The Last Look…
Posted by Colin Lambert. Last updated: February 22, 2021
Sometimes genius comes from the most unexpected direction. The sharp eyed amongst you will note this column has a new name after one week (and no, it does not signal it will become a one subject feature). It has changed because I really wanted Any Other Business for a section that will be exclusive to our newsletter every week, rounding up some of the less prominent news items from the week and bits and pieces that got my attention.
I was pondering a name out loud and bemoaning the loss of my beloved And Finally… when my wife, who knows nothing about FX or the intricacies of the industry said, “why don’t you call it The Last Look?”
So, with a bonus for Mrs L and a smack on the forehead to me for not thinking of such an obvious title (which I hope you agree works on many levels!) let’s get to it. I am somewhat mystified by the arguments of those who think internalisation is a threat to the well-being of the FX market.
Obviously, this leans into the transparency debate, but if, as so many people are happy to state ad nauseum, we really do care about the end customer – the genuine hedgers – how can it be bad?
If I were in a corporate treasury that cared about these things, or on an execution desk at an asset manager, my best execution policy would definitely look at the high internalisers and their relative lack of market impact. Of course, the LPs would have to be genuine internalisers, not a firm that takes the order and skews everywhere to “internalise” the flow – equally, the LP should not leak information into the market by skewing internal pools only but which are used by other professional counterparties. If we are talking genuine internalisation, this really looks after the major hedgers.
I wonder if the first trades on Siege FX last week prompted people to get involved because mechanisms like that, and FX HedgePool in the FX swaps market, are quasi dark pools, which in turn are, after all, a form of internalisation. It’s all about the flow not hitting the public or “lit” market. The two aforementioned mechanisms have been supported by buy side firms either disaffected by the shift of so many banks to an agency style model where the algo offers the most likely best execution (where it used to be risk transfer); or who are just keen to add another execution channel. Either way, these firms want to keep their business quiet.
A few weeks ago The Full FX discussed an FCA report in the UK which looked at dark pools and their benefits (and drawbacks). The report concluded that the lack of transparency offered by dark mechanisms could translate into less market impact. This is clearly recognised by the larger buy side firms (as well as algo execution providers) and doesn’t sit well with the anti-internalisation segment.
Why is that exactly? At the risk (perish the thought) of being a cynic, those making the most noise are probably those firms without customers and who rely upon price feeds from elsewhere to ‘pip’ top of book (using last look if they can I am sure) or need more volatility to make money speculating. They could also be traders whose model is very much predicated upon catching an early (small) trade and then jumping in front of the main order. Fair enough, they are allowed to do this, but why should they have a free ride? Good trading should be identifying market trends (preferably for more than 10 seconds), not identifying “good” flow.
Pricing wider is not really an option for many critics whose model involves being top of book in the smallest possible amount with the intention of picking up market intelligence whilst taking the least amount of risk
So, as you can tell, my sympathy is not with the critics of internalisation, especially because I suspect they sit in another of my least favourite FX baskets, recyclers. My sympathy remains with the end client, and the best way for these firms to execute a big ticket in these markets is where less of it goes public.
If the critics are worried about pricing, it’s not as though there isn’t enough market data out there. Non-last look venues like Matching, EBS Market, LMAX Exchange and Cboe FX’ ‘firm’ mechanism, offer sufficient data from diverse sources that can power a pricing engine if that what people are worried about. Flow doesn’t match off perfectly, there will always be exhaust to these public, and lit, markets.
Equally, there are plenty of firms who aggregate the leading LPs, who can make sure that the pricing is kept onside and if that isn’t enough then independent TCA firms exist to provide the appropriate analysis.
I understand that in the electronic world being off with one’s pricing can be expensive, but whose fault is it if the pricing is wrong? One reason we have more flash events in FX markets is because computers are not going to do one thing that humans do to help alleviate market stress – guess where the market is.
If an automated LP is concerned they don’t know where the market is then they can price wider, or, if they have an interest, place a bid or offer in the market. Of course, pricing wider is not really an option for a firm whose model means they have to be top of book in the smallest possible amount, with the intention of picking up market intelligence whilst taking the least amount of risk.
I would argue that genuine internalisers, who are the modern-day risk warehousers, are adding far more to the market ecosystem than those making all the noise about the process. Equity markets are coming around to the benefits of transparency of conduct rather than order flow, so why make this noise now in FX?
It’s not as though the critics of internalisation can’t continue to play in the markets, there are plenty of models to go around, it’s just they will be limited to certain venues, with no last look to protect them – and that means their trading skill will be what gets them through. I’ll leave it up to the readers to ponder where that leaves these firms…