A phd in physics or stats will not make you successful. Plenty of hot shot traders around here without those. Creativity and a TON of hard work will get you there. Ive taken a ton of advanced math classes in college and can honestly say I havent used any of those skills in finding trading edges. A basic understanding of probability is enough. Your on the right path... keep coming up with ideas and testing them. Its just a LOT of work. Ive personally coded up several hundred system ideas before I found any that worked. One thing I have learned is this. If its something easy to test mechanically, or somewhat obvious, its pretty much guaranteed not to work. For example... testing MA crossovers, even tweaking them to death, is something a zillion traders have already tried, so dont bother. OOOOOOOooooooooorr, your gonna have to take it to a whole new level to find an edge. peace axeman
things to consider: 1. a good system can't always be backtested accurately. e.g. you don't know if you got a fill, unless you assume slippage every time (not always the case). 2. tradestation et al are not flexible enough. think outside the box. 3. use several systems at once, and take the signal when they all confirm each other. disclaimer: i do not claim to have a successful system.
NihabaAshi You're right, it would be naiive to expect someone to dump the answer in my lap, or to expect someone to elaborate heavily on a unique system that works. But a lot of ET posters seem to enjoy helping along newcommers. Whether that's benevolence, or kharma, or they're not actually sharing anything valuable - I don't know. By the way, I've never fully understood the supposedly basic concept that sharing a working trading system kills it. I can visualize that an extremely short term system will stop working if others see it and perpetually seek ways to enter and exit faster... but I can also visualize intermediate and longer term systems that work in a self fulfilling manner via others making the same trades. axeman - was there a particular author or book that you found most useful?
vanilla2, Couple of questions. What markets are you looking at? Are you looking to daytrade, position trade or what? What kind of returns are acceptable? Suggestions. Volatility breakout systems are proven longterm winners. Typically 30-40% win/loss, 50% drawdown and consistent profits when traded across a portfolio of commodities. Most big CTA's use some form of this. TradeStation used to include a canned one called channel break or something like that. I agree that simple indicator-based systems will not work consistently. They will be killer in one market environment, then turn around and kill you in another. I do think a divergence system would work. check out the thread on MACD divergence scan. He even has code for it. For the stock indexes and particularly for daytrading, I think pattern-based systems are the way to go. I would examine the opening gap , the range breakout and the retracement patterns. Good daytrading systems are rare, and tend not to trade very often, certainly not multiple times in a day. It would be worthwhile to get ahold of Larry Connors book on VIX-based system, or just take the www.tradingmarkets.com free trial and get into it from the trading course there. There are a lot of systems detailed on the wealthlab site. I would go through them and look for ideas, then backtest them yourself. Do us a favor after you have done the backtesting and come back here and report the results, good, bad or outstanding. Good luck.
There are tons of profitable strategies out there. The problem is to define a profitable strategy. Good Luck, it's a neverending education.
Re: Advice from the experienced; I am very experienced, I almost made losing money a career and a favorite pastime. What works for me while trading ETFs is to develop my strategies using the underlying cash data. Its smooths out the noise, so for example if you trade QQQ develop your strategies using $COMPX. Rather than MA, I like breakout strategies, take a look at Tradestation's built-in strategy ChannelbrkoutLE, SE and also add close at end of day. Apply these three to a 2 minute chart of $COMPX. Use that as a starting point, the default setting of 20 is fine. The best book I can recommend is Van Tharp's book 'Trade your way to Financial freedom'. It has a corny tittle but the concept of expectancy is priceless. Good luck. Here is some food for thought: http://keplerweb.oeh.uni-linz.ac.at/trading/moneyMan.htm http://www.eis.pl/kr/ELZ/index.html
AAAintheBeltway Thanks so much. I will definitely share when I find something worthwhile. I have set out to trade either SPY, DIA or QQQ, with trades as short as an hour and as long as 2-3 days, primarily because of their relatively limited intraday movement compared with individual stocks which could wipe me out fast. Are commodities potentially better because they're poised for a long term rally and more trending patterns as the equity bear deepens and people move from intangible to tangible assets? Is there a deep discount commodity broker that I can access with a relatively small capital base?
Well, think about. If you were a stock trader, would you want 1000 other traders jumping into your alerts with a thousand shares each? It would kill your system, so why share it? I don't trade indexes, but I suspect the same fear exists there. A million shares hitting the market on a signal that is commonly followed would pretty much negate the system.