Shorting and voting

As some people will be aware, I've been looking into changes in voting turnout at companies in trouble. Short version: turnout has fallen at several such companies, in some cases significantly. Why this might be is another question. Since several of these companies were also been heavily shorted there's a potential explanation. Perhaps investors lending shares to those shorting were not recalling them in order to vote.

So I've had a look at public shorting data (which I know is limited) and plotted this against turnout. The charts below are v basic. I've just looked at the short position on (or shortly prior to) the date of the relevant meeting. Really I ought to put more shorting data in, as obvs the shares don't get lent out and shorted immediately before meetings.

Anyway, it looks like there is a bit of a correlation, although the Debenhams example does not show that at all. So it's possible that something else is going. Perhaps changes in the register explain it - as companies get into trouble big institutions who are more likely to vote are no longer there in numbers. Or perhaps some investors just give up - if you're an indexer who has to hold the shares, but know the company is doomed why vote? Why not at least get a bit of lending income? So not a straightforward story.

PS - Premier Oil obvs not in the same category as the others. I included it as the story about the massive short in the company is in the press today.

Dividend trades and voting

Obviously, I've not been blogging much lately, but something I've been spending a bit of time looking at is dividend arbitrage. In it its most well-known version, this is the practice of shifting stock around the ex dividend in order minimise tax payable. There is a variety of different trades, and at one end of the spectrum some of them are the subject of legal cases. (great piece on Macquarie's involvement here).

What all of them involve is stock lending. I recommend having a read of this paper from Richard Davies IR, which opened my eyes to the scale of lending that is going on in the UK. 20% to 30% of stock is going out on loan around the ex dividend dates of major UK PLCs, which immediately makes me think about the potential governance impact.

It's very hard to pin down who is involved. But one thing that can happen with large movements in stock is that they trigger regulatory announcements because voting rights thresholds are crossed. These are are the TR1 notices, which appear with the title "Holding(s) in company" if you look on sites like Investegate.

So one of the things I did was look at TR1 notices issued around the ex dividend dates of a few companies. And I can see some, Blackrock in particular seems to be triggering them on a regular basis. These appear to show a shift in allocation of voting rights a few days before the ex dividend date and back again a few days after it.

Where it gets particularly interesting is when the ex dividend date is close to the AGM. If this shuffling of stock involves lending some to another party then there *might* be an impact on voting turnout if shares aren't returned in time to vote. I have identified cases where voting turnout has gone down (very significantly in one of them) when the ex dividend date has been close to the AGM date.

I can't say for certain if the stock-lending is a) linked to a dividend trade or b) resulting in lower voting turnout. But it's a bit of a coincidence.

What business are asset managers in?

Another issue I am starting to look at in more detail is the overlap between stock-lending, shorting and other activity. I was recently pointed in the direction of some of BlackRock's literature on stock-lending and it fed my sense that at least parts of its business are not necessarily all about asset management as we might typically understand it.

Here are a couple of excerpts:

The first bit shows you how it works (in this case for Exchange Traded Funds). You deposit money with Blackrock in an ETF invested in a stock index. Blackrock lends some of the stock (for a fee, obvs) to an institution and uses the collateral provided in return to potentially get a return from a money market fund. Blackrock splits the income generated with you and offsets this against the management fee. In some cases the lending income you get is larger than the management fee.

What's Blackrock's cut of the lending income? About a quarter to a fifth according to the blurb:

(incidentally, Blackrock looks like it takes a bigger cut of lending income from European funds, but that's another story)

So if we assume that the 73.5% figure applies to the top 3 Russell 2000 ETFs in the list, then that suggests that the total lending income from these funds is 35bps, 27bps and 19bps respectively. That would make total lending income 40%+ higher than the management fee in the first two cases, and 21% lower in the third. If that 73.5% is correct then Blackrock's cut of the lending fee (26.5%) adds a reasonable chunk to their overall income.

iShares Russell 2000 Growth ETF - Management fee (24) + lending fee (9) = 35bps
iShares Russell 2000 ETF - Management fee (19) + lending fee (7) = 26bps
iShares Russell 2000 Value ETF - Management fee (24) + lending fee (6) = 30bps

There may even be a dribble more income for Blackrock as I assume that they would use their own money market fund to stick collateral in, so perhaps another management fee there?

