The risks of Exchange Traded Fund (ETF) investing have been brought into stark relief by a volley of warning shots from the G20’s Financial Stability Board (FSB), The International Monetary Fund (IMF), and the Bank for International Settlements (BIS).
While much of the alarm has centred around exotic synthetic ETFs, physical ETF risks could also be storing up a shock for the financial system, according to the FSB and IMF.
Often described as plain, vanilla funds, physical ETFs are so-called because they physically own the bulk of the securities of the index they aim to track.
The problem is what they do with those securities next.
Many physical ETFs run security lending programmes, renting out securities to hedge funds et al, who like to make short-selling mischief with them.
Securities are lent for a fee which, if we’re lucky, is split between the ETF manager and the fund shareholders (you and me).
ETF risk alert
The short-sellers get a ready supply of securities to bet against the markets and the ETF makes some easy money that can offset its tracking error.
Everybody’s happy? Everybody except the FSB which warns:
Thin margins on plain-vanilla physical ETFs create incentives for providers to engage in extensive securities lending in order to boost returns… Securities lending… may create similar counterparty and collateral risks to [those incurred by] synthetic ETFs. In addition, it could make the liquidity position of the ETF fragile, by challenging the ability of ETF providers to meet unexpected liquidity demands from investors, particularly if outflows from ETFs become significant under severe stress.
So supposedly simple, original flavour ETFs potentially expose investors to the same unpalatable dangers as those tutti frutti synthetic ETFs. Specifically:
Counterparty risk – In this case the chance that you’ll never see your borrowed security again, because the borrower goes under before they give it back.
Collateral risk – The danger that the collateral the borrower posted to lend your security (to cover the counterparty risk) turns out to be worth a lot less than the value of the original security. Alternatively, perhaps you can’t get your hands on the collateral at all, due to a bankruptcy bunfight with the counterparty’s administrators.
The nightmare scenario envisaged by the BIS and FSB is that – in the midst of a market panic – doubts about counterparty viability and collateral value trigger a run on the ETF. (I explored that chain reaction more fully in last week’s synthetic ETF piece).
Don’t touch physical ETFs with a bargepole?
Completely avoiding physical ETFs would be an over reaction, to say the least.
Firstly, we’re compensated for the risk through the securities lending income returned to the ETF, in much the same way that we’re compensated for the heightened risk of investing in emerging markets by the expectation of greater returns.
The key question is are we’re compensated enough? You can only answer that question by knowing how much of the bunce is going into your pocket. I’ll return to this in a sec.
Also, securities lending isn’t unique to ETFs. Almost everyone is at it. Index funds, active funds, pension funds – securities lending has been going on for decades without major disaster.
Neither do these reports suggest that their apocalyptic visions will come to pass. The authors stress that they are highlighting concerns over the unintended consequences of rampant ETF growth, the mutation of the product, and possible interactions with contemporary trading developments.
By highlighting these physical ETF risks, they’re pointing out potential weaknesses and flashpoints in the system. Flaws to be addressed now by regulators, investors and the industry, rather than by blackened survivors raking through the rubble.
It is the work of regulators keen not to be caught napping again quite so soon after the credit crunch.
What should happen next?
If the security lenders have us surrounded, then we need to know what they’re up to. And we need to be able to compare those activities in a meaningful way: Apples with apples – rotten or otherwise.
The ETF industry has barely moved a muscle on the issue so far. A fund’s factsheet and website page should contain the following information, in my view:
1. Does the ETF actually lend securities?
Not all do. You can check which iShares ETFs lend out securities on the individual fund’s page in the latest annual report (but not the interims). If no securities lending income is reported then you’re in the clear. Although that doesn’t mean that fund can’t lend in the future.
Credit Suisse’s latest semi-annual report says they didn’t indulge in any securities lending, although they’re perfectly entitled to according to their prospectus.
If you’re in any doubt about the position then contact your ETF’s provider and ask them.
2. Are we being well rewarded for the risk?
iShares split securities lending income 60:40 in favour of the investor. They dine out on 40% of the income. The other 60% takes the edge off an ETF’s costs.
