SMSFs – new cost data reveals an expensive status symbol

For the past couple of years we’ve been having an ongoing discussion with one of our clients about SMSF cost data released by the ATO and relied on by much of the industry in the debate about costs.

Our client’s view was that average costs of SMSFs released by the ATO were significantly understated. Because pension divisions are tax exempt, and therefore costs are not deductible, the ATO was not collecting data for SMSF pension divisions in the annual tax returns on which the statistics are based. As a result, those costs were just disappearing – a big deal given that a substantial portion of SMSF assets are in pension phase.

It wasn’t clear to us whether this was right – it was such a huge oversight – but SMSF tax returns were changed for the 2012-13 year to collect all cost data deductible or not, which would settle the question. And just before Christmas, the ATO released the 2012-13 SMSFs statistics.

It looks like our client was right. The “official” average costs of managing an SMSF have risen considerably, casting ever more doubt on the view that SMSFs are a cost-effective way of managing your super – arguably unless you have at least $1m, and ideally considerably more.

So let’s have a look at the numbers. Today’s chart shows:

Average SMSF operating costs for different size ranges for 2009-12 (ATO) (grey column)
Average SMSF costs for 2013 using the new approach (ATO) (light blue column)
Adjusted SMSF costs for 2013, excluding costs typically not counted by APRA funds (Tria) (dark blue column)
Typical member costs of large APRA fund default / simple super products (Tria) (purple line)

So SMSF operating costs are much higher than was previously thought. That said, the ATO numbers are not an apples-with-apples comparison with APRA funds, so we need to adjust them:

– The ATO’s operating costs for SMSF are understated in that they don’t capture fees and costs in underlying structures such as trusts, and other investments that SMSFs make.

– However they are also overstated as they include deductions that would not be considered MER-style costs by APRA funds: eg interest, depreciation, and insurance premiums. Looking at prior year taxation statistics, these have accounted for ~30% of SMSF deductions.

Considering past taxation statistics, our view is that you would deduct ~25% (net) from the ATO’s 2013 SMSF costs to create a reasonable comparison with APRA funds – the 2013 adjusted figures in today’s chart.

So where do we land?

Arguments typically made about SMSFs being a way to save on fees in APRA funds are pretty badly undermined.

Compared to alternatives such as retail simple super or not-for-profit defaults – which average 0.8% pa plus $6 per month – you need to be in the $1-2m band before average SMSF costs are attractive.

Equally, claims of SMSFs being cost effective at $200,000 are clearly rubbish, on average experience at least. At that level, average SMSF costs are 2-3% pa – a multiple of any major not-for-profit or retail simple super product, and equal to or higher than legacy retail products.

The other issue that the new data exposes is the gap between what SMSFs theoretically cost, and what they actually cost. For example, a typical advertised pricing structure for a new SMSF is:

Establishment $880
Ongoing administration $2,200
Ongoing regulatory $699

Amortising the establishment cost over 10 years, that adds up to about $2,987, or ~0.4% pa on a 750K SMSF. That sounds pretty good.

But from the chart we can see that adjusted average costs for SMSFs of 500K–1M are actually ~1.01%, which on a 750K SMSF would be $7,538 – more than double!

Where’s the gap?

Legacy SMSFs: older SMSFs are probably paying higher costs on average than newer SMSFs.

As with anything, there’s the advertised price, but when you add on all the extras – keeping in mind SMSFs are a complex product – the actual price can be much higher. Maybe SMSFs are not much different in this way to other consumer marketing bugbears such as cars and airfares.

The new data indicates that in many cases, when a member switches from a collective fund to an SMSF, their overall fees and costs go up significantly. They may reduce investment costs, but often incur a raft of new accounting, administration, and regulatory costs which far exceed the savings they think they are making.

In the growth of SMSFs, members may be clawing back fees from asset managers, but there is an even more substantial transfer of wealth going to the SMSF industry, with little evidence that SMSFs outperform over the longer term. The price of control is high indeed.

A couple of reminders about SMSF data. There are over 500,000 SMSFs out there, but it’s a highly fragmented market. Firstly, most of what passes for SMSF research is based on relatively small samples, and the risk of sample bias is considerable. The only comprehensive SMSF sample occurs when they put in their tax returns, which of course is once a year, and well after the end of the relevant financial year. That’s why comprehensive SMSF data is very much a lag indicator – we’ve only just got June 2013 data; ie it’s 18 months behind at best.

Posted In: Trialogue