chinman.com
Public Zone 公開區 => Bookwyrm 書蟲天地 => Topic started by: chin on 25 July 2009, 02:08:32
-
The term CRW was invented in the US horse racing market, describing people who use software and sophisticated algorithm to scan the pools and odds to find mispriced bets.
Less sophisticated gamblers thought CRW teams have access to odds in the normally "blind" pools like Trifecta & Superfecta, where the odds are not displayed. They often cry foul and accuse the CRW teams of "seeing other people's hand" before betting.
In reality, scan is the wrong word, because the software does not really "scan" because the pools are blind. Project and bet against the projections is more accurate description of CRW teams.
And then I found this article today. The big boys DO scan and cherry pick, perhaps with some projections to boot.
AND we are not talking about betting horses, but the largest gambling pool in the world - the stock markets.
-
Before quoting the New York Times article on the stock market equivalent of CRW, I would like to relay a personal experience.
In around April 09, we received a visitor G*** who is the head of a specialized proprietary trading unit in a very large & well known investment bank. My friend brought him to our office to see our program trading or CRW. In those few hours, I found out that:
- His unit trades the bank's own capital.
- His algorithm can project share prices a few hours ahead.
- His unit made "large number of small bets" in his own words.
- His unit issue 4 or 5 million orders per day, but only a few hundred thousand orders would be matched!
- His unit accounted for 4-5% of all trading in the Japanese market!
- They are in other Asian markets as well. Tradings are all electronic, of course.
- My friend told me G's unit is among the most profitable units in his bank.
The above sounds very much like a CRW to me.
Then I see this NYT story.
-
Compare the highlighted to my points above.
The strategy described in this NYT article is very similar to front running. The only difference is that Front Running is when a stock broker is cheating his own client, while this strategy is cheating(?!) everyone.
---
http://www.nytimes.com/2009/07/24/business/24trading.html?_r=1&ref=global-home
Stock Traders Find Speed Pays, in Milliseconds
By CHARLES DUHIGG
Published: July 23, 2009
It is the hot new thing on Wall Street, a way for a handful of traders to master the stock market, peek at investors’ orders and, critics say, even subtly manipulate share prices.
It is called high-frequency trading — and it is suddenly one of the most talked-about and mysterious forces in the markets.
Powerful computers, some housed right next to the machines that drive marketplaces like the New York Stock Exchange, enable high-frequency traders to transmit millions of orders at lightning speed and, their detractors contend, reap billions at everyone else’s expense.
These systems are so fast they can outsmart or outrun other investors, humans and computers alike. And after growing in the shadows for years, they are generating lots of talk.
(http://graphics8.nytimes.com/images/2009/07/24/business/0724-biz-web2TRADING.jpg)
Nearly everyone on Wall Street is wondering how hedge funds and large banks like Goldman Sachs are making so much money so soon after the financial system nearly collapsed. High-frequency trading is one answer.
And when a former Goldman Sachs programmer was accused this month of stealing secret computer codes — software that a federal prosecutor said could “manipulate markets in unfair ways” — it only added to the mystery. Goldman acknowledges that it profits from high-frequency trading, but disputes that it has an unfair advantage.
Yet high-frequency specialists clearly have an edge over typical traders, let alone ordinary investors. The Securities and Exchange Commission says it is examining certain aspects of the strategy.
“This is where all the money is getting made,” said William H. Donaldson, former chairman and chief executive of the New York Stock Exchange and today an adviser to a big hedge fund. “If an individual investor doesn’t have the means to keep up, they’re at a huge disadvantage.”
For most of Wall Street’s history, stock trading was fairly straightforward: buyers and sellers gathered on exchange floors and dickered until they struck a deal. Then, in 1998, the Securities and Exchange Commission authorized electronic exchanges to compete with marketplaces like the New York Stock Exchange. The intent was to open markets to anyone with a desktop computer and a fresh idea.
But as new marketplaces have emerged, PCs have been unable to compete with Wall Street’s computers. Powerful algorithms — “algos,” in industry parlance — execute millions of orders a second and scan dozens of public and private marketplaces simultaneously. They can spot trends before other investors can blink, changing orders and strategies within milliseconds.
High-frequency traders often confound other investors by issuing and then canceling orders almost simultaneously. Loopholes in market rules give high-speed investors an early glance at how others are trading. And their computers can essentially bully slower investors into giving up profits — and then disappear before anyone even knows theywere there.
High-frequency traders also benefit from competition among the various exchanges, which pay small fees that are often collected by the biggest and most active traders — typically a quarter of a cent per share to whoever arrives first. Those small payments, spread over millions of shares, help high-speed investors profit simply by trading enormous numbers of shares, even if they buy or sell at a modest loss.
“It’s become a technological arms race, and what separates winners and losers is how fast they can move,” said Joseph M. Mecane of NYSE Euronext, which operates the New York Stock Exchange. “Markets need liquidity, and high-frequency traders provide opportunities for other investors to buy and sell.”
The rise of high-frequency trading helps explain why activity on the nation’s stock exchanges has exploded. Average daily volume has soared by 164 percent since 2005, according to data from NYSE. Although precise figures are elusive, stock exchanges say that a handful of high-frequency traders now account for a more than half of all trades. To understand this high-speed world, consider what happened when slow-moving traders went up against high-frequency robots earlier this month, and ended up handing spoils to lightning-fast computers.
It was July 15, and Intel, the computer chip giant, had reporting robust earnings the night before. Some investors, smelling opportunity, set out to buy shares in the semiconductor company Broadcom. (Their activities were described by an investor at a major Wall Street firm who spoke on the condition of anonymity to protect his job.) The slower traders faced a quandary: If they sought to buy a large number of shares at once, they would tip their hand and risk driving up Broadcom’s price. So, as is often the case on Wall Street, they divided their orders into dozens of small batches, hoping to cover their tracks. One second after the market opened, shares of Broadcom started changing hands at $26.20.
The slower traders began issuing buy orders. But rather than being shown to all potential sellers at the same time, some of those orders were most likely routed to a collection of high-frequency traders for just 30 milliseconds — 0.03 seconds — in what are known as flash orders. While markets are supposed to ensure transparency by showing orders to everyone simultaneously, a loophole in regulations allows marketplaces like Nasdaq to show traders some orders ahead of everyone else in exchange for a fee.
In less than half a second, high-frequency traders gained a valuable insight: the hunger for Broadcom was growing. Their computers began buying up Broadcom shares and then reselling them to the slower investors at higher prices. The overall price of Broadcom began to rise.
Soon, thousands of orders began flooding the markets as high-frequency software went into high gear. Automatic programs began issuing and canceling tiny orders within milliseconds to determine how much the slower traders were willing to pay. The high-frequency computers quickly determined that some investors’ upper limit was $26.40. The price shot to $26.39, and high-frequency programs began offering to sell hundreds of thousands of shares.
The result is that the slower-moving investors paid $1.4 million for about 56,000 shares, or $7,800 more than if they had been able to move as quickly as the high-frequency traders.
Multiply such trades across thousands of stocks a day, and the profits are substantial. High-frequency traders generated about $21 billion in profits last year, the Tabb Group, a research firm, estimates.
“You want to encourage innovation, and you want to reward companies that have invested in technology and ideas that make the markets more efficient,” said Andrew M. Brooks, head of United States equity trading at T. Rowe Price, a mutual fund and investment company that often competes with and uses high-frequency techniques. “But we’re moving toward a two-tiered marketplace of the high-frequency arbitrage guys, and everyone else. People want to know they have a legitimate shot at getting a fair deal. Otherwise, the markets lose their integrity.”
-
The 6th paragraph in the above article refers to the 7-Jul news below. It sounded to me started as employment dispute.
Apparently Goldman's profit making algorithm is almost raised to the level of US national security matters, while the morality of this practice may still in question. Not only Goldman was too big to fail, it has to be extremely profitable too. 官商勾結 is only natural.
http://www.nytimes.com/2009/07/07/business/07goldman.html?scp=3&sq=Sergey%20Aleynikov&st=cse
Ex-Worker Said to Steal Goldman Code
By GRAHAM BOWLEY
Published: July 6, 2009
He is no John Dillinger, no public enemy No. 1. But Sergey Aleynikov nonetheless masterminded a dazzling bank theft, the authorities say, and he did it without brandishing a gun or cracking a vault.
Instead, he cracked — or, rather, hacked — the secrets of Goldman Sachs, according to federal agents.
Until a few weeks ago, Mr. Aleynikov, 39, was a computer programmer at Goldman, whose prowess in trading has long made it the envy of Wall Street.
But over five days in early June, the authorities say, he stole proprietary, “black box” computer programs that Goldman uses to make lucrative, rapid-fire trades in the financial markets. Their value, experts say, could be incalculable.
Mr. Aleynikov, however, will not get a chance to use those secrets. He was arrested by federal agents on Friday evening, as he got off a plane at Newark Liberty International Airport. He has pleaded not guilty to charges of theft of trade secrets and transporting them abroad.
The case, as detailed in a federal complaint filed in court in the Southern District of New York, throws a spotlight on the multimillion-dollar technology that is increasingly employed by the world’s biggest banks to gain an edge in financial markets.