What I’d like to know is would anyone else give me a 70:30 deal, or better, to compensate me for the risk that is all mine, not iShares.
Investors certainly aren’t getting any love from HSBC’s growing range of physical ETFs. Their latest annual report baldly states:
The Investment Manager is entitled to retain any fee received for such [stock-lending] services for its own benefit and not for the benefit of the relevant Sub-Fund.
As we’ve seen, Credit Suisse’s ETF securities lending programme is on hold. They don’t mention what the revenue split would be if they cranked it up.
3. How much income is being generated?
I’d like to see securities lending income presented as a clear benefit to the investor. For example, what percentage did it knock off the TER?
iShares is the only provider I know of that coughs up details on security lending income for individual ETFs. As mentioned, you can find the info buried deep in the annual report (though not the interims) which reveals:
- Lending income generated in pounds.
- The value of securities each ETF has on loan.
- How much collateral is covering the loan.
You can compare these figures with the ETF’s Net Asset Value (NAV) to work out how exposed the fund is.
It would be much more investor-friendly if iShares (and everyone else) provided the information for every ETF as per this snapshot (choose the institutional settings to view) of selected funds:
This info gives you an instant gut-feel for how hawked around the ETF is as the loans are presented as a percentage of the fund’s NAV. The average on-loan percentage is crucial, too.
We passive investors don’t want to fritter away our lives checking this stuff daily.
4. Basket case
There’s been plenty of woolly calls for more ‘transparency’ regarding the actual collateral basket holdings of ETFs. The idea is that a rummage in the basket enables investors to perform ‘due diligence’ on their collateral risk exposure.
That’s all very well, but the information has got to be readily useable by retail investors. Too often transparency or disclosure means a provider pumping out a spreadsheet containing hundreds of listings that defy analysis by the average punter.
I’d like to see a simple rating system devised that would enable the collateral basket and the credit worthiness of counterparties to be easily assessed. Perhaps using the rating agency methodology of AAA and so on down the food chain.
No doubt this would be proclaimed expensive, but it’s not as if the ETF providers’ risk managers don’t already know what they’re getting themselves into. (I hope!)
These reporting requirements should be consistently presented for ease of comparison across providers. They should also apply equally to other UCITS funds marketed to retail investors, too. Why single out ETFs when my pension fund is merrily renting out my shares too?
Where’s my risk analysis department?
My biggest fear is not that physical ETFs have suddenly become radioactive and I’ve lost my asbestos gloves. Risk abounds, so I just need to know what I’m getting myself into.
My biggest fear is that the reports prompt a raft of new regulation, or the ETF industry makes pre-emptive moves to put their house in order, but in such a way that the little guy is still left none-the-wiser.
Given that we’re all increasingly thrown onto our own mettle to provide for our future, we need the risks of investable vehicles, including physical ETF risks, to be fully understood by the average Joe.
Take it steady,
Er – perhaps I’m reading the wrong table or something, but on my reading of the iShares document you link to (I’m looking at the pages numbered 4 and 5 and headed London Stock Exchange), an asterisk beside a fund indicates that it has UK Distributor status, which sounds like a different thing from participating in securities lending. Please can you clarify?
I also interpreeted this passage
as an indication that lending applies to those with superscript 2 – which woudl account for the TER 0.35% (0.15%) – which I take to mean that the underlying TER is 0.35% but because fo this dodgy carry-on the TER has been reduced to 0.15% for the relevant reporting year
Sorry! I misinterpreted the small print. That factsheet doesn’t help at all. Instead, you can check the individual fund page in the annual report and see if any security lending income is reported. I’ve edited the body text in line with the above. Thanks for keeping me straight!
1. Thanks for another great post, TA.
2. How exactly does stock lending take place here? That is, do all the relevant ETFs from a given ETF issuer act independently in this regard? Or, is stock lending effected at the level of ETF issuer?
3. Your call for “a simple rating system”: yes, but can we really trust the rating agencies and their assessments now in light of recent history?
