HFT exchange BATS ex CEO resigned but may consult but IEX and Flash Boy Brad Katsuyama not invited to the Washington regulation party
Some highlights from this New York Post article:
Garrett, chairman of the House Financial Services subcommittee on capital markets, invited execs from 18 firms that touch on every aspect of high-frequency trading, from brokers Citadel and Virtu to exchanges like Nasdaq and BATS.
Bill O’Brien, the ex-president of BATS, left his job last week after Schneiderman’s office and the SEC pressured him to step down following a televised fight with Katsuyama, in which O’Brien mistakenly claimed there was no difference in the speeds of BATS data feeds, according to two people told of his departure. Neither BATS nor O’Brien would comment.
Garrett’s panel won’t touch on the main thesis of Lewis’ book — that some traders wield ultra-fast connections essentially as surveillance on the market, bidding up prices before others can get to it. Instead, it will focus on fees, pricing feeds and other trading technology.
Fix8 – A high performance alternative to quickfix and commercial FIX offerings
Tuesday, Apr 22, 2014, 7:30 PM
LiveStreram Online meeting Where ever you can get online Toronto, ON
9 Researching Traders Went
Fix8 – A high performance alternative to quickfix and commercial FIX offerings, for Linux, Windows and OSX.- what is Fix8? brief history of the project – why do I need it? do I need it? – is Fix8 a turnkey solution? a framework? a library? a toolkit? – what do I need to build a FIX adapter using Fix8 – Windows – how? – support and help? who can I …
Bats: The Epic Fail of the Worst IPO Ever businessweek.com
From Bloomberg Businessweek1 day agoLike CommentFollow Flag MoreMarty Kulas likes this1 comment Follow AshishAshish Mudgil • The article seems to incriminate all automated trading systems/ HFT in general.. I thought that is extrapolating the potential impact/ causes too much when investigations are still pending…
The sequence of events for BATS stock on Mar 23 as we know in public domain is as follows –
1. 10.45.26AM ET – last trade @ 15.25 @ BZX exchange 2. 10.48.41 AM ET – BZX report problems with symbol range A – BF (that includes AAPL stock as well) 3. 11.07.25 AM ET – BYX declares self-help against BZX (stops routing to BZX) 4. 11.28.22 AM ET – NASDAQ say it will cancel all trades that were executed on its trading platform for BATS stock between 11.14.00 – 11.15.00 AM ET 5. 11.5326 AM ET – BZX say “all issues” at Z exchange were resolved 6. 11.55.41 AM ET – BYX revokes self-help against BZX and starts routing orders to BZX 7. 12.21.45 PM ET – BZX say it will cancel all open orders in symbol range A-BFZZZ @12.40 PM ET 8. 12.50 PM ET – BZX resumes trading for all symbols (but BATS stock) in range A-BFZZZ 9. 1.03 PM ET – BZX says that it will further delay quoting and auction for symbol BATS and will advise further updates later.
We need to remember few things –
– If there really was any colluding between HFTs and BATS, then why would there be a crash at all at BATS? If there was a collusion, the whole scenario should have been carefully carefully and played our perfectly per some type of (hypothetical) common understanding…. The fact that the stock saw problems during IPO opening auction, should give more credence to the fact that BATS themselves were not involved in any ‘gaming’…
– BATS auction was supposed to end on March 28. So no money or shares would have exchanged hands when the problem occurred on day-1, that too even before the BATS stock could move to continuous trading session. BATS later declared it will cancel all executed trades that took place during the opening auction.
– The actual nature of reasons that caused problems with BATS IPO is yet not known. So it will not be correct to incriminate HFT and directly compare it with May 6, 2010 flash-crash just yet. It may very well be the case that BATS’ algorithm for handling execution priority for various order types during opening auction did not work as expected. This was also the first time that LLOO order type or ‘volume based tie breaker’ approach was in action for a corporate IPO in BZX.
– SEC have said they will examine if there was indeed any direct relationship of HFTs causing the crash of BATS stock price during opening auction. No wrong-doing has yet been concluded.
I think its only fair to wait for the findings to emerge before drawing conclusions
High Frequency Trading HFT certainly trashed BATS share price pretty quickly , but ..
HFT certainly trashed BATS share price pretty quickly: but there really was a problem with BATS’ trading platform as well. We can even make the argument that HFT just made everyone realise that problem faster: and yes, we do want markets to reveal such information as fast as they can
Charts of the day, flash crash edition
What you’re seeing here is the price of shares in BATS, at 11:14 this morning. The white spots are trades: there are 176 of them altogether. They start just below the IPO price of $16, and then just fall lower and lower and
High Frequency Trading HFT Flash crash at BATS renews market concerns
The crash caused BATS to halt its IPO on what should have been the proudest day in the exchange’s history.
Time is money… No really… Time is money!
This Infrastructure error type event – the Flash Crash is the reason external timestamping is key to market integrity. Timestamping from uniform time (NIST UTC) , and time data sourced at the picosecond resolution level which can be applied to the outside of all exchange practices and operations is the answer. This is simply tied to the mindset that you cannot take it apart unless you can track it in real-time.
But there is more – because the timestamps themselves can be bonded this can refocus the liability model outside the matching/trading or exchange engine operations. In fact with proper time-request practices individual exchange events could be carried in time-request and time-response instances from the reference clock used. Flash Crash instances can be undone easily as MARKET INFRASTRUCTURE ERRORS and no-harm, no-foul declared…
That means bonding is also easily applied in amounts ranging from millions to even billions of dollars which the HFT and Hedge operators need for their systems. And all of it simply because of ultra-precision timestamping – timestamping at market-wire speed!
The power of that final statement speaks for itself – the reference clock not only serves for calibration but event authentication from a forensic perspective.
I should point out – I am NOT talking about using NTP for just timesetting – I am talking here about using it as a time-stamping service which runs at wire-speed. Most people would tell you this is impossible – for us it seems the obvious solution since today the issue is how events are logged across very-tightly coupled infrastructures.
This solution allows for actual use of NTP as the time-stamp feature for end-use events allowing individual trade or consummation events to be actually time-stamped through the time source reference clock itself…