A really slow MA on a long timeframe. Open up a futures index chart and throw on a 300 period WMA. Color the bars based on slope. Awesome, right? Huge swings worth thousands on a single contract. Too bad you will get chopped out so often (a holy grail system can have losses; it just has to make money). Back in my indicator days I tried many variations of this, different MAs, different time periods, none were profitable (reversing them wasn't profitable, either; wrap your head around that). I tried things besides MAs, too. Bid/ask graphs, random market sentiment indicator type things, a bunch of stuff I coded up myself. They all basically gave you the same info. I kept thinking everyone else was trying too hard. Just use a fairly slow chart (5 minutes, 30 minutes, whatever) and ride the waves. One problem is that if the indicator is slow enough, you often won't exit a position anywhere near the maximum profit level. But you can't afford to exit too soon, like with a fixed profit target, either, because then you'll leave a lot on the table. Position scaling and/or trailing stops are not the solution to this problem. It's one of those things that works great sometimes, but you won't know beforehand if it will work great or not (trend or chop). Too bad it doesn't work. It'd be a great way to save on commissions. A few trades each week. Sometimes more, sometimes less. Oh well.
I've always thought it would be cool if the holy grail was just a high 50% winning system trading intraday 20-30 times per month with both profit targets, stops, and daily and portfolio monthly level targets. Oh, and make sure it actually is an arbitraugable model that can be genetically optimized.
Does that describe the system you're trying to sell for $8.5M? And whether it does or not, how did you decide on $8.5M as an asking price? Where do you market such a pricy item?
Run the MA and scale out at some fraction of the ATR or RV (lookback) of your average holding period. 1/3 at 1SD; 1/3 at the ATR; 1/3 at 2xATR.
No. 8.5M is 75% win percentage, 2.5-3RR, 2-3 trades per month, and scalability. What was described in my post is $250 per $25k block per month, earns about 10% per month as I said, so for 1% of the account per month, do I need to ask if there isn't a value proposition or something going on here? The $250 does use the for sale system, but as far as that goes, the system above has been the primary success of the holy grail below.
Good to see you back, Beau. Autumngold doesn't have any of your numbers beyond May12. How have you done since?
All of those calculations would probably give the model enough wiggle room to find a profitable result, and even the "1/3" could be a parameter between 0 and infinity, but I guarantee it won't be infinity.
At 278% up in May for 2012, I had been waiting a semi-annual calculation so I could focus on trading myself. Then PFG, the percentages became irrelevant as a proprietary account so I gave the codes to Robbins and they've been trading since August. The sites I reported to haven't seen any results since because it's personal and not proprietary capital, and I was tracking both the proprietary trading results where we use multiple units, and letting world cup track the base unit. Now I'm letting world cup represent the base unit, and the capital at PFG's been frozen for a long time. What piddly amount I get after making $100k+ in a month just left me enough to keep trading my QuantMaster program. You asked about my results on AG, and since those trading results were representative of my CTA's proprietary trading, once the capital was frozen, I was left with just my personal funds, and not wanting to start over those programs terminated. The losses exceed 75% even if they want to make me feel good that they did give me 30% back since the loans defaulted and even though they were without recourse I'm almost oblivious to it compared to how easy it has been for me to make profits since handing off the trading algos.