Electronic Trading Caused Meltdown: NYSE CEO (not a trader error)

Source: CNBC

The cause of Thursday’s nearly 1,000-point stock market meltdown was electronic trading, not a “fat finger”, New York Stock Exchange Euronext CEO Duncan Niederauer told CNBC Friday.

Niederauer was referring to reports on Thursday that a trader error caused the market's sharp swing by typing a "b" for billion instead of an "m" for million.

“The rest of the market has to respect us and not trade through us,” Niederauer said, referring to the “60 to 90 second” period when stocks such as Procter & Gamble and Accenture , fluctuated violently.

“Everybody has to be concerned about this technology. How fast is too fast? When is enough is enough? That’s what we’re trying to figure out,” Niederauer said.

Niederauer did not choose to get into a feud with Nasdaq CEO Bob Greifeld, who blamed the NYSE for the error. “Let’s stop the fingerpointing and move the ball forward. We’re not walking away from this but let’s be constructive.”


Read more: http://www.cnbc.com/id/37017292

Hmmm. Casino...or NYSE/NASDAQ...either way, mobsters are going to end up with your money, so you might as well have a good time while you're getting plucked.

we would never have had a financial meltdown.

It seems something in addition to electronic trading caused this problem.

We just had months of an HFT driven meltup based on positive sentiment created by massive financial bailouts, government stimulus and imaginary "green shoots". Now we get to see what happens when sentiment shifts back toward the other direction and all of that deleveraging and the erosion of liquidity starts to work against the interests of our infinite-growth based political and economic models.

And I even stayed in a Holiday Inn

has to be a Holiday Inn Express!

:)

Stock market sentiment is usually a contrary indicator, however. Historically the market tends to go up when sentiment is most negative, and down when it is most positive. That said, DU would be a very poor place to measure market sentiment, IMO, because the mood here ALWAYS trends negative.

Bottom line: When making investment decisions, it's a good idea to ignore anyone who claims to know what the market will do in the short-term. Focus on things like asset allocation, investment expenses, and market fundamentals.

What you're saying assumes everybody is using the same trading algorithm, which certainly is not the case. It seems to me there was a catalyst that set this off. Trading programs may have then reacted to that, but I am still not convinced of the cause.

The basic principles behind these programs is much the same across the board, though of course exact implementations vary enormously. One is to repeatedly offer the same share up for sale at an ever reducing price until a buyer bites and then fill that order. Rinse, repeat, or wait a little for the price to percolate through and start again. Flip side offer an ever increasing price starting at a low floor and wait for a seller. Either way do it at incredibly high speed, with the added bonus of an advance peek at how the market is responding a few seconds before the common people get to see.

The programs are also self evolving, which mean not even the programers themselves know exactly how their creations make their decisions any longer. The programs are capable of using any data their translation layer can tease into a useful form. And they will have their own web spiders out there constantly seeking out more.

This could be caused by an unforseen interaction between a sellbot and buybot. Or it could be something as ridiculous as a lost COBOL stop record (at least one key field all '8's) that somehow got interpreted as a sell order for 888888888 shares at 0 cents autocorrected to a minimal allowable price of 1c.


High Frequecncy Trading should be outlawed, as should the advance peek at the market the big boys get. All the advantage falls to a handful of the biggest traders.

One trade per stock per 5 seconds or so (per trader/client) should be as fast as it gets. As fast as a good human broker might be able to manage.


And the "marketplace" itself should be a black box that simply matches buyers to sellers. Each simply inputs stock name & type, min & max volumes and their min & max unit prices. Equal min/max unit price transactions are "advertised" in the end of tick data dump if unfulfilled at the end of the tick. Perhaps include a time limit.

Every x seconds the sell orders are processed (lowest to highest price) amd matched against buy orders according to strict, and universally applied rules. Price advantage always goes to the buyer (something about the idea of the buyer willing to pay the most gets to actually buy at the minimum the seller is willing to accept seems particularly appealing). In the event of ties where the total buy volume exceeds sell volume each buyer gets equal shares (up to their max). Once no further overlaps can be found, or all orders of one type (buy/sell) are cleared the next batch of orders is dropped into the "pit" and on the tick the next round begins. At the end of each round a data dump consisting of "advertised" prices for unfulfilled orders and all individual transactions reported in the simplest possible form x paid for y shares of z, but without identifying who paid who.

Oh and the black box only accepts (e)cash. No options. Any and all promisory notes are between the buyer and his financial institutions and other third parties.

To be honsest, I know naff all about the stockmarket, (just what incidentally falls in as I read about subjects which reference the market) but something in me says that rules along those lines would serve to reduce volatility by shifting much of the risk to the seller, but at the same time flattening out big sales to many customers and to level the playing field giving exactly equal access, to exactly the same data AND trading oportunities, to both mom & pop retirement portfolios and Goldman Sachs.

during periods of high volatility. Program trading now accounts for 2/3rds of the volume in the markets. When these programs decide it is time to sell, there is no one left to take the other side of the trade so these types of sudden drops are exactly what critics of high frequency trading have predicted would occur. Many observers noted that over the past year an ever increasing number of hft programs were merely chasing momentum, using an easy-money strategy which all but guaranteed sell-side panic when employed en masse, such as it was.

It's herd mentality and when the herd changes direction, it can lead to sudden drops, stampedes even.

The past year has seen a reinflation wave of the previous bubble, and now we're starting to see the next deflation wave. Lower highs and lower lows all the way down. It may take years to reach a bottom.

Thanks!

'“Everybody has to be concerned about this technology. How fast is too fast? When is enough is enough? That’s what we’re trying to figure out,” Niederauer said.'

Sounds like there's a need for government regulation.

keep that in mind as you circle down the toilet bowl of history.

yesterday?