If the actual level of stock-lending income in some of these cases is higher than value of managing the portfolio of assets, that suggests that there must be quite a lot of lending going on. Does it also suggest that the funds don't hold own the stock in the index they notionally track for prolonged periods? I don't know anything like enough to comment sensibly.

But what is quite obvious is that this looks like a nice business to be in. You take the clients' money to put in, say, a US smallcap ETF. You loan the stock out to someone who wants to borrow (for whatever reason) and take a decent cut from the income from renting assets bought with your clients' money. And you might even be able to make a few crumbs managing the collateral the lender gives you while the stock is out on loan.

Questions this raises for me are what is Blackrock really selling (and what are you really buying), and how does it see the business? It would be really interesting to see how much of the stock is out on loan at any time, because to me it looks a bit like in some cases Blackrock runs a stock-lending service facilitated by having an asset management offering. If the total income generated (though not Blackrock's cut) is higher from the former than the latter, how is it seen internally?

It's also important in the current context where asset management fees are in the spotlight. Based on the fees above, it looks like they are giving the asset management away for free for some ETFs - a very tempting offer. In reality they are actually earning more than the headline management fee, from two sources. And it only hangs together because the client provides the money to make it work and because lending is lucrative.

Some of this reminds me of the origins of asset management. Back in the 70s and even into the 80s, asset management was not a big deal in a lot of markets. Banks provided it as an add-on service alongside more profitable areas of work. Peter Stormonth Darling's book on the history of Mercury Asset Management is fascinating on the early history of that manager (which ultimately ended up as part of.... Blackrock). Warburgs were pretty much willing to give the business away to Flemings at the end of the 1970s, but couldn't. It was only when they realised that they could get away with charging a lot for the service that the industry really took off, and created what we see today.

Perhaps now that fees are (finally) under serious scrutiny we'll see more of this kind of activity as managers look to try to make money elsewhere to keep the fees they charge down. But as I say, it does raise some questions about what the business really is. And this is without even getting into the question of Blackrock's numerous short positions, and where the stock comes from for them.

Stock-lending and shorting, again

I've been thinking a bit more about the potential conflicts of interest around stock-lending and shorting in light of the RD:IR paper I blogged about a couple of weeks back. It suggested that passive managers are in part able to offer lower fees because they make money through stock-lending.

I was initially interested in the way that this might facilitate shorting on the cheap by the active side of the same business. If you're managing passive assets on a large scale on an AUM-based fee then I think you can probably lend to your active side safe in the knowledge that even if they call it right the decline in one stock isn't going to rock the boat (NB - I'm not saying shorting makes share prices decline!). I think that is still worth digging into, just to find out a) how extensive lending by passive managers is b) how much the passive clients make out of it and c) if cross-group lending is done, and, if so, if it's charged at a cheaper rate than to external clients.

But what interests me is what other issues might arise. I am reminded of this bit from an old Takeover Panel paper about the conflicts created for counterparties to derivatives in bid situations:

First, the Code Committee believes that, notwithstanding the contractual arrangements between them, a counterparty will usually know the derivative investor’s likely wishes and therefore it would be naïve to assume that the counterparty (who has no economic interest in any hedge securities it holds but who does have an ongoing client relationship with the investor) will act without having some regard to those wishes. In addition, as indicated in paragraph 3.3 of PCP 2005/1, the Code Committee understands that it is frequently the expectation of a long derivative investor, notwithstanding the terms of the documentation, that his counterparty will ensure that the securities to which the derivative is referenced are available to be voted by the counterparty and/or sold to the investor on closing out the contract. If the counterparty does not hold any such securities (because, for example, its book is balanced by an offsetting short derivative), the investor would normally expect the counterparty to acquire the necessary securities, even if that resulted in a cost to the counterparty.…..the Panel continues to encounter situations where holders of long derivative positions behave as if they were shareholders and, more importantly, situations where investment bank counterparties enquire of investors with long derivative positions as to their preferences in terms of bid outcomes in order that the counterparties may take those preferences into account;

