MrScalper - Yep. Thank you for the feedback regarding being long winded. The short take away for me on backtesting is that whatever I assume about fills and slippage is only that, an assumption until I try something IRL for reference.
Backtesting is of limited value, and once known is fine. One of the biggest problems can be ignoring the big picture and concentrating too much on the immediate action. It is unwise to try and capture ticks, unless you are on a level playing ground!
Ok, back again. In keeping with the feedback, I'll keep it short (er). - After an initial shock that my plan was crap and that I was not going to be making markets I started to dig in to TA and read up on systematic trading. I used Tradestation to automate some of what I developed and to manually trade along side of the systems. - It took from 1998-2003 before I ran my first automated system on the ER2, EMD and ES. In the interim period I lost money on day-trading options and futures. - I became profitable for the first time after letting a system run for a few weeks. After that I started trading along side of the systems, keeping records and occasionally posting P&L on the P&L threads. - A big shift happened for me after reading Atlas Shrugged (mostly because I was impressed with ET's Rearden Metal and wanted to understand the handle). I learned that I had a hangup with winning that I continue to work on. - I trace most of my trading issues to psychological/behavioral issues and focus on addressing those. - I intend to start discretionary trading with very small size in the coming days/week and will post results here. - Eventually I will start work on systems and automation again and will also post on that topic here as well.
In the mean time here is the kind of things I document and trade: LEAD LAG I look for correlated markets that seem to be in sync . I try to identify the leader/laggard with respect to the individual markets trading at or through a reference point. I then trade the lead lag relationship like a spread without ever actually measuring the spread and I usually only take one side. Instead of looking a a traditional spread, I compare where each market trades relative to a reference point like VWAP or the days High or Low... Here is PDF that includes a link to a link to a screen video I used to explain to trading buddies that are swing/trend traders. - Sorry PDF haters - Alex
Fully automated trading is very hard for consistency. If the people with bags of money can't crack it, then odds are neither will you or anyone else. You are correct to focus on YOU, as all your problems start and end there. Half the battle is being in the right place at the right time.
Yes. Odds are stacked against everyone including those that are successful. That said, someone will win. People with bags of money and the willingness to crack it do crack it. That said, the find it and automate it part is not hard for me but finding a system that produces results I can trust/tolerate has been hard. In my experience, many systems, markets and timeframes are required to achieve something that can compete with good discretionary trading (smooth equity curve, high risk adjusted return, small tail risk). People with the funds to diversify effectively across asset classes/exchanges/markets/strategies/timeframes etc. have an advantage and likely have a team and have also accepted that many different types of work and skills are required to pull off automated trading. It's not something most individuals are willing or able to do alone. Virtually every system I have created has gone through extended periods of terrible performance. The ones I have run have shown periods of good enough performance to be worth the risk until they weren't. Inefficiencies that make for high Sharpe approaches don't last. Some of the robust stuff that does last, like long term trend following requires users to endure very long periods of deep drawdown and the feeling that "maybe this time it's finally broken"... I'm in the short term, does not last forever and might never work again camp. I accept this. The biggest impediment and challenge is how to recognize it's over without giving back all of the profits. I have a crude approach to this but I have not yet developed a satisfactory and effective way to systematically turn systems on/off according to their performance. So, rather than expecting that that I'll develop a perfect set of systems and a system for managing them, I use system development as a way to stay in tune with current market regime/tendencies and to develop a framework and set of reference points or expectations about different methods. I may end up with something I have willingness to run and automate but that's not my primary objective. I find my discretionary trading is much better when I am testing ideas and monitoring systems concurrently. I am not looking for a holy grail or trading advice. I have confidence in my trading skills and abilities. Sometimes I am very happy with my trading results and behavior. I'm working on understanding where my outlier behaviors and risk taking come from because I can trace 80% of my aggregate $ losses and associated drama to a very small small number of bad and un-planned trades . I plan to continue chipping away at my limitations and issues and think that posting here, sharing ideas and results and receiving feedback will bring up my issues and help me to work on them. I appreciate the feedback.
You do realize that you are saying you are no different to most that try! Diversification, is really a trap and there are much better less riskier ways. In your long post there is one line that jumped out. Do you know which line..actually over 2 lines?
I do not know. I do have interest in receiving your feedback if you would like to tell me which lines jumped out. Thank you, Alex
"The biggest impediment and challenge is how to recognize it's over without giving back all of the profits." This is actually quite easy. What makes it hard is the timeframe you have chosen to trade. If you are not locking in profits, then you are not trading correctly. If you are waiting for price to hit a target, you are not trading correctly. If you wait for price to hit a stop, you are not trading correctly. You are trading correctly when your results show bigger winners than your losers, and you withdraw your profits from your trading account and keep them.
My objective for this thread is to contribute ideas for people to consider and check out if they like, to think out loud and receive feedback. So far the most meaningful feedback for me has been that the posts are long. I do not believe there is a right/wrong approach to trading without an objective to evaluate results against. The topics of bet size, expectancy, win size, loss size, trade frequency, signal accuracy and orders are all methodology dependent can ideally be considered with an objective measure in mind. I can confirm that it has been hard for me to create automated short term (intraday) systems with edges that persist, but I already said that... So to repeat: All the intraday systems I have successfully coded have had finite shelf lives of 6mo -5 years (including test and out of sample periods). It's still worth developing them because of what I learn. From experience, I expect them to eventually fail and I never know where I am in the life-cycle of a strategy so trading them is risky. If they break down soon after trading starts it's a losing proposition. If they last long enough to have profits greater than the eventual cutoff, it's a winning proposition. Either way, during periods I have been developing/testing/running systems I have had significantly better results with my discretionary trading. Have other people seen improvement in their discretionary trading and market feel as a side effect of system development? Thank you, Alex