Hi Tommor,
In one post you say “Update - I have abandoned the price:200EMA requirement - not reliably indicative” and another you say “started to ignore the 200EMA.”
Am I right in thinking that EMA(200) is totally gone from your system now?
Yes it is, it seems for my trading its just too long-term an indicator to be relevant. Enormous price swings can occur one side or another of the 200 without breaching it within the 5-20 day holding periods I’m generally looking at so its out now.
Makes sense. Thank you for your response.
Hi Tommor,
You have a few secondary requirements that look at Weekly values, e.g. “At least last 4 weekly bars not pierced by 50EMA?”
When you talk of the “50EMA” in this secondary requirement, are you still referring to the 50EMA calculated in the Primary Requirements, which I assume is on the Daily chart? Or do you have a different 50EMA drawn on the Weekly chart, and that is what the 4 weekly bars must not pierce?
Its the daily 50EMA, so in some chart programmes you must be careful the software does not automatically re-set the EMA period to weekly when you switch the bar or candlestick time period to weekly.
Perfect. Thank you for the clarification.
Hi Tommor,
I’m trying to wrap my head around this rule.
My assumption is that this rule looks for (to avoid) a ranging market, and that less overlap is better.
My interpretation of this rule, thus far, is that you find the most recent weekly bar. We could call that PRIOR.
Then you start looking back in time, bar by bar, on the weekly chart. So, you compare PRIOR to PRIOR - 1 to see if there is overlap. Then you compare PRIOR to PRIOR - 2 to see if there is overlap. Etc.
Now this is where I got lost.
How many bars back do you test? Exactly 4 bars?
So, do you award +1 (bullish or bearish) so long as PRIOR doesn’t overlap all 4 bars?
1 Like
I count the most recent weekly bar as Week 1, or the Prior as you call it. I look back and see if the Prior has a range overlapping with as few as possible of the 3 previous consecutive weeks before it. If the Prior overlaps with just 1 week or maybe 2 or all 3, that’s a sign that price is moving with the trend. The more weeks the Prior overlaps, the weaker the movement. so if the Prior overlaps with the previous 7 or 10 or 17 successive consecutive weeks, then price is pretty much stuck in a range and not moving. Although its possible some other definitions of a trend might allow this chart to be called a trend and give it some positive points, I don’t want to get into a market that is moving forwards so slowly.
3 Likes
So, using your nomenclature of Week 1, it seems that overlap in Week 2, Week 3, and Week 4 don’t really matter as long as there is no overlap in Week 5, Week 6, Week 7 . . . Week 17.
It seems that an overlap of Week 1 to Week 2—but no overlaps other weeks beyond that—may suggest slowing of the trend (as though the engine has lost steam) and possibly even the end of the trend. Although sometimes it’s just a lull/consolidation in the same trend.
This is an interesting rule.
As always, thank you for your response.
Well the indications you look for from your chart should match your strategy, which doesn’t have to be a replica of mine. so you might use very different MA time periods, you might attach more significance to price rise over a certain period than I do in % or pips, you might be finding that just a couple of weekly closes on the same side of a given MA gives as good an indication as 10 or 20.
My aim was really to show that long-term trend-following can be profitable and that there are ways to give an objective higher or lower value to any trend.
Onwards and upwards.
Yes, definitely. This system provides a good substrate. My initial thought is to understand the rules fully, codify them, then use it to select instruments. That’s when I will hammer out a specific strategy—although having a SL 1 to 2 ATR(20) below/above, no TP, and pyramiding seems straightforward enough.
Regards
2 Likes
Hi Tommor,
My question is: what charts do you consider when considering the “clear majority?” Are you considering just currency pairs based upon the major currencies? Or do you also consider more exotic pairs based upon currencies like PLN, SGD, NOK, SEK, DKK, MXN?
My first investigation had me looking at currency pairs based only upon the 8 major currencies. I get these 28 pairs.
"EURUSD", // major
"GBPUSD", // major
"USDJPY", // major
"EURGBP", // major
"EURJPY", // major
"GBPJPY", // major
"USDCAD", // major
"CADCHF", // major
"USDCHF", // major
"CHFJPY", // major
"EURCHF", // major
"GBPCHF", // major
"CADJPY", // major
"EURCAD", // major
"GBPCAD", // major
"AUDUSD", // minor
"NZDUSD", // minor
"AUDCHF", // minor
"NZDCHF", // minor
"AUDCAD", // minor
"AUDJPY", // minor
"AUDNZD", // minor
"EURAUD", // minor
"EURNZD", // minor
"GBPAUD", // minor
"GBPNZD", // minor
"NZDCAD", // minor
"NZDJPY", // minor
Your rule states, “a clear majority of other pairs with same base currency.” So, in the case of EUR, there are seven charts to consider. OK. But in the case of USD, there are only three charts to consider.
Three doesn’t give a good statistical sample. What does “majority” mean in that case? 2 of 3?
You are being too literal. It is legitimate for this purpose to regard pairs in which USD is the counter currency by inverting the pair from the way it is usually expressed/offered as a market. I only take account of the major currencies of which there are 8, so there are 7 pairs which can be expressed as if the USD is the base currency.
Groovy. Now it’s completely clear.
Do you use a stop loss?
I always use a stop-loss. A good method I found is to use 2 or 3 x ATR20. I have even used 5ATR20 for a little while but 3x seems good. The nuisance is having to keep re-setting the stop as the ATR varies.
But the idea is to close the position manually if the stop is not hit but the TA deteriorates such that the trend is about to end. There is no harm getting out of a trend when it pulls back as long as you get back in when it resumes.
2 Likes
How are the results using this strategy over the past year?
1 Like
Thanks for bringing this up, unfortunately I understood nothing
1 Like
Hi ronnie - I don’t share trades or results like win ratios, profit % etc.
I could say it has worked better than it could have done for me but not as well as I would have liked. Thing is, if you’re a trend-follower you already have a strategy, or you will take elements of mine and use just those. If you’re not a trend-follower, the best trend-following strategy ever designed will not be right for you.
2 Likes
Two more strategies which are maybe even more boring. They’re still trend-following but they are simpler and shorter-term so perhaps they’ll be better suited to those traders who like quicker in-out action.
50EMA slope - Set a buy order at the day’s high if it is lower than yesterday’s. Set a stop-loss at yesterday’s low or entry minus 1.5 x ATR20. Size the position to a 1% account capital risk. Exit at the third successive higher close or the at the close of any day with the fifth successive higher high.
20/50EMA sequence plus slope - Buy at the close of any day on which the 20EMA is above the 50 and both are sloping upwards. No SL. Do this on 10-20 pairs including the shorts, but use minimum position sizes. Exit at the close the next day. Repeat if the EMA set-ups are still in place. Useful for exotic pairs with high margin requirements.
Don’t watch the news or the open positions.
1 Like