4. Aren’t all physical ETFs other than those using full replication exposed to stock selection risk?
Have you investigated plain vanilla index funds too? A very quick scan of the HSBC Index Funds prospectus, where I have been contemplating putting some money, reveals that stocklending is allowable in those products too, so presumably the same risk warnings apply.
A few Devil’s Advocate points and follow-ups to The Accumulator’s super post, which together with the previous synthetic ETF one I think are perhaps the best primers I’ve read on this subject since it emerged in the past few weeks.
Firstly, the ETF provider/manager may well rebate some of the money made by lending to reduce annual fees, but I’m skeptical that many of us can really take a view on whether that compensation is adequate for the risks they are taking by lending out the stock.
Say the TER falls from 0.6% to 0.4%. Is that a great or a bad reward for these risks? I for one don’t have a clue either way.
Also, keep in mind the risk/rewards could be extremely asymmetric — i.e. Say in our lifetime only 0.1% of ETF holders ever suffer a problem, and perhaps 0.01% of them blow up and lose the invested funds completely. (I’m giving this as a purely hypothetical example, NOT making any suggestion at all that it reflects the likelihoods).
This might make ETFs very safe overall. But if you’re one of the unlikely 0.01% and your ETF blew up because of something that had nothing to do with your chosen asset allocation strategy – you were just trying to track an index at reasonable cost – and that perhaps you didn’t even know it was doing, you’re unlikely to be mollified. Especially if you had a lot/most of your wealth in it.
This is one of the reasons why I always raise my eyebrows when even some of my favourite authors (e.g. Mike at Oblivious Investor) say they have their net worth in 2/3 index funds or similar. You want to diversify away from the unexpected and unlikely when the impact is potentially, personally, catastrophic.
As I’ve written before, most of the successful rich people I’ve known or studied diversify widely. In contrast, nearly all the worst horror stories I’ve read about, from split caps and banking failures to Bernie Madoff, whatever, begin “I thought it was a safe place to put all my money”.
I know I’m a stuck record on this, but there you go. 😉
Personally I think we should do as the rich do and over-diversify to some extent, even away from entirely passive, collective vehicles, once we have significant funds to carry the costs.
But I know that’s an extreme view – and I do think the odds of losing your money through non-synthetic ETFs are really very low. As I hear it, these bodies are warning of dangers to the system as much as to individual investors. e.g. If £1 billion of ETF money is temporarily suspended and inaccessible, that’s a shocking development with diverse potential consequences, even if everyone invested does get their money in a few weeks.
If I recall correctly, some ETCs stopped trading for a bit when Lehman Brothers went under, as it was the counterparty to those securities. I don’t recall the exact details, but they may offer a clue.
@ Alex – 1. Cheers! 2. iShares quote lending income per fund, so we can speculate that it’s at the fund level in their case. No idea what standard industry practice is on that score. 3. It’s hard to trust anyone. Supposedly the rating agencies have tightened up procedures a great deal since being exposed by sub-prime. At least in this instance, we’re talking about transparent baskets of equities and bonds that have to match certain regulated criteria rather than the latest in black box financial engineering. The main point still stands – how can a retail investor judge a basket of obscure securities that changes every day? 4. For the purposes of this article, I’m only concerned with the risks identified by the regulators’ reports.
@ Dilbert – Yes, I tried to make this point in the article, along with active funds, pension funds and so on. Regulators aren’t sounding the alarm bells about other vehicles, perhaps because (a) security lending exposure is particularly high in the ETF industry or (b) because security lending combined with the instant liquidity demands of an ETF vehicle is particularly combustible or (c) because they were only looking at ETFs in this report.
@ The Investor – Re: TER idea. The main thrust of that one is to give investors a comparable sense of how much they are benefitting from security lending. If an investor is choosing between two comparable funds: Fund A = 0.4% TER and Fund B = 0.5% TER then he might as well plump for Fund A, all things being equal. But if security lending transparency meant that he could see that Fund A gained 0.2% from lending and Fund B gained 0.05% then I can see that being useful information that might alter the calculation somewhat.
