Subscribe to our RSS feed
EconLife.com connects economics to everyday life, current events and history.

Tag Archives: flash crash

stock market...ticker...16564_3.29_000011656552XSmall

High frequency trading might be an elephant in the closet.

Representing perhaps 51% of all daily trading volume, speedy trading hits the headlines only when something goes wrong. But since I’ve been reading Frank Partnoy’s Wait: The Art and Science of Delay, and particularly enjoyed his chapter on the topic, I wanted to share some of what I learned.

Let’s start by imagining 25 milliseconds. Whereas Warren Buffett’s Berkshire Hathaway has owned Coca-Cola stock for more than 25 years, a high frequency trader might own Coca-Cola or a security that relates to it (called a derivative but let’s not think about that now) for 25 milliseconds.

Here is a simplified version of how it works:

Listening to its algorithm, the computer identifies a flood of buy orders for a certain security. It just has to zoom in and purchase those shares. The trick though is doing it before anyone else does. And then, within maybe 25 milliseconds, it has to sell before anyone else does to all of those buyers it had previously identified. The goal? Make a penny or less for each share. Added together? Many thousands of dollars and trades.

Fundamentally, high frequency traders are looking for market volatility. For currencies, futures, options, bonds, when markets are gyrating up or down, price anomalies occur. A stock, for example, might suddenly be cheaper than expected or there might be that deluge of buy or sell orders. At that moment, the super speedy computer swoops in to take advantage and make a penny or 2 for each trade.

Speed is crucial because the very act of buying and selling changes the price. If everyone tries to buy at a lower price, together they push the price up. To beat the increase, you have to be the first one. Similarly, if everyone sells simultaneously, the price of the security declines. Again, to make your pennies, you have to lead the pack. You have to have the fastest computer.

And that takes us to location. You not only need speedy computers but also, you need to be close to stock markets’ servers. Seven years ago, one California firm saw its profits slipping. Slicing several milliseconds off trading time, a move to NY solved the problem.

A New York location, though, was only the beginning of a race. Not cheap, computer “co-location” is now sold by stock exchanges. Nasdaq meanwhile offers a high speed data package and feed for $25,000 a month.

Our bottom line? High speed trading has created a list of concerns. Mistakes are exaggerated. Front running a deluge of buy orders might be harming smaller retail investors. Human wisdom is absent just when you might need it. But still, high speed trading could indeed be a huge source of liquidity–of buyers and sellers–just when stock markets need them.

So, what to do? Is high speed trading the elephant in the closet that requires more attention?

Sources and Resources: For more discussion of high speed trading and the sources of many of my facts, I recommend this NY Times article and Frank Parnoy’s Wait.  You might also enjoy this recent Bloomberg discussion of the 2010 “Flash Crash” and this CBS 60 Minutes segment on high frequency trading.

A hat tip to Josh Goldberg!

Posted by: adminEcon
Tags: , , , ,
Comments (0) Add a Comment

Soon after the New York Stock Exchange (NYSE) opened yesterday, the prices of more than 140 securities moved irrationally. Some were up 150%. Though no one yet has said why, we do know that fast trading computer programs—high-frequency trades—were the source.

In his book, Wait: The Art and Science of Delay, Frank Partnoy, tells us why we had a problem. Partnoy, a former investment banker and current professor of law and finance says that, “…in most situations we should take more time than we do. The longer we can wait, the better.”

In a chapter on “Superfast Sports,” he explains how Jimmy Connors and Chris Evert were unusually fast at physically responding to a speeding tennis ball. Their rapid physical response gave them more time to think beforehand.

Reading Partnoy’s chapter on high-frequency trading, I thought about stock market history. Since Wall Street’s first traders gathered under a buttonwood tree to buy and sell securities, the time to think has decreased. The onset of the telegraph, the ticker tape, and the telephone sped and spread the pace of transactions. Then, during the 1990s, computers began to handle more orders.

You can see where we are going. Technology accelerates the speed of orders and diminishes “thinking time.”

Today, we have reached the point where people need not interact directly when they buy and sell securities. Representing 70% of US stock trades, high-frequency trading transactions among computers occur in milliseconds. Partnoy suggests that even in a millisecond world of demand and supply, as with Connors and Evert, slight delays could benefit all participants.

And that returns us to yesterday’s “glitch.” Is delay a part of the solution?

This Reuters article and the Washington Post discuss high-frequency trading problems.

Please note that content was edited after posting.

