Most of us are comfortable with the idea of diversifying across geography. We’re all too aware that countries can rise and fall through the league of nations, that the Chinese Dragon can catch a cold or the Land of the Free can turn into the Land of the Hideously Over-Valued.
Diversifying across the sources of investment return is a road less travelled.
Academics have isolated the main factors that drive equity returns. And we can invest in them just as with countries or continents – piggybacking their unique properties to hopefully reinforce our portfolios whichever way the wind blows.
The easiest way of doing this is with a multi-factor exchange traded fund (ETF). These combine factors like value, size and momentum into one handy package.
In a previous thrilling installment, we stared into the multi-factored eye of iShares FactorSelect MSCI World ETF. Now we’re gonna sidle up to a couple of ETFs that follow the Scientific Beta Developed Multi-Beta Multi-Strategy index.
That’s right. You asked for it!
Okay, maybe you didn’t but I’m gonna tell you anyway because multi-factor exposure offers:
- A chance of outperforming the broad market over time.
- Better risk-adjusted returns through exposure to less correlated equity asset classes.
Two very nice outcomes, even if the majority of passive investors might rather just be woken up in 20 years time when their vanilla portfolio has finished rising in the oven of compound interest. (Which is absolutely fine by us, of course).
Welcome to my laboratory
The Scientific Beta ETFs are a potent potion of value, size, momentum, and low volatility.
These factors should mix well together because correlations have been historically low between:
- Value and momentum
- Size and momentum
- Size and low volatility
- Low volatility and momentum
The Scientific Beta Developed Multi-Beta Multi-Strategy index – apart from being stuffed with more keywords than a pornography website page circa 1999 – tracks 1,600 equities that tilt towards the above factors across the developed world.
You can check its diversification chops for yourself on the factsheet.
Two ETFs currently track variations on this index:
1. Fundlogic Morgan Stanley Scientific Beta Global Equity Factors ETF (GEF) tracks the equal weighted (EW) version of the index.
This means the index is evenly divided between the factors, so its fortunes are less likely to be swayed by any particular one of them.
2. Amundi ETF Global Equity Multi Smart Allocation Scientific Beta ETF (SMRU) tracks the equal risk contribution (ERC) version of the index. This mechanism is designed to minimise tracking error regret – the malaise that makes investors want to sell up when their factor tilt underperforms the broad market.
Scientific Beta’s backtesting suggests that the equal risk index sacrifices a performance smidge (0.3% per year between 2003 and 2013) in exchange for less volatility (1% less per year between 2003 and 2013) in comparison to its equal weight cuz. Since they’ve gone live, there’s scarcely a whisker between them.
Now, our multi-factor pair may have longer names than a Welsh village, but luckily they don’t charge by the word. Their Ongoing Charge Figures (OCFs) are comparatively modest at 0.4%.
You’d pay 0.5% for iShares FactorSelect ETF, by comparison – but the cheapest, plain, developed world ETF is less than half as dear at 0.15%.
Will you get what you pay for?
So are these multi-factor ETFs worth it?
Sadly, there’s too little data to go on to answer that question.
However an early comparison of their holdings versus plain, developed world rivals does reveal a slant away from the consensus.
The difference they can make is confirmed by this Trustnet performance chart that matches the last 24-months worth of results from our multi-factor ETFs (A and B) versus two plain, developed world ETFs from HSBC (C) and Vanguard (D).
Ignore the graph if you like, and drop your gaze to the numbers below.
A year ago our multi-factor upstarts had just put in a 12-month stint of slamming their no-frills rivals. They’d beaten HSBC’s ETF by over 3% and Vanguard’s by well over 2% over the period. Albeit the year was no great shakes – even the multi-factor ETFs barely broke even.
But 12-months later and the vanilla crew reigned supreme – beating the Smart Beta duo by over 4%.
What does this tell us? Only that the the Scientific Beta formula does actually create products with returns that differ from regular ol’ world ETFs. They should therefore add some diversification to your equity mix.
What we absolutely cannot infer from 24 months worth of data (no matter how tempting) is which of these products will turn in the best results over the long-term.
And we can’t tell how they might perform under different economic conditions, either.
Good signs
The factor metrics used by Scientific Beta to select its index holdings are straightforward and supported by independent research. This gives me more confidence that they haven’t just sent their data miners into the bowels of history to pluck out some shiny backtests that won’t be replicated in real life.
