flash crash and algo trading

모두 잘 알고 게시겠지만 2010년 5월 6일 미국 시장에서는 Flash Crash 라고 부르는 사건이 발생했습니다. 약 5분만에 다우지수가 600포인트 넘게 빠졌다가 다시 올라온 사건입니다. 그리고 9월 30일 SEC가 사건의 추이 및 원인에 대한 보고서를 제출했습니다. 보고서는 http://www.sec.gov/news/studies/2010/marketevents-report.pdf 에서 보실 수 있습니다.

보고서에서는 사건 자체는 시장의 취약성에 의해 발생했지만 그 등락 속도는 알고 매매와 관련있다고 언급합니다.
본 글에서는 어떤 식으로 알고 매매와 flash crash가 관련되어 있는지 보고서의 executive summary 부분을 살펴보겠습니다.

일단 사태가 어떤 식으로 진행되었는지를 보겠습니다.

… By 2:30 p.m., the S&P 500 volatility index (“VIX”) was up 22.5 percent from the opening level, yields of ten-year Treasuries fell as investors engaged in a “flight to quality,” and selling pressure had pushed the Dow Jones Industrial Average (“DJIA”) down about 2.5%. Furthermore, buy-side liquidity in the E-Mini S&P 500 futures contracts (the “E-Mini”), as well as the S&P 500 SPDR exchange traded fund (“SPY”), the two most active stock index instruments traded in electronic futures and equity markets, had fallen from the early-morning level of nearly $6 billion dollars to $2.65 billion (representing a 55% decline) for the E-Mini and from the early-morning level of about $275 million to $220 million (a 20% decline) for SPY Some individual stocks also suffered from a decline their liquidity.

At 2:32 p.m., against this backdrop of unusually high volatility and thinning liquidity, a large fundamental5 trader (a mutual fund complex) initiated a sell program to sell a total of 75,000 E-Mini contracts (valued at approximately $4.1 billion) as a hedge to an existing equity position.

… This large fundamental trader chose to execute this sell program via an automated execution algorithm (“Sell Algorithm”) that was programmed to feed orders into the June 2010 E-Mini market to target an execution rate set to 9% of the trading volume calculated over the previous minute, but without regard to price or time.

… on May 6, when markets were already under stress, the Sell Algorithm chosen by the large trader to only target trading volume, and neither price nor time, executed the sell program extremely rapidly in just 20 minutes.

유럽사태와 유로화 급락으로 변동성이 급등하고 주요 선물 및 ETF의 유동성이 급격히 빠진 상황에서 4.1B의 매도 주문이 알고 주문으로 나왔습니다. 브로커리지 알고 전략은 여러가지가 있지만 기본적으로는 VWAP과 같이 거래량을 어느 정도 따라가게 합니다.

HFTs and intermediaries were the likely buyers of the initial batch of orders submitted by the Sell Algorithm, and, as a result, these buyers built up temporary long positions. Specifically, HFTs accumulated a net long position of about 3,300 contracts.

기본적으로 HFT 전략은 market making 전략, 즉 시장에 유동성을 공급하는 전략입니다. 따라서 초기 crash 초기에는 이들이 물량을 받아 주었습니다.

However, between 2:41 p.m. and 2:44 p.m., HFTs aggressively sold about 2,000 E-Mini contracts in order to reduce their temporary long positions. At the same time, HFTs traded nearly 140,000 E-Mini contracts or over 33% of the total trading volume. This is consistent with the HFTs’ typical practice of trading a very large number of contracts, but not accumulating an aggregate inventory beyond three to four thousand contracts in either direction.

그러나 변동성이 증가하면서 HFT algo market-maker들이 손절하고 포지션을 정리하기 시작했습니다. 여기에선 나오지 않는 이야기이지만 algo market-making 초기에는 꽤 리스크를 크게 감수했다고 합니다. 하지만 2008년 사태를 겪으면서 algo market-maker 들이 큰 손실을 입고 보수적으로 운영을 시작했다고 합니다. (automated trader 기사 참조) 따라서 stop—loss 시 포지션을 정리하는 속도도 빨라졌겠죠.