Do similar issues arise for passive managers who stock-lend? Say you're a regular lender to funds doing the merger arbitrage trade. Say a big contentious hostile bid comes up and your lending clients have taken a big punt shorting the acquirer in the expectation that the bid succeeds. Is there any tension in how you decide to respond to the deal? Big passive managers are going to be long in both target and acquirer. I'm pretty sure there must be an optimum outcome in this situation, but if your positions are passive (so you're not being judged on performance) I suspect you've got freedom of movement to decide how to respond.

If you oppose the bid, and it fails, it's quite possible that stock-lending clients will lose a bundle. If so, will they come back to you as a borrower next time? I have heard anecdotes about borrows being sniffy about lenders who want to recall to vote, so I could imagine some conflicts / client pressure.

Another area to dig into...

Shorting and stock-lending

Via LinkedIn (sorry!) I came across this interesting paper from RD:IR on what's going on with regards to stock-lending in the UK. Definitely worth a quick read for anyone with an interest in this area.

A couple of headline - the overall level of lending of UK stocks decreased after the crash, but has been rising since 2015. However this is driven by lending of FTSE250 stocks rather than FTSE100. At least part of the explanation appears to be Brexit, with lending in the FTSE250 increasing significantly post June 2016 (and the FTSE250 being generally a better representation of "UK" companies).

Although the paper makes clear that not all lending is to facilitate shorting, nonetheless it's clear that is what the bulk of it is for. As such there is a snapshot of who is doing the shorting:

As at 20 November 2018, there were 609 disclosed short positions greater than 0.5%, totalling £13.4bn. The biggest “shorters” in UK Plc were Marshall Wace with £1.4bn, AQR Capital Management with £1.3bn, and BNP Paribas with £925m total shorts open and disclosed to the FCA. BlackRock, across its various global entities, held £896m in open short positions above 0.5%. 

This is based on FCA disclosures, which means the actual levels will be higher (because there will be a lot of short positions between 0 an 0.5% and some of them will be held by the big players). As noted in the previous example of Kier Group, the total shorts were more than double those disclosed in the FCA list, according to other data sources.

Also, as a comparison, here are the largest 20 UK shorts disclosed on the FCA list as of Friday.

Marshall Wace, AQR and Blackrock are all in there, though no BNP Paribas.

Another interesting point in the paper is the the interplay between quant funds and passive managers, and, in particular, the importance of the latter in facilitating lending:

One of the key reasons for the ability of passive investors to charge lower fees than active managers is the additional revenue stream of securities lending. This activity may seem contradictory to the classic view of investment, as you would invest into a certain sector/index via an EFT or index fund in the hope that your investment will increase in value. However, the fund manager or custodian for that fund may well at the same time be providing the market with shares from that fund’s portfolio to allow other investors to go short in that portfolio’s constituents. 

The rise in passive investing via index and quant funds is fuelling shorting and stocklending. Quant funds are a driving force in the activity, with algorithms automatically opening short positions because of the sheer size of the lending market. At the end of 2017 the global market of tradable assets stood at more than $20tn, with over 10% of this total being lent out. The European market alone generated revenues of $2.6tn. Expectations are that 2018 produced even higher figures and 2019 more so. Of course, it is not just passive investment managers that are providing the liquidity in the lending market, with active managers seeking additional returns to cover falling fees in the ultra-competitive asset management market.

Once again, my eye is drawn to Blackrock. They are both a major passive manager and a major player on the short side. If they want to short stocks they need to borrow them, so does the passive business lend to the bits of the business that go short, and if so do they charge the same rate they would do to other managers looking to short?

I keep returning to this, because something about all this makes me uneasy and I'm not quite sure what it is and why. Anyway, one for another day.

PS - I am pretty much convinced that the FCA should disclose all short positions that are reported to it. I appreciate this will make for a bigger list (and thus more admin). But 0.5%+ is a big hurdle.

