01.18.11
Stock Lending: the current state of play
The UK financial regulators recently conducted a consultation regarding the topic whether there were conflicts of interest between asset owners and their managers, regarding short-term and long-term perspectives, and which practices tended to promote one at the expense of the other. I received an inquiry from a prominent academic regarding stock lending, and asking me whether comments regarding it were warranted in responses to the consultation. This gave me occasion to make an updated review of the ongoing issues in this area. This is an amended version of my reply:
Dear Professor C_____:
In response to the question whether it is worthwhile to discuss stock lending practices in the context of a UK consultation regarding short-termism versus a long-term investment perspective, I would summarize the current state of lending activity as follows:
The biggest problem with stock lending in this context of short- vs. long-term perspectives is that it provides a powerful inducement not to vote on issues which are otherwise deemed important by the investor in question—either the asset manager or the beneficial owner. This can produce conflicts between owners and managers, particularly when a major part of the managers’ business model is built around margins from lending, as it often is with index fund managers. Related issues are that it sometimes interferes with other engagement activities, may on rare occasions threaten the integrity of the shareholders’ meeting, and that lending is not quite as risk-free as it has often been portrayed to trustees. But the real stumbling block from the viewpoint of the lender is the effect upon voting.
The ICGN Securities Lending Code, which addressed these issues from an international perspective, called for more transparency in lending activity. In particular, lenders should (1) implement a clear policy regarding these potential conflicts between voting and lending, and make it public, particularly to beneficiaries, (2) that the income from lending be reported separately from investment income, and (3) that all parties to the lending process pledge not to lend shares in suspicious situations, when they have reason to believe that someone is borrowing shares in order to influence a vote. This third point has been much ridiculed by the agents and custodians, but in fact there have been cases where loans suddenly went ‘special,’ (i.e. someone was suddenly willing to pay a significantly higher margin than usual) with unusual or rigid delivery terms, in the days just before the record date for a controversial vote. Borrowing shares for the principal purpose of voting them is banned in the US by Reg. ‘T’ and frowned upon in many other markets, but it is not usually illegal, and is in any case difficult to police. The lending industry operates on the basis of ‘see no evil, hear no evil,’ and this may sometimes be an inducement to overlook the obvious.
In general, our attitude in drafting the Code was that lending is a useful activity which improves market liquidity, but that the tail shouldn’t be allowed to wag the dog. In and of itself, we see nothing wrong with some market participants profiting from a decline in a share’s price: it is a part of price discovery. Also, many loans have nothing to do with constructing short positions, but are rather for covering fails and the like, or for creating hybrid hedged positions in which the ultimate borrower may also be net neutral, or even net long. And although the most frequent function is indeed to construct a short position, often the bet doesn’t work or has only limited effect, and the lender has lost nothing on the capital side either. But this isn’t always true, of course, and sometimes the short sale might add to the weight of selling putting downward pressure on the share price.
With reference to the UK consultation: One good thing about the UK which we don’t enjoy in the US or Canada is that the agenda is posted long before the record/reconciliation date, and therefore, investors have an opportunity to recall shares after seeing whether there is anything to be voted upon they actually care about. One problem however, in the UK, which is not usually a problem over here is that there is normally a totally separate chain of authority for authorizing and transmitting voting instructions from that of portfolio management, including decisions whether a particular vote affects shareholder interests. The right hand is not normally allowed to know what the left hand is doing. This could be addressed by re-writing the contracts to allow for direct communication between managers and custodians, but to the best of my knowledge, no one has done so.
A problem on both sides of the Atlantic is that lending is more and more being done by third parties, many of whom have authority to lend shares without informing the beneficial owner of specific loans. Thus, a beneficial owner often gives out orders to vote its shares, not knowing that the majority of them have been lent, and may not be able to be recalled in time. The procedures involved in the process could use updating, but those most involved are most resistant to any changes in the system, or any further investment in technology.
Since the margins on most loans are very thin, and the construction of short positions by prime brokers is very lucrative, it could be argued that the lenders are throwing away their votes for too little. For this reason, a few asset owners have withdrawn from lending entirely, as not being worth the cost. But there are many others willing to step into the breach.
A danger particularly of the U.S. system is that lenders always have the temptation to augment their returns by reinvesting the collateral in riskier paper. This was highlighted by the problems some pension funds had in 2008, when they discovered that their agents had reinvested the collateral from loans in risky and less liquid instruments, and then had to return the collateral on short notice without being able to cash the investments they had made, or only at large losses. Several public pension funds in the U.S. had put some of their collateral into SIVs! So the riskiness of the portfolio was actually increasing, relative to the equity position.
Another problem to emerge in 2008 was counterparty risk. Since a loan is essentially a contractual, unsecured agreement with a prime broker (the collateral is posted by the actual borrower, the prime broker acting as its agent), there is some risk that if the counterparty goes belly up during a loan, part of the transaction may be tied up in bankruptcy proceedings. In the US this risk has apparently been dealt with by changes in the law to explicitly protect such agreements, but to the best of my knowledge, in the UK it has not. There were instances of lenders who had ‘their’ shares (i.e., the commitment to buy them back in) frozen for some months after the Lehman bankruptcy, but who still had to return the collateral.
In answer to your students’ question, yes, individual portfolio managers often want to know whether someone is betting against their long positions, especially when these are in mid- or small-cap stocks which may be more easily influenced by a few big shorts. They sometimes ask to recall, when they know that such shares are being borrowed against their own positions. However, if the lending institution is on average a long-term holder, they are usually told by their senior managers or trustees that the income is significant, and that over the longer term, the effect on prices is not significant. They are also told that the process is a ‘riskless’ source of cash income (nothing is riskless, but never mind) and that they should shut up. The greatest conflict, in my opinion, is not between asset owners and asset managers, but between those (senior managers, directors, trustees, and of course anyone in the lending apparatus) who have an interest in maximizing annual income, and those (portfolio managers, corporate governance staffers, and concerned beneficiaries) who are concerned about the capital return, either directly (the PMs) or indirectly over the long-term (governance types).
Feedback from the other parts of an asset owner or manager would usually act to mitigate some of these problems. But lending is usually treated separately as a back-office function, and flies beneath the radar of most decision-makers. Those administering the lending are (a) usually not involved economically with any other part of the investment process, and (b) always rely upon the argument that even a relatively piddling amount is worth vastly more than a ‘worthless’ vote. Asset owners and especially their trustees have to decide beforehand whether votes (other than on sale of the company) are worth anything to them, and instruct their lenders accordingly. Most don’t.
I don’t necessarily want to stir up a controversy between active UK lenders and governance advocates, but the share-lenders’ lobby has been very aggressive, and includes several asset owners which otherwise bill themselves as proponents of good governance. I think many such funds have failed to integrate the issues with potential conflicts to governance policy because internal advocates of lending grossly understate some of the problems. Indeed, the difficulty with ameliorating many mechanical difficulties which could be solved without suspending most lending activity is because the position of lending advocates is that there should be no encumbrance upon lending at all.
Incidentally, although it won’t provide lenders with substitute income, there is developing an alternative to the stock loan: the single-stock future, especially when incorporated into an Exchange for Physical contract. It’s cheaper, simpler, and doesn’t require anyone to borrow stock. (Disclosure: I am an independent director of OneChicago, the single-stock futures exchange.)
I look forward to discussing the subject further with you, if you are interested.
Best regards,
Andrew