The Sell Algorithm used by the large trader responded to the increased volume by increasing the rate at which it was feeding the orders into the market, even though orders that it already sent to the market were arguably not yet fully absorbed by fundamental buyers or cross-market arbitrageurs. In fact, especially in times of significant volatility, high trading volume is not necessarily a reliable indicator of market liquidity. What happened next is best described in terms of two liquidity crises – one at the broad index level in the E-Mini, the other with respect to individual stocks.

The combined selling pressure from the Sell Algorithm, HFTs and other traders drove the price of the E-Mini down approximately 3% in just four minutes from the beginning of 2:41 p.m. through the end of 2:44 p.m. During this same time cross-market arbitrageurs who did buy the E-Mini, simultaneously sold equivalent amounts in the equities markets, driving the price of SPY also down approximately 3%.

Still lacking sufficient demand from fundamental buyers or cross-market arbitrageurs, HFTs began to quickly buy and then resell contracts to each other – generating a “hot-potato” volume effect as the same positions were rapidly passed back and forth. Between 2:45:13 and 2:45:27, HFTs traded over 27,000 contracts, which accounted for about 49 percent of the total trading volume, while buying only about 200 additional contracts net

그러자 가격 하락 속도가 빨라지고 손절 기준이 다른 HFT algo market-maker 끼리 포지션을 돌리는 폭탄 돌리기 현상까지 발생하면서 거래량이 급증합니다. 거래량을 따라가는 브로커리지 알고 주문의 주문량도 덩달아 증가하며 양의 피드백이 걸립니다.

Between 2:32 p.m. and 2:45 p.m., as prices of the E-Mini rapidly declined, the Sell Algorithm sold about 35,000 E-Mini contracts (valued at approximately $1.9 billion) of the 75,000 intended. During the same time, all fundamental sellers combined sold more than 80,000 contracts net, while all fundamental buyers bought only about 50,000 contracts net, for a net fundamental imbalance of 30,000 contracts. This level of net selling by fundamental sellers is about 15 times larger compared to the same 13-minute interval during the previous three days, while this level of net buying by the fundamental buyers is about 10 times larger compared to the same time period during the previous three days.

HFT 들이 제공하던 유동성이므로 빠지는 속도도 빠르고 일관성(?)이 있습니다. 어떤 속도로 물량이 빠졌는지는 아래 그래프를 보시면 알 수 있을 것입니다.

201289814.png?AWSAccessKeyId=0ZRYP5X5F6FSMBCCSE82&Expires=1291195323&Signature=tbpqrhz01bo7U3dSDqxTqjIG5Ow%3D

At 2:45:28 p.m., trading on the E-Mini was paused for five seconds when the Chicago Mercantile Exchange (“CME”) Stop Logic Functionality was triggered in order to prevent a cascade of further price declines. In that short period of time, sell-side pressure in the E-Mini was partly alleviated and buy-side interest increased. When trading resumed at 2:45:33 p.m., prices stabilized and shortly thereafter, the E-Mini began to recover, followed by the SPY. The Sell Algorithm continued to execute the sell program until about 2:51 p.m. as the prices were rapidly rising in both the E-Mini and SPY.

사태는 CME의 시장Stop에 의해 진정됩니다. 가격이 반등하는 와중에도 Sell 브로커리지 알고는 매도를 계속합니다. ^^;


여기까지는 보고서 요약부의 일부분입니다. 자세한 내용이 이어지므로 관심이 있으신 분들은 보고서 원문을 살펴보시기 바랍니다.

알고 매매 이야기를 하면서 flash crash를 말씀드리는 이유는 지금까지의 설명에서 보신바와 같이 flash crash의 진행에 있어서 알고 매매의 두 가지 큰 카테고리 - market making 와 brokerage algo order 가 큰 역할을 했다는 사실을 이야기 하기 위해서입니다. 자료들에 의하면 이미 미국시장에서는 전체 주식 거래량의 80% 이상이 이 두가지 알고 매매에 의한 것이라고 합니다.

현재까지 우리나라 시장에서는 brokerage algo order가 큰 역할을 하지 못합니다만 언제까지 이러한 상황이 계속될지는 두고봐야 할 것 같습니다.