This page has been archived and commenting is disabled.
From The Horse's Mouth: GSET's Sofianos On Program Trading And Intraday Liquidity
Goldman Sachs Electronic Trading's own George Sofianos (ironically caught here discussing execution shortfall costs... but wait, wasn't HFT costless) providing his objective view on PT and volatility from the days when he was merely a member of the largely unaffiliated New York Stock Exchange.
Soon to be covered - "Sonar"
Harris__Sofianos_And_Shapiro-Program_Trading_And_Intraday_Volatility -
- 5756 reads
- Printer-friendly version
- Send to friend
- advertisements -


What happened at 85 Broad after 4:30pm today?
if anything happened, you will be first to know
The top 50 whales have their liquidity with UBS.
The top 50 whales have their liquidity with UBS.
Going back 15 years to a paper based on data that is more than 19 years stale.
2 take aways:
1) Program trading is nothing new.
2) "Since little or no average price reversal occurs in the 30 minutes after program trades, program trades do not seem to cause excess volatility."
Given that this paper was written in an era where there were no ETFs to arbitrage with the futures markets, I'm not sure that there is going to be much relevance to todays market place.
you read the whole paper in 24 minutes? i am impressed.
lol
Ok... so what's the problem? I can only assume you've posted the program trading piece because you think its an example of possible manipulation, but I just don't see it. As for the more recent limit orders piece, I read that when it was first published, and still cannot believe what a pile of crap it is. That "research" is riddled with biases. Sofianos should know better.
I was about to basically post this...
Very non-telling study as far as I can tell (from a quick scan). Unimportant conclusions, impractical linear regressions, stale data, no large crashes or bubbles in the sample data, etc...
Is it fair to assume that market forces' dynamics are now greater than they were 20 years ago when the study was done?
Yes - but on both sides of the trade.
He's looking at the movement of the cash market when program traders come in to arbitrage the futures market with a basket of stocks in the cash market.... from an era where bid/ask spreads were larger, commissions were higher, and you couldn't readily short at will.
Once ETFs came into existence (right around the time this paper was written, but certainly not widely used until much later), it became much easier to execute the overall arbitrage, bringing the futures market much more in line with the cash market.
Now... I would love to see a paper on recent data doing the same type of analysis that Sofiano did, showing the inter-play between the futures, ETF and cash markets, and understanding how long it takes for things to mean revert.
The closest I have come to finding one is the following 2002 MS Thesis from Ron Tarantino at NYU:
http://w4.stern.nyu.edu/emplibrary/Ron_Tarantino_honors_2004.pdf
I would love any links to papers on the topic that are more recent.
OT- is marla workin' it soon? down robots!
as soon as the market starts to slide south again program trading will get the blame it deserves. until then, joe sixpack loves pt. nothing to see until the masses are upset, then the show will begin and the usual suspects will be rounded up and hung in the public square.
Irrational behavior is the surest way to defeat colocated I7's running algorithms.
Yes,
If you are willing to lose a fortune in the process, you can make machines go over the cliff with you.
this shit is just noise, nothing else
thank you for providing the paper TD
I guess you can call it? Seriously, rationalize please.
not the article, the paper; it is irrelevant in every single point and it makes to much " noise" and i don't see any real value in it to help me with my investments ( which i currently have 0 ) or to further clarify the state of the economy ( not the fucking market) ... it is more of a academic research filled with theoretical presuppositions and invalid conclusions .... and as one poster above said; it is also completely biased ...
Is it Lobbyist?
Clue: build a huge net neutral position, then let one side go and vola, TRAP profits will rain down on GS, JPM, ect.
Morning conf call coordinates the teritory to reduce collateral damage.
This study was conducted when trades on the NYSE were done manually , shorts were tick restricted , the minimum tick was 1/8th , the bulk of program order flow was arbitrage against the futures and most importantly, specialists and brokers made deep markets. Contrast that to what we have today and I think we know what the conclusion is. Good Night.
Email | Print | Share | Reprints | Single Page [-] Text [+]
1 of 1Full Size
MARKET NEWS
July jobs data lifts stock indexes | Video
Oil slips as U.S. jobs data boosts dollar
Stocks' rally hinges on consumers and Fed
More Business & Investing News...
Featured Broker sponsored link
By David Lawder
WASHINGTON (Reuters) - U.S. Treasury Secretary Timothy Geithner formally requested that Congress raise the $12.1 trillion statutory debt limit on Friday, saying that it could be breached as early as mid-October.
"It is critically important that Congress act before the limit is reached so that citizens and investors here and around the world can remain confident that the United States will always meet its obligations," Geithner said in a letter to Senate Majority Leader Harry Reid that was obtained by Reuters.
A Treasury spokeswoman declined to comment on the letter.
Treasury officials earlier this week said that the debt limit, last raised in February when the $787 billion economic stimulus legislation was passed, would be hit sometime in the October-December quarter. Geithner's letter said the breach could be two weeks into that period, just as the 2010 fiscal year is getting underway.