Stock-lending snippet

spotted this on page 50 of the PBR:

The FSA has been reviewing the governance and risk management of stock lending in the market. The Government welcomes this work and will work with the FSA and market participants as necessary to help develop thinking in this area.

A snippet on stock-lending

Interesting blog survey on stock-lending levels.

BT pension fund halts all stock-lending

According to a report on Thomson Investment Management News. Here's the key section:

The BT Pension Scheme (BTPS) has suspended all stock lending on concerns that short-sellers using the shares could further hurt market sentiment, two sources close to the pension scheme said.

The BTPS, Britain's largest pension scheme, in September placed some 20 British and global financial companies on its list of stocks it had barred from lending, but has now widened the ban.

"The BT Pension Scheme stopped all stock lending not just on financial stocks some months ago. Cannot say if the (decision) is permanent but it is still considered to be prudent at this time," one of the sources said.

This, I would suggest, is a pretty big deal, as BT is the biggest pension fund in the UK. Some of the feedback you get from in-house staff at funds is that such moves are driven by trustees, and it's sometimes an ill-informed decision. But the BT trustees are a pretty competent bunch, and surely recipients of much more specialist and expert advice than many others out there. So this must be quite a significant decision.

It may also suggest that the theoretical assumptions around stock-lending and shorting are beginning to shift and part of a wider reassessment of views about how markets work - something that came up in the Turner Review. The Turner Review of course asked whether decisions about shorting should take into account the dangers of market irrationality - suggesting a rather broader conception of how financial markets operate. Could BT's decision suggestion something similar developing in relation to stock-lending?

I'll briefly restate my own views here. I'm not particularly bothered by shorting in principle - because I think that it's just a set of trades so no better or worse than any other - but nor do I buy the argument that it plays as positive function as it enables negative sentiment to be expressed. A (sort of) comparable argument would be that borrowing money to buy shares is a good for markets as it allows vakuable positive sentiment to be properly expressed. And all that extra trading must be (in aggregate) a cost to someone, somewhere.

Stock-lending does provide an income for long-term investors, but it also confuses the ownership question. First, you may lend to someone who shorts the company - is that in your long-term interest? Second, you may lose the voting rights, meaning that the oversight of corporate governance is reduced. So I think there are questions for investors to ask (and some may conclude that actually the lending income is worth more).

These are by no means straightforward issues, but it does seem that sentiment is turning away from the simple mantra that shorting and stock-lending are good for liquidity and market efficiency.

Select committee stuff - stock-lending

Following on from my previous post, here are some bits and pieces from the ISLA's submission to the committee's inquiry into the banking crisis. I have to say I find the ISLA's stuff really interesting reading, even if I don't agree with some of what they come out with. The text below is from this doc - pages 48 onwards.

Executive summary:

• In general, share prices are driven by changes in the underlying fundamentals of companies and not by the actions of particular groups of buyers and sellers.
• Short selling did not cause the falls in bank share prices since 2007.
• UK financial share prices have continued to fall and remained volatile following the imposition of restrictions on short selling by the Financial Services Authority in September. Independent academic research has found that the restrictions in the United Kingdom and elsewhere have had no discernible effect on subsequent movements in the prices of financial shares.
• But, based on research by the London Stock Exchange, those restrictions have reduced liquidity in the market for financial shares and raised trading costs for investors.
• They have also constrained the ability of investors to hedge their positions and led to significant compliance costs for market participants.
• ISLA welcomes the planned consultation by the Financial Services Authority on regulations for short selling. Any new regulations should be linked to clear regulatory objectives, follow full and open consultation, be evidence-based, have a sound legal basis and be subject to rigorous cost: benefit analysis, as required under the Financial Services and Markets Act 2000.
• ISLA also welcomes the initiatives by IOSCO and CESR to harmonise regulation of short selling internationally, as differing rules create confusion and compliance costs for market participants.

From a bit further in:

Short selling and share prices

4. Academic research has shown that restrictions on short selling reduce market efficiency and liquidity. Studies have found that allowing short selling:

• means prices adjust more quickly to new information about fundamentals;

• decreases the likelihood of price bubbles;

• leaves unchanged or even reduces the probability of price crashes;

• may lead to higher equilibrium prices: because investors have greater confidence that prices are fair and therefore require lower returns to compensate them for risk.

