At night, the Bunds, Gilts, DAX and JGBs (the latter on the TSE, not LIFFE as there is very poor liquidity) are good for day trading programs.
-$344 today......... not a good day, but not overly horrible either.. its the first real loss for the bot in a week..... still not over my screwup from the other day which is keeping the bot underwater right now...
Frost - we're all intrigued by your bot, so could you provide a bit more detail....like the number of trades for the day, and the long or short positions taken ?
syswizard, When I get some time I'll generate some backtests to get those statistics.... from what I remember is the avg num of trades is around 4..... The proportion of long/short trades is fairly even.... Over the past 4 months only a very SLIGHT edge in total profit from long trades.... but havent looked at those statistics in a while, short trades may actually have the edge in total money earned lately now.... The bot has just liquidated for the day. +$85 today.... fairly uneventful day...the past week or so the bot has been pretty hot... although you would never know that from looking at the totals because of my screw up of-$744 early in teh week... remove that number from the equation and the bot woudl be fairly positive overall I belive.. will get those numbers put together as soon as I get some time to go back and add it up..
Actually My totals are a little off since the bot has been live, after 26 trading days the bot is currently at: -$648 including my -$744 screw up I made. -$104 if you assume the -$200 the bot was at when I could have easily liquidated after the screwup that day... Either way after 26 trading days the bot is still in the red...... but all in all,especially past week or so the bot is returning to more normal operation....... not the performance i was expecting when I started this journal... but considering its made around 100 trades in this time period thats not THAT horrible... Hopefully next few weeks continue to be good and it can get back into the green
Frost, I was curious about something: for your entries and exits, do you use market or limit orders? Merry Christmas, RoughTrader
Frost - hate to keep bothering you with questions, but from your backtesting, how does the 26 days compare to the AVERAGE drawdown length of time ? Also, do you know the standard deviation of this length ? I'm just wondering that if you get to 2+ standard deviations from the average, it won't be looking too good....
Good question RoughTrader. Since frosty says he has almost no slippage, I suspect he uses stop orders. market orders would be about a tick per round turn. Limit orders would be an average one tick per side due to unfilled orders.Am I right?
For my system that is currently in development, I am using limit orders to get into a position and market orders to get out of a position (because according to my rules, I need to get out no matter what).
Actually, stop orders are the same thing as market-if-touched (MIT) orders. They simply convert a predefined level into a market order when the level is touched. Limit orders mean no slippage if executed as such, but you're not always guaranteed a fill. This can cause synchronization problems for automated strategies. I think what you're referring to is MIT orders. Basically, a limit order converted to market if not filled the instant price touches the limit level. RoughTrader