There is a fundamental misunderstanding of marketplaces and where liquidity comes from. And asset managers' fears of high-frequency trading arise primarily from this misunderstanding. Matt Hurd clears up some of the misconceptions about liquidity provision, the dangers of HFT, speed, and market efficiency.
Liquidity misconceptions in finance are some of the annoying itches I have wanted to scratch for a while. This is a meander that scratches that itch.
I won’t meander anything newsworthy for most market professionals here. If you’re a professional, I’d save your time. Nevertheless, this may be a reasonable reference to point to if you’re asked about liquidity.
The question of liquidity is very much a question of dance. The point of trading is to engage with another party so that two transactional needs are met. Utility is the goal. Utility is also the gaol [This is the the Australian/British "gaol" which is jail in American and often now jail in Australian too, though the original "gaol" is still used. I guess I could have used prison, but I liked the literary twist on the preceding use of "goal." Sorry for any confusion 😉] constraining a liquidity hunter's options. Just as it takes one to provide a cliche for another one to read, it takes two to tango. If your trading friends, or marketplace, do not supply you with liquidity, you have to leave them behind, “Cause your friends don’t dance and if they don’t dance, Well they’re no friends of mine.”
Marketplaces allow a more efficient risk transfer and price discovery process to take place. Without a marketplace, you are in danger of trading at an unfair price or being not able to trade at all. Despite all the warts of capitalism, markets provide the best way for parties to trade, despite the motivations of the participants.
Generally, none of the participants is trying to help any other participant. They trade for their own interests. The magic of the market is that they work despite those motivations. The misunderstanding of this simple fact is the cause of much strangeness and inefficient enterprise in modern markets.
Asset manager versus HFT
Asset managers are wary of HFT firms, particularly private trading firms (PTFs). An asset manager views with suspicion the utility an HFT gains from trading with them. Very few losing days in the HFT bank account seems unnatural. An asset manager feels he is being ripped and feeding this continuous profit the HFT enjoys.
One way for an asset manager to avoid an HFT is to find a marketplace without an HFT. One minor trouble is such venues are difficult to find. Even at IEX the most prolific traders are HFT. If you’re an asset manager trading at IEX it is more likely than not that your counterparty is an HFT. HFTs are prepared to risk more and earn less at the best price, which is why they win such trades. Buying low and selling high is still the only way to make money. An HFT is more prepared, in many ways, to roll the trading die. The law of large numbers gives them a good chance of making positive cash flow every day even if their success rate on your individual trade with them is only 51-55%. The asset manager doesn’t realise there is a fairly high chance the HFT may lose on their individual trade.
But, but, I want to earn that spread …
OK, go build your own marketplace then and see how well that works for you.
A bunch of asset managers get together and decide that they’d be better off without proprietary trading, especially those “nasty” HFT firms.
From Reuters 5th October 2015 story, “Luminex ‘dark pool’ enlists 73 members ahead of trading launch”:
Luminex plans to only allow institutional investors, such as mutual fund companies, to trade on its platform, snubbing the broker dealers and proprietary high-frequency trading firms that own and generate much of the trading in other private trading venues.
Luminex’s 73 subscribers, which are listed on the company’s website, include Vanguard, TIAA-CREF Investment Management, AllianceBernstein, Eaton Vance, Goldman Sachs Asset Management, Greenlight Capital Inc, and J.P. Morgan Investment Management. (luminextrading.com/)
The consortium that owns Luminex and collectively manages 40 percent of U.S. fund assets said it started the trading platform with the aim of lowering transaction costs and eliminating the types of profit driven conflicts of interest that have been seen in some existing venues. Any profits made by Luminex will be invested back into the company to further reduce trading costs.*
The Reuters article implies some motivation may have stemmed from illegal activity such as that from ITG:
Recent enforcement actions involving the trading platforms include a $20.3 million SEC settlement with brokerage firm Investment Technology Group on charges it ran a secret trading desk that profited off of confidential customer information within its dark pool.
Not an unreasonable motivation.
Luminex has recently touted its May 2018 results: “Luminex May 2018 Recap”:
Wow, that sounds impressive. The $10 billion traded in a month sounds like a big number, doesn’t it? An efficient exchange might make $0.0001 per trade. That is $20,000 per month at an average stock price of $50 for such a Luminex notional size. Not bad, but it is not easy to run the technology of an ATS, cope with the regulatory requirements, and pay your staff on such a budget.
The National Market System (NMS) exchanges traded 6,875 billion dollars of notional value for the same period. Luminex traded the equivalent of just 0.15% of the exchanges, which is only a partial picture of the liquidity landscape.
Let’s have a look at their tier one stock performance compared to the other 30 odd somewhat ATS peers:
That doesn’t look so great. The percentage market share of the ATS space is just a paltry 0.45%, with a ranking, by volume, of 26th. It’s hard to tell if that record-breaking memo is something to shout about or a job preservation exercise.
Here is a chart of the changing market share landscape in ATS land:
That’s right. Luminex is somewhere down there in the noise floor in the bottom of that chart.
What has gone wrong at Luminex? Nothing really. This is my point. They wanted to create a mono-culture client trading pool and succeeded. It is an understandable reaction to the fear that mongers peoples’ minds against HFT and proprietary trading. However, such an enterprise is going to find success a difficulty. Success often requires heterogeneity. Dare I say it? It takes two to tango:
Pearl Baily – Takes two to tango
Luminex looks to have done a good job with admirable enterprise. However, it represents a fundamental misunderstanding of marketplaces and where liquidity comes from.
Asset manager fears
There are several notions of wrong. First, anti-HFT is a misunderstanding of who your competitors are. Your competitors are other asset managers. An asset manager should be worrying more about selling a stock to another asset manager. They both think their opposing actions are a sensible medium- or long-term solution. Both can’t be right. If you sell to an HFT, both can be right and make money, as there is a time arbitrage there.
Not only does such a time difference support an asset manager’s and HFT’s difference in utility, there is a more fundamental aspect to the utility of every trade. Let’s meander through this thought.
Take your passive buy at best that gets traded through. If you’re an asset manager you’ll grumble about being traded through. You’ll imagine it was one of those nasty HFT firms and be pissed off that you could have bought the stock a tick cheaper. However, you’ll be happy as you wanted the trade and shrug your shoulders as, at an average price of $50 today, a tick of one cent is just 0.02%. If you’re in the asset manager game to only make 0.02% on your medium- or long-term trades please send the investors their money back; you’re in the wrong game. Annoying but not crippling.
If you’re an HFT and you get traded through, that is an existential crisis. Your margins are so thin that if such adverse selection happens too often you’re simply out of business. You have to take the risk and take such trades on the chin to play at all, but this is a careful calibration. The law of large numbers that guarantees your daily profit also guarantees the immediacy of your death if you get such calibrations slightly wrong. It’s a tough game for an HFT.
Such a “traded through” case is quite different for both types of market participants. It is an existential threat to an HFT, and an annoying mosquito bite to an asset manager.
The other case of the trade not happening at all is also quite different for both. For an HFT if its bid does not get lifted, it’s simply a shrug and they move on. For an asset manager it is a huge pain as they want the stock and now they are chasing it higher and not getting their portfolio at the right price. An HFT can feel a bit of that pain if they are long inventory, but it is quite different to the pain the asset manager endures.
Again, we see different utility for the outcomes. Your trade not filled is a shrug for an HFT and quite the pain for the asset manager.