A friend of mine is a top trading system developer. I asked him for his thoughts on high-speed trading, and he responded as follows:
“Without really diving into the entire issue with both feet, my random comments would not do justice to this topic. I write code, not books, as you know. I might say that Goldman Sachs (GS) is not the only player in the box. If Sergey Aleynikov [a former Goldman employee accused of stealing trading software] could get code out of Goldman, there are probably others. I’d hate to be on the (GS) receiving end of a really dedicated counter-predatory ‘market harvester.’ I doubt that the markets will become the war zone with collateral damage with the magnitude that [some believe]. [Hedge fund] LTCM had Nobel laureates on board and they blew up.
“High frequency trading (HFT) is the most profitable category on Wall Street right now, however. News is still moving the markets, and machine-readable news can be hooked up to algos. By the way: GS does not have our algo unless they are stealing patents. We make this algo [commercially] available to ALL traders, and we are not alone. Circuit breakers will probably increase in use and regulation will probably increase.
“Regulators are not real happy with GS, in my humble opinion. The Bloomberg-tapping trader in Bermuda shorts trading FOREX is simply being replaced by the shorts-wearing FOREX trader running his own personal algo. More sudden market moves have happened due to fat-fingered errors than algos. The leverage unwind set upon us by us helping people get into houses that they could not afford has hurt us more than GS’s algos. We can do much by voting in leaders that have common sense. Who really wanted to ‘Equalize Housing’ in an unequal world?
“My advice: Work hard, act smart, be careful and if in doubt…stand aside. Merton-1936 Law of Unintended Consequences #1 reason….Ignorance. #2 reason: Error. Nothing has changed, we just have ‘evolved.’ Back to the GP code now….”
Goldman’s plunge tore through a major midpoint support without evincing even a hint of a bounce. I now expect the stock to fall to at least 145.23, possibly pausing at 147.80, the midpoint sibling of the target. _______ UPDATE (12:54 p.m. EDT): Goldman’s overnight low was 147.81 – tradable, perhaps, since it occurred just an hour before the NYSE opening.
I’d flagged a midpoint support at 152.80 as a place to try and get long, but you should do so only via “camouflage,” buying on the first ABC uptrend that occurs on the very lesser charts after 152.80 is approached within 5-10 cents.
It’s been a while since we looked in on this stock, which appears headed for 164.05, an important Hidden Pivot off the daily chart. Traders should look to do their buying using camouflage on a small abc reversal uptrend from near 152.80, the midpoint of the relevant C-D leg.
We haven’t done anything in Goldman in a while, but a moderate rally today could provide a good opportunity to get short with a tight stop-loss. Officially we’ll offer 200 shares at 136.83, stop 137.01, but if you want to use puts instead, try the August 125s. You should pay no more than 2.80 for them, however, and the stop-loss will still apply. _______ UPDATE (11:11 a.m. EDT): A short from 136.92 was stopped out minutes after entry for a tiny trading loss, although one could have partially covered the trade as low as 136.51. That’s how low GS dipped initially after opening on a $1+ gap. The actual high was 137.17, implying that higher highs lie ahead, but for now the stock has relapsed to a so-far low of 134.76. While it may seem as though our stop-loss was too tight, the pattern was sufficiently precise to justify the one we used.
It’s been a while since we pondered this limping lump of brick dust, but it’s not a happy sign for bulls that the stock has slipped so easily beneath the Hidden Pivot support shown in the chart. What it suggests is that Goldman will fall anew, this time to Hidden Pivot support at 122.25. Camouflage shorts are encouraged from these levels, but if you’ve got the patience to wait, bottom-fishing at 122.25 could provide the kind of low-risk opportunity that we thrive on.
Longer-term charts suggest Goldman will grope its way down to at least 115.61 this summer, but it’s too early to tell whether a lesser but even nastier pattern in gestation at this moment will wreak the maximum damage of which it is capable — i.e., a pounding down to 94.19. It’s shown in the accompanying chart, and the uncertainty yet remaining concerns whether there is a pop yet to come that will invalidate the existing point ‘C’.
I’m not buying any of this — literally — since it is coming off a 133.81 low that lies almost $2 beneath the lowest low I could have projected using the hourly chart. That suggests sellers will soon resume control, even if it would only take a print today at 144.31 to temporarily cede the advantage to the madmen. Incidentally, a pullback from just above 144.31 could provide camouflage for traders who want to jump on the stock for a presumptive short ride north.
We’ve been using a downside target of 135.35, but I am now recommending bottom-fishing if and when Goldman falls to another Hidden Pivot at 137.69. I’ll update with more detailed instructions following a print within $1 of that target, so check back if you’re interested. The target would be hit, ideally, following a slow grind lower, but if it comes in conjunction with a stock market avalanche, you should back away. _______ UPDATE (2:40 p.m. EDT): The stock has traded down to 137.55 so far. For your guidance I will establish a tracking position of 400 shares. A 137.44 stop-loss is advised for now, and you should take profits on half if GS trades up to 138.69. _______ FURTHER UPDATE: The stock finally bottomed at 136.60, stopping us out for a $100 trading loss – but also warning that, despite the nearly $2 bounce from the lows, sellers are not yet done with Goldman for the moment.
GS – Goldman Sachs (Last:136.89)
by Rick Ackerman on June 16, 2010 2:57 am GMT
We talked of shorting Goldman for a possible ride down to 122.25, and this rally might provide a good opportunity to do it. It projects to 138.71, so let’s offer 200 shares short at 138.65, stop 138.95, day order. You can substitute options, but the stop-loss will still apply. ______ UPDATE: Before diving $1.50, Goldman topped at 138.26 — not quite high enough to get us short. Cancel the trade, but keep in mind that the very bearish target remains in force if you want to take the initiative.