So, writing up a little stock market piece. I noted that the US, OTC, price of a share was wildly out of line with hte London, AIM, price of that same stock.


Then on closer examination that OTC price was the grey market, which isn’t a market at all, it’s simply a record of the last recorded trade price, which might have been weeks, or months, ago. So, no arbitrage.

$100 bills on the floor do exist but not for long.

Which leads to a larger musing. Of course there is arbitrage between, say, ADRs and London stocks. That’s why the prices do move in very near lockstep. The people tasked with creating the ADR itself are asked to move either way, creating or unpicking them, buying or selling at either end, in order to make those prices move in very near lockstep. This happens in large amounts and at very fine margins.

So, what about less liquid stocks? AIM that’s on the Pink Sheets say? It’s possible to imagine – imagine – that the margins are wider here. That’s rather what less liquidity means.

It’s also true that the world is reducing trading margins. Using Transferwise the commission to move $10,000 to £ might be 0.1% with no spread. Robin Hood (and Schwab etc) now have commission free trading.

Which leads to, well, what actually is the price mismatch between such more thinly traded stocks? That will depend upon whether there’s anyone arbitraging institutionally of course. It will also depend upon the details of the nuts and bolts of buying on one exchange and selling upon another. The nitty gritty of proving ownership, of whether it’s actually an ADR and thus some formal change needs to be made to turn it into a London stock, or is it actually the same thing etc.

At which point, well, does anyone know. The real details of these markets?

5 thoughts on “Arbitrage!”

  1. Any conversions between ADR’s and equities are handled by the prime broker, who holds the instruments on client’s behalf in a nominee account, so there aren’t any questions about ownership. And if one buys an ADR, the settlement will happen in say T+2 or whatever at one’s prime brokerage account, so there aren’t any questions whether it’s actually an ADR or something else because unless the two counterparties agree on all the details of the trade it won’t settle and there is an unmatched trade to be investigated.
    Prime broker will offer advice on the procedure, timetable, actions/forms needed etc and will liaise with client’s back office. But I don’t think an article on “arbitrage” should go into too much depth when it comes to back office procedures, I think the trading and markets bit ie the trade itself, is more relevant.

  2. Hi Tim. I know a lot about this.

    Some (most) DRs trade very tightly with the underlying. Others actually don’t. For example many Russian DRs listed in London vary by 1, 5, or even 20% from the underlying stock at point in time.

    There are varied reasons. A premium for holding a more liquid line. A premium for holding a line that is the one included in a benchmark. A premium for a perceived less risky trading venue (specifically with regards to counterparty risks and suspension/capital controls risk).

    In most stocks the DR premium/discount is effectively arbitraged out by funging the DR into the underlying or vice versa. But sometimes they can’t be. The classic reason is a DR program that has hit its allocated capacity. Or another curious feature of DRs is that it’s often a flat fee to funge, so on stocks with a low nominal value it can be prohibitive. You often have to bear some market risk during the funge as well (which takes a few days normally), if you can’t short the other side and with many of these stocks you can’t get the inventory or when the DRs are dislocated the cost of the borrow is prohibitive. Then sometimes there are also issues about market timing; what looks like an arb on screen can’t be done as one of the markets is shut. Then of course commission is generally low, but dealer bid/ask spreads can be surprisingly wide (pps, not bps in the less liquid stuff) and often the prices you see on screen are really only indications of interest and have no real volume behind them.

    So apparent arb opportunities are way more common than people generally realise. But those that are truly implementable are quite rare. But they do happen sometimes; best I saw actually done on a nearly pure arb was 9% for taking 4 days market risk

    I know you question was also partly about the same DRs trading in different venues. The obstacles are similar.

    Then when you have DRs and underlying that are persistently not anchored tightly (due to things like the full program I mentioned before) you can often take a view on the premium/discount and trade that, as it wanders around but has a loose anchor that tends to keep it in relative price channels. It’s not really arbitrage, you might call it stat arb if you were being charitable… Tends to move with risk and liquidity generally.

    If you want to know more, don’t use my email address entered on the form, it’s just a dummy.

  3. Yes, I would like to know more. Please email me (timworstallATgmailDOTcom) with an addy that I can get to and I’ll explain.

  4. As Oblong said, most arbs are not really implementable. I don’t even really remember any classic arbitrage trades in the 90’s and 00’s when I worked in hedge funds. Say buy something there and sell the same stuff here, convert and match up, pocket the difference. Spreads were much more common, even ords and prefs, long the prefs short the ords, or pairs trading, or bond yield spreads.

Leave a Reply

Your email address will not be published. Required fields are marked *