At the most basic level, it is often said that Fear and Greed drive The Market. Now that X% of trading is done by machines, are Fear and Greed still the duo through whose actions All can be understood?
I hope some people chime in on this thread. I think it could be interesting. My 2 cents. Macro economics, trading instrument marketing/perception, and randomness are explanatory factors. Fear/Greed enter into macro economics by supply and demand. I don't need to explain to this audience why and how. Fear/Greed can be used to market by the unscrupulous, but there are other buttons to push as well. Fear of missing out on the next big thing. Does that really fall under Fear or is that Greed? It certainly relates to ego and the phantom pain of missing opportunities. The contingent nature of market participation creates an instability. The macro economics and market participants create stability. We have all seen how abruptly a paper airplane can change directions due to its lack of rudder. It makes sense that an instrument, detached from its underlying would also go crazy. Professional Traders, thus, do server an important function.
Lets see.... When I first started trading at Schonfeld...fractions were used. Then the mistake of decimals thus killing spreads and making the market more "Efficent". Early 2000, HFT and Quants start to develope Computures to trade. Thus, taking most of the human emotion out of their trades. Flashforward to present.. close to half the volume (which is very low right now) is controlled by non-emotional computures. The markets are far more efficent than when the move to quote in decimals happened. Bottom line, Retail traders are a dying breed unless they have a "Machine", like my old friend Nitro. If you sit and bang at the key board, thinking we are back in the SOES days you'r sitting on your death bed or making less than an average person at a Call Center. Keep in mind the "Flash Crash" was not caused by Fear or Greed. It was caused by automated "Support" numbers that were broke, thus automated sell orders that were absent any emotion kicked in. Enough Said. The Fear and Emotion for individual Retail traders may still be there but the overall "Fear and Emotion" of the collective in the markets is being drained. The one bright side for all the struggling traders is that "Humans" program the Computures. Humans must adjust the "Models" based on data. So, there is still a fraction of "Fear and Greed".
On another recent thread ... http://www.elitetrader.com/vb/showthread.php?s=&threadid=213076 ... Landis82 wrote ... ... Are these "magnified movements" just the super high-speed expression of fear and greed of the trading machines' programmers/strategy architects? (In which case, perhaps price action can still be looked at with many of the interpretative tools of old ...) ... Or are fear and greed dying as market forces? (so that we [or more accurately, our trading bots!] need new ways of interpreting price action)
It's still humans who make the programmes. And they have one goal. To make money, and of course they fear losing money.
As long as Humans are still involved in the Markets, as Quants programing and Model making based on the persuit of profit....and "Fear" of loosing, Human emotions will be part of the equation. That being said, Fear and Greed will not be a good gage as real time movement. Maybe longer term moves....maybe "Directional" confirmation. However, real time sale and buy orders based on real time Emotions is not much of the equation anymore.
Take the following two, oversimplified examples as bases for a comparison. A = = = = Imagine a world with no trading machines (ah! What nostalgia! The good old days!), and imagine a breakout. Suddenly price breaks from its prior trading range. The skilled/lucky traders anticipate this. For many of the rest (the unskilled/unlucky), the emotions run high, and they are afraid (fear) to be missing out on the dayâs Big Move (greed), and so they jump in after the break out (against their own better judgement, i.e. they will kick themselves afterwards for having done it again ...), just in time to take over the positions of the skilled/lucky traders as the move dies (and then reverses). B = === Now imagine a world where only trading machines battle it out. Are the machines going to be âafraidâ when price breaks out? Are the machines going to act like the unskilled, emotional traders who (out of fear and greed) bought just before the move died? Are machines ever going to act against their better judgement? If not, then subsequent price action in Case B would differ from subsequent price action in Case A, wouldnât it? i.e. In Case B, there are no "suckers" to take the skilled traders out of their positions before the move dies ...
I like your analogy and I think that in short-term movements price discovery is very much like this. Especially in FX. I can't believe there are retail traders who actually try to trade FX (or anything) on a second-by-second basis. I sometimes try and it never really works well. That is indeed like trying to predict the direction of a paper airplane. Over the long run, however, the impact of computerized trading is highly overrated around here (and in general). Do you really believe that computers decide the direction of the market? Get real. For directional approaches human decisions are still made. The algos are mainly to be thought of as cute little helpers (like elfs, or goblins depending on yr view of algos) who make sure the positions are acquired at favorable price levels once those decisions are made and especially to do so without moving the market too much. The huge volume in HFT is more focused on predicting the paper airplane (Which apparantly they have found a method for doing) than the long-run. I'm not sure if you pay attention to what those quants are doing but after 2008 they pretty much unanimously said "we were wrong" (I find it interesting that before most major crashes commercial short interest in index futures went up dramatically whereas before 2008 they were massively long). And they will never be right, unless you can predict human behaviour in general with mathematics which I don't think has been done. I don't understand why they still pay these people a salary in finance.
The "machines" don't have a better way of knowing beforehand whether it would be a good idea to fade a move or not, but they'll be the fastest to the table when it becomes obvious it is. Also not all breakouts should be faded like yr example suggests. There will always be suckers, even if it's efficient machines vs inefficient ones.
Most of the volume in algorithmic trading, I believe, is a matter of patterns and relative valuation. Pattern: for whatever reason, a certain setup will elicit a predictable enough response that front-running is possible. Relative valuation: a run-up of the QQQQ without a run-up of MSFT won't remain so play with the odds. That's fine, but it doesn't explain, high prices of Chinese stocks, copper, Facebook, gold, real estate, Madoff investments. (Ordered, IMHO, from Macro-Economically based to Perceptionally based.) The CFTC COT data show that sometimes there are groups which take a definite directional position. When extreme, these can be leading indicators.