Posted by: adminEcon
Tags: , ,
Comments (0) Add a Comment

15836_8.21_000007073004XSmall

Do you remember the ‘flash crash‘? On Thursday, May 6, for close to 20 minutes, markets everywhere wildly fluctuated.  It began at 2:23 when certain stock prices started moving oddly. With Apple trading at approximately $250 a share, at 2:44, the stock plunged $23 while at another moment one share was selling for $100,000. At 2:47, Accenture PLC shares dove to 1 cent from $40 a share. During that day, the Dow Jones Industrial Average opened at 10,862, dropped to 9869.62, and then closed at 10,520. Some say it felt like a roller coaster.

Financial markets are not supposed to feel like roller coasters. Instead, at the NY Stock Exchange, for example, different “specialists” oversee the buying and selling of different firms’ stocks. Historically, the specialists’ job was to maintain an “orderly market” by buying or selling the stocks themselves when price was not gradually moving up or down. So, if everyone wanted to sell a stock and there was no one to buy it, the specialist (in theory) stepped in to buy it temporarily so that price could change smoothly. 

Now though, with computerized trading, worldwide markets selling the same companies’ stocks and bonds, and a group of firms called “quants‘ that speed trade based on complex computer models, it appears to be impossible for one group to maintain an orderly market. Yes, much of the time, markets tend toward rational ups and downs. However, during the ‘flash crash’, they did not.

The Economic Lesson

Why should we care?

Accenture says it all. If one stock, for no apparent reason, can drop from $40 to 1 cent in seconds, then investors will be less willing to allocate their savings to stocks. However, our market economy needs dependable financial markets for savers and businesses. We need to invest in order to save for college, for retirement, for emergencies. Correspondingly, businesses need investors’ money as start-ups, when they expand, and for everyday operations.

Knowing that the continued possibility of a ‘flash crash’ diminishes investor confidence, a final report from the SEC and Commodities Futures Trading Commission should be completed during the next several months.

Please note that for Accenture PLC and other firms that experienced an erratic stock fluctation on May 6, those trades were canceled.

 

Posted by: adminEcon
Tags: , , ,
Comments (0) Add a Comment

15643_3.29_000011656552XSmall

No one has figured out the cause of the “flash crash”. Some background information might help, though, when we try to understand what happened.

  1. On May 6th, the Dow Jones Industrial Average plunged close to 1000 points, then made up some of its loss, and closed down 348. Most of the drop happened from 2:40 to 3:00.
  2. Historically, the New York Stock Exchange (NYSE) was the epicenter of buying and selling stock. Yes, there were other stock markets but the NYSE was king.
  3. Now, the NYSE is only one of many stock markets in which NYSE listed companies’ stock trade each day. For example, before 2003, Procter & Gamble shares were traded almost exclusively through the NYSE. Now, as shown on an Economist chart titled “Flash Crash Mish Mash,” Nasdaq, ArcaEx, Direct Edge, and BATS, and others also trade NYSE listed equities.
  4. Simplifying considerably, we can say that not only are stocks individually traded, but also “packages” of stocks are traded, and beyond that, people trade stocks that are based on stocks. And beyond all of this, not only are people buying and selling, but also, trying to make pennies on each transaction, we have computers doing very speedy trades with other computers.  In fact, these high frequency trades now represent more than 60% of daily stock trading.
  5. Fundamentally, stock prices fall when the number of shares people (and computers) want to sell exceeds the amount people want to buy. If there is an avalanche of sell orders, the NYSE has slow down rules that are designed to let it maintain an orderly market. The other markets have (or don’t have) their own rules and no one necessarily has the same rules.

Can you see why no one knows why the Dow dove suddenly? And yet, to maintain investor confidence, regulators need to prevent it from happening again. 

The Economic Lesson

The first rule of “Investing 101″ is, “Maintain an orderly market.” Orderly means that upward or downward price changes should unfold in steady increments; it means that trading will be halted whenever the buy or sell side becomes unmanageable. For several centuries “specialists” in trading posts on the floor of the NYSE were charged with maintaining orderly markets in the securities when they matched buyers and sellers. On May 6th, with no new news about the company, consulting firm Accenture’s stock tumbled 99% to one penny from above $40 a share and then returned to prior levels. Most of the drop occurred between 2:40 at 3:00. For Accenture, the market was not orderly.

Posted by: adminEcon
Tags: , , , ,
Comments (0) Add a Comment