Tagged: Flash Crash RSS

  • Steve Eberbach 12:55 AM on May 17, 2010 Permalink
    Tags: Flash Crash   

    If “flash” does not refer to the practice of studying “flash” order flow to the likes of JPM, Goldman (who allegedly does over 60% of the trading on NYSE, I would attribute the sudden loss of bids to the more modern type of technical analysis and execution called “high frequency trading”. That means trading much more often, in smaller lots, for tiny profits. Like, the old saying: “If you can’t make your numbers on margin, than make it up on Volume”.

    Much of that fast trading is called “latency arbitrage”, or profiting on tiny differences in price predictable if your data feed is faster than somebody else’s, and your orders can be placed on the exchange faster than somebody else’s. It helps it you have no fundamental need to buy or sell other than to play the tiny fibrillations of price. I call it “Trader Tetris”. It’s like those old computer games where you “take out” blocks of whatever impedes your progress to the higher score.

    One way to arbitrage latency is: rather than get a technical analysis “signal”, then place your order, then go to the back of the line at the exchange and wait for your order to be filled, you put bids and offers on the book above and below the NBBO inside bids and offers, and with an exchange co-located trading server, you cancel whichever side is opposite the side you want to trade. Since you created the hole before other traders’ orders were “up to bat”, you have made a short term impact by removing support or resistance before others can even see it or respond to it. The price can gap right through the hole on low volume. You can collect the market maker rebate as a bonus even if you happen perchance to trade with your own dark pool subsidiary at break even. By canceling one of two opposites instead of placing a one sided order, your latency is reduced because you do NOT have to wait for your standing orders to be canceled! The remaining standing orders are also like the tetris blocks with which you stop your competitor from bidding or offering quickly before you do. Voila! Now you see them, now you don’t! The bids and offers can disappear instantly as a “reverse order flow”.

    I think people got used to the liquidity provided by these high frequency traders, and when the foreign exchange markets got too crazy, the high-volume fast stock traders just pulled their standing market maker liquidity and there was a sort of “vacuum” suddenly. That would leave a huge hole in the depth of market and set off an algorithmic chain reaction, particularly if it happened so fast that standing stop loss orders placed below the expected trading range for those unlikely “black swan” events got hit for no sound technical or fundamental reason other than a misjudgment of the impact of automated programmed trading under unforeseen and untested circumstances.

    I also think that the instability was exacerbated by a lack of normal investment buying order flow compared to the currently normal (lately of greater proportion due to ‘HFT’ and latency arbitrage than previously normal) volume of in and out trading order flow, probably combined with lack of supervision of trades by human traders, who certainly have more common sense than computers which are issuing orders as fast as they can, even faster than the exchanges can digest them.

    I am not a very experienced trader; I come from a systems engineering background. I happened to have recorded (and saved without refreshing the data) that flash crash tick by tick (Tradestation actually kept up very nicely, except for one minute or so near the climax) To me as an engineer, the price action looked like a classic system overload saturation event, or “clipping” distortion due to lack of dynamic amplitude handling capacity, and resulting loss of the usual negative feedback stabilizing efficiency normally existing in the markets.

    If there are any conspiracies here, I will guess that they might soon come to light in the government hallways when lawmakers try to find new rules to better “regulate” the newer (“unfair”) computer trading technology and the resulting opportunities for deploying new trading strategies which can scoop the markets. If you ask me, real conspiracies are what we watch competing for the Stanley Cup in the playoffs lately. I’m from Michigan, I was rooting for the Detroit Red Wings.

    Ah, we also watch the thrilling bareback rides of unbridled free enterprises and their “robber barons”!
    When the biggest government-funded investment banks boasted publicly of their quarterly “perfect” runs of no days with trading losses, I thought to myself: “Pride goeth before the fall”……

    Personally, I vote for prohibiting abuse of excessive leverage and abuse of “fake” liquidity (courtesy of TARP, even) that is so transient and practically unusable that it has no real benefit, or even is hazardous to the physical economy, and then diverting the evacuation and cleanup bill from the perpetrators to the innocent taxpayers when the reactor core melts down from thermonuclear runaway like Chernobyl.

    But it could already be too late to stop that now. I learned THAT back when I “minored” in Economics in college back in the early ’60s, and after then got my MBA at U. of Mich. They came up with the Glass Steagall act(s) only 30 years before my college days when the horses had already escaped from the barn, so “it” would not as likely happen again. (deflationary collapse due to over-leverage and (risk) abuse of bank depositors’ funds, among other things) and “The Rest of the Story”.

     
  • Teresa Lo 1:42 AM on May 16, 2010 Permalink
    Tags: conspiracy, , Flash Crash,   

    Flash Crash: Is Technical Analysis the Smoking Gun? 

    Sometime in the 1990s, Bill Gates achieved his goal of putting a computer on every desk. Since then, i.) academic studies have noted that the edge once attributed to indicator-based technical trading rules has practically disappeared, and ii.) it’s become obvious that most professional managers do the same thing as everyone else.

    Much has been written about the so-called Flash Crash. From the start, my thesis has been that there is no conspiracy. If anything, it was the proverbial packed nightclub that caught fire. Since Friday’s revelation of the timeline of trades executed by the investment firm Waddell & Reed, we may finally have the smoking gun:

    Gary Gensler, chairman of the Commodity Futures Trading Commission, said in congressional testimony Tuesday that regulators were focusing on one particular trader in the market for E-mini futures as part of the commission’s investigation into the flash crash.

    Gensler said the trader in question entered the market at around 2:32 p.m. ET on May 6 and finished trading by around 2:51 p.m. ET. He said this trader and others had executed hedging strategies of similar size previously.

