High Frequency Trading

相信未來數個月後，「High Frequency Trading」一詞將會成為投資世界經常談論的題目。

據了解，「High Frequency Trading」是大行交易員運用的程式買賣系統，以自動模式為交易員以更快和更有效率的方法進行買賣活動。

傳统交易方法是交易員在「Real-Time」報價器面前望著買賣(Bid & Ask)的供求盤路而決定下單所放的價格。

相反，程式買賣系統則善用高速電腦的優勢，以較短的時間自動為交易員尋找買家或沽家，盡快完成買賣。另外，一些程式買賣系統亦會智能地從買賣盤落以及成交量的變化而作出交易(當然跟據交易員事先所定下的「Trading Rules」)。

據「FT/alphaville」一篇名為「The Cold War in high frequency trading」的文章，現時全球大有70%的股票買賣活動是由「High Frequency Trading」的程式買賣系統所完成。

然而，電腦主導的系統有時亦會有意無意地造成一些意想不到的結果，尤其是在收市前一分鐘內的時段內，而市場參與者變得「快來快住」之際。近排的例子就正如2009年7月15日滬深300 ETF (2827 HK)的收市價，或者是7月16日東方海外(316 HK)的收市價...........

始終，金融市場只不過是人類心理情緒大上的大落的反照，千百年來並沒有隨著科技的發展而有所改變。

--------------------------------------------------------------------
尾市烏龍盤 標智ETF飈51%
(明報; 16 Jul, 09)

內地股市做好，帶動本港中資股走勢大旺，挾着A股概念的A股交易所買賣基金（ETF），亦全線報捷，其中更以標智滬深300ETF（2827）的升勢，最令市場嘩然。滬深300ETF昨日收報53.75元，大升51％，這個「驚人」升幅，全因尾市有經紀錯誤輸入「烏龍盤」。撇除這個「人為錯誤」，滬深300ETF相應的每股資產淨值，應為35.86元。

港交所（0388）昨晚已向證券經紀刊發通告，提醒他們留意有關情况，因今天開市股價會相應調整，而市場應以ETF的資產淨值作參考基礎。滬深300ETF昨日跟隨A股升勢向上，但全日都在30多元水位上落，直至下午3時59分，卻由37.85元，彈升至53.75元，這是因為有一個叫價53.75元的買賣盤成功配對。港交所發言人表示，知悉有關情况，亦就此聯絡相關經紀，但相關經紀沒有申報「錯價交易」，是以有關交易沒有被取消。

程式「不問價買」致買貴貨
據了解，以「高價」購入9600股滬深300ETF的經紀，來自某大型經紀行，有關經紀確是在程序上出錯，致出現「烏龍盤」，惟由於只涉及50多萬元，金額並非龐大，故有關經紀只當自己「買貴貨」，沒有要求取消及申報「錯價交易」。據悉有關經紀昨日一直以人手方式入盤，但在尾市改以系統程式入盤，由於程式設定了「不問價買入」，恰巧又有人掛出53.75元的沽盤，因而令這個「莫名」盤成功配對。

--------------------------------------------------------------------
The Cold War in high frequency trading
(FT/alphaville; 8 Jul 09)

This April 2009 quote from Rick Bookstaber, with its overtones of Cold War mutually-assured destruction, seems eerily prophetic given recent events involving the alleged theft of Goldman Sachs’ proprietary high-frequency trading code by a programmer of Russian origin:

. . . high frequency trading is embroiled in an arms race. And arms races are negative sum games. The arms in this case are not tanks and jets, but computer chips and throughput. But like any arms race, the result is a cycle of spending which leaves everyone in the same relative position, only poorer.

While Bookstaber, who quite literally wrote the book on financial technology and quantitative trading strategies, was emphasising computer chips rather than code, the sentiment is easily transferable. Banks spend millions developing the types of proprietary trading programmes Serge Aleynikov stands accused of stealing as well as the actual hardware that enables them.

In fact, high frequency trading has boomed in recent years, to reportedly account for an estimated 70 per cent of average daily trading volume. The technology essentially involves automated tick-by-tick, high turnover buying and selling, which relies heavily on the speed and sophistication of the network and computer systems involved. These are lightning fast and voluminous trades to take advantage of minute and fleeting changes in prices. Speed is obviously essential — as is beating the competition to an arbitrage opportunity.

It’s probably (hopefully) not quite front-running, but it’s not far off in terms of speed.

In fact, to give you an idea of just what kinds of times we’re talking about, jck at Alea has dug up the following, showing order latency from BATS — the time it takes to execute a trade. BATS is the fastest of the alternative trading platforms, though the NYSE recently cut its own order latency to five milliseconds.

--------------------------------------------------------------------
How high-frequency trading works
(Marketplace; 14 Jul 09)

One reason for the success of financial institutions like Goldman Sachs is the emphasis on high-frequency trading. Jill Barshay explores how fast technology can provide a cutting edge in the marketplace.

Steve Chiotakis: Later this morning Goldman Sachs is expected to release a big profit report. Even while other big banks are still struggling. One reason Goldman is succeeding so mightily is its emphasis on something called high-frequency trading. Jill Barshay reports.

Jill Barshay: In the old days, traders hit a computer key to buy or sell stocks.

Sang Lee is the managing partner of the Aite Group. He says that's old school.

Sang Lee: By the time I push a button to trade, it's not inconceivable that a machine could have already sent, you know, 2,000 orders out there.

There are two parts to high-frequency trading: the PhDs who write trading programs and the high speed computers that execute the trades. The faster your computer, the quicker you can get to a buyer or a seller, and the more money you make.

Lee says these computers dominate stock market activity and play a vital part.

Lee: Machine driven trading has literally become some defacto liquidity provider in the marketplace. Which is a very important role to have.

High-frequency trading may make the market more efficient, but it's not necessarily fair. If you can't afford a team of PhDs and the most up-to-date computers, you may find giants like Goldman Sachs beating you at almost every trade.