You can read all about how the indexes are constructed on Scientific Beta’s website.
In fact, Scientific Beta go to greater lengths than most to strip their indexes bare for all the world to see. Indexes should be see-through like chiffon, so this is another encouraging sign, especially as I’ve been unable to find dissenting voices picking the methodology apart.
The equities in the index are not cap weighted but strained through more filters than sewage. Initially they are equal-weighted, then they’re deconcentrated, decorrelated, risk-weighted and risk-adjusted to keep your portfolio smelling sweet.
I’m poking fun because the methodology is a jargon fest and I have no way to unravel it. In Scientific Beta’s defence they’ve documented each stage in detail. The gist is that the process is designed to increase the diversification effect.
Scientific Beta’s own pitch for why you’d invest boils down to:
Such a Multi-Strategy Index is suitable for investors who do not hold a strong view on what is the best strategy over the relevant investment horizon and wish to protect themselves against uncertainty by i) diversifying strategy specific risks and ii) smoothing their outperformance across distinct market conditions.
The Investor’s Field Guide wrote a great piece on why investors looking for outperformance should take a strong position in defiance of the market consensus.
While true, this requires you to do three things that are tough:
- Take a contrarian view.
- Be right.
- Stay the course for as long as it takes to become right. That could mean spending years grappling with the consequences of being wrong.
The Scientific Beta ETFs do not offer concentrated contrarianism. They allow you to wander off the beaten track ever so slightly. It’s a detour to grandma’s house that shouldn’t take you too deep into the forest. (That’s what the big bad wolf of stockpicking is for.)
We can see evidence of Scientific Beta’s pitch in the chart above. The multi-factor ETFs move in tandem with the performance of the plain developed world product, but they’re offset enough to make your portfolio a little different, hopefully in a positive direction over time.
So why wouldn’t you?
Aside from adding considerable complexity to your investing life, there’s a couple of particulars about these ETFs that may give passive investors pause.
Firstly, both are synthetic ETFs. They don’t physically hold the equities in the Scientific Beta index. Instead they use a financial contract known as a total return swap to match the index return.
Personally, synthetic ETFs don’t worry me. They do others. Either way, you should know the risks you’re exposed to.
In particular, financial regulators have warned of the potential for conflicts of interest when ETF providers and their total return swapping counter-parties are arms leading from the same financial octopus. (We’ve written more about such risks, if you’re keen).
It appears to be the case that Morgan Stanley is the counter-party to the Morgan Stanley ETF. See pages 172, 384 and 393 of the annual report.
And there’s another eyebrow raiser on page 39 of the ETF’s prospectus.
Morgan Stanley states:
When Morgan Stanley acts as broker, dealer, agent, lender or advisor or in other commercial capacities in relation to the Fund, Morgan Stanley may take commercial steps in its own interests, which may have an adverse effect on the Funds.
Also…
It is anticipated that the commissions, mark-ups, mark-downs, financial advisory fees, underwriting and placement fees, sales fees, financing and commitment fees, brokerage fees, other fees, compensation or profits, rates, terms and conditions charged by Morgan Stanley will be in its view commercially reasonable, although Morgan Stanley, including its sales personnel, will have an interest in obtaining fees and other amounts that are favourable to Morgan Stanley and such sales personnel.
It’s as well to be aware of the bargain you’re making when investing in any product.
Amundi is backed by the French bank Credit Agricole and the counterparty of their Scientific Beta ETF is BNP Paribas. The ETF is domiciled in France which means you’ll be exposed to withholding tax if it distributes dividends.
According to the dividend policy, the fund manager is at liberty to decide whether dividends are paid out or reinvested in the ETF. It doesn’t say on what basis the manager will make this decision.
The final point to note, if you’re interested in what other investors think, is that the Amundi Scientific Beta ETF has been the most successful in attracting paying customers.
As of January 2017 it has acquired $536 million in assets, compared to $246 million for Morgan Stanley’s version and $270 million to iShares FactorSelect ETF.
Over to you
Ultimately, these are complicated investment products that require much due diligence before they can comfortably slot into anyone’s portfolio.
Every investor needs to decide whether the underlying principles of factor investing make sense to them and whether the available investment products are likely to do a good job.
All I can do is lay out some of the issues worth thinking about and link to further research that may help.
Personally, I do invest across the factors. I do this as much in the hope of diversifying my portfolio as in squeezing any more performance from the financial toothpaste tube.
