≡ Menu

Have the Powershares FTSE RAFI ETFs done the business?

It’s never a good idea to invest in a product that you don’t understand or can’t get good data on.

And while the ideas behind the Powershares FTSE RAFI ETFs aren’t so hard to grasp, getting a handle on how they’ve performed is like a Soviet show trial in numbers – the truth is hard to find.

Why should you be bothered with the RAFI ETFs? Well, they could offer UK passive investors a chance to capture the value premium, via a so-called ‘smart beta’ strategy known as fundamental indexing.

  • The value premium offers investors the chance to amp up returns.
  • Fundamental indexing theoretically fixes some of the problems of market cap indexing, particularly the tendency to load up on overvalued equities.

Better still, the RAFI ETFs have been available in the UK for over five years now, so we can pit the claims of fundamental indexing against some hard numbers.

However it turns out that getting Powershares, Bloomberg, Morningstar and Trustnet to agree on the performance data about these ETFs is like hoping for consensus at a UN Climate Change Conference.

What a mess!

Pick a number, any number

The following table shows the various returns quoted for the Powershares FTSE RAFI Developed 1000 ETF (PSRD).

Data source 1-year return % 3-year return % 5-year return %
Powershares 16.04 7.86 6.04
Bloomberg 26.16 6.76 6.29
MorningStar 23.22 4.08 4.24
Trustnet 18.5 2.8 3.4

Source as stated

So that’s about as clear as John Prescott then – lots of different numbers that would ideally be the same.

  • The figures date from 3 May, except for Powershares which quotes from March 31.
  • Bloomberg doesn’t say whether dividends are reinvested. The others all do.
  • MorningStar and Trustnet clearly state the returns are annualised. The other two skip the details.
  • Powershares’ performance data isn’t available on its retail site. You have to masquerade as a professional client to access such privileged information.

Trustnet’s numbers are clearly off as they’re lower than its figures without dividends reinvested. Trustnet’s net return figures match MorningStar’s total return numbers, bar rounding error, so we’ve at least got some kind of match.

Ultimately, it’s not good enough and I wish Invesco Powershares would present clear, updated, annualised figures on its own website, so that investors can see what these ETFs are really capable of.

Performance anxiety

Inconsistencies also bedevil the other three RAFI ETFs I checked out, namely:

  • PSRU – UK equity
  • PSRW – All-World including emerging markets
  • PSRM – Emerging markets

In the case of PSRW and PSRM, Powershares doesn’t quote performance data for either fund, despite the fact they’ve been available for over five years. Instead it quotes the index performance.

That’s sneaky because the indexes don’t include the fund’s actual costs, which drags down performance like a lead weight.

The RAFI Emerging Markets ETF is known to have suffered significant tracking error due to costs. (Presumably that’s why PSRW and PSRM switched to a synthetic index replication strategy after a couple of years).

A lack of performance transparency is enough to make me give up on a fund there and then. There are already enough potential grey areas and grey hairs associated with investing, without creating extra room for doubt.

But as I said earlier, value funds are hard to come by in the UK. It would be great if I could find strong evidence that these ETFs work, so let’s persevere.

Battle of the value trackers

What I really want to know is that PSRD performs well against other value trackers. I know that value funds can lag the market for many years, so I need reassurance that my value pick isn’t a duffer.

The following index trackers all offer varying takes on the International1 Large Value strategy. The exception is the L&G fund, which is a Large Blend international tracker that acts as a proxy for the market.

How does PSRD match up?

Fund 1-year return % 3-year return % 5-year return %
PSRD – Bloomberg numbers 26.16 6.76 6.29
PSRD – MorningStar numbers 23.22 4.08 4.24
Dimensional Int Core 21.6 7.6 6.9
Dimensional Int Value 23.6 5.2 2.8
DBX STOXX Global Select Divi 24.5 10.9 5.4
L&G Int Index Trust I 21.2 8.0 6.3

Source: As stated or Trustnet

On the whole, it hasn’t been a great five years for value equities, so if PSRD performed as per the Bloomberg numbers then I’m interested. Much less so if the MorningStar (and Trustnet) numbers prove true.

I can’t get a clear signal from the numbers though, and the same story repeats itself for the UK RAFI ETF – PSRU – which seems OK by some lights and slothful by others.

The All-World and Emerging Markets ETFs don’t look great by any yardstick over five years, and have clearly suffered at the hands of reality. Perhaps the synthetic approach will turn things around, but the lack of live data on the site is hardly reassuring.

Case not proven

Here’s the problem. Fundamental indexing is something of a novelty act in comparison to tried and tested market cap investing.

