king-logo

>

For the past several years Street operators have assumed that the computer jockeys who were being employed by proprietary trading departments on The Street were developing algorithms that would find other algorithms that represented buyside orders so prop desks could trade against those orders.

Another trading prop that has been occurring for years is certain firms feed their electronic trading systems into prop desks so traders can see in real time money flows into and out of stocks and groups.

However recent revelations are forcing the Street to consider the possibility of automated front-running on an unfathomable scale. The two ‘front-running’ issues are: 1) ‘queuing’ [of orders] – finding orders loaded into a system, particularly limit orders, and trading against them; and 2) ‘latency’ – discovering and then front-running electronic orders or a penny or more by exploding the latency or lag in execution.

HFT (high frequency trading) is being done on every electronically traded item on a global basis. Ergo, firms could be making pennies a few billions times per day…It was imperative for the NYSE and other exchanges to price securities in pennies to disguise ‘HFT’ & to provide ample trading opportunities.

While the Street is percolating with anger and curiosity about ‘High Frequency Trading’ there is also frustration and astonishment that the media, regulators and our duly elected are not addressing what could be the biggest financial abuse story of our times, if not history.

Though the blagoshpere is all over the ‘HFT’ trading story an important piece of the puzzle has not been publicized enough. Few people realize that exchanges actually pay firms to trade against order flow when they act as a SLP – ‘Supplementary Liquidity Provider’.

Exchanges will pay firms ¼ of a penny if they ‘provide liquidity’ when an order appears in their system. This is extra incentive to front run order flow… Theoretically a firm could ‘scratch’ all day and profit. The NYSE Euronext (Oct.24, 2008): A newly announced pilot program will establish Supplemental Liquidity Providers (SLPs), a new class of upstairs, electronic, high-volume members incented to add liquidity on the NYSE.

The program will reward aggressive liquidity suppliers, who will complement and add competition to existing quote providers.

o SLPs will be obligated to maintain a bid or offer at the National Best Bid or Offer (NBBO) in each assigned security at least 5 percent of the trading day.
o The NYSE will pay a financial rebate to the SLP when the SLP posts liquidity in an assigned security that executes against incoming orders. The goal is to generate more quoting activity,
leading to tighter spreads and greater liquidity at each price level.
o SLPs will trade only for their proprietary accounts, not for public customers or on an agency
basis….

Zero Hedge (May 5, 2009): Previously Zero Hedge observed the rather curious integration of Goldman Sachs within the fabric of the NYSE’s program trading environment, which, by their own admission, has everything to do with Goldman being the (monopoly) actor in the NYSE’s Supplemental Liquidity Provider program. I highlighted that the program was set to expire on April 30.

Today, unsurprisingly, the NYSE posted a notice of a proposed rule change extending the SLP program another six months, until October 1, 2009 (this does not change my commitment to providing weekly NYSE program data). I appreciate our readers’ existing and future feedback in this matter…

The full text of the comments by Jeffrey S. Davis of the NASDAQ Stock Market LLC is presented below, but here are some very relevant snippets:…

“NYSE fails to explain why proprietary liquidity is more valuable than agency liquidity, or why proprietary liquidity should be favored over agency liquidity. NYSE claims that the proposal is designed to prompt liquidity provision but it simultaneously disqualifies large liquidity
providers…

In NASDAQ’s view, these irregularities reveal that NYSE’s true motivation for the SLP Proposals is to discriminate among its members and to burden some members’ ability to compete with NYSE…”

And who is the one and only beneficiary of this rampant disregard for almost 80 years of market regulatory practice? Who is it that has now become the de facto provider of “market liquidity” which however has much more sinister connotations when reading through the comments of not just some blogger but the NASDAQ Stock Market itself?

When a firm has an 87.5% trading accuracy record, something unnatural is occurring. And we wonder why the buy side has been so docile and malleable when the money is being derived from them!

Over the past decade the move to electronic trading and pricing in pennies was heralded by Street insiders as a means to improve liquidity for clients. This appears to be a deception. Virtually every facility benefitted proprietary trading at a select few firms. Who’s the patsy?

Anyone with a modicum of industry experience understands that ‘providing liquidity’ is at best a euphemism for front-running order flow. Anyone that regularly ‘provides liquidity’ will go broke.

GS jumped yesterday on this: (BN) Goldman Sachs Trading Revenue May Beat 2007 Record. Goldman Sachs Group Inc. is on track to beat its 2007 trading-revenue record, enabling it to boost compensation by an estimated 64 percent from last year, according to Bank of America Corp. analyst Guy Moszkowski. Goldman Sachs has “unmatched risk-taking/risk-management skills in a market that strongly rewards these because of decline in competitor risk appetite,”…Six months ago, Goldman
Sachs was supported by $10 billion from the U.S. Treasury and relied on government guarantees to issue debt. Moszkowski predicts the company will reap $26.45 billion from trading this year, a gain from
$25.36 billion in 2007 when the firm shattered Wall Street profit records.

