I think there’s a misunderstanding. I referring to the latter. Hypothetical if we are short after the 20 goes below 50 were would the anchor be? HL or HH
Wrong, the first picture the fibs are wrong. I never said the high before stochastic leave early. I don’t understand why is it so hard to draw fibs.
Rules for fibs for a buy
-
Look at the lowest point when 20 EMA was still under 50 EMA.
-
Look for highest point after the 20 EMA crossed above the 50 EMA. On one condition that high has to have occured before your entry.
-
On MT4 select your fibonacci and draw from the high to the low so that the 100 is at the high and the 0 is at the low.
They are three steps that everyone must apply blindly without thinking.
Philip, so it would be this Fib then right?: 301 Moved Permanently
From the low to the Highest High… yes? Considering that between this HH and the lower high, a buy signal was triggered?
I wrote down the following rules for drawing the Fibonacci retracement that I refer to when I enter a trade (maybe it’ll help):
[U]Setting Fibonacci Retracement[/U]
Long Trades:
Once you’ve entered a long trade you need to insert the Fibonacci retracement lines onto your chart.
• Use the low immediately preceding the EMA cross and the high after the stochastics retracement.
• Go from high to low. This will give you the following levels: 100 at the high and 0 at the low
• Make sure the levels are set correctly (i.e. 1.272, 1.618, etc.)
• Once price closes above the 1.00 level move the stop loss to break-even
• When price closes above the 1.272 level move the stop loss to the 1.00 level
• When price closes above the 1.618 level move the stop loss to the 1.272 level
• Etc.
Short Trades:
Once you’ve entered a short trade you need to insert the Fibonacci retracement lines onto your chart.
• Use the high immediately preceding the EMA cross and the low after the stochastics retracement.
• Go from low to high. This will give you the following levels: 100 at the low and 0 at the high
• Make sure the levels are set correctly (i.e. 1.272, 1.618, etc.)
• Once price closes below the 1.00 level move the stop loss to break-even
• When price closes below the 1.272 level move the stop loss to the 1.00 level
• When price closes below the 1.618 level move the stop loss to the 1.272 level
• Etc.
1 Like
Yest exactly.
Hello everyone, could someone please list all the Fibonacci retracement levels they use for this system? Also, i’m wondering how far back in time should I go when looking for highs and lows? I use Metatrader 4 if that helps. Lastly, I have attached a picture of USD/JPY 4 hour chart. On March 18th around 4:00 is when the EMA’s crossed and stoch matched it. The trade went down later after going up a bit. Would I stay in the trade until the EMA’s crossed again?
Hi Philip,
I think we should all know by now how to plot the fibs once we enter a trade, (highest/lowest point before cross of Ema’s to lowest/highest point after cross. wait for the stoch to retrace and cross over/below the 80/20 zones then enter.
There is a situation on USD/JPY at the moment where the Ema’s have crossed waiting for retracement and cross now of stoch 80 zone, which would give us our entry and could then draw fibs for our targets…
There is a candle that spiked down just before the cross that may become the low point of the trade, would you still stick with the rules here and still go for the low after the Ema cross or would you take this lower candle into consideration…
Thanks Philip
tecky89
Your stoch should be set at 14,3,1 yours is wrong, all the information about fib settings are on post #3.
[QUOTE=“techy89;690698”]Hello everyone, could someone please list all the Fibonacci retracement levels they use for this system? Also, i’m wondering how far back in time should I go when looking for highs and lows? I use Metatrader 4 if that helps. Lastly, I have attached a picture of USD/JPY 4 hour chart. On March 18th around 4:00 is when the EMA’s crossed and stoch matched it. The trade went down later after going up a bit. Would I stay in the trade until the EMA’s crossed again?[/QUOTE]
The fib levels can be found in a previous thread post, and I don’t see a trade entry on your chart. I suggest you go back to the beginning of the thread and read/reread all of Phil’s instructions to get a better understanding of this system.
Philip, good news/bad news. First the good news -> I was able to implement obtaining Weekly RSI(14) filter from 4H bar data, as we had previously discussed. Courtesy Jalapenofan, I tested the same code base that I used for Dow-30 testing (with the addition of the Weekly RSI filtering) against 4H bar EUR/USD data from 2009-2015 (through early March, 2015). Since my testing platform doesn’t natively support FX, I took trades like they were stocks, and used 10% of my starting $250,000 for each trade taken. The bad news is that the performance, albeit not a whole lot of trades during this period, was pretty dismal. The total number of trades taken during this time period (13), are attached in a text file (please convert it to a .XLS or .XSLX extension, to view in Excel).
Overall Trading Performance:
Annual Raw Returns:
For those of you who cannot access the Trades Log attachment, since the list is small, here’s a graphic of the Trade Log:
This is a very small sample set, but the data does span 5 years. Please look at the trade logs to verify the correctness of trades taken, and let me know if there are any trades that you took in 2014 or 2015 (thru early March) that didn’t make it on this trade log list.
P.S. to Jalapenofan, if you can provide me with maybe 3 other CCY pairs over the same period of time (4 hour bars) that would be great, if you have the time. Thanks!
What if without RSI filter?
Hi Doug, I am slightly lost here. Are you trading the original system or have you adjusted anything? The reason why I am asking is because of the lack of activity. I see a trade on Nov 5th and Nov 25th. i’m sure I could find a fe more…
Yes, just notice that I like to have mental stop. Only a [B]close[/B] beyond the extension levels would have me close the position. Merely touching it doesn’t count.