The greatest argument against this move is simplicity. If you’re an exponent of the art of KISS then… as you were.
Take it steady,
The Accumulator
Great article, thanks. Not only informative, but for me it also passes the Mark Kermode 6-laugh test for a comedy film (http://witterpedia.net/wiki/index.php?title=The_Six_Laugh_Test). As for multi-factor ETFs, whilst they may pass lots of filters, they don’t pass my BS filter. I would suggest that if you add enough filters you’ll just end up with either an overly expensive total market tracker, a clogged filter or a polished t*rd. 🙂
A lot to digest here – thanks. First thing I notice is the differing portfolios on the two etfs. The Amundi country weighting mimics a Global index, however the Fundlogic is not only benchmark-unaware country-wise, but also seems to have quite concentrated holdings. It looks very un-passive, but to my mind quite appealing. On with the research…
A question on synthetics – I note that the Portfolio Turnover on the Fundlogic MS ETF is HIGH – this could add on some hefty non-TER charges on a Physical ETF, but with a synthetic – is this all included?
The MS small-print is sobering. The generous view is ‘lawyers made me do it’, but it is a good reminder of why I tend to go for small firms where I trust the bosses to do right by me (e.g. Buffett, Lindsell Train, Capital Gearing etc.)
Thanks again for making something so complicated become relatively palatable. (n.b. I’m not ruling that further research may make me agree with Jeff above…).
As more and more investors use easily available multi-factor funds (a few years ago there wasn’t any) a contrarian needs choose other strategy like single factor.
Interesting thanks. There are two unknowns here – will this strategy increase returns, and will it do it with the same or better risk. Personally if I need more returns to meet my goals I would prefer to just increase my equity/bond ratio slightly. Yes, more risk/volatility – but at least it’s going to do “what it says on the label” (well probably, obviously!).
This is probably a case where synthetic ETF is a good call. All that rebalancing away from market weighted to the various filters would add some big trading costs.
@tom. Not sure what data you looked at for holdings. But make sure you were looking at the index holdings and NOT the ETF holdings (which are just the collateral for the swap). Some vendors and data sites mix them up.
BTW the Amundi one is a French FCP so if in taxable it is a bit of nightmare for tax reporting being considered a transparent fund – all interest/dividend/expenses deemed directly on the investor, CGT opaque though and reporting status treatment of gains.
Presumably the Amundi fund pays withholding tax on the dividends that it receives from around the world, then a UK investor would pay it on any distributions from the fund. Please correct me if I am wrong.
Withholding tax rules in general are both perplexing and impenetrable. I’m hoping Trump will bin them.
Soicowboy,
“Withholding tax rules in general are both perplexing and impenetrable. I’m hoping Trump will bin them.”
Depends whether the taxes are brown or white?. He’s more likely to tax foreigners more rather than less, I imagine. We, after all, are sucking the lifeblood from the States, aren’t we?
Seems there’s an awful lot of science being applied to something which is rather primordial, being more successful than the next bloke. I know I can’t form a judgement as to whether the smart beta refinements will improve my future risk weighted returns so am not inclined to try. Given the vast universe of investment choices available, and the strong dependency on getting asset allocation right, would any benefit outweigh the sensitivity to variations in broad allocation anyway? I suppose more choice is good but it adds to my confusion and diminishes clarity which I never have much of.
Hmm,
0.4% OCF vs 0.15% for a perfectly good Developed World ETF… without real evidence of any advantage..?
Think I’ll stick to plain vanilla thanks..!
SMRU isn’t domiciled in France but in Luxembourg, so no withholding tax, right?
I’m curious, has anyone looked at some of the newer multi-factor ETFs, such as FRGE, HWWD, EFIW, LYXW, WOFM? What do they think?
Just to be aware that multifactor will diversify strategy compared to 100% in global market cap weight equities but the available multifactor ETFs tend to reflect the cap weight national equity market concentrations, so they are unlikely to diversify geographical concentration risk.
So, if one happens to be concerned about the concentration risk in US equities in a global tracker (~65% US stocks’ weighting in 2024 for Vanguard Global All Cap Fund) and decides to dial down their US exposure a little bit, then moving some of that US allocation to the available multifactor ETFs won’t currently help in that aim (although it will give you exposure to a non cap weight strategy, of course).
This makes sense as the US is overweight in Momentum characteristics presently but underweight in Value, so it’s washing out to roughly the same concentration for the US in multifactor funds as with global cap weight trackers.