Remember this Bloomberg story from May? Goldman Sachs’s $100 Million Trading Days Hit Record Goldman Sachs Group Inc. reaped more than $100 million in trading revenue on a record 34 separate days during the first three months of 2009, up from the previous peak of 28 in last year’s first quarter… The first-quarter number was almost double the total for all of 2005…

Zero Hedge: Citadel Joins The Program Trading Industrial Espionage Fray, Sues Malyshev And Teza The gloves are now completely off in the escalating program trading fiasco that was started by Goldman’s former Sergey Aleynikov. Oddly, while Zero Hedge was fully expecting the Teza injunction to come from Goldman, it seems Griffin was more than happy to burden himself with that task. Hopefully Citadel is not faced with a case of reverse discovery and forced to document the 40% returns that it generated compliments of Malyshev when all its other groups on average lost around 50% in 2008.

Bloomberg: Citadel Investment Group LLC, the $12 billion hedge fund firm founded by Ken Griffin, sued three former executives and the firm they founded, Teza Technologies LLC, claiming violation of non-competition agreements…

Teza described itself in a July 6 e-mail as a “formative” firm that is neither trading nor investing. Named after a river in western Russia, the Chicago-based firm was co-founded by Misha Malyshev, Jace
Kohlmeier and Matt Hinerfeld. All were named in the complaint…

Malyshev worked at Citadel for almost six years and until February was its head of high-frequency trading… He was on the team that ran a $1.8 billion tactical trading fund that uses computer model to make trades every few seconds. The fund climbed 40 percent last year, while its main funds tumbled 55 percent.

Category: Markets, Think Tank

Please use the comments to demonstrate your own ignorance, unfamiliarity with empirical data and lack of respect for scientific knowledge. Be sure to create straw men and argue against things I have neither said nor implied. If you could repeat previously discredited memes or steer the conversation into irrelevant, off topic discussions, it would be appreciated. Lastly, kindly forgo all civility in your discourse . . . you are, after all, anonymous.

12 Responses to “King Report: HFT”

  1. bergsten says:

    Whose pocket are these “quarter penny rebates” coming out of? (sorry, gang, I just don’t believe money comes out of thin air.)

    I read somewhere that “market reporting” revenue (e.g. Yahoo Finance) subsidizes these rebates (though charging by the LENGTH of the tape seems peculiar). If so, that’s a whole lot of charge.

    Wherever it’s coming from, it yet another “tax” on the system, in this case, by “system” I mean “us.”

  2. bergsten says:

    And another thing. Market makers (allegedly) provide liquiditywith a pool of real, actual stock.

    This SLP stuff seems like a subtle form of “naked”, well not shorts exactly, but some sort of virtual (nice word for nonexistent) equities.

    Or am I completely off base here?

  3. gregh says:

    Some trader advice discusses personal ownership of bad trades, accepting your loss as either a bad trade, or a side-effect of your system…something to learn from etc. But this makes that kinda hard to swallow – was my trade that bad or was some algorithm calculating how easily it can droop down to my stop limit?

  4. Bruce in Tn says:

    What if Goldman weren’t around? If they were Lehmanned?

    Would there be more trading and fewer bankers in government? Would there be a fall in the cynicism titer? Seems like years ago we broke up AT&T into Ma Bell and a bunch of regional phone companies…anyone remember the term “trust busting”?

    …just a thought…

  5. KidDynamite says:

    Come on King, you have many of your facts wrong again.

    “Over the past decade the move to electronic trading and pricing in pennies was heralded by Street insiders as a means to improve liquidity for clients. This appears to be a deception. Virtually every facility benefitted proprietary trading at a select few firms. Who’s the patsy?”

    I don’t know how anyone can argue that the costs of trading (including spread) have not been DRASTICALLY reduced.

    “Few people realize that exchanges actually pay firms to trade against order flow when they act as a SLP – ‘Supplementary Liquidity Provider’.”

    completely wrong. the ENTIRE point is that if you trade against order flow, ie, REMOVE liquidity, you pay, you don’t get paid a rebate. You get a rebate when order flow trades against you. It’s not just semantics, it’s entirely different.

    “Anyone with a modicum of industry experience understands that ‘providing liquidity’ is at best a euphemism for front-running order flow. Anyone that regularly ‘provides liquidity’ will go broke.”

    this couldn’t be more wrong. I suggest Mr. King sits down with someone who has ever worked on a sell side program trading desk for 15 minutes and clears up his vast misconceptions.

    ““NYSE fails to explain why proprietary liquidity is more valuable than agency liquidity, or why proprietary liquidity should be favored over agency liquidity.”

    it’s not – all liquidity is equally valuable. The SLP program requires the participants to provide a certain amount of liquidity. It’s almost impossible to come up with a conclusion that this is harmful.

