGS + Small Caps wholesale over 5y, but ShareSight doesn't report tax withheld on wholesale funds.

Does the following hypothesis look correct?

If I invest $100,000 in a pooled fund and ignoring dividends, over 10 years grows to $200,000 for a capital gain of $100,000 and I still hold it, the super fund has to reserve $10,000 (10% tax on long term gains) and show a balance of $190,000?

This $10,000 deferred capital gains is held in cash until one of the following
a) Another member converts to pension which allows the deferred gain to be wiped and the cash invested, adding to the reported performance at the time of conversion.
b) The fund has to sell shares for liquidity purposes at which point the deferred gain is converted to a real gain and cannot be ever recovered.

According to UniSuper when I quizzed them, even with the March selloff and $10K withdrawals, they had sufficient liquidity to avoid b). I expect active funds regularly realize capital gains through trading, but we are talking index funds here so it is really all about liquidity vs wiped gains.

This whole scenario reminds me of DSSP in AFI. It can be advantageous as long as you never sell, but if you do, you would have been better off in the DRP. 40 years is a long time to stick to an investment and I wonder if for most people pooled funds offer better flexibility as you can switch without the guarantee of full capital gains being realised.

The analogy here is that you might be slightly better off in direct ETF's instead of pooled funds, but only if you never ever sell. If you ever sell ETFs, the full 10%-15% capital gain is guaranteed to be realised by the individual.

For pooled funds, you may have some deferred gains, but it appears other members converting to pension phase mostly wipes them out for the benefit of all members with the added benefit of being able to change asset allocation without additional capital gains.

In fact, the retirement transfer bonuses on pension conversion could be worse for members as they stop deferred gains from being continuously wiped and returned to all members during their accumulation phase where they can compound for decades. At least SunSuper limit the bonus to $4800 so the majority of wiped gains can be compounded for the benefit of all members.

 

Hi @Hockey Monkey

You have motivated me to do a comparison for myself.

I used Suncorp 7 year Australian index vs VAS in SMSF and got similar finding to you.

Used 7 years June 2013 to June 2020 as that was the longest complete period, I could get actual unit prices for on the Suncorp site (they provide 10 year history from todays date)

Suncorp calc was easy once obtaining entry and exit unit rate and they matched the published 7 year figure.

VAS was more complicated, I modelled re-investing dividends through DRP but couldn’t account for things like AMIT adjustments and other tax deferred amounts or foreign tax credits etc but I wouldn’t think the amounts would be significant.

End result was that Suncorp ending value was higher than VAS holdings by a couple of thousand dollars. Add in the deferred capital gains tax you would have to pay in the SMSF if you sold up and VAS in SMSF was approx. another $2300 behind Suncorp. The SMSF would have had a Net income tax benefit of $4,600 which basically brings them back to line ball on an after-tax basis.

So, this comparison suggests that Suncorp is flowing returns from tax deferred amounts back into the indexed pool bucket. This seems to also be confirmed by the fact that their after-tax returns are very similar to the before tax returns of the index itself.

If the pooled super funds are allocating income from assets backing deferred liabilities back to the option buckets that create the deferred liabilities than the potential drag is neutralised. If fund flows in and out can be used to avoid realising gains at the fund level than the activities of active switching and active management within the fund can also be neutralised. (but what happens when fund flows change as the super environment matures?)

My comparison between Suncorp Indexed and direct via SMSF shines a positive light that the potential drags are not impacting the Suncorp Aus share indexed option to date, if anything the pooled taxation might be creating a small benefit. What the future tax benefits or drags will be could only be gauged by understanding the fund taxation and allocation rules, investment outcomes and member accumulations/de-accumulations.

The comparison gives you ex-post facto that drag hasn’t been consequential to date in this individual comparison, what is need is pooled super allocation transparency to determine the issue ex ante. Its too late to get to the end of your accumulation to identify a poor actual outcome only then. Current comparisons only tell you so much about the future.

Cheers

 

Awesome @dunno, thanks for doing your own comparison to confirm.

Yes, I ignored the small amount of foreign tax credits and AMIT adjustments as well.

Fortunately we don't need to wait till the end of accumulation to find out the result, it is possible to track this over time and take action if things start to tip in favour of direct ownership.

Any thoughts on if it might play out any differently for International Shares?

 

Not really, My solution is to avoid the potential drag.

The same potential for deferred tax drag is there probable even more so as more return comes in the form of capital growth. (The Capital gains in previous Aus index comparison was very limited over period in question) Question is whether the fund allocation methods are neutralizing the potential, without allocation transparency, insight on that can only be obtained by doing a comparison against your chosen fund/option and as you say monitoring it over time.

I’m lazy and I try to limit my time online to 1 Hour a day, so I’ll leave you to the complexity of trying to chase down data and do an accurate comparison.


Cheers