I had a look at the breakthrough system of yours, the "SuperBands with linear regresion analysis" and it has shocking -49.38% maximum drawdown !!!!!!!!!!!!!! :eek: I wouldn't trade such system even with "your money". Oh my God, your system will take your equity down to a half, and you're claiming to be someone that knows what he's talking about. I guess if I'd question your comments then it would make me "the most uninformed or simply arrogant trader" as you have pointed out, so I'll rather keep my comments to myself.
Hello to all, and thank you very much for all your advice! Mr. Consistent suggested that I start a blog where I can start talking about my newbie process. I've already written two articles describing how I got started here: http://startuptrader.wordpress.com I am inviting all experienced, profitable traders to jump in and offer your advice, as you have time, and all beginners like myself to read and ask questions. Hopefully this blog will help the newbies of trading go from booksmart to demosmart!
I've sent you a PM, we first need to discuss some parameters, because I decided to stay away from a situation where everyone will jump in and ask questions. I'm not at the spiritual level of people like Mother Teresa who would help every Dick, Tom and Harry. I came to the conclusion that private blog should be the way to go, let's leave others here on Elite trader forum, here they can amuse themselves, no need to invite anyone
Yeah, they did beat the S&P by several hundred thousand basis points now. None of them may be the holy grail, but they are as close to it as you can get without making your own system. I haven't seen you on there in awhile. I guess you moved on.
It's funny that you think I endorse it. Why else is it freely available then? It's a timeless experiment, but one that has managed to beat all of the ten thousand plus systems ever to be published on WL. If you ever get a chance, it looks pretty good on the NAZ100 with 1% of sizing. The drawdown is from 10% per trade up to 100% position sizing, which, obviously is something I don't recommend. That is to say, 100% in a single stock is not recommended, but it can be done with ETF's as you essentially hold 100 securities or more at once. Since you probably aren't well versed in understanding these statistics on WL, I'll save you the trouble. SuperBands, with Linear Regression Analysis since 1/1/1996 on the NAZ100 Long + Short Starting Capital $100,000.00 Ending Capital $1,102,590.63 Net Profit $1,002,590.63 Net Profit % 1002.59% Compared to the S&P's profits of 37%, this is oustanding, and would beat BRKA over that time period as well. Annualized Gain % 19.81% Say whatever you want. Good luck finding a system with this long of a backtest and that high of an APR Exposure 8.89% Cash Interest $127,905.48 Margin Loan Interest ($5,561.16) Total Commission ($231,451.11) 0.01 per share Number of Trades 29,285 high statistical significance Avg Profit/Loss $34.24 Avg Profit/Loss % 0.75% Avg Bars Held 1 Winning Trades 16,396 Winning % 55.99% Gross Profit $2,503,521.50 Avg Profit $152.69 Avg Profit % 4.42% Avg Bars Held 1 Max Consecutive 169 Losing Trades 12,889 Losing % 44.01% Gross Loss ($1,623,275.20) Avg Loss ($125.94) Avg Loss % -3.91% Avg Bars Held 1 Max Consecutive 225 Max Drawdown ($153,103.25) Max Drawdown Date 11/20/2008 Max Drawdown % -23.87% Max Drawdown % Date 8/31/1998 Wealth-Lab Score 169.5453 RAR 222.7115 Profit Factor 1.5423 Recovery Factor 6.5485 Payoff Ratio 1.1307 Sharpe Ratio 1.2275 This is probably where experience tells you something. Put about this way. If this is the most profitable system ever to be published, then it is the benchmark against which you should compare yourself. This tells me a long term sharpe above 1 is very good for systems. Dghost Annual Volatility % 15.8595 Excess Return 19.4676 Ulcer Index 5.5967 WL Error Term 13.491 WL Reward Ratio 1.4683 Luck Coefficient 7.5119 This is low for a system so old. Pessimistic Rate of Return 1.4147 Equity Drop Ratio 0.2617 A very simple system. Buy in 0.25% increment starting at 5.5% below the 10,1.5 bollingerband if the linear predicted value is greater than the band. Sell on the open the next day. Maybe it's not so simple, but it's done perfectly well for 12 years, compared to the S&P's NP of 37% in that time. I feel like my other systems are better on a risk adjusted basis. Anyway, if you can get a system that looks this good "and can actually be traded" you'll be rich. That's the only goal really, is to be rich. This one has so many technical challenges that it's just plain easier to time the market than sitting at a computer all day. Swing trading daily systems have made me more money in two years than in three years of my start percentage wise. So definitely need to have a system. You just won't know your risk at all without one. Again, that's the best backtest people can produce in WL and still be verifiable. I'm sure there's others reading this that can throw out their systems measures, but we won't actually know if they aren't just saying if priceclose(bar+1)>priceopen(Bar+1) then buyatmarket(bar+1,'');
So, dipshit, then why did you answer?? And congrats, you've created 2 money losing systems. Brilliant! Don't quit your day job kid.
Watch the market a lot (and all inputs into the market), then notice patterns linking the inputs to future price behaviour. Form a hypothesis using observation and pattern-recognition ability/creativity, then test it. If it seems to work, try it with real money, using adequate and robust risk control. Rinse and repeat.