In theory, fundamental indexing is the superior strategy over time, but theoretical advantages can be overwhelmed by the costs and the practical difficulties of real-world application.

I need more reassurance, not less, to take the plunge. If I can’t tell whose numbers to trust – and the ETF providers aren’t making matters crystal – then I’m going to err on the side of caution and leave these funds alone.

I dare say that other funds are afflicted by inconsistent data, too. But bold claims have been made for fundamental indexing and so the supporting evidence should be placed squarely in the hands of investors – assuming the evidence is out there.

Take it steady,

The Accumulator

  1. Developed world equities including the UK, but not emerging markets. []

Receive my articles for free in your inbox. Type your email and press submit:

{ 10 comments… add one }
  • 1 vanguardfan May 14, 2013, 10:36 am

    Thanks for a useful and informative post, as ever.
    Can I just clarify – in your second table, is the L&G fund there to provide a benchline comparison with a plain vanilla market cap fund? For me, that’s the relevant practical comparison for the question that needs to be answered – is there additional performance from the value/fundamental index fund compared with the market cap? If not, keep it simple.
    At present the data are far from convincing. What I find fascinating is how attractive we all (me included) seem to find the promise of ‘something better’. I am increasingly of the view that our own human psychological weaknesses are our worst investing enemies! (Which would logically imply that the more mechanistic you can make your strategy, the better).

  • 2 Rob May 14, 2013, 12:26 pm

    Welcome to the dizzy world of collective performance analysis.

    If it’s this hard to measure returns imagine the problems of truing to calculate risk adjusted returns.

  • 3 Mike May 14, 2013, 2:35 pm

    I find it amazing that the finance industry gets away with making it so hard to compare products. If any other market were this secretive there would be a huge investigation – for example, you can easily compare cars’ miles per gallon directly or the efficiency of various washing machines. How much longer do we have to put up with this nonsense?

    The fact that this data is so imprecise instinctively tells me that these aren’t such a good buy . Even allowing for costs on a regular FTSE tracker can make a big difference and chasing the value premium is not simple, otherwise everyone would do it all the time.

    What I’d like to see is a comparison based on a monthly purchase of each of these funds within a suitable wrapper (ISA or SIPP) over the last 5 or 10 years. This would include all costs and dividends and be benchmarked against purchases of an equivalent low-cost fund tracking the high yield index or the FTSE 100/AS. Candidates could be the Vanguard FTSE UK All-Share or iShares ISF.

    My suspicion is that they would not show much of a premium and even if they did, this would revert to the index over time…

  • 4 DMDave May 14, 2013, 4:16 pm

    Thank you for another well thought-out piece. It’s interesting to see there is such a big difference in the published performance of fundamental index.

  • 5 Luke May 14, 2013, 4:29 pm

    I read the article, thought ‘that sounds more complicated than it needs to be’ and then checked my Vanguard Lifestrategy holdings. I think I will stick to fundamental indexing 🙂

  • 6 pedro May 14, 2013, 6:52 pm

    Have you looked at the Schroders QEP Global Active Value Fund (over 1000 holdings) and QEP Global Core Fund ? I think it might do a similar type of “thing” to the RAFI series and may be an alternative ??. For disclosure I am not invested in either product.

  • 7 The Accumulator May 15, 2013, 7:50 pm

    @ Vanguardfan – yes, that’s right, the L&G fund is there as a vanilla market cap comparison with the fundamentals. Very good point about our ceaseless search for greener grass. It’s all too easy to fall into the trap of thinking the real party is going on wherever we’re not.

    @ Mike – you could probably simulate such a comparison in Morningstar’s portfolio tracker, a lot of work though to insert monthly purchases over 5 years. You’re right about endemic obfuscation of data though of course it’s widespread practice – mobile phone companies and energy companies being classic examples. I recently read a story about the absurd lengths car companies go to when testing their cars for fuel efficiency. Removing wing mirrors and taping over the gaps in the body panels were just two of the ridiculous ways they gamed the numbers.

    @ Pedro – I haven’t looked at those funds but I will. Thanks.

  • 8 Steve May 17, 2013, 9:48 am

    I looked at the EM ETF on the Powershares website. There is some information, but I couldn’t see any geographical info, i.e. % China, Malaysia, Brazil etc.
    Maybe I’m looking in the wrong place. Any clues?

  • 9 The Accumulator May 17, 2013, 5:17 pm

    Hi Steve,

    You’ll find it in the factsheets section of the professional client part of the site:


  • 10 Steve May 17, 2013, 8:43 pm

    Thank you very much!!


Leave a Comment