Goldman divulged little about the trading programs on Monday, though court documents related to the case said the code that Mr. Aleynikov was suspected of stealing allowed the bank to “engage in sophisticated high-speed and high-volume trades on various stock and commodities markets.”
The software generated “many millions of dollars of profits per year” for the bank, the documents said.
Mr. Aleynikov joined Goldman in May 2007 and was a vice president for equity strategy, but announced his resignation after little more than two years.
He was, he told Goldman, joining a new trading company, which various news reports said was in Chicago. He said he would earn triple the $400,000 salary he commanded at Goldman.
But, just before he left, according to the complaint, Mr. Aleynikov used his desktop computer at Goldman’s New York offices to upload a stream of code to a Web site hosted by a server based in Germany.
Later, he downloaded the files again to his home computer, his laptop computer and to a memory device.
He was caught when the bank noticed the surge of data leaving its servers — and despite his prowess as a highly paid programmer, his activities were recorded even though he tried to erase his programming commands because Goldman kept back-up records.
When confronted by federal investigators at the airport in Newark, Mr. Aleynikov, a naturalized American citizen who immigrated from Russia and now lives in New Jersey, insisted that he had intended to collect “open source” files on which he had worked and only later realized he had copied more files than he had intended.
On Monday, Goldman Sachs refused to comment publicly on the attempted theft. A person familiar with the bank said it had since “secured its systems.” This person, who asked not to be identified, given the confidential nature of the programs, insisted that the theft had had no effect on Goldman Sachs’s business or on that of its clients.
However, at a court appearance in Manhattan on July 4, Joseph Facciponti, the assistant United States attorney, told a federal judge that Mr. Aleynikov’s supposed theft posed a risk to United States financial markets and that other people may have had access to it, according to Bloomberg.
“The bank has raised the possibility that there is a danger that somebody who knew how to use this program could use it to manipulate markets in unfair ways,” Mr. Facciponti said in the court, according to Bloomberg. “The copy in Germany is still out there, and we at this time do not know who else has access to it.”
Bruce Schneier, the chief security technology officer for British Telecom and an expert on computer security, said this type of corporate crime — of a former employee leaving a company with data he should not have — occurred quite regularly. But he agreed that Goldman’s systems had worked well in stopping Mr. Aleynikov.
“This is an example of a system of detection and response working,” he said.
But computer experts expressed caution on the value of the code outside the bank.
Peter Niculescu, a partner at Capital Market Risk Advisors, an advisory firm specializing in risk management and capital markets, said computerized trading had become increasingly important drivers of revenue growth within banks over the last 10 years.
But he said stealing a bank’s trading code did not necessarily guarantee riches, because running it somewhere else was not easy without, for example, a bank’s databases or links to customers.
“If you have the code, but not the database then it is of limited value,” he said. “It is not easy to transfer the code and run it somewhere else.”
Mr. Schneier said, “It is certainly possible that if you knew what the big guys were doing you could anticipate it and make money.” He said that if a rival bank in the United States had been approached to buy the software, it would most likely have called the police, but a seller might have had better luck abroad.
“It is worth a lot less in the U.S. than you might think, but in countries that are more lawless it could have value,” he said.
-
CRW (see 1st message above), high-frequency flash trading (see 3rd message), Statistical Arbitrage (link here (http://chinman.com/index.php?topic=154.0)) - all have one thing in common.
When you have a statistically sound long term edge, bet small but bet a lot, the margin is low but return is very high.
-
CRW sounds interesting. But I'm wondering whether these opportunities really exists. The racing company will set the pools such that no one can make a fortune out of it.
--
cmchan
-
CRW sounds interesting. But I'm wondering whether these opportunities really exists. The racing company will set the pools such that no one can make a fortune out of it.
--
cmchan
CRW only works in pari mutuel pools, like HK Jockey Club, US, Japan, etc.... where you are betting against other gamblers, not the Jockey Club.
CRW does not work in place where bets are placed with book maker, like UK, because you are betting against the bookie.
-
After some googling, I finally understand what you meant.
These are quite some terms that one needs to be really 'in' this field. I can see that there might be some cases where you gets bits of profits from the 'rounding' of odds, or from the rules that governs the min. payout.
-
After some googling, I finally understand what you meant.
These are quite some terms that one needs to be really 'in' this field. I can see that there might be some cases where you gets bits of profits from the 'rounding' of odds, or from the rules that governs the min. payout.
There is no official definition of CRW so I don't know what can be googled. But what I had in mind in this thread is much more than profit from rounding errors. It's generally algorithms, usually very mathmetically sophisticated constructions, to arbitrage mispricing against a more accurate prediction (of price/probability/event/whatever).
People have been doing arbitraging for ages, what's new now is the application of computers to scan the real or projected mispricing in a very fast and efficient (and by some preceived as unfair) manner.
The examples I cited in this thread - CRW in racing, Stat Arbitrage & now High Frequency Trading, are all sort of CRW in my opinion. That computers running sophisticated algorithms scan the markets for mispricings. And one shared strong characteristic is the "bet small but bet many" concept.
-
This subject continued to interest me in the last few months, and after reading more books, I realized that this is an entirely new area in finance that I have never heard of - market microstructure. Some of the studies in this area focus on liquidity discovery through advance mathematics, clever algorithms & superfast computers.
In one of the books I read, the author claimed that now about 50% of the trading volume of NYSE listed shares are done by about 10 firms whose name most people won't recognize & the trades were made by computer algorithms instead of human instructions.
One of these little known automated trading firms is ATD - found by a finance professor David Whitcomb. In an 2002 testimony to SEC, ATD was said to be trading 80 million shares in 300,000 trades per day (or at 267 shares per trade, that's LOTS OF SMALL ORDERS!) According to ATD web site, in 2006, they traded 6% of NYSE & NASDAQ volume. By 2007 they routinely traded 300 million shares per day! (There is no update on these figures after Mar 2007.)
The following article in today's New York Times tells the story that majority of tradings of shares listed in NYSE are not longer traded in NYSE. Instead they are traded on ECNs - private computer networks for trading.
After reading more about high speed automatic trading on these "dark pools", I realized that I may have judge too soon to said the traders are front-running. However, I am still highly suspicious of firms who provide the trading platform, provide the liquidity, AND ALSO trading on house accounts. Maybe for their trading customers who trade large blocks, the liquidity & minimized market impact out weights the harm from potential conflict of interest?!
http://www.nytimes.com/2009/10/15/business/15exchange.html?ref=global
Rivals Pose Threat to New York Stock Exchange
By GRAHAM BOWLEY
Published: October 14, 2009
For most of the 217 years since its founding under a buttonwood tree on Wall Street, the New York Stock Exchange was the high temple of American capitalism.
Behind its Greco-Roman facade, traders raised a Dante-esque din in their pursuit of the almighty dollar. Good times or bad, the daily melee on the cavernous trading floor made the Big Board the greatest marketplace for stocks in the world.
But now, even as the Dow Jones industrial average topped 10,000 for the first time since the financial crisis sent it tumbling, the exchange and its hometown face an unsettling truth: the Big Board, the symbolic heart of New York’s financial industry, is getting smaller.
Young, fast-moving rivals are splintering its public marketplace and creating private markets that, their critics say, give big banks and investment funds an edge over ordinary investors.
Some of the new trading venues — “dark pools,” the industry calls them — are all but invisible, even to regulators. These stealth markets enable sophisticated traders to buy and sell large blocks of stock in secrecy at lightning speed, a practice that has drawn scrutiny from the Securities and Exchange Commission.
These upstarts are utterly unlike the old-school Big Board, which is struggling to make its way as a for-profit corporation after centuries of ownership by its seat-holding members. Last year, its parent company, NYSE Euronext, lost $740 million.
Wall Street’s judgment has been swift and brutal. Since January 2007, the share price of NYSE Euronext has lost nearly three-quarters of its value, even though stock trading over all has soared.
While the exchange has been under assault since the beginning of the decade, its decline has accelerated in recent years as aggressive competitors have emerged. Today, 36 percent of daily trades in stocks that are listed on the New York Stock Exchange are actually executed on the exchange, down from about 75 percent nearly four years ago. The rest of are conducted elsewhere, on new electronic exchanges or through dark pools.
The old Big Board was far from perfect. Its floor brokers — who occupy a privileged, and potentially lucrative, niche between buyers and sellers — have sometimes enriched themselves at their customers’ expense.
But changes inside the exchange’s grand Main Hall are startling. For decades, the New York Exchange was the kind of place where sons followed their fathers onto the trading floor. But half of the jobs there have disappeared over the last five years. Many of the 1,200 or so remaining workers retreat quietly to their computers shortly after the opening bell clangs at 9:30 a.m.
The Big Board has been forced to close one of its five trading halls, and it has repopulated two others with business from the American Stock Exchange, which NYSE Euronext bought last year. The Main Hall — the soaring, gilded room opened in 1903 — can seem little more than a colorful backdrop for CNBC.
“It has not been pretty,” said Benn Steil of the Council on Foreign Relations in New York. “All the big established exchanges around the world have experienced the same phenomenon, but the New York Stock Exchange has taken the biggest beating.”
It is a remarkable comedown for the New York Exchange, and for New York. Once the undisputed capital of capital, the city is struggling to retain its dominance in finance as the industry globalizes. “Wall Street” seems to be no longer a place, but a vast, worldwide network of money and information.
