Hi Wide Tailz, Wondering if you have implemented above in your live trading and if yes how is the live equity curve coming up? Thanks
Hello! On balance, the long only, stock algos are basically flat (some are up some are down), the SPY algo is mildly up, and my discretionary trading is way up. I have decided to pour my heart into tech analysis based discretionary trading from here on out (I have found my Answer to the trading mystery), while waiting for more blind forward test data from the bots. I have not given the bots any real money, but the SPY one will probably be the first to go live. It has been doing the best and is also the simplest. It's just a moving average envelope with finely tuned settingz... Rodney King was right on the money - my OOS tests were not totally blind and you only get one WFO attempt before contaminating the resultz. But so far I'm happy with the algos. They're messy but they work. Very crude.
Ze latest action ....... Whipped around a bit and flat for the past few months, but doing what it was designed to. This bot makes huge money when the market gets wild, and goes thru mild DD's when things are more quiet. Moving average envelopes.
Hi Wide Tailz, From reading this forum topic it sounds like you and I are attempting to do the same thing - Except you are doing it for your retirement and I am doing it for my family. And it sounds like we're taking a similar'ish approach. I thought it might be wise to get in touch and maybe we can help each other with what seems like a common goal ? I saw your forum posts a few day ago when I was searching for info on Portfolio Maestro. I have a simple trend following strategy that works very well with certain stocks (JRCC, FSLR and other big trenders). Your comment a while ago saying "some stocks are just too noisy to be tradable".. is exactly the conclusion I came to - after a great deal of time spent studying charts in TC2000. I've tried out a few other strategies but most are not giving the results I'm looking for. At least so far.. Poor rewards or bad drawdown.. or they only work well with selective optimizing.. So far the simplest trend follower is the only one I would go forward with. At the moment. I've only been programming easyLanguage for a week so I'm still using rather crude methods to configure the strategies to do exactly what I had in mind. But they work enough to prove or disprove approaches. I'm now putting more effort into finding the right stocks rather than the world beating strategy. And I'm not adverse to the idea of manually marking up the charts with trend lines etc to get it to do what I want.. Without me having to sit here and click buy/sell all day. I have several things on my TODO list. 1. Scanning methods to find more stocks like JRCC, FSLR etc. The stocks with the best trends and least noise. My girlfriend is a statistician and now that I'm able to save out the minute interval data from TradeStation she is going to analyse it with R. 2. Adding trendline support to enable/disable strategies at key price areas. 3. Adding trendline support to generate trades as price crosses trendlines. Probably using colour coding on the lines to configure the actions. And very short term moving averages to approximate price.. Both of these are mainly to take the donkey work out of discretionary trading. (and the stress) 3. Automatic trendline markup. Taking the basic built-in Tradestation "Trendlines Automatic" indicator and improving it. Perhaps with ideas from here (http://markplex.com/free-tutorials/tutorial-4/). 4. Bayesian stats (and other) methods.. Aimed at detecting character changes in stocks by detecting changes in their statistical modelling parameters. To notice when a stock enters a new phase - which may mark the beginning of a new trend. Again, with the help of my girlfriend. I'm UK based, Edinburgh - working full time on this stuff. Get in touch if you want to collaborate, I tried to send you a PM but it said I was forbidden - I'm new to this forum so that's probably why. You have some pretty excellent looking equity curves in this thread and I wondered if they turned out to be generally successful, or if they were the result of over-optimisation. Or some other problem. I'm trying to find strategies that can be relied on to make the maximum returns with the lowest possibility of draw down. In your recent posts to this thread you say you are only running one algorithm on SPY - I'm very surprised by this. Or maybe I misunderstood. Why not run on other symbols ? I think the SPY has a pretty poor risk to reward ratio. Short trends and lots of noise. I'm curious about the chart you posted today using Moving Average Envelope. Is that the built in indicator ? How are you using it ? I think it might be similar to how my favourite strategy operates. (buying and selling based on gradients of short timeframe moving averages.) I can post the code if you like If anyone else here is interested in collaborating then feel free to contact me. Kind Regards, -Jason
Greetings! Thanks for sharing your thoughts and I'm sure if you post a few more times, your PM privileges will kick in. I have a different background from most system traders (mechanical / aerospace engineering) and it may be why I'm doing things they normally wouldn't. If you are afraid of curve fitting, my optimization efforts were probably a bit over the top for what you may consider valid. The process I used was (from what little info I found) fringe industry standard: make the program, optimize it, test it out of sample, then dump it all into a walk forward tester. Finally, let it run forward for enough time to confirm that the equity curve characteristics are still there in the future, and keep re-optimizing it on a set schedule. Note what is missing: testing one program on several different stocks / futures / markets. I don't believe in this. I think it is an invalid assumption to believe that if a system works on several markets then it must also work the future. Working in the future is what is important, and has nothing to do with working on several unrelated instruments. My 'over optimized' algos mostly passed this process, but some of the more filtered stock buyers are still in forward test because they don't trade as often as the SPY system. In my opinion, the only rule that matters is that the algo must make money in the future in the actual market. You can try to go with the simplest thing that works if it helps you believe that it won't break down in the future...... or you can go with something that you have proven to work in the future, which was made to trade the particular instrument for which it is applied. And yes there are professional fund managers using trend following algos of this type..... The way I see it, why use a pair of channel lock pliers when you could get much better results with a torque wrench and set of sockets?
Thats a bit disappointing. I was hoping you had more success with devising other trading strategies - I was hoping for something that does quite well in a range bound environment - to save me the effort of manually drawing trend lines and marking up charts for discretionary trading.. (or semi discretionary with the computer monitoring the markup on the charts.) What I find amazing is that what I'm building isnt already part of TradeStation. Or maybe I just havnt stumbled across it yet.. I need to put more effort into developing some alternatives to just the trend follower I have. Since it relies on finding the right stocks to fit it in order to work. And that has its own difficulties. And I dont really trust any backtest on a strategy that doesnt produce at least a few trades a month - since very few stocks worth looking at have trends that extend that far without backtracking on themselves to a lesser or greater extent.. Unless you only look at strategies that relate to much longer term moving averages such as the 100 or 200 days.. or 200 week... But what validity can they have if they cant be properly tested.. Plus if the logic of the strategy is to be of any use then it has to be capable of being proven on historical charts. If it cant be applied under a given set of circumstances then it doesnt have validity - or it cant give the required confidence and confirmation that it can be applied to the future data with any sense of optimism.. or certainty The markets may have changed in recent years due to HFT machines etc.. But they have not changed so much that nothing that worked in the past is still applicable. We just havent been able to describe the rules and method thoroughly enough.. I agree with what you say about the only test being future prices.. But I do think that certain rules are followed repeatedly.. and finding ways to fit those patterns is what is useful. I will get on with my studies. Thank you for your help Kind Regards, -Jason
Yes, I'm not after the grail, but an insect that can survive and bring me the $$$ over time. I found one that I particularly liked, mutated it, then set the group of them loose on suitable environments so that they would tend to balance each other out. Portfolio Maestro proved that the theory of firing on many cylinders does indeed make a smoother power delivery....... My life's passion isn't trading systems, unfortunately. I'm a lot like a Russian engineer - designer of very simple, crude, functional solutions that no one would show off at an expo.