I know what you’re saying about do as the rich do, but for those of us on more meagre investing rations, it’s tough to own property on five continents (except in a fund!) along with a heap of Picassos and Krugerrands in the bank vault. Funds are our means of diversification. And for many punters, the complexity of owning half a dozen different funds is onerous enough. Still, I think you’re right to err on the side of paranoia. That’s why I split some of my bigger asset allocations between multiple comparable funds from different providers e.g. FTSE All-Share trackers.
@Accumulator — Ah, as a comparison between two otherwise similar funds then that definitely makes sense.
I hear what you’re saying about diversification. There’s no doubt it costs, unless it protects you and saves you money of course! And even a cost if say 0.5% a year is going to look horrible compounded.
I think your approach of splitting funds between providers is definitely part of the solution for pure passive investors who can’t stomach that cost. Maybe between types of entities, too — e.g. A bank is different to a life assurance company. Once you hit a six figure portfolio, it might even be prudent to try to open an account overseas!
Of course the downside of spreading your eggs about is one is more likely to be hit, too. While operating within the insurance schemes etc, it probably won’t pay to be *too* paranoid.
That said I hold shares in four different places, as a for instance! 😉
TA, I fully understand your response to my question in para 4.
My only point was: while we’re right to hypothesize about potential risks for physical ETFs, we shouldn’t forget that anything other than full replication must, by definition, entail tangible risk for this type of ETF. Such risk due to composition difference from the relevant index is easy-to-understand – and avoidable. Hurray!
How would you avoid it?
1. By only using full replication – not partial replication!
2. But I know full replication has several potential disadvantages. For example, it’d surely be expensive for indices comprising many constituents because of costs incurred in rebalancing such an ETF. Also, some stocks and markets are more liquid than others, of course.
3. I feel that synthetic ETFs have a key advantage here: we can track any index very accurately without having to solve the problem of how to identify an effective strategy for partial replication. I sometimes think people assume all physical ETFs use full replication, when of course they don’t.
As an aside, synthetic replication in general is an interesting / potentially dangerous area. There are mutual funds in the US claiming to semi-match certain hedge fund strategies (much cheaper!) by holding often different assets that just happen to have historically been closely correlated. I fear the worst!
Those funds you mention: they do things differently in the US, don’t they? Well, that’s what my investment advisor, Mr Madoff, told me.
@ Alex – Heh. I was hoping you were going to say that you ran a sophisticated strategy of owning the shares that index funds found hard to reach. Then we could have had a right old ding dong 😉 Of course, you’re right that I’d rather be packing a full replication ETF than a partial. But I find, given the state of the ETF market, I often don’t have the choice for the markets I want to track. I have to take what I’m given regardless of niceties like the replication strategy. But, as you say, it’s important to know there are different strategies available, on the rare occasions when our cup runneth over.
Incidentally, did you see the new fishing ETF released in the US? Ticker EATX. Just gotta get hold of that one 😉
No, I hadn’t seen EATX. Clever ticker that for an ETF on the global food industry though! Not “fishing”, as you said?
What we really need are: ETFs on ‘phishing’…now, that seems to be a buoyant sector…Er…
FISN – that’s the fishy one.
I think I’ll leave that one in the water, thanks.
Meant to say – love the idea of a phishing fund. I can just imagine turning up at a shareholder meeting and demanding they sort their grammar and spelling out in order to increase returns.
Does anybody know if there is hidden risk with the gold supposedly held for a physical gold ETF?
@John — There’s a lot of hedge fund money in those funds, and I presume they’ve done their checks, but there’s no guarantees. I can think of scenarios that wouldn’t be helpful, that’s for sure, even beyond Doomsday Day scenarios where your ETF Gold proves totally irredeemable compared to a few Kruggerands buried beneath the kitchen floorboards… 😉 You could always consider diversifying some of your gold horde into a physical provider like Bullion Vault.
Not a personal financial advice style recommendation, just another option for you to consider.