The Big Board says that it is fighting back — and that its hybrid of computers and human traders can beat the new rivals. It slashed commissions and developed its own purely electronic exchange, Arca, in Chicago. Arca has captured about 11 percent of the market for Big Board-listed stocks. It is also winning business in areas like derivatives.
“What’s going on here is a reinvention,” said Lawrence Leibowitz, head of United States markets and global technology at NYSE Euronext. “How can you bring this institution forward into the 21st century?”
Proponents of the new exchanges and private trading systems contend that ordinary people benefit from the technologies whether they know it or not.
“Competition has benefited the average investor,” said William O’Brien, chief executive of Direct Edge, one of the new exchanges. “Their broker has so many choices available, on or off exchanges, anywhere in the world, and they can get their order executed in less than a second.”
Critics maintain that only the most sophisticated players are benefiting, able to execute their trades seconds before smaller investors and in private.
“There are tools now that certain investors have that give them an advantage over other investors,” said Joseph Saluzzi, who trades equities for institutional investors and hedge funds at his boutique brokerage, Themis Trading.
The Securities and Exchange Commission is beginning to take notice of such complaints, opening investigations into the new type of trading venues and promising action. It is worried, for example, that dark pools, with their scale unknown, could destabilize the market.
Unlike the Big Board, the new electronic exchanges are virtually unknown outside financial circles. Direct Edge, the largest, is in Jersey City. Another, the BATS Exchange, is based in Lenexa, Kan. Both are only about five years old. But each now accounts for about a 10th of daily United States stock trading.
In its fight to survive, the Big Board is building a new data center in New Jersey and another outside London. The Main Hall is being overhauled, in an attempt to lure business back to the floor. There is even a new coffee shop, Outtakes.
Even so, the world still watches — literally — what happens on the New York Stock Exchange. Twenty television networks broadcast live from the exchange, in nine languages.
But whichever way the market goes from here, many see a difficult road for the Big Board. The competition is unlikely to let up.
“There has been a sea change,” said Sang Lee of the Aite Group, a financial services consulting company. “I don’t envy what any of the exchanges have to do.”
-
A good video explaining high frequency trading and the basic idea behind statistical arbitrage, by the founder of a high frequency trading fund.
http://link.brightcove.com/services/player/bcpid1827871101?bctid=57418375001
-
The Technology Review article where the above video was found.
http://www.technologyreview.com/computing/24167/page1/
Also attached is the white paper "Toxic Equity Trading Order Flow on Wall Street" mentioned in the below article. The Toxic paper explains how the high speed frequenct guys are squeezing the pennies out of other investors, and that the high frequency guys are chasing rebates. (This is very familiar and similar to anti-rebate arguments in the US racing industry.)
Trading Shares in Milliseconds
Today's stock market has become a world of automated transactions executed at lightning speed. This high-frequency trading could make the financial system more efficient, but it could also turn small mistakes into catastrophes.
By Bryant Urstadt
If Manoj Narang is about to bring down the markets, he's certainly relaxed about it. Narang, who wears a goatee and wire-frame glasses, is casually dressed in a brown shirt and dark gray sweatshirt. Sitting on a swivel chair with one leg tucked under the other, he seems positively composed, especially for a man who has just bought and sold 15 million shares with a total value of $600 million. For Narang, however, such volume represents just the start of a normal day. Though it's about noon on a Friday morning, he has barely begun.
Narang is the head of Tradeworx, a hedge fund and financial-technology firm that makes purely automated trades; all decisions are reached and acted on at near light speed by computers running preprogrammed algorithms. "Actually, we run two businesses," he says. "The first trades in and out of shares in about a second and holds them for an average of two or three days. That's the medium-speed fund. The high-speed fund could make thousands of trades a second and holds them for a matter of minutes."
By the end of the day, his computers will have bought and sold about 60 million to 80 million shares, with the heaviest activity in the last hour of trading, from three to four in the afternoon. Tradeworx and similar firms around the country will race to close billions of bets that hinge on things like tiny differences between the prices of shares in an exchange-traded fund holding the S&P 500 and the individual shares that make up the same index. The profits go to the company with the fastest hardware and the best algorithms--advantages that enable it to spot and exploit subtle market patterns ahead of everyone else. At the end of a typical day, the Tradeworx high-speed business holds no shares at all. Come Monday, Narang will look to trade millions more shares. It seems like a lot, and it is, but Narang estimates that he's probably only somewhere in the middle of the top 50 traders by volume.
Just five years ago, automated trades made up about 30 percent of the market, and few of those moved as quickly as today's trades do. Since then, however, automated trading has become much more widespread, and much quicker. Narang acknowledges starting his ultrafast group as a defensive maneuver when he began to notice faster traders eroding the performance of his medium-speed strategy. Now the medium-speed fund is adopting the techniques he developed in the ultrafast fund.
TheTabb Group, a consultancy based in Westborough, MA, estimates that high-frequency automated trading now accounts for 61 percent of the more than 10 billion shares traded daily across the numerous exchanges that make up the U.S. market. Tabb estimates profits from high-frequency trading in the first nine months of last year at $8 billion or more. With the rise of automation, the bulk of U.S. stock trading has moved from the once-crowded floor of Manhattan's New York Stock Exchange (NYSE) to silent server farms run by exchanges and broker-dealers across the country: the proportion of all trades that the NYSE handles has shrunk from 80 percent in 2005 to 40 percent today. Trading is now essentially a virtual art, and its practitioners put such a premium on speed that NASDAQ has considered issuing equal 100-foot lengths of cable to the brokers who send orders to its exchange servers. (Though Narang and his team program their algorithms on PCs in their own office, actual trading is done through brokers' servers located on the premises of an exchange--NASDAQ, the NYSE, and dozens of others.)
The NYSE itself is just finishing construction of a 400,000-square-foot data center in Mahwah, NJ. The new complex, slated to open in the spring, will have enough computing power to handle every trade on every market in the world, though the exchange will probably have trouble grabbing much of that business back. Hardware used at the facility will operate at a 40-gigabyte-per-second standard, enabling it to handle as many as a million messages a second. (The limit, in many cases, is not the speed at which the information travels but the ability of switches to route it quickly enough.)
The Tradeworx office in Red Bank, NJ, a wealthy shore town about an hour from Manhattan, is a far cry from Wall Street. A quieter place would be hard to imagine. About a dozen employees, most of whom graduated from top-tier schools with degrees in science, math, or engineering, work largely in silence. On this morning, as those 15 million shares come and go, the Tradeworx staff says hardly a word.
In Narang's office, the shades are drawn, the better to read a large monitor on one wall. There are no tickers scrolling by, no flashing updates on the value of the Dow Jones index. Narang's strategy is "market neutral," meaning that when it works--and it usually does--he makes money no matter which way the market goes. His profits don't depend on whether share prices rise or fall; instead, he relies on a set of algorithms that can find and instantly take advantage of tiny, fleeting movements in trading activity. On the wall across from the window is a whiteboard filled with code: a scribbled flowchart in different colors, with variables and occasional amounts in boxes and the words buy or sell. In the middle of the monitor is a large number in a box, going up and down but mainly up. That is Tradeworx's profit so far that morning.
A crash waiting to happen?
High-frequency trading has become controversial, with critics charging that traders are manipulating the market, taking advantage of the little guy, and even courting a full-scale financial meltdown. The critical voices are growing louder and more united, and they're reaching higher up the rungs of power. In September, Senator Edward Kaufman (D-Delaware), speaking on the floor of the U.S. Senate, worried that the United States was moving toward a situation with "one market for huge-volume, high-speed players, who can take advantage of every loophole for profit, and another market for retail investors, whose orders are seemingly filled as an afterthought." The Securities and Exchange Commission has recently proposed a rule to eliminate one controversial tactic of high-frequency traders: the "flash trade," in which exchanges alert designated traders toincoming orders. Critics call it a variation of front-running, an old (and illegal) practice that involved traders buying and selling in advance of large orders.
But accusations of unfairness are not the only issue. Any trend that becomes as dominant as high-frequency trading should be studied to consider potentially serious side effects, warns Paul Wilmott, the publisher of the quantitative-finance journal Wilmott and the founder of the mathematical-finance diploma program at the University of Oxford. "High-frequency trading is the latest bandwagon, and everyone is jumping on board," he says. "Wall Street always piles on to the next thing, and it always blows up."
Wilmott, a self-described "instinctive contrarian" who correctly warned in 2000 that the derivatives market was dangerously unstable, sees particular threats from this trend. The increasing dominance of algorithmic trading and the growing speed of execution, he says, could cause tiny price changes to snowball, rolling down the hill at exponentially increasing speed--either because the machines are trading too fast or because too many funds are trading in the same style. "The potential is there for a crash to happen quite quickly," he says.
Bernard Donefer, who oversaw electronic trading at Fidelity Investments from 1996 to 2002 and now teaches about information technology in financial markets at Baruch College's business school in New York, worries that high-speed algorithmic trading will lead to a smaller-scale version of the crash of 1987, when the market dropped 22 percent in one day. Many now blame that crash on simple automated "portfolio insurance" systems, which were meant to keep a fund's holdings from losing more than a preset amount of value by automatically selling shares when the price dropped by a certain amount. They had their roots in a practice used by floor traders: the "stop loss" order, which initiates the sale of a given share if it falls below a given price. But the herd of computers issuing stop-loss orders created a stampede that pushed the then-dominant floor traders to sell as well. Donefer worries that if such a sell-off happened now, it would happen many times faster.
