Have you checked out <a href="http://www.wealth-lab.com/" target="new_win">Wealth-Lab</a>? Though the code (ChartScripts) is in Pascal/Delphi for the trading systems, its not too difficult to understand. You can test most of the available systems via the website, then convert anything interesting to TS code. Be on the lookout for code that peeks ahead / cheats as anybody can upload stuff there. Smitty
Here's the outer limits on ESmini Just pencil your efficiency in along with your perception of the market's efficiency...LOL Change the ESmini prformance h/L too if you wish. You can use this on any yellow brick road as the starting potential to find out where you might be able to trade. It sure doesn't take much to get in a nice ball park.
oops one cell (h17 had a factor (*h13/100) missing, use this. Some other with same thing too . Sorry.
yea, but to get those "income streams" you gotta decipher 2000 pages of Jack-erese crap (which very few ever managed to do
I think purely systematic trading is very hard, because it is inevitable you will experience a drawdown in the future greater than the maximum shown in historical testing. In the real world, what do you do when this happens? It would be very hard to stick with the system when this happens, because you will be asking yourself is this just a drawdown or has the market changed? The basis of my trading is a selection of mechanical systems, but I put a discretionary filter over the top of it. The reason being that systems are good for modelling patterns and the like, but can't take into consideration other things such as news, economic announcements, political situation, sentiment etc. I am constantly tweaking my systems, incorporating new ones and dropping off ones that haven't been working too well. Combined with my discretionary analysis as the final filter on when to do a trade, I am attempting to stay ahead of the curve and adapt to the current market. I would probably make money just sticking to the system religiously (albeit with larger drawdowns), but I would be nowhere near as confident of long term longevity than with my current approach of constant adaption + market reading experience. For example, the current war situation has had an impact on my systems. Instead of just sticking my head in the sand and trading through it, I have saved a fair bit of angst and dollars by cutting back trading size, and focussing on shorter time frames than I usually do. With regards to trading ideas, no one is going to give you their tried and true models. I made a post a few weeks ago outlining typical CTA systems. These probably aren't viable for an individual trader unless you have a fairly large account. If you are developing systems for a single market, I would definitely look at intermarket patterns. Hope this helps.
oh yeah but arguably that's just a more striking name for the first ("fooled by curve-fitting") phase. well, one way of looking at it is to fathom that there's a collective "other side" to your trades, meaning that the system will work only as long as the "other side" can afford the losses. here's what i do: i start with the assumption that the lifetime of my system is going to be finite, and i downsize whenever i hit a drawdown. this way i'm effectively giving away a chunk of profits as an insurance that i won't be losing it all once the edge falters. - jaan
here's what i do: i start with the assumption that the lifetime of my system is going to be finite, and i downsize whenever i hit a drawdown. this way i'm effectively giving away a chunk of profits as an insurance that i won't be losing it all once the edge falters. best approach i ever heard. peace
Have you considered doing it as it approaches a wash? How about before it approaches a wash? How about when you have a breakout on a channel? How about when the trend meets resistance and forms a formation top? Etc?
Would you look at the equations and tell me how to improve them. They are very rough and just give the compounding potential. I would like to change how the efficiency connects. As a first pass all I did was use it a a multiplier of the total potential. Do you see a way to use it to connect it to profits and to cycles? My first thought was to put the user effeciency with profits and the market efficiency with cycles (exponent). I think that by checking the various fractals you find out where to trade. There are human limits so you chuck those. Like the 1 min would have to be mechanical and computerized. By adjusting the 5 min to the ratio of trades ordinarily taken compared to every period available, I think that helps define stuff. Putting DELL on the sheet as an equity is interesting.