Stops & Profit Taking I am not certain whether this has been discussed earlier in the thread. Assuming one understood entry requirements stated in the very first post, then the next step it to decide upon stops & profit targets. So firstly I would try and work out an average historical range expansion when following this specific pattern. Would it be reasonable to expect this average range to be 2 pts, 3, 4, 1.5? Let's assume that all signals are taken off the 2 min MACD as long as 10 min MACD confirms rules set out by T28 (10 min MACD <0 then sell 2 min MACD downside X, 10 min MACD >0 then buy 2 min MACD upside X). I personally think that the rules can be better defined, of course T28 made no suggestions re: spoon feeding, so it's up to an individual to refine initial info. For instance MACD distance from 0 line for a X to qualify as a valid signal or would all Xs on 2 min be considered a valid signal as long as 10 min MACD 'approves'. Therefore, I would try and establish that also. Assuming all 2 min MACD Xs are to be considered a valid entry signal, let's assume that a maximum stop is 2 points and an average NET expansion range after the X is 3.5 pts average (based on backtesting 200 Xs) basing profit target on whenever MACD crosses its MA line in the opposite direction to the time of entry. In the attached image 1 - sell & 2 - cover. So assuming after back-testing an average expansion range was 3.5 pts, next calculation is an average % of winning trades. Was it 50%, 60%, 45%. IMO only afterwards one can conclude whether this mech way would be profitable or not. If it is discretionary, then it could be profitable for T28, BE for T27 & a losing one for T26.
Mech is a method killer Romik, Mech implies the markets never change, that they are static and every situation is a clone of the last one, does anyone believe that? Not to mention the best traders develop intuition which is the subtle signals the subconscious picks up from market action but not apparent to self conscious awareness
Here is back test #2. The parameters are now hopefully closer to what the thread starter had in mind. Data set: Russell 2000 E-Mini Future (symbol ER2), 44190 2-minute OHLC bars (includes after-hours trading data) Data set dates: 09/07/2006 to 12/06/2006 (3 months when Dec 2006 contract was the front month contact) Start of trading session: 9:30 EST every business day End of trading session: 15:55 EST every business day Overnight position carry: Not allowed (open positions are closed at the end of the day) Indicators used: 1. 2-Min MACD(12, 26, 9), referred to later as MACD2. That is, 12 2-minute bars for a fast EMA, 26 2-minute bars for a slow EMA, and 9 2-minute bars for an EMA of MACD itself (which is the so called "signal line"). 2. 10-Min MACD(12, 26, 9), referred to later as MACD10. That is, 12 10-minute bars for a fast EMA, 26 10-minute bars for a slow EMA, and 9 10-minute bars for an EMA of MACD itself (which is the so called "signal line"). Trading costs assumptions: 0.10 bid/ask spread, 0 slippage (very optimistic), $2.40 commission per side. Long Entry Rule: When all four of the following conditions are met: - MACD10 is above zero - MACD10 is above its signal line. - MACD2 is above zero - MACD2 is above its signal line. Exit from Long Entry: When at least one of the four conditions above is not met. Short Entry Rule: When all four of the following conditions are met: - MACD10 is below zero - MACD10 is below its signal line. - MACD2 is below zero - MACD2 is below its signal line. Exit from Long Entry: When at least one of the four conditions above is not met. Results: 800 trades for a loss of $5030 per contract over the test period. Below is a chart for one of the days (October 17, 2006). The circled letters on the charts indicate at which time and price a position was taken: "S" for short, "L" for long, and "F" for flat. The lower two charts show MACD10 and MACD2 (see legend).
Nonlinear, I brief glance over the chart and corresponding B&S times suggest that you are misinterpreting the proposed set-up requirements. If 10 min MACD < 0, then only sell 2 min MACD downwards X If 10 min MACD > 0, then only buy 2 min MACD upwards X
Trader28, Can we talk about the MACD indicator for a second? When I read through the ET posts a lot of people talk about the macd indicator. And I believe most people like the indicator. It's pretty straight forward. However, it seems like everyone has different settings that they use: For example...just to name a few... 1) 12-26-9 2) 3-10-16 3) 9-5-10 4) 5-6-13 5) 5-8-10 6) 3-8-13 7) 5-10-34 So my question is why 12-26-9 as the settings? Why not these other settings? I guess I'm confused about the settings! Could you shed some light on the settings? Thanks snarlyjack
1. Correct: Both 10 min MACD and Signal < 0 line, 2 min MACD not a perfect cross downwards, but considering time of morning and 10 min MACD could have been traded via discretion; 2,3,4,5 Error: 2 min MACD already below 0 line, it has to be above and cross signal downwards for it to have been a valid sell; 6. Error: 10 min Signal (MA) is still below 0 line and 2 min MACD already above 0 line; 7. Correct: 10 min MACD and signal both crossed and are above 0 line and 2 min MACD crossed signal & 0 line from below 0 line upwards
Reading the very first post, it seems to me that the basic principle for entry is to wait until MACD crosses the zero line and until MACD crosses its moving average. I am not sure what "X" stands for. Look, I thought I was doing you a favor by back testing your own method. If you don't want to see the results, I'd go away -- it's your blog and your money at stake. On the other had, if you are interested, and willing to be a little more constructive, I'd be happy to test the parameters that you believe would define a profitable system.
I just don't believe in it, it wont translate the way you think because in the real world there are far too many variables, you are trying to bottle trading... nothing against you, just cant be done I use an 8 period Triangular Moving Average with 2+ horitontal shift to anticipate zero line crosses... how are you going to program that? You wont get a formula out of this, only a method you can TRADE