While such "forced selling" can be the result of forethought (misguided or not), it can also start with a mistake: pressing an extra button (what traders call "fat-finger syndrome") or botching the code that drives an automated algorithm. In 2003, shares of Corinthian Colleges, a company that manages for-profit educational institutions, plummeted when faulty code or human error caused a computer to begin selling shares its user did not have. The system had been programmed to sell if the security returned to the price at which it had been bought. When that time came, the computer sold the shares the customer held and just kept going. In 12 minutes, it sold short nearly three million shares at prices from $57.50 all the way down to $39.50. In a market dominated by high-frequency trading, such glitches could mushroom within seconds.
Even some high-frequency traders worry about what Donefer calls "algos gone wild." John Jacobs, the COO of the New York City-based Lime Brokerage, wrote the SEC in 2009 to voice concerns over the proliferation of brokers who allow major clients to engage in high-frequency trading without validating their margins--that is to say, without making sure they actually have enough money to back a trade. Lime provides high-speed market access and order validation to hedge funds and other traders, some of whom cannot, or don't want to, place their own servers on an exchange floor. In his position, Jacobs regularly sees algorithms executing more than 1,000 orders a second. At that rate, one algorithm trading the wrong way could execute 120,000 orders in two minutes. At 1,000 shares per order and an average price of, say, $20 a share, that's $2.4 billion inunintended trades. In his letter, Jacobs warned of "the potential for trading-induced multiple domino bankruptcies." He cautioned that "unrestrained computer-generated trading has the potential to create catastrophic economic damage to the U.S. national market system."
Penny Pinching
The players in high-frequency trading are many and varied. Some are institutional investors like pension funds, endowments, and mutual funds; others are brokerages or trading desks at banks, using the banks' own money. Enormous hedge funds like the Citadel Investment Group in Chicago use these techniques, and so do startups like PhaseCapital in Boston, which began trading with just the partners' money in the spring. Designated "market makers"--traders licensed by an exchange to create a stable market in a security by making it available to both buyers and sellers in an orderly fashion--use high-frequency strategies to fill orders and to hedge positions, constantly rebalancing inventory so as not to get caught with too many or too few shares. And the field will only grow. Companies now offer high-frequency packages that include software, brokerage hookups, and as much consulting as you can afford.
Indeed, in many ways, practices associated with high-frequency trading have become a routine part of how the market operates. When a customer places a trade through a Charles Schwab account, for example, that order is likely to be handled by a high-speed algorithm. Institutional traders like Fidelity, which buy large blocks of shares for their mutual funds, use algorithmic trading to split their enormous orders into blocks of 100 to 300 shares so that other traders don't recognize the true demand and take advantage of that knowledge for their own profit.
Hedge funds with high-frequency operations, like Tradeworx, work between and around the institutional traders and the market makers, and against each other, attempting to profit by anticipating the moves of others. Their reliance on statistical patterns and quantitative analysis has won them the name of "quant funds." (A quant fund typically holds a portfolio derived from statistical analysis, but its trades may take place over months as well as microseconds. Though most high-frequency funds are quant funds, not all quant funds trade at high frequency.) The explosion in high-speed automated trading has engendered a massive buildup in technology; Renaissance Technologies, a hedge fund based in East Setauket, NY, boasts that its computing power is equal to that of the Lawrence Livermore National Laboratory.
Just one example of what speed can do explains a lot about how high-frequency trading works and why it angers some observers, as Joseph Saluzzi and Sal Arnuk, the principals of the New Jersey-based Themis Trading, made clear in their 2008 white paper "Toxic Equity Trading Order Flow on Wall Street." Imagine that a mutual fund enters a buy order, telling its computer to start by offering the current market price of $20.00 a share but to take any asked price up to $20.03. A high-speed trader, Saluzzi and Arnuk explained, can use a "predatory algo" to identify that limit by "pinging" the market with sell orders that are issued in fractions of a second and canceled just as fast. It might start at $20.05 and work its way down to $20.03, canceling and reordering until the mutual fund bites. The trader then buys closer to the current $20.00 price from another, slower investor, reselling to the fund at $20.03. Because the high-frequency trader has a speed advantage, he is able to do all this before the slower party can catch up and offer shares for $20.01. This speedy player has found the buyer's limit, gathered up and sold an order, and snipped a few pennies off for himself.
Liquidity and Order
Picking up all those pennies can be risky, Narang says, but he makes what he considers an important distinction. "There is risk, definitely, but quant funds like us take it all," he says. "If a quant meltdown happens, it won't affect the retail investor."
Narang turns to his computer and brings up two graphs, superimposing one on the other. The first shows the erratic up-and-down crawl of the S&P 500, the value of the largest 500 companies in the United States, over the last six years. The second shows Tradeworx's profit and loss over the same period. It is a steady march up; in Tradeworx's worst year, it made 15 percent. "All [high-frequency] funds have a profit-and-loss line like this," he says. Then he magnifies the graphs to show just the weeks around August 2007, when many quant funds self-destructed as they sold off their portfolios to meet increasing margin calls (see "The Blow-Up" November/December 2007). In those days, his P&L dropped by 7 percent, and many other funds saw similar losses. But the S&P 500, overall, was little affected.
"And here's the second quant meltdown, in January of '08," Narang says, zooming out and then in on another blip in the graph, showing the value of the S&P 500 when a second, albeit smaller, dislocation occurred. "It's tiny. You can hardly see it. That's because funds running quantitative strategies are mostly market neutral. When we take a position, we're always balanced somewhere else, and when we unwind, it doesn't affect the market either." By this he means that forced selling by quant funds may be painful for the funds themselves, but that pain is barely reflected in the market, because the funds' long and short positions--positive and negative bets on the direction of given securities--cancel one another out. "We don't take from the retail guy," he says. "We make the market more efficient. Things are better for the retail investor because of high-frequency trading."
Narang, and academics like Donefer, say that high-frequency traders are making money by delivering a service: liquidity. In today's highly decentralized market, defenders say, their systems are simply the most efficient way to match buyers and sellers. And because they can capitalize on small differences between the prices at which a seller is willing to sell and a buyer is willing to buy, those differences stay small. The upshot is that retail buyers pay a little less to buy a share and can sell it for a little more. Indeed, since electronic trading has come to dominate the market, spreads between buying and selling prices have decreased dramatically, and so have fees. Ten years ago an investor might have paid $150 in fees to trade 500 shares with a broker, facing a spread of maybe a dime on each share. Today's retail investors pay $10, with spreads of a penny or so in most big stocks, and most of their trades are filled almost instantly.
Understanding how high-frequency trading improves liquidity explains a lot about why many such traders do well when the market is plunging or volatile, as it was last year. "We don't make volatility happen," says Narang. "We reduce it, but it is how we make our money. We create order. When the markets are disorderly, we make a lot of money, but we are doing it by restoring the markets to order."
If Narang is right, the new ways are good for the retail investor. But the argument that high-frequency funds improveliquidity, as if they were providing a public service, is disturbingly reminiscent of the justifications offered by hedge funds and banks that created complicated derivatives in the years leading up to the recent crash. When things went bad in that case, the liquidity disappeared--along with many of the funds invested in them, and much of the investors' money. And this type of history doggedly repeats itself. Wilmott, for one, is not convinced that high-frequency trading is useful to the economy. "People have to say things are fine because they're being rewarded for it," he suggests.
At least for now, though, things are calm, and the spreads are narrow. After lunch, Narang's day at Tradeworx starts to get busier as hundreds of high-frequency funds jostle to close out their positions to their best advantage. Narang says good-bye at the door, his words the only sound in the quiet office. On the wall behind him, Tradeworx's daily profit-and-loss line still ticks up and down, but mostly up.
-
Seems you're drag by this topic for quite a while. Surprisingly from the link above there is a video by this guy Manjo Narang, link here (http://link.brightcove.com/services/player/bcpid1827871101?bctid=57418375001).
The Technology Review article where the above video was found.
http://www.technologyreview.com/computing/24167/page1/
-
[hide]
Recent events bring up this topic again. :)
Let's re-read and see what's changed.
[/hide]
-
There is a court case about computer wagering on HKJC. A professor in CUHK was sued by someone who alleged that the professor used a computer program co-owned by the plaintiff to bet in HKJC. The plaintiff seeks injunctive relief to stop the professor using the program and account for the profit.
The chamber hearing ordered that the professor should disclose his account of earning from using the computer program during the period from 2004 to 2012. The professor appealed for the decision but the ruling was upheld.
This is now reported in newspaper today. Based on the information from newspaper, it seems that the case still has long way to go as it is still under discovery stage.
The public may later know how much money can be made by such computer program, when the case goes to trial.
Case no.: HCA 2352/2012 Bruce James Stinson vs Gu Ming Gao
-
Is Gu MG the one mentioned by Chin last time that he didn't have manual confirmation on placing bets?
-
[hide]
I think so. This case has been going on for 2 years or so.
I am interested to read the case file also. Is it on LII already?
Another case I am interested in reading is the Yeung Kar Sing (once owner of Manchester City Football Club) money laundering case. Do you know the case number or where to find the doc?
