This is the final post in a series of three looking at the potential risks that ETFs may pose to the stability of financial markets. The previous two discussed synthetic and leveraged ETFs. Now we’ll take a look at the practice of securities lending.
Many active traders, including hedge funds and prop traders at investment banks, engage in short selling when the they believe a stock is about to fall in price. For example, if a company is trading at $20, a short seller might borrow shares and sell them on the open market for that price. If the stock falls to $18, the trader can then buy it at this lower price, return the shares to the lender, and pocket a profit of $2 per share (before costs).
Firms that borrow shares need to get them from somewhere, and the most common lenders are mutual funds, ETFs and pension funds, who use securities lending agents as intermediaries. When a fund lends shares to a short seller, it collects a fee for doing so.
The problem
The main concern about securities lending is common to all funds, not just ETFs: namely, the borrower may default on the agreement to return the securities to the lender. While all developed countries have regulations that require the borrower to provide collateral, the strictness of the regulations vary. According to the UK site Fundweb, in Europe “the assets offered as collateral are often less liquid than those that are borrowed. In the event of a market run, ETFs could find it hard to pay back clients if they have large amounts of illiquid assets on their books.”
A second problem applies to ETFs more directly. Unlike active mutual funds that charge fat management fees, index ETFs have low margins. There may a greater incentive for their managers to overindulge in securities lending—and therefore take greater risks—in order to squeeze out higher profits.
There is also a question about the fairness of securities lending. The revenue collected from borrowers is typically shared by the lending agent (which may be affiliated with the fund provider) and the fund itself. Industry standard seems to be that investors in the fund get about 50% or 60% of the revenue, which can offset some of the management fee and lower the fund’s tracking error. That’s a good thing, but investors might reasonably ask whether their share should be higher. After all, they own the shares, and they are the ones taking all the risk when they’re lent to a short seller. Why, then, are they only getting half the reward?
What Canadians need to know
As with the concerns about synthetic ETFs, the major problems with securities lending are in Europe, where the practice is more common and less strictly regulated. Canadian mutual fund regulations require that no more than 50% of a fund’s securities be on loan at any time. The collateral must be of high quality and must have a value at least 102% of the loaned shares.
You can learn the extent of an ETF’s securities lending by checking its Management Report of Fund Performance and its financial statements (look under “Statement of Operations”), both of which you can find at SEDAR. In most cases, you’ll find that the revenue is very small. For example, the iShares S&P/TSX 60 (XIU) has about $10 billion in assets and earned all of $157,000 from securities lending in the first half of 2011. The iShares DEX Universe Bond (XBB), with $1.7 billion under management, raked in a whopping $25,000 during the same period.
iShares explains in its financial statements that they split securities lending income 60-40, with the larger share going to the fund and the smaller going to the lending agents. The lending agents are affiliates of BlackRock, which owns iShares. You can read more about iShares’ securities lending practices in page two of this document.
Claymore employs a third party, CIBC Mellon, as its securities lending agent. In an email to me, Claymore president Som Seif explained that their collateral requirements are even stricter than what the regulators require. They insist that all collateral be valued at 105% of the securities loaned, and that “this amount must be in cash and cash equivalents, and cannot be invested in riskier securities in effort to generate higher returns.”
Unfortunately, Claymore does not disclose how securities lending revenue is divided between CIBC Mellon and the funds themselves, saying only that “it is in line with other players.” From looking at the ETFs’ financial statements, it seems to vary from fund to fund. For example, the Claymore Canadian Financial Monthly Income (FIE) shows some revenue from securities lending (about $40,000 annually). However, several others—including the Claymore Canadian Fundamental (CRQ), Claymore BRIC (CBQ), and Claymore Equal Weight Banc & Lifeco (CEW)—engage in significant securities lending, but no revenue from this activity appears in their 2010 financial statements.
The most generous ETF provider in this respect is Vanguard. According to its website: “Unlike other firms that allocate a significant portion of lending revenues to their management companies, Vanguard returns all lending revenues, net of broker rebates, program costs, and agent fees, to the funds. Other securities lenders may divert up to 50% of the revenues derived from their securities lending programs.”
With Canada’s strong regulatory framework and the relatively small amount of securities lending that goes on in ETFs, I don’t think the practice is unduly risky. However, as an investor I would demand full transparency when it comes to revenues: since the unitholders own the stocks in the funds, they have a right to the majority of the profits. Or at the very least, they have the right to know who receives the revenue, after which they can decide if they’re comfortable with the arrangement. Jason Zweig of The Wall Street Journal made these arguments back in 2009 and again last month.
This is especially true if the ETF in question has a significant tracking error. I expect a fund to trail its index by an amount equal to its management fee, but if the tracking error is more than that, then revenue from securities lending might be used to close that gap. If it ends up in someone else’s pocket instead, then that tells you something about the fund manager’s priorities.
I agree that there should be a regulatory requirement for transparency. I’m not sure how you ensure the fund gets its fair share of the proceeds from securities lending. Perhaps require that they use an independent agent? Or that the fees for the agent be stated in addition to the MER.
I wondered how much secrurities lending by ETFs goes on. Apparently, not much in Canada. I’d certainly be interested in periodic (yearly?) updates if you’re willing to do the work. Then I’d know whether I should start being concerned.
Very informative post!
Despite the stronger regulatory enviroment, there are still huge gaps in disclosure by all the ETF companies.
If they lost money in security lending, do you really think they’ll put a net negative number on their filing documents? It’ll just show up as poor performance relative to the index. Claymore, for example, uses obscure indices and doesn’t even allow comparisons of their fund with the index it tracks on their website.