Firstly thanks for a fantastic effort. It is interesting that it was only profitable in 2015.
The file you sent me doesn’t open in text or excel, there is some gibberish writing there so I can’t follow the trades.
What I can do though is since they were 13 trades I can go through the chart and look manually for now. Then I’ll give you the trades I took and may be present a chart for illustration for a particular trade you want to see or something.
Reverse the RSI, when it closes above 66.6 sell and when it closes below 33.3 look for buys.
Let me explain my thinking behind the RSI and what could have happened wrong. The idea is RSI measures the avg. gains to the avg. losses over n period. When that ratio is 3:1 and higher the RSI hits the 66.6. when its 1:3 and lower it starts to hit the 33.3. by 3:1 I mean the up move was three times more than the down move over n period.
So something called range-analysis was developed by someone called Anthony Cardwell. The idea is price should find resistance in two levels of RSI; 66.6 and 80 (similarly 33.3 and 20 are two areas of support). When price enters between the 66.6-80 area (33.3-20), it is at the strongest moment of a trend.
I wanted to apply that idea on the monthly. Because we would know that price is in a long-term vicious trend. That suited the 4 HR since the avg trade length on that time frame is 14 days (there are 20 trading days a month on average.)
With applying the weekly however its possible that we got the opposite effect; Let’s say that RSI weekly closed above the 66.6. By the time that trend was translated to the 4HR exponential moving average, price had already found resistance and moved in the opposite direction.
So there are two options: either try the weekly with 1 hour so its more responsive (which would be a lot of work). Or just reverse the 66.6 and 33.3 since the weekly time frame will not reflect a strong trend anyways. Note: This is still a hypothesis, wait for me to trade the weekly chart as your test and come to a conclusion.
So may be switching the rules would be easier. I will look at the trades now and will probably backtest the monthly RSI manually.
Thanks bro.
I need someone have can actually read the trade log to have a look at this chart and confirm if the two winning trades were during the two marked periods.
If not then Dough, where the two winning trades on a period between February 1 2010-July 5 2010 and August 18 2014-March 23 2015?
If my hypothesis is correct, then (without the need to back test manually on the weekly) the two trades should have taken place in either or one of the two periods I outlined.
Furthermore, if my hypothesis on RSI monthly (and a potential relationship with volatility I said earlier I was studying. Your system would have missed the trade opportunity from February 2010- July 2010 and got the two winners in the period from August 18 2014-March 23 2015. This is because RSI monthly closed below 33.3 only during that period.
Am I correct?
If I’m wrong let me know so I can start trading the 4 HR chart from 2009 and see how I would have traded the system.
I’m sorry I’m bombarding the thread with posts but a million ideas are flying into my head.
I’m worried reversing the RSI won’t actually work. Because at one moment, the weekly RSI should align with the monthly RSI for that vicious trend and we would lose all our money.
I’m worried that it will have to be the monthly RSI.
I’ll start backtesting on all timeframes and report back I guess.
EDIT: that is assuming I’m right about the above posts lol. I’m still waiting for an answer from someone who managed to read the trade log.
Hey guys, sorry that I set off the alarm bells here. This model is still a work-in-progress, but I think it has improved with time, thanks to Philip answering my annoying questions repeatedly. Just a couple of things, hopefully answering some of the questions from the last couple of pages:
-
Not sure why BabyPips doesn’t allow you to attach .XLS files - probably worried about being responsible for being the culprit of potentially spreading viruses. That said, I rename my .XLS file to be .TXT, before attaching it. You can download the file, rename it to have a .XLS extension, and then open with Excel;
-
For those unable to do #1, I updated my original post today to include an image of the Trade Log, since there were only 13 trades;
-
The results on the EUR/USD were based on the EXACT code base that I used on my revised Dow-30 Daily test, over 32 years or so, with ONE EXCEPTION. On the Dow-30 Daily data test, I used an RSI(14) Monthly filter that took longs when Philip’s algorithm told us to, AND when RSI > 66.67(for Longs) and when RSI < 33.33 (for Shorts). On this 5-year EUR/USD 4-hour Bar data, I changed the RSI(14) filter from Monthly to Weekly, using these same threshold values;
-
The coding to calculate RSI Weekly on the 4-hour bar chart, requires a lot of computer power. That is, it takes me about 15 minutes to run it on this 5-year data each time. So, I can easily make mods to thresholds and such, but just be aware that it takes some time to run each iteration. Philip, I’ll run any RSI filtering simulations that you suggest;
-
Shemski, thank you VERY MUCH for suggesting that trades should have been taken on 11/5/2014 and 11/25/2014. This helps confirm that we’re looking at the same charts and algorithms. To answer your question, both trades would have been taken when looking at things in isolation, but the RSI filter noted above kept us out of those trades. Also, the 11/5/2014 trade for example, would not have even been considered w/o the RSI filter, since a 10/27/2014 Short would not have been covered until 10/30/2014 8:00 AM, so my model would have ignored the 10/29/2014 EMA cross-under as a potential start to a new trade, since a position was currently active when this new EMA cross-under occurred (does that make sense to you?);
-
Leonlee0116jl, good question on the RSI filter removal. I just finished running that simulation and here’s what I got:
Overall Trading Performance:
Annual Raw Returns:
Image of Trade Log (hope you can read this):
I don’t understand. So you are saying you could have done it on the RSI monthly all along? Then what was the point of changing it to weekly?
Hey Dough thanks for updating the trade log. Please refer to my post no. 596. The result of the trade log where exactly what I anticipated.
The RSI monthly filter will work.