[/hide]
-
[hide] Maybe. It is premature to publish the case, as what the news said was just an interlocutory application.
You can find the ruling on Lii about those reported by the newspaper now.
If no one applies for private hearing, it may be searched from LII after finalized. [/hide]
-
[hide] Go to HKLII. Search Carson Yeung. There are a number of cases, including those interlocutory application and JR.
The main one should be DCCC860/2011 [/hide]
-
Hang, Yeung CarSing is closely related to Mrs Zhu. 金利豐
-
[hide]
Go to HKLII. Search Carson Yeung. There are a number of cases, including those interlocutory application and JR.
The main one should be DCCC860/2011
Thanks!
I am book marking here.
http://www.hklii.hk/cgi-bin/sinodisp/eng/hk/cases/hkca/2012/230.html?stem=&synonyms=&query=carson%20yeung
Verdict
http://www.hklii.hk/cgi-bin/sinodisp/eng/hk/cases/hkdc/2014/219.html?stem=&synonyms=&query=DCCC860/2011
Sentence
http://www.hklii.hk/cgi-bin/sinodisp/eng/hk/cases/hkdc/2014/223.html?stem=&synonyms=&query=DCCC860/2011
[/hide]
-
There is a court case about computer wagering on HKJC. A professor in CUHK was sued by someone who alleged that the professor used a computer program co-owned by the plaintiff to bet in HKJC. The plaintiff seeks injunctive relief to stop the professor using the program and account for the profit.
The chamber hearing ordered that the professor should disclose his account of earning from using the computer program during the period from 2004 to 2012. The professor appealed for the decision but the ruling was upheld.
This is now reported in newspaper today. Based on the information from newspaper, it seems that the case still has long way to go as it is still under discovery stage.
The public may later know how much money can be made by such computer program, when the case goes to trial.
Case no.: HCA 2352/2012 Bruce James Stinson vs Gu Ming Gao
http://www.hklii.hk/cgi-bin/sinodisp/eng/hk/cases/hkcfi/2014/631.html?stem=&synonyms=&query=stinson%202352/2012
===
BRUCE JAMES STINSON v. GU MING GAO [2014] HKCFI 631; HCA2352/2012 (2 April 2014)
HCA 2352/2012
IN THE HIGH COURT OF THE
HONG KONG SPECIAL ADMINISTRATIVE REGION
COURT OF FIRST INSTANCE
ACTION NO. 2352 OF 2012
____________
BETWEEN
BRUCE JAMES STINSON Plaintiff
and
GU MING GAO (顧鳴高) Defendant
____________
Before: Hon L Chan J in Chambers
Date of Hearing: 1 April 2014
Date of Decision: 2 April 2014
_____________
D E C I S I O N
_____________
1. This is an appeal by the defendant against an order of specific discovery made by Master de Souza on 27 November 2013.
2. Master de Souza ordered that:
“(1) The Defendant do within 7 days from the date of this Order produce for inspection by the Plaintiff and his Solicitors at the office of the Defendant’s Solicitors at … the following documents referred to in Paragraph 8 of the Defendant’s Defence and Counterclaim dated 18th February 2013, namely:-
(a) The statistical model using mathematical concepts and language and reduced to a series of written mathematical formulae referred to in paragraph 8(a) therein;
(b) The variables generation program (together with written operating instructions) written in a publicly available ‘MatLab’ software capable of receiving Hong Kong Jockey Club and ‘RaceMate’ data feed and converting such data into a series of specific race meeting variables referred to in paragraph 8(b) therein;
(c) The computer values program written in the publicly available ‘Excel’ software (together with written description of and series of commands used) capable of converting specific race meeting variables into numerical values to be assigned to each of the horses listed to start in each race meeting referred to in paragraph 8(c) therein;
(d) The written formulae enabling such numerical values to be represented as probability estimates for each of the four bet types: Win; Place; Quinella; and Quinella Place, referred to in paragraph 8(d) therein,
and permit the Plaintiff and/or his solicitors and/or agents to peruse the documents so produced and to make notes of their contents;
(2) The Defendant do within 7 days from the date of this Order,
(i) make an affidavit stating whether any document specified or described below has at any time been, in his possession, custody or power, and if not then in his possession, custody or power when he parted with it and what has become of it; and
(ii) produce for inspection by the Plaintiff and his Solicitors the following documents in the possession, custody or power of the Defendant:
(a) all statements and records of all betting accounts on horse racing with the Hong Kong Jockey Club under the name of the Defendant and/or his nominees or agents on his behalf (including, but not limited, to Leung Man-kit, Lam Yuk-fai and Siong Liying Ivan) covering the period from the start of 2004/2005 racing season to the end of 2011/2012 racing season; and
(b) all statements and records of all betting accounts on horse racing with the Hong Kong Jockey Club under the name of the Defendant and/or his nominees or agents on his behalf (including but not limited to Leung Man-kit, Lam Yuk-fai and Siong Liying Ivan) covering the period from the start of the 2012/2013 racing season onwards;
(3) The Defendant do within 7 days from the date of this Order supply the Plaintiff with copies of the documents referred to in paragraphs (1) and (2) herein on payment of reasonable charges; and
(4) Costs of and occasioned by this application including any costs reserved in respect thereof be to the Plaintiff with certificate for Counsel to be taxed if not agreed in any event.”
Background
3. By an agreement dated 27 January 2004, the plaintiff and defendant agreed to form a partnership to establish and operate a horseracing betting operation. The operation was to develop and use a mathematical model to provide accurate probability estimates for Hong Kong horse races.
4. The defendant’s obligation under the agreement was to provide, develop, and maintain the mathematical model. He and those working under him were called “the modelling team”. The modelling team at all material times had and the defendant still has sole control of and access to the mathematical model as developed from time to time. The modelling team included the defendant, Leung Man-kit, Lam Yuk-fai and Siong Liying Ivan.
5. The plaintiff, on the other hand, was responsible for managing all other aspects of the partnership business, including the provision of capital and seeking outside investment. The plaintiff was also responsible for paying all costs incurred by the business. The plaintiff also had people working under him and his team was called “the management team”. The management team did not have access to the mathematical model.
6. Under clause 15 of the agreement, the mathematical model, including all of its components and their continuing development, constitute the property of the business, and both the plaintiff and defendant undertook that they would not deal with, sell or share with others the model.
7. Since around 2007 and 2008, the model had been able to produce accurate estimates and generated profits for the business.
8. There is no dispute that the defendant had been making private bets, but there is a pleaded dispute on whether he did so with the use of the mathematical model and, hence, in breach of the partnership agreement and his obligations as a partner. The plaintiff alleged that the defendant’s private betting had gone on to a much bigger scale around the 2010/2011 racing season, but the defendant refused to discuss with the plaintiff about the defendant’s private betting. The plaintiff has also adduced evidence alleging that the defendant had placed private bets through nominee betting accounts held by members of the modelling team as well.
9. By an email dated 27 July 2012, the defendant gave notice to the plaintiff to terminate the partnership there and then. The plaintiff then proposed that both the plaintiff and defendant should walk away with a working copy of the mathematical model, but the defendant refused the proposal.
10. The plaintiff then launched this action on 21 January 2013 for delivery up by the defendant of the model and other reliefs.
11. The defendant in paragraph 8 of his defence and counterclaim referred to the mathematical model, and pleaded that it comprised of four discrete elements. Particulars of the elements were also given. The plaintiff then served a notice to produce documents dated 12 July 2013 on the defendant under Order 24 rule 10 for the production of the mathematical model for inspection, but the defendant refused to comply with the request in the notice.
12. By an amended summons dated 14 November 2013, the plaintiff applied under Order 24 rule 10 and 11 for a production of the mathematical model and its discrete elements as referred to in the defence and counterclaim. The plaintiff also applied for discovery of the defendant’s betting records under Order 24 rule 7.
13. On 27 November 2013, Master de Souza made the orders sought. The mathematical model and its discrete elements are the subject matters of paragraph 1 of the order, which has been cited above. The betting records are the subject matters of paragraph 2 of the said order.
14. The defendant then gave notice of appeal. The execution of the order of discovery was then stayed by consent.
The appeal
15. On this appeal, Ms Tam, leading counsel for the defendant, submitted that the contents of the mathematical model are of little relevance to the pleadings, and the defendant may be prejudiced if he should disclose the same to the plaintiff before the partnership is wound up.
16. This submission is pregnant with an implication that the plaintiff would abuse the model by exploiting it for its own benefit if he should be in possession of it.
17. I am not in favour of this unsupported attack on the good faith of the plaintiff.
18. Counsel further submitted that the model is not the subject matter of the dispute between the parties, but is property owned by the partnership. Hence, it should not be delivered up to the plaintiff now, but should only be made available after the taking of the partnership account so that it could then be sold or disposed of properly.
19. Alternatively, counsel submitted that even if the model is the subject matter of the partnership agreement, its contents are irrelevant to the main issues to be resolved in the action, and has no bearing on the result of the action.
20. Counsel also submitted that the plaintiff already had access to the outputs of the model through computers, but I think this is neither here nor there.
21. The defendant, however, acknowledged that one major issue between the parties is whether the defendant had used the model in his private betting.