  6. Simon says:

    If we don’t get too bogged down in details and just look at trends and the fundamentle nature of things it’s hard not to reach the conclusion that short term trading is a betting contest. Any where you look where there is large scale betting insiders become insiders because they are good at it and have an advantage maybe that is only knowledege but could easily be more. As the games evolves they will try to enlarge that advantage in any way they can. It would surprise me greatly if the big players in the trading betting game did not have a material advantage reasonably well concealed from the majority.

    It would also surprise me greatly if they had any advantage at all on the longer time horizons better associated with investing. Lesson avoid the short term betting game and stick to longer term investing.

  7. Moss says:

    Sounds a lot like card counting except with better odds.

  8. bergsten says:

    @StatArber
    I frankly don’t know the process intimately enough to provide specific directions for frontrunning (and wouldn’t do so in public if I DID).

    Having said that, THEORETICALLY, any entity (authorized or otherwise) “between” your order and the book can CONCIEVABLY insert their order(s) ahead of (and/or delay) yours. I’d expect this behavior more of those handling retail orders (as there are more chances due to the number of intermediaries) — in passing, I’ve always been a bit uneasy about the concept of orders filled “in house” first.

    Otherwise, it’s either some sort of race (the time between intermediaries is small but non-zero, so perhaps someone watching/communicating at picosecond speeds can get in first), or diddling/rules about order priority (there ARE some rules here, e.g. filling small odd lots to keep these from being frozen out, etc.). I don’t KNOW all of the rules but like most rules, there have to be some favoring the big guys.

    Or, this business of “instantaneous” order placement/cancellation to statistically pick off orders in process.

    Absent all of that, there is ANTICIPATORY “front-running” — watching stock behavior to predict and order based what’s likely to happen next. I think that’s called “trading.”

    Has it occurred to anybody that the furor over the GS code theft is that they’re EMBARASSED by the code?

  9. bergsten says:

    Speaking of “front-running” I managed to “reply to your post before you sent it”!! (look at the order of posts).

    Clever, custom programmed, co-located, fast-networked on-site servers at Barry’s place (rackspace.com if you’re curious)?

    Nope. A “bug” — I’m on PDT, you are (apparently) on EDT. My timestamps are three hours earlier than yours — WordPress is sorting by local time instead of “universal” time.

    So, here’s another chance at frontrunning — who sets an order’s timestamp? If it’s the originator, mayhem is certainly possible!

  10. StatArber says:

    I have a quick question about this stuff. How exactly is this front running stuff supposed to work?

    My understanding of how all this works:

    Market Makers (“liquidity provider”) send limit orders to the market (quickly, canceling, replacing, whatever they want, very quickly). These orders are routed to the exchange’s computers and form the order book.

    ——
    Active orders:

    Other traders send orders to the market. Again, these orders are routed to the exchange’s computers (i.e. these orders are only seen by other market participants after being filled or placed in the order book). If the order is active, some market participant (probably a market maker, from above) collects the spread (and a liquitidy rebate) and takes the other side and the trader gets his fill (and pays a liquidity demand fee to the exchange — this fee is higher than the rebate for providing liquidity).

    At no point do market makers “see” active orders before they are filled. If this is the case, then how is it possible to pursue

    “2) ‘latency’ – discovering and then front-running electronic orders or a penny or more by exploding the latency or lag in execution.”

    If what I describe above is not true, what is the error I have made?

    —-
    Passive orders:

    Other traders place passive limit orders in the order book alongside other market makers. Once the order is placed and is in the order book, other market participants (including market makers/high frequency traders) can adjust their quotes accordingly. Clearly, the new limit order entering the market reflects demand for shares. Should large numbers of limit orders be placed on one side of the order book, it is likely that the price of the security will increase/decrease (supply and demand at work). High frequency traders/market makers, by their very nature, are the first to react and therefore are most likely to profit.

    Is this what the post refers to when speaking of

    1) ‘queuing’ [of orders] – finding orders loaded into a system, particularly limit orders, and trading against them;

    ??

    Again, this seems to me to be the only possible explanation. If this is not the case, where have I err’d?

    I don’t see how this should be an issue for the marketplace. There has always been someone reacting first to changes in order book dynamics. In the old days, this was the floor specialist when traders had to phone in orders and got delayed quotes and fills. It should not be news that people trading with more latency will have less information about the liquidity dynamics of the market.

    —–

    So either I have made mistakes in my description of how the market works (very possible, as I really only have a passing knowledge of these issues), or maybe there is some subtle issue that I have missed.

    Thanks.

    –Stat

  11. bergsten says:

    Yo, Stat. Like how I answered your comment BEFORE YOU EVEN POSTED IT?
    How’s THAT for front-running?

    p.s. Speaking of front-running, the Themis Trading paper I posted a link to here (which I GOT from here days before) is the subject of a John Mauldin newsletter article, and boy is he pissed about it! This pseudo-journalism thing is just so cool… http://www.frontlinethoughts.com/gateway.asp

  12. bergsten says:

    @Stat Sigh. I was so busy gloating, I forgot to mention the obvious — my “post before your post” is ABOVE your post in this thread — you have to scroll up above yours to see it (them actually).