    The Tipping Points

    What is the smoking gun, you ask? I say it was the execution of stop loss orders amassed at levels based on popular technical trading rules:

    1. Investment guru William J. O’Neil wrote “in his second book 24 Essential Lessons for Investment Success, “find it gut-wrenching and hard to admit” they were wrong when a stock loses money, but they must overcome that emotion and sell anyway – an essential move if a stock has lost between 7 and 8 per cent from your purchase price.”
    2. Popular financial blogger and portfolio manager Mebane Faber’s timing model “mechanically buys (sells) an index when it crosses above (below) its 10-month simple moving average”, according to CXO Advisory. SSRN shows that Faber’s paper has attracted 138,520 abstract views, 51,516 downloads and is number 4 in download rank of all time. His book, The Ivy Portfolio: How to Invest Like the Top Endowments and Avoid Bear Markets, features the 10-month moving average model. It’s Amazon.com Sales Rank is #19,207 in Books.
    3. Stan Weinstein’s classic, Secrets For Profiting in Bull and Bear Markets has advised using the 30-week moving average for timing for decades.
    4. Finally, there is the ubiquitous 200-day moving average that is on plotted on every chart, a line Louise Yamada once defended the use of based on “observations noted over time.”

    (More …)

     
    • Lyle 6:32 PM on May 16, 2010 Permalink | Log in to Reply

      Teresa, I am surprised no one has commented. You make it sound so easy, but I KNOW it is not. I want to thank you for being the one, like in an old Hollywood movie, wearing the white hat. Or as I think of it a bit more; think of the last Hollywood Die Hard remake. You Know “The Guy,” that makes things safer for anyone listening. You can’t help everyone just those that are listening. Thank you for being too nice.

  • Mick 10:03 PM on May 14, 2010 Permalink
    Tags: Flash Crash   

    The media has finally grasped the concept that there was no “fat finger trade” nor “Million/Billion mixup” that caused the extreme volatility in last Thursday’s market (May 6th, 2010).

    What really happened, of course, was that different systems of “circuit breakers” on the NYSE and other, electronic exchanges, buckled under the weight of massive sell orders. One must understand that nothing happens in a vacuum these days. If selling occurs in one place, it will generally cause selling in another place because of index arbitrage or other relationships between stocks or futures. Consider this example: e-mini S&P’s plummet, so an index arb (or even a computer programmed to do index arb) buys them at a significant discount to fair value. To actually capture that discount, the stocks that comprise the underlying index have to be sold. Thus, cheap e-mini’s cause selling in (nearly) all $SPX stocks. This is effectively what happened in the Crash of ‘87 as well: portfolio insurers sold S&P futures well below fair value, and index arbs bought them and sold stocks. Eventually, the discounts got so big that the index arbs ran out of capital (or were shut down by their risk managers), and the discounts persisted for days. The market was already down a lot last Thursday when some extremely heavy sell orders were entered. Whether they were in the e-mini S&P futures or in Procter & Gamble really doesn’t matter much, since selling in one place looks like a discounting/arbitrage opportunity to computers programmed to look for such things.

    The NYSE took a “breather” to let some humans look at the sell orders, and that created the problem that we saw as a 500+ point drop and recovery in the Dow in a matter of minutes. When the NYSE shut off its electronic systems (for 85 seconds, they say), orders were automatically shuttled off to other electronic exchanges. In many issues, these other exchanges were completely unprepared for this order flow. There were few bids. So, as you’ve heard on the news, some stocks plunged all the way to 0.01 in price. Meanwhile, some high frequency traders (who also provide a great deal of liquidity) shut down, either because they were already losing a lot of money, or because they feared they were about to.

    This also brings up another matter, which is that one can’t always assume that liquidity will exist, no matter what the situation is, but that’s really a matter for another time…

     
    • Teresa Lo 2:46 AM on May 15, 2010 Permalink | Log in to Reply

      A few weeks ago, I told my daughter that the market cap of a company is essentially zero in a panic there are no bids. We simply cannot take the current bid and multiply it by all of the shares issued and outstanding to come to the “value” of the company. LOL.

      Years ago, one of my friends at the office was really pissed off at a deal. He called up the guy and screamed, “I’m going to hammer your bids down to zero,” which is exactly what happened last Thursday afternoon. This sort of thing happens in penny stocks all the time. When the sellers come out and the bids start getting weak, there is a Minsky Moment when you have to decide because if you don’t sell now, there are literally no bids below. But selling also means taking out the bids and the whole thing becomes a vicious circle, hence the reason we must sell longs when the going is good.

      In 1987, the cascade was pinned on portfolio insurance [DOWNLOAD PDF] while this episode may have something to do with dynamic hedging, which was — wait for it — the brainchild of one Nassim Taleb. See ETFs and the ‘flash crash’.

      Would be ironic if Taleb was his own Black Swan, sort of like I’m My Own Grandpa.

    • Teresa Lo 3:04 AM on May 15, 2010 Permalink | Log in to Reply

      NY Times has an article out on it now.

    • Mike Sinder 12:30 AM on May 16, 2010 Permalink | Log in to Reply

      I wasn’t around to see it, an interesting video showing the ‘cascade and some excited individual calling. Some pretty interesting spreads and price movement. The link:

      http://www.youtube.com/watch?v=9OWFaGNN5hU

c
compose new post
j
next post/next comment
k
previous post/previous comment
r
reply
e
edit
o
show/hide comments
t
go to top
l
go to login
h
show/hide help
esc
cancel