22. Mr Yan, leading counsel for the plaintiff, submitted that delivery up of the model is important and relevant to the issues as the defendant intends to engage an expert to study the contents of the model and the defendant’s private bets to see if the bets were placed with the benefit of the output of the model. For that purpose, it is necessary to have possession and use of the model.
23. I think the stance of the plaintiff is reasonable, and it is necessary to have the model delivered up by the defendant to the plaintiff. I also do not think that there should be any condition to be imposed on the plaintiff for the model to be delivered to him unless the defendant be subject to the same.
24. For these reasons, I am inclined to dismiss the appeal in relation to the production of the mathematical model.
25. At the end of the hearing yesterday, the parties jointly proposed that they would work out an undertaking for the custody of the model in the event that I should dismiss the appeal in relation to the model. However, the good intentions of the parties failed to bear fruit, and they have not been able to reach agreement on the undertaking.
26. In the premises, I dismiss the defendant’s appeal on paragraph 1 of the order. I further vary the mathematical model referred to in paragraph 1 of the order to include all versions of the model and its discrete elements developed by the defendant and/or the modelling team from time to time up to the date of delivery up.
27. Regarding the betting records, Miss Tam attacked strenuously on the strength of the plaintiff’s claim against the defendant on private betting. She also referred to the defendant’s denial on oath that he had undertaken private betting with the use of the model in question. However, this is not an application to strike out this claim of the plaintiff on any basis. There is also no such application. Hence, this is a claim that needs to be resolved. Documents relevant to this claim should likewise be produced in discovery. The records of the defendant’s private betting are obviously documents relevant to this claim and should be disclosed.
28. Furthermore, it is the plaintiff’s case that he needs these records for examination by an expert to find out whether they show a pattern that is related to the output of the model. This is indeed a reasonable argument.
29. Finally, Ms Tam said that the defendant’s private betting records to be produced should be limited to the racing days when the partnership had placed the bets. For racing days that the defendant had placed the bets but not the partnership, there would be no record of the partnership to compare with. Hence, she submitted that the defendant’s betting records for those days need not be produced.
30. Mr Yan, in reply, however, pointed out that the study by the expert would have to include all the defendant’s private betting records regardless of whether they were records of bets on racing days when the partnership had placed no bet, as the study is to discern a pattern which would be produced by the mathematical model.
31. I agree with this submission too. But, I would limit the production of such records up to the date when the defendant should perform and comply with paragraph 1 of the order as varied by me.
32. In the premises, I dismiss the appeal as a whole. I also lift the stay of execution, if it has not expired by its own terms.
33. I further make a costs order nisi that the defendant do pay the plaintiff the costs of this appeal with certificate for two counsel.
(Louis Chan)
Judge of the Court of First Instance
High Court
Mr John Yan, SC, leading Mr Dominic Pun, instructed by Lily Fenn & Partners, for the plaintiff
Ms Winnie Tam, SC, leading Mr Jason Yu, instructed by Baker & McKenzie, for the defendant
-
[hide]
Some observations from the above decision on the appeal.
1. The modelling platform is MatLab, data source RaceMate, Excel maybe used for pre-processing. It's not an integrated system, probably still very labor intensive like CX Wong's???
2. However the above was from Gu's defense and he could be just describing an old system to fool the other side. I imagine Stinson had not had access to the real inner working of the models.
3. There is no mention of how the bets were placed. If last season's $30m mistakes were indeed from Gu, the bet placement system was not mentioned in the defense.
4. They bet Win, Place, Q, QP only?!?!?! Again maybe these are only what Gu would tell Stinson.
5. I am also surprised that in this partnership they didn't have non-compete, minimally exclusivity limiting personal betting.
[/hide]
-
Is Gu MG the one mentioned by Chin last time that he didn't have manual confirmation on placing bets?
Turned out it was not him... someone else from CU. Just how many professors are doing this! ;D
-
[hide]
Yep. This is an application for discovery. Subject to the perusal of the document discovered, there may be more information revealed, maybe including the change of all pleadings.
It still has long way to go.....
[/hide]
-
Recent development of the Stinson vs Gu case. Booking here for easy access for my friends. :)
http://legalref.judiciary.gov.hk/lrs/common/search/search_result_detail_frame.jsp?DIS=102847&QS=%2B&TP=JU
HCA 2352/2012
IN THE HIGH COURT OF THE
HONG KONG SPECIAL ADMINISTRATIVE REGION
COURT OF FIRST INSTANCE
ACTION NO 2352 OF 2012
____________
BETWEEN
BRUCE JAMES STINSON Plaintiff
and
GU MING GAO (顧鳴高) Defendant
____________
Before: Hon Au-Yeung J in Chambers
Date of Hearing: 19 October 2015
Date of Decision: 26 February 2016
________________________
D E C I S I O N
________________________
INTRODUCTION
1. This is the Plaintiff’s appeal against the Registrar’s decision refusing to order non-party discovery. The issues in the appeal are necessity and relevance of the discovery.
BACKGROUND
2. The Plaintiff and the Defendant (a statistics professor), have formed a partnership (“the Partnership”) since 2004 to establish a horse racing betting operation using a Mathematical Model which provided accurate probability estimates for Hong Kong horse races.
3. Under the Partnership Agreement, the Defendant was responsible for a modelling team that provided, developed and maintained the Mathematical Model. The Defendant and the Modelling Team at all material times had sole control of and access to the Mathematical Model as developed from time to time. The Defendant and one Leung Man Kit (“Leung”) were part of the Modelling Team. The Plaintiff did not have access to the Mathematical Model.
4. Under the Partnership Agreement, the Mathematical Model constituted the property of the Partnership. Both parties agreed that they would not deal with, sell or share the Mathematical Model with others.
5. Since around the 2007 racing season, the Mathematical Model had been able to produce accurate estimates and to generate consistent profits for the Partnership Business.
6. It then came to the attention of the Plaintiff that the Defendant had, allegedly in breach of the Partnership Agreement and the Defendant’s obligations as a partner, been making private bets using the Mathematical Model. It also became apparent that the Defendant’s private betting had gone on to a much grander scale around the 2010/2011 racing season.
7. By an email dated 27 July 2012, the Defendant gave notice to terminate the Partnership Business.
8. The Plaintiff proposed that each party should walk away with a working copy of the Mathematical Model, and requested the Defendant to deliver up a copy of the same; but the Defendant refused.
9. The Plaintiff issued the Writ on 19 December 2012, seeking the delivery up of the Mathematical Model, damages for breach of the Partnership Agreement and account of profits made by the Defendant from private betting using the Mathematical Model on his own account.
10. The Defendant does not deny private betting. The issue is whether he had (on the Plaintiff’s case) used the Mathematical Model (based on a logit model), or (on his own case) the probit model.
11. On 27 November 2013, upon the Plaintiff’s application and after a contested hearing, Master De Souza ordered the discovery of the betting records of the Defendant and/or his nominees or agents on his behalf (including but not limited to Leung, Lam Yuk Fai and Xiong Liying Ivan). The discovery was to cover the period from the start of the 2004/2005 racing season to the end of 2011/2012 (ie 8 racing seasons).
12. The Defendant’s appeal was dismissed by L Chan J on 2 April 2014. However, the learned judge limited the discovery period to the time of delivery up of the Mathematical Model.
13. By summons dated 4 April 2014 (“Defendant’s Variation Summons”), the Defendant sought to vary or correct the Order of L Chan J. The draft order attached to that Summons expressly provided, in §1 thereof, that the Defendant should provide discovery of the Defendant’s betting records including those of “his nominees or agents on his behalf, if any, …”
14. The Defendant’s Variation Summons was dismissed by L Chan J on 15 April 2014 as unmeritorious.
15. On 19 June 2014, upon the Defendant’s application, Deputy High Court Judge Seagroatt ordered discovery against the Plaintiff of all statements and records of all accounts maintained with the Hong Kong Jockey Club (“HKJC”) from 27 January 2004 to 27 July 2012 in respect of horse betting activities “undertaken by individuals on behalf of the Partnership Business.”
16. Meanwhile, the Defendant only disclosed about 3 years of his own records between 13 April 2011 and 18 December 2013 in purported compliance with the Order of L Chan J. His excuse was that that was what the HKJC had supplied to him. The Defendant, denying private betting through nominees/agents (collectively “agents”), did not disclose any betting records of the alleged agents.
17. HKJC had confirmed that it was possible to provide betting records up to 7 years from the date of the order. The Plaintiff drew this to the Defendant’s attention. The Plaintiff himself had in fact disclosed 7 years of HKJC betting records to the Defendant pursuant to the order of Deputy High Court Judge Seagroatt. Despite these, the Defendant has not produced further betting records.
THE PRESENT SUMMONS
18. The Plaintiff took out the present summons seeking discovery against HKJC (“the HKJC summons”) of:
(1) all documents showing details of all betting made by and/or on behalf of the Defendant for each race from the commencement of the racing season in 2004 to the present (“class 1 documents”);
(2) statements and records of all betting accounts of the Defendant relating to the betting under (1) above (“class 2 documents”);
(3) statements and records of all betting accounts for each race from the commencement of the racing season in 2004 to the present under the name Leung Man Kit (“class 3 documents”);
(4) statements and records of all other betting accounts relating to betting made on behalf of the Defendant under (1) above (“class 4 documents”).