And about the Cass paper I've mentioned previously:

8. The main findings were:

• No strong evidence that restrictions on short selling changed the behaviour of stock returns. Stocks subject to the restrictions behaved very similarly both to how they behaved before their imposition and to how stocks not subject to the restrictions behaved.

• Comparing behaviour across countries where the nature of the restrictions differed, no systematic patterns consistent with the expected effect of the new regulations, i.e. no evidence of a reduced probability of large price falls.

• No sign of any detrimental impact of the constraints in terms of reduced efficiency of pricing.

• Regression analysis suggested that changes in stock returns were driven mainly by other factors affecting the financial sector as a whole rather than the restrictions on short selling. That is, some systematic changes in the behaviour of financial sector stocks could be discerned, but no strong evidence of a systematic impact of the restrictions could be identified.

Just one obvious point, two of the bits of evidence submitted appear to contradict each other. This -

Academic research has shown that restrictions on short selling reduce market efficiency and liquidity.

And this -

No sign of any detrimental impact of the constraints in terms of reduced efficiency of pricing.

I really can't get my head around this efficiency argument (which is why I'm not surprised that the Cass paper found no evidence of 'inefficiency' as a result of the ban). I'm sure there is a certain level of liquidity required in order for price formation to be broadly 'efficient' but I can't believe that long-standing equity markets like the UK's don't already have this, and that we therefore need short-selling in order to facilitate it. There's something else lurking at the back of my mind that bothers me about this efficiency argument that I can't quite articulate yet. I'll come back to it.

In the meantime if anyone has any thoughts on this one - particularly making the efficiency case - I'm all ears.

NAPF and stock-lending

I'm think this headline - NAPF chairman urges schemes to restart stock lending programmes - might oversell what the NAPF chair actually had to say, which was:

"Don't abandon your securities lending programs. Having stock to lend and borrow is crucial for efficient markets."

But still I find the pro-stock-lending stance from a trade body meant to represent pension funds - not custodians - a bit odd. For one, arguably he's endorsing a particular strategy - why not tell his members they must do commission recapture? Secondly pension funds also remain split over this issue. In addition shouldn't the membership give the steer to the representative body, not t'other way round?

And what about sentence two - is having stock to lend and borrow "crucial" for efficient markets? Is there evidence that markets without stock-lending are less efficient, and in what sense do we use the term? The existence of stock-lending implies more trading, which means more costs. So that might actually be less efficient - if you think extra trading simply increases frictional costs - from a pension fund point of view, surely?

A bit of a strange perspective all round, especially in light of this.

Myners on shorting and stocklending

Yesterday saw some interesting banter in the House of Lords. Given the FSA's decision to let its ban on the shorting of financial stocks lapse, the Archbishop of York asked the following rather pertinent question:

To ask Her Majesty’s Government what steps they are taking to ensure that the lifting of the ban on the short selling of shares in financial companies does not adversely affect the market.

The Government's pointman on this kind of thing is, of course, City minister Lord Myners, and he provided some good commentary in response (though anyone expecting a denunciation of shorting will be disappointed). I'd recommend reading the whole transcript of the section of the debate on shorting, but the key part in it is where the link to stock-lending is made. Here's what Myners had to say about it:

It is important to note that just about every major pension fund and every major endowment in this country is in some way or another involved in short selling. However, my noble friend made a fundamental point about stock lending practices. I have asked the FSA to look at whether those practices are sufficiently understood by practitioners and subject to appropriate regulation.

Obviously this right up my street, and it's great to hear that the Government wants the FSA to take a closer look. I've droned on about this a few times in the past, but there are a few issues for supposedly long-term shareowners if they lend out stock. Fundamentally there is the question of whether it's counterproductive if stock is returned at a lower a value, but what about governance issues such as the use of voting rights? I'm not sure many pension fund lenders bother to recall for the purpose of voting, for example. So it will be interesting to see what any FSA review covers.

Very encouraging stuff!