Quote from CyborgTrading: You can't automate gut feelings, and this is why traders will always have an important role in the decision making process. Having worked for IT of investment banks, the gut feelings of most people is not worth automating. Algorithmic trading is rapidly replacing "traders" who are expensive, error-prone, slow, etc. Almost all traders lose or break even. Some who "win" have lousy stats, are an accident waiting to happen, are short term lucky, make small amounts, etc. This is why I think gray-box tools have such an important role in the future of trading. You have the best of both worlds: The execution speed and accuracy of a machine and the discretion/adaptability of the human trader. Proof or evidence would be a lot more convincing than opinion. For example, do you sell your website services, because you are not a lucratively profitable trader? Nothing personal, but human traders are long on belief and short on evidence. I just have a strong bias against "gut feelings."
Here's some proof for you, there are a lot more examples on our youtube page: http://www.youtube.com/watch?v=ngwpcJy7CTA&feature=channel_page This is one of our market making algorithms that the trader launches at their discretion. I'm not opposed to black-box systems, designed by a Quant. I just think that making traders an integral part of designing systems/monitoring them is important. You can't put everything into a computer model. There will always be variables that weren't accounted for or that cannot be understood by a machine.
Quote from CyborgTrading: Here's some proof for you, there are a lot more examples on our youtube page: http://www.youtube.com/watch?v=ngwpcJy7CTA&feature=channel_page This is one of our market making algorithms that the trader launches at their discretion. If you have some success, then bully for you. But examples are not proof. I'm not opposed to black-box systems, designed by a Quant. I just think that making traders an integral part of designing systems/monitoring them is important. You can't put everything into a computer model. There will always be variables that weren't accounted for or that cannot be understood by a machine. Again, most of what gut feelers and human traders do, is lose money, make mistakes, etc. etc. There is a host of trader wananbes such as here on ET, who are poster children for why human traders should almost always not be allowed to touch real money. Black box systems are not necessarily the ultimate, but if something really does work and is quantifiable into algorithmic trading logic, then letting a person come in and interrupt its trading logic rarely makes it better. As I said, I saw quite a few seats on the IB's exchange floor that emptied out and were seated with IT and other company personnel. The world is moving rapidly into programmed logic and decisionmaking. The flash trading recently quarreled over in the news was an example. Most trading is being done by computers, not by people. That percentage grows every day. The number of trading floors that have shut their pits and gone electronic is loud testimony for where everything is going. Almost every human "trader" the average person will meet such as on ET, is not profitable, prone to gross mistakes and sudden large losses. I would put the % of longterm, lucrative traders at considerably lower than 1%. Just the turnover on ET from 5 years ago (user IDs) is very much in agreement. A few make money and everyone else dreams. As I have said before. 100% of the new traders believe they will be in the 5% of "winning" traders (more likely, 1% or so)
I think though it is important to distinguish between what most retail traders (like ET's population) do, which is speculation, and what traders on the floors of investment banks do. Most of the money that IBs make trading is not from speculation, but from making markets. They might speculate on the side but there are usually strict limits on that. So it makes perfect sense to automate a human market maker especially in simple products like vanilla FX, then you can replace 10 traders (and their bonuses) with 1 computer or something like that. For the retail speculator, they can have an automated system that can speculate for them, or they can use gut feel/intuition, or some combo. But it is unlikely that a retail type trader can competitively make markets.
TraderZones I completely agree with basically everything that you said. For those who are still smashing keys trying to beat the curve the situation is becoming increasingly gloomy. That is why we provide tools allowing traders to automate their winning Ideals. Adding automation to discretion allows traders to execute with speed, accuracy, and scale, while maintaining the instincts and adaptability that has been the downfall of so many black boxes. Cyborg Trading is not in itself a winning strategy. If you automate bad principals you are not much better off. As a rule traders should always be wary of pre-packaged trading strategies, if you can just press a button and automatically make money, why would it be for sale?! The demand for tools that increase a trader efficiency, however, is increasing and the tools themselves are improving because the needs of the discretionary trader are just as organic as the markets. Cyborg Trading gives discretionary traders the means to fight back against institutional volume. The direction of the market is undeniable and it is the reason that tools like Cyborg exist. I think that the only way that our opinions differ is that we believe that there is still an integral role that the human mind can play in successful trading.
I generally agree with your post. But we more differ in that I see little evidence that the vast majority of people can or will ever have the ability to make longterm profits with good stats. IF a person somehow manages to find an edge, then: 1) they are wise to automate it as much as possible 2) I will add, that even an automated system should be watched. I observed a lot of "autotrading" on some of the websites that offer "systems." And the feedback on autotrading (piping your or someone else's logic to the broker), was trades not being filled, getting out of sync, unexpected results and occasional blowups. So I think the role of a person is more to research ideas, vindicate them through extensive walk-forward/sim trading, ensure they have good trading stats (PF, Sharpe, DD, etc.), and automate anything they can. But the trader should still be around to monitor the results, in case of unexpected activity of the automata.
Hey TZ, STFU! Read the players d*ckhead, read the players! PF, Sharpe, DD.....WTF! Keep draggin' the welfare assclown! Another Bollinger butt boy trying to explain how to trade,......Yeh right!
TraderZones, In the end, if you can fully automate a strategy...thats great...I think that is the end goal for sure...who wouldn't want a program to sit there and print money all day....and yea if you find an edge, automation will allow you to get RICH from it...rather then just make a few bucks manually exploiting it. But a program will only do what it is programmed to do. Machines can't create their own strategies...and this is the reason that traders need to be an integral part of the process....even if they're simply there to help recognize a new pattern, inefficiency or assist the machine. So basically...I think we agree on everything
TraderZones, you seriously need some sample posts for your usual rants, at least for the "lousy stats, waiting for blow up momentary winners" and for the "commissions, spread, slippage" negative expectancy blah blah... It would save a lot of time to work on the floor of your IB... Cyborg, your products seems interesting. Grey box definitely have some value.