19. Both before the Registrar and this court, HKJC took a neutral stance. It does not object to production of the betting records of the Defendant and Leung for 7 years from the date of the order of the court.
20. The Defendant vigorously opposed the HKJC summons despite the prior order of L Chan J. It is the Defendant’s case that his private bets did not involve using (a) the Mathematical Model; and (b) agents.
21. The Registrar dismissed the HKJC summons.
22. In this appeal against the Registrar’s decision, the scope of discovery sought has been greatly narrowed. The Plaintiff only seeks discovery of betting records of the Defendant and Leung but not other agents. The bone of contention is relevance and necessity.
23. Ms Tam SC has framed 2 issues:
(1) Further discovery of the Defendant’s records is unnecessary;
(2) Discovery of agent’s records is irrelevant and unnecessary.
24. Since the filing of the appeal, the Plaintiff had entered into further correspondence with HKJC. HKJC took the stance that it could only preserve records (dating back for longer than 3 years) if the Defendant and Leung did not object to the same. The Defendant did not object to the preservation of those records, but he did not consent to disclosure of the same.
25. After the hearing, I have ordered HKJC to preserve the documents sought pending handing down of this decision.
26. Before I analyze the case further, I wish to deal with a point on evidence. The Plaintiff has purportedly put in the affirmation evidence (in relation to another summons dated 27 November 2014), which was filed after the evidence in this HKJC summons was closed.
27. To adduce evidence in such manner reflects a lack of discipline. It is not open to the Plaintiff to submit that the affirmations were filed only after this appeal was launched and hence fell within Order 58, rule 1(5). I shall not rely on the evidence in relation to the other summons, except Professor Fan’s affirmation not objected to by Ms Tam SC.
LEGAL PRINCIPLES ON NON-PARTY DISCOVERY
28. Under s.42 of the High Court Ordinance, Cap 4, the Court has a discretion to order disclosure of documents by a non-party. It is for the applicant to show that the documents sought are in existence, in the possession of the respondent and are relevant to the issues, and that discovery is necessary for fairly disposing of the cause or matter.
29. The test of relevance in an application under s.42 and O.24 r.7A is the same test as that which is applied for other types of discovery under O.24, namely documents which may contain information which may enable a party either to advance his own case or to damage that of his adversary. These include documents which may fairly lead a party to a train of enquiry which may have either of those consequences: Peruvian Guano case (1882) 11 QBD 55; Tullett Prebon (Hong Kong) Limited v Chan Yeung Fong Nick & ors (unrep, HCA 2197/2009, 9 June 2011) at §74 per To J; Chan Tam Sze v Hip Hing Construction Co. Ltd, (unrep, HCA 1931/1988, 17 October 1989) per Bokhary J at p.5.
30. Even if the court is satisfied with the above requirements, it must still exercise its discretion bearing in mind that disclosure orders against third parties are exceptional. It should not be used as a fishing exercise for documents nor be speculative. It should not be oppressive to witnesses: Tullett Prebon v Chan at §77 and Re Estate of Ng Chan Wah HCAP 5/2003 at §16 per Chu J (as she then was).
31. The following factors should be considered on an application for third party record:
“First, how important is the information to the issues? Second, has the applicant taken appropriate steps to obtain the information within the proceedings before seeking disclosure from the third party? Third, would it be sufficient for the court simply to draw adverse inferences on the basis that the party from whom the information was sought within the proceedings has failed to supply their information? Fourth, what is the nature of the relationship, if any, between the parties to the proceedings and the third party? Fifth, if disclosure is necessary and proportionate wheel [sic] the editing of documents protect private information?” (Tullett Prebon v Chan (supra) at §85 per To J, citing a decision of Hartmann JA, as he then was).
32. There is no dispute that the documents sought (dating back 7 years) are in existence and that HKJC is in possession, custody or power of the subject records of both the Defendant and Leung.
ISSUE 1 – FURTHER DISCOVERY OF THE DEFENDANT’S RECORDS
Relevance
33. In his decision dated 2 April 2014 (§28), L Chan J decided that the Defendant’s betting records were relevant and that the Plaintiff needed these records for examination by an expert to find out whether they showed a pattern that was related to the output of the Mathematical Model. I agree with the learned judge’s views.
Necessity to fairly dispose of the cause or matter
34. Ms Tam SC, however, contends that relevance is not by itself sufficient. Her submissions fall under 3 limbs:
(i) It is the Plaintiff’s own case that he already has sufficient evidence to prove the central issue on the Defendant’s betting;
(ii) The full 7 years’ betting records only go to quantum and are unnecessary at the present stage;
(iii) Discovery of the full 7 years’ betting records will be a disproportionate burden to the HKJC, the Defendant and the court.
That it is the Plaintiff’s own case that he already has sufficient evidence to prove the central issue on the Defendant’s betting
35. Ms Tam SC submits that the Defendant has already produced 3 years’ HKJC records amounting to 1,931 pages. On those materials, each party’s expert agreed that it was sufficient with the present discovery to prepare an opinion.
36. Ms Tam SC pointed out that Professor Fan (Defendant’s expert)[1] said that the statistical data from one racing season was “more than enough statistically” to draw his conclusion.
37. On the other hand, she points out that Mr Ziemba, the Plaintiff’s expert, was able to come to a confident, unqualified opinion that the Defendant’s private betting was done using the Mathematical Model.
“Even for the 2011/2012 season alone, there is a lot of data to be analysed in [the Defendant’s] Betting Records and the Partnership Betting Records and that data alone is sufficiently large and sufficient for me to reach the conclusions that [D’s] bets were made with the Model and that he must have placed bets other than those shown in [the Defendant’s] Betting Records through other accounts such as that of [Leung].”
Such opinion was repeated in various places in his report.
stated that further disclosure of the Defendant’s betting records was necessary to arrive at a more accurate opinion, so Ms Tam SC submits.
39. With respect to Ms Tam SC, Mr Ziemba was responding to the Defendant’s contention that no matter how much study and comparison was made between the Defendant’s and the Partnership’s betting records, the opinion that the Defendant’s private betting was done using the Mathematical Model was inconclusive. There clearly was a dispute in opinion between the experts.
40. Further, Mr Ziemba has made it clear that what he had said in his affirmation was not intended to be a detailed analysis and comparison but that such detailed analysis and comparison would be done in the expert report which he would be preparing.
41. It is not for the court to decide on which expert’s approach or conclusion is correct at this stage. Mr Ziemba has not stated that he did not require 7 years of the Defendant’s betting records. It was justified for him to study as many of those records as possible to identify the Defendant’s betting pattern, to prove the Plaintiff’s case. It was not a fishing expedition.
42. Besides, the Defendant should not be allowed to circumvent L Chan J’s order through a back door by purportedly relying on disputed expert’s views. I reject the first limb of Ms Tam SC’s argument under Issue 1.
The full 7 years’ betting records only go to quantum and are unnecessary at the present stage
43. Discovery relating to damages or profits will not be ordered until all issues of liability have been determined and the Plaintiff has elected whether to claim damages or an account of profits. Such discovery will be “premature, must impose trouble and annoyance upon the Defendant to no purpose if the verdict be in his favour on the question of liability, and ought not to be granted”. See Auto-Treasure Ltd t/a Albert Jewelry Creation v Noble Diamond Ltd t/a Noble Jewellery & anor [1992] 1 HKC 117, at 119D-G and 120E-F, per Godfrey J (as he then was).
44. In an action between partners, there will be no award for damages and the Plaintiff’s only remedy is the taking of partnership accounts: Yau Wah Hing & anor v Yuen Kay Ming (unrep, CACV46/2012, 19 March 2013) per Lam JA at §§62-63; Leung Wing Yiu v Siu King Yuen & ors [2003] 2 HKLRD 21 per DHCJ Lam (as he then was) at 28‑29.
45. Accordingly, there will inevitably be a separate trial of liability and taking of partnership accounts in the present case. Ms Tam SC submits that the full 7 years’ betting records are relevant only to quantum and so discovery of them is pre-mature.
46. With respect, I disagree. At the trial, the court has to find whether the Defendant had breached the Partnership Agreement and when it was breached. These are issues on liability. If, during the subsequent stage of account and inquiry, the Defendant says that he had not used the Mathematical Model for the rest of the 4 years, the Master will have no finding to assist him or her. The Master may not be able to order further discovery as to when the breach had occurred.
47. I am not satisfied that the undisclosed HKJC records are only relevant to the issue of quantum.
48. In any case, even if the documents are not necessary at this stage, it is necessary to order HKJC to preserve those documents, otherwise it would erase them in accordance with its internal policy.
Discovery of the full 7 years’ betting records will be a disproportionate burden to the HKJC, the Defendant and the court
49. Ms Tam SC points out that according to HKJC, providing the Defendant’s betting records up to 7 years would require significant administrative effort. The Defendant has already disclosed 1,931 pages of betting records for a 3-year period. The parties, their lawyers and their experts have already incurred significant time and costs analyzing those records. Adding another 4 years of records will add significantly to that burden and is unacceptable in light of the underlying objectives in Order 1A of the RHC.
50. Now that the scope of discovery has been narrowed, HKJC has not suggested in this appeal that the administrative effort would be too onerous for it. In fact, it has produced 7 years of the Plaintiff’s records. Given my finding that the 7 years of betting records are relevant to the issue of liability, any administrative “burden” and costs cannot override the need to do justice.