The latest request comes as the Treasury is ramping up borrowing to unprecedented levels to fund stimulus and financial bailout programs and cope with a deep recession that has devastated tax revenues.
It is expected to issue net new debt of as much as $2 trillion in the 2009 fiscal year ended September 30 and up to $1.6 trillion in the 2010 fiscal year, according to bond dealer forecasts.
The request to increase the debt limit will likely raise the ire of Republicans who have accused President Barack Obama of runaway spending. They may try to hold up the legislation in effort to win concessions on Obama's health care reform plan.
Geithner urged Reid to not let politics hamper U.S. credit-worthiness and said he looked forward to working with the Nevada Democrat to secure enactment of legislation on the debt limit as early as possible.
"Congress has never failed to raise the debt limit when necessary. Because members of both parties have long recognized the need to keep politics away from this issue, these actions have traditionally received bipartisan support," he wrote. "This is clearly a moment in our history that calls for continuation of that tradition."
Dear Timmay,
I'll give you an increase on the federal credit limit if you give me an audit of the federal reserve.
XOXO,
Taxpayer
Timmay says " Well, just to be safe and to keep me from coming back real soon, I think about $300 trillion more on the debt ought to do it. OK, see you in 6 months for the next debt raising session.."
They are gonna hang Geithner.
and a modern gaming comp with quad sli running custom trading code is gotta be faster than the fastest commercial supercomputer in '89
nevermind the packetsniffing, flash orders and colocation.
its pretty darn dated.
This is not limited to stocks. I was a Prime options market maker in MBS in 1987 on October 19th. Mortgage bankers were going insane trying to hedge. Back then we actually performed a function--hedging for mortgage bankers and mutual fund call writing for the most part. Split-fee MBS options blew up during this time--shutting down more than a few desks--and we were a select bunch that traded these puppies. Our back office manager suffered a nervous breakdown after the crash. If you understood the delivery mechanisms in the 80's of MBS you would understand why.
You think program trading wasn't an issue? Telerate only had room for a 2 digit move in the Dow. When Black Friday was down 105, it looked grim. Bloomberg wasn't even a competitor then--Telerate owned the bond market real-time data. Monday was outrageous down 400, up 250 from there, and down again--I could charge over 100% vol in a binomial model during the crash as bonds were going insane. Program trading, HFT--it's all the same beast. It's evolution. The past is always a great lesson for the future. Because history, inevitably, repeats itself.
Nice perspectives...
Back in those days we were told of the wonders of dynamic hedging. Today the spin is "microstructure ".
Program trading is just like everything else. It works 'till it don't. You can just ask John Meriwhether, or Myron Scholes about that one. It's only a matter of time before GS gets a LTCM pulled on it. I'm pretty sure Meriwhether and Myron moved on and imploded two other hedgefunds.
"Millions saw the apple fall, but Newton was the one who asked why."
-Bernard Baruch
Not to forget that Isaac Newton was also Master of the Mint in 1699 and changed the British currency forever. He was a gold bug.
> It works 'till it don't. You can just ask John Meriwhether, or
> Myron Scholes about that one. It's only a matter of time
> before GS gets a LTCM pulled on it.
I doubt it. They are runnning too many scams simultaneously that they'll survive easily if one or two of them blow up in their face. The only thing that could have killed them was the banking crisis of '07-'09, and they got Bernanke to bail them out.
All mortal things must die.
Unfortunately for us, Goldman is a vampire with a permanent I.V. tube hooked up from the Federal Reserve.
15 years ago chips were 32 bit and there was no fiber.
Folks, the government is borrowing trillions to give to banks to buy stocks. that is the essential nature of the run up. the liquidity programs can game the system working on the volume weit=ghted average by buying a lot, bringing un buyers, then dumping bringing down. all end up around the average. has to do with the patterns of how the buying and selling is done around the average. manipulations seen in the pumping and dumping.
Geitner needs debt levesl raised for more market manipulation
Check out the caption on Geithner asking congress to increse the federal debt limit http://www.numisex.com/blog/1/2009/8/geithner-asks-congress-increase-fed...
The paper seems to prove that money can be made by arbing the bid-ask spread without causing undue volatility (duh).
Proof of bid-ask spread manipulation would be the order of the day....hft and all that:)
40muleteam borax
This comment doesn't even make sense. Tyler D, I think you need a more effective way of screening for intelligence.
The central banks have decided to only manipulate the market with 400 billion tons of gold per year instead of the previous 500 billion tons of gold per year. A 20 percent reduction in manipulation quota. I guess to celebrate their 72nd bank failure which happened today.
Who needs HFT to manipulate the market when you can just dial up the Treasury Secretary and get him to do your bidding?
http://money.cnn.com/news/newsfeeds/articles/reuters/MTFH71375_2009-08-0...