@Andrew F: Even if it doesn’t become a regulation, I think ETF providers should disclose this voluntarily. If it adds value, as they claim, there’s nothing to hide.
@Michael: I will keep an eye on the activity of the primary funds in my Model Portfolios. I don’t expect this to become a problem, but it’s worth noting whether the practice gets more common.
@Sean: As I understand it, losing money through securities lending is very unlikely, given the collateral in place. But in theory, I imagine you’re right: investors would likely be on the hook.
Re: Claymore, while you can’t track this on the site, you can learn the fund’s tracking error versus the indexes in the semiannual Management Reports of Fund Performance.
Isn’t the most likely scenario, for a borrower to default, that the security went way up while the borrower was betting on it going down? Say that the security went 30% up; in such a case, there would be little consolation for ETF/Mutual Funds (MF) investors to get the 102% in collateral, as this would still represent a 28% loss!
So, when securities go down, ETF/MF investors assume all the risks, but when they go up, they also risk losing the appreciation because of security lending. This looks like a lose/lose situation for the ETF/MF investor. I don’t see much reasons for doing such lending other than so that the fund managers (and their security lending broker friends) make more money.
I really don’t like this. I thought that Couch Potato / index investing was about tacking the index (minus minimal management fees), not about risking part of the positive returns to chase interest on lent money.
Am I looking at this too negatively?
@CCP Fan: You’ve misunderstood how securities lending works. If a fund lends a stock to a short seller and the stock goes up, the borrower is still on the hook to return that stock to the fund. The fund is not giving up any upside whatsoever. The other key point is that the collateral is kept at 102% on a daily basis. So if the loaned stock rises in price today, then the borrower must post more collateral to make up for that.
The worst-case scenario is that the borrower cannot return the borrowed securities and there is a problem with the collateral. If the fund insists on the collateral being cash, then this is an extremely low-risk proposition. It would be like you lending your neighbor a watch worth $100 and him giving you $102 in cash to hold until he gives it back. The potential problems in Europe come about when the collateral is not cash, but some illiquid asset. Now it’s like lending your neighbor a $100 watch and him giving you a $102 share in an apartment building he owns, and the building would have to be sold before you could get your money. Now, if he takes off with your watch, what are you going to do?
@CCP: Actually, CCP Fan’s question made sense to me. He/she is looking for a scenario where an ETF could get in trouble with securities lending. If the lent stocks were to run up quickly, this would trigger a call for more collateral. If the borrower can’t come up with more collateral and can’t return the stock then the ETF is out of luck.
@Michael: Yes, this is a possibility, though I don’t know how big of a risk it is in practice since the collateral is marked to market daily. If at the end of the day the borrower cannot post more collateral for a stock that has gone up, presumably the lending agent would immediately take the collateral and buy back the stock. They would have 2% to 5% leeway, so the stock would have to shoot up more than that in one day before it was even an issue.
The lending agent presumably keeps a close eye on its borrowers to make sure they are in a position to handle these daily movements, but clearly that’s not a guarantee.
@CCP: If the lender sells my watch for $100, and after selling it, it was discovered (by its new buyer) that it was a historical watch worth $100000, then I wouldn’t be happy to keep the $102 share in his apartment building, instead of getting back my now $100000 watch. That’s what I’m trying to say.
I understand that it’s not necessarily a “big risk”, but it is a risk nonetheless that I had no intention to take when buying ETFs/mutual funds. It seems, though, that I don’t have a real choice in the matter, as they all do it.
I guess that the only alternative would be to buy my own stock portfolio, but that’s impractical as I wish to have wide diversification (Canada, US, international, and emerging markets).
I meant: “If the borrower sells my watch”, of course.
Thanks for having taken the time to reply to my message and for your very informative blog!
@CCP Fan: I realize you’re exaggerating, but a stock cannot suddenly increase 1,000-fold in one day, which is what would need to happen. Securities lending is not new, and I’m not aware of any situation like this resulting in a significant loss. It seems a rather remote risk, and even if it did happen, it would likely result in a very small loss.
For the record, not all ETFs engage in securities lending—there are actually many that do not. You can always check the funds’ literature to see if your funds do it.
http://canadianfinancialdiy.blogspot.com/2010/02/three-etfs-that-have-no-securities.html
@CCP: You’re right; it is quite unlikely for a security in a prominent index (S&P 500, etc.) to move up much more than 2% in a single day, to a point where both the borrower would be in trouble and the 2% security would represent a significant lost. It just that, before reading your blog post, I had not realized that such risk existed.
Thanks for your teaching!
All three posts regarding ETF Risks highlight the strength of our country’s securities framework. However, I don’t think their is an appetite for more disclosure rather investors want to be able to access information when they see fit. The point about tracking error could focus on an acceptable range allowed and those that exceed should disclose why.
Dan, superb post! You are really educating people on ETFs – myself included ;)
If a borrower puts up $102.00 cash to secure the $100.00 in borrowed stock, why wouldn’t he simply buy the stock for $100.00 and save the extra $2.00?
@Don: Remember, he’s not paying $102 for the $100 stock, he’s just posting it as collateral, with the expectation that he will get it back. His goal is to borrow the stock when it is trading at $100 and then sell it on the open market for that price. If it falls to $95, he will then buy back the shares and return them to the lender. He pockets $5 on the deal, and he gets his $102 back. I’ve simplified this, obviously, but that’s the basic idea of short-selling.