51. In summary, the Defendant has not established any of the grounds in opposition in respect of Issue 1. His discovery under L Chan J’s order was inadequate. It is necessary and fair to order HKJC to make discovery of the Defendant’s betting records for 7 years to cure the inadequacy.
ISSUE 2 – DISCOVERY OF AGENT’S RECORDS
52. It is the Plaintiff’s case that the Defendant had been betting through agents of which Leung was one. The bases are as follows:
(a) The strategy which the Defendant had devised for the Partnership Business using the Mathematical Model was to place bets on 4 betting options, namely Win, Place, Quinella and Quinella Place. It proved to be highly successful. It was inconceivable that the Defendant only bet on Place and Quninella in his private betting, especially since his private betting was about 3 times more than the Partnership Business’ bets. Based on the opinion of Mr Ziemba, the Plaintiff contends that the Defendant must have been placing bets on Win and Quinella Place through other HKJC accounts of his nominees.
(b) The Plaintiff’s witnesses made statements to the effect that the Plaintiff and his staff were aware of the alleged private betting conducted by Leung purportedly on behalf of the Defendant from 2004 to 2012. Leung was also seen logging into his betting account which showed extraordinarily high balances (from HK$1.3 million to HK$1.7 million) for someone of his income.
(c) Despite earning a modest salary before resignation from the Partnership in July 2012, Leung and his family members have in recent years been able to purchase 7 properties with substantial amounts of money.
(d) It is the Plaintiff’s contention that Leung has been working with the Defendant in placing bets using the Mathematical Model, and was rewarded handsomely by the Defendant. Such incomplete records of the Defendant’s betting account which have been disclosed by the Defendant showed that the Defendant had made over HK$56 million of profits through betting with his own betting account alone. It is the Plaintiff’s case that the Defendant must also have made more profits through other betting accounts including Leung’s.
(e) The Plaintiff invites the court to draw adverse inferences from Leung’s alleged evasion of service.
53. Despite the Defendant’s denial, those bases do disclose a case that Leung has been an agent of the Defendant.
54. However , it is Ms Tam SC’s case that:
(i) Betting through agents has not been pleaded;
(ii) Discovery of betting records of agents is not necessary for the fair trial of the action as the Plaintiff claims to have sufficient evidence even without third party discovery;
(iii) The issue of Leung betting on behalf of the Defendant is peripheral to the central issue concerning the Defendant’s use of the Mathematical Model in his private betting;
(iv) This case requires strong case management to prevent proliferation of non-essential issues.
Lack of plea of agency
55. Relevance is one of the 3 pre-requisites to the court’s jurisdiction to order discovery. In the context of relevance, the issue must be one identified in the pleadings. See Paul’s Model Art GMBH & Co KG v UT Ltd & ors [2006] 1 HKC 238 at §§24(1)(b) and 25 per Cheung JA.
56. I will add that a matter does not become an issue simply because it is hotly contested in affirmations in interlocutory proceedings, witness statements or expert reports.
57. Agency must be expressly pleaded. Discovery of documents relating to whether an alleged agent had taken secret profits was refused on grounds of relevance where the party seeking discovery did not plead agency: Vastfame Camera Limited v International Freight Express (HK) Limited & ors (unrep, HCCL8/2003, 29 April 2004), per DHCJ Muttrie at §§6-9.
58. The importance of proper pleading has been emphasized by the Court of Final Appeal in Kwok Chin Wing v 21 Holdings Limited (formerly known as GFT Holdings Limited, Capital Prosper Limited and Rockapetta Holdings Limited) & anor [2013] HKCU 2272 at §21:
“21. … The basic objective is fairly and precisely to inform the other party or parties in the litigation of the stance of the pleading party (in other words, that party’s case) so that proper preparation is made possible, and to ensure that time and effort are not expended unnecessarily on other issues:- Wing Hang Bank Limited v Crystal Jet International Limited. It is the pleadings that will define the issues in a trial and dictate the course of proceedings both before and at trial. Where witnesses are involved, it will be the pleaded issues that define the scope of the evidence, and not the other way round. (emphasis added) In other words, it will not be acceptable for unpleaded issues to be raised out of the evidence which is to be or has been adduced. As the Court of Appeal remarked in Wing Hang Bank Limited v Crystal Jet International Limited:-
“(2) In a trial, particularly where evidence is given by witnesses, it becomes extremely important that each side knows exactly what are the live issues. Where issues are sought to be introduced that have not been adequately or properly pleaded, amendments must be sought unless the consent of the other party or parties has been obtained. It will simply not do for unpleaded issues to be ‘slipped in’ when evidence is being given in the hope that the other side is not sufficiently alert to object.” ” (emphasis added)
59. The lack of earlier challenge against relevance does not confer on the court a jurisdiction that it does not have.
60. In the present case, there cannot be any dispute that the agency issue was never expressly pleaded but it was hotly contested in the witness statements and expert reports.
61. Mr Yan SC relies on paragraph 7 and its particulars of the statement of claim which pleads that the Defendant has converted the Mathematical Model to his own use and made private profits by utilizing the Mathematical Model to place bets “on his own account and not for the Partnership Business”. Mr Yan SC submits that those pleas mean that the Defendant had bet for his own benefit instead of for the Partnership’s.
62. In my view, a fair reading of those pleas does not include betting through an agent. As the case of agency is based on inference from matters in paragraph 52, there is all the more reason to plead it before discovery.
63. Then Mr Yan SC submits that Master de Souza and L Chan J had regarded the betting records of Leung as relevant and necessary to the resolution of the issues of this case. No one raised the issue of lack of plea of agency in those 2 courts. However, I note that L Chan J’s decision of 2 April 2014 did not explain the relevance of Leung’s records.
64. Mr Yan SC relies on the Defendant’s Variation Summons to show that the Defendant had not sought to remove the provision about discovery of betting records of his alleged agents. I do not consider that as relevant. The hearing before L Chan J had ended. The draft terms of variation was a desperate attempt to change the order and could not be taken as an admission by the Defendant that the alleged agents’ records were relevant to the claim.
65. I find that the lack of agency plea is fatal to the application for Leung’s records.
Other limbs of arguments under Issue 2
66. I deal with paragraphs 54(ii) to (iv) for completeness sake. If betting through Leung as an agent had been pleaded, discovery of Leung’s betting records would clearly have been necessary for the same reasons given in respect of the Defendant’s betting records. Betting through an agent would not be peripheral to the issue of which model the Defendant used in his private betting. It would be in itself an act of breach of the Partnership Agreement. In that case, it would not be a question of case management but that the discovery ought to be ordered for the fair trial of liability of the action. Leung has denied on oath that he had betting records of agents. That was conclusive on the issue of possession in the context of discovery but not the issue of agency. In respect of discovery, the Plaintiff would have been entitled to seek discovery from HKJC.
Lack of service on Leung
67. I deal with a procedural issue for completeness sake. Mr Yan SC confirms that the HKJC summons had not been served on Leung and there has not been an order for substituted service. I consider that it would not be fair to order HKJC to disclose Leung’s records anyway without affording him the opportunity to address the court before an order is made.
CONCLUSION
68. I allow the appeal to the extent that HKJC should produce all documents showing details of all betting made by the Defendant for each race counting from the commencement of the racing season in 2004/2005 to the date of delivery up of the Mathematical Model as ordered by L Chan J. This is a modified version of classes 1 and 2 documents.
69. I dismiss the appeal in relation to classes 3 and 4.
70. Subject to these orders, the order of preservation of records which I have imposed on HKJC on the date of the hearing is discharged.
COSTS
71. I agree with Ms Tam SC that the scope of this appeal was much reduced from that of the HKJC summons. The Plaintiff is partly successful to the extent of the Defendant’s betting records but not Leung’s. The Defendant has unreasonably contested the part relating to his records despite L Chan J’s order and HKJC’s stance. I therefore make an order nisi that there be no order as to costs of this appeal.
72. On a nisi basis, HKJC shall be awarded costs of this appeal on indemnity basis: (a) such costs are summarily assessed and allowed at $180,000; (b) half is to be borne by the Defendant. The other half shall be borne by the Plaintiff in the first instance but the ultimate liability shall be reserved to the trial judge.
73. As regards the hearing before the Registrar, the Defendant should be awarded 50% of the costs. I make an order nisi setting aside the Registrar’s order for party and party costs so that: (a) the Plaintiff do pay the Defendant costs of $250,000; (b) half of the costs awarded to HKJC in the sum of $160,000 shall be borne by the Plaintiff; (c) the other half in the sum of $160,000 shall be borne by the Plaintiff in the first instance but the ultimate liability shall be reserved to the trial judge.
74. I thank counsel for their assistance.
(Queeny Au-Yeung)
Judge of the Court of First Instance
High Court
Mr John Yan SC and Mr Dominic Pun, instructed by Lily Fenn & Partners, for the Plaintiff
Ms Winnie Tam SC and Mr Jason Yu, instructed by Baker & McKenzie, for the Defendant
Mr Nicholas David Hunsworth (solicitor advocate) of Mayer Brown JSM, for the Hong Kong Jockey Club
[1] The use of this report was not opposed to by Ms Tam SC.