Very interesting. What stop were you using, underneath the extreme candle?
I believe the 5 lot will reduce stop out loss and drawdown, as stopout loss is only on 1 lot instead of two. It should also greatly increase profit. Let’s see.
Yes, lots of cross currents in these statements. Let’s define terms and that will help.
When most people talk about trading “with the trend”, they mean the trend of your home time frame. Some, including me consider trading with the trend as the trend of your home TF and the TF above it. In fact, if two or three TF’s above you are in the direction of your trade, even better. So yes, that precludes or prevents counter trend trades, since by definition, those trades would be going against the trend of at least your home chart. I simply don’t take trend trades if my home chart and the TF above it aren’t in the same direction. I just wait until they are.
But most of Tymen entries are counter trend, so what to do? There are at least two possibilities. If you have a valid CBL on a buy counter trend trade, odds are very high it will run up to a positive position (at least at least it will pay the spread). If you take the buy trade and you are trading counter trend you have a very high probability of it moving into positive, but a low probability of it making a long O-BB walk on your home chart. This is why I said this trade is just like a scalp trade. You want to let it run as far as it will and exit with the most pips possible. You never want to let a CT trade go positive and then turn back negative. You never let these counter trend trades turn from winners to losers, and if it goes to loss right after entering, then you don’t let it run more than a few pips beyond the spread before closing it. In that sense, you are treating them like a scalp.
A very few of Tymen’s trades will be “with the trend” by my definition of being with the trend on my home page and the TF above. Those are great entries and can give you a long O-BB walk. but they are rare.
There’s nothing wrong with trading short counter trend scalps and I’ve made 100 pips a day steadily doing that, but there is something else you can do. You can wait until this counter trend move is complete and use it as a good entry point to the main trend on your home time frame. The highest probability way I’ve found to do this is to look at lower TF’s and look for a valid CBL retracement entry on a lower TF that is going in the direction of your trend on your home chart. The closer to your home chart the better.
So say, the trend of your home chart and the TF above is down and you want to get a trade down. You know that just jumping into the middle of an existing trend is risky. But you can minimize that risk by finding the best entry you can as close to your Home TF as you can. Say you are on the 1H chart. Look above at the H4, is there a valid cbl there? Is the daily chart trend in the same direction? It happens some times and is a very good entry.
If not, and you don’t want to trade the daily or weekly, then move down. Is there a valid CBL on the 30 minute? Is one likely to come up soon? If so, that’s the best retracement entry you are likely to get. As you move further away from your home chart, the risk increases. A 15M CBL for a retracement entry is pretty good. A 5M cbl is weak and a 1M is too horrible to consider without other factors lining up in your favor.
So what’s best? You have to be nimble. The best possible way to trade this set-up is to enter the home TF counter trend CBL and drain every pip you can out of it, then look for a stop and reverse. Exit with all the pips you can and watch lower TF’s and see if you can get a valid CBL entry in the direction of the main trend on your Home chart. If it’s a good one and it’s probably the best CBL retracement entry you will be able to make for how long you are willing to wait (patience is a virtue, the longer you wait, the better the retracement entries), then take it and learn. You can use a CBL off an inner or mid band, but they aren’t nearly as reliable as the Tymen CBL off an outer band and a lower TF, so I’d advise against it.
The good side is, using a CBL on a lower TF to enter an existing trend on a higher TF, you don’t have to risk as much on stops, usually. You may get stopped out once or twice, but when you get in, the pips are so great with the 5 lot on a sustained trend that it doesn’t matter. Be consistent and track your results. Don’t try this in choppy markets as it won’t work. If you have different results than me, we’ll resolve any differences, OK?
Any questions?
I got some initial results from the 5-lot system in backtest.
:eek:
More detail in about 12-16 hours.
Thanks for the comprehensive answer, Graviton.
I’ll try and get some more experience of these concepts in demo. My live account is not the best place to learn.
I suppose my real challenges right now are:
-
Still unwilling to take a SL once it gets too large (focussing on the $ rather than the pips)
-
Trying to put all of this knowledge into a process map cum flowchart which will force me to reduce the emotion element.
I like to check my understanding now and then just to ensure I haven’t misinterpreted something.
So am I right to say that to conclude which way the higher TFs are trending, we see if the PA is above, below or around the middle BB, just like how the scoring is done for the pairs?
Therefore we don’t:
-
Look at the direction of the middle BB which is a lagging 20SMA indicator to indicate the direction of the higher TF trend.
-
Attempt to draw a trendline under or above the candles of the higher TF to deduce the higher TF trend of the most recent PA because that PA could be just a retracement against a trend in the other direction. ie. 2 or 3 downtrending candles (lower highs and lower lows) above the mid BB would qualify as an overall uptrend still.
The 5-lot system has exposed testing problems.
I’m using MT4’s “M1” timeframe data that you download from the Metatrader website through the History form in MT4. What MT4 does in that case, is look at the “volume” for the M1 candle, and makes up that many fake ticks using a fractal algorithm to match the candle. Open, High, Low, Close.
What this means is that any EA that does anything “intra M1 candle”, including hitting stop losses is not being truly tested, and its behaviour will differ from one run to the next even with the same setup parameters.
This affects the 5-lot algorithm significantly, as your stops and additional lot entries come thick and fast - and relatively close together. Depending on the particular fake ticks MT4 generates, you can end up with results that are wildly different.
For example, one series of runs using 2010 data for USDJPY 1HR, changing no parameters between runs, resulted in the following final profit percentages:
17%, 17%, 7%, 7%, -1%!, 1%, 1%
The reason for this, is that with the 5-lot system you have a small fraction of trades that run on to the full 5 lots for big pips. But if they get stopped early by fractal noise it can affect the result massively.
While it is encouraging as the bias is definitely >0, it makes it impossible to determine if a tweak or improvement actually does any good.
The classic Tymen trades have profit and stops further apart, and are therefore not as badly affected by this fractal tick “noise”.
Anyway, here’s the very first trade picked up in Jan 2010;
$28 Classic Tymen Trade
$595 5-Lot All In, 20 pip, no stop widening.
I shall post more results when I get proper data.
Hey NB, you’re right that MT4 does have some operational limitations - but I don’t think should be taken too seriously.
You’re right that the “fractal noise” can cause problems with setting stop losses, as they can happen at the wrong time and take you out early - whereas you would still be in a trade in reality.
This is, however, a very good simulation of chaos theory, which I believe is widely employed within the technical trader community.
Given that markets can act in an infinite variety of ways in the future, the MT4 simulator does a very good job in telling you not to optimize your results to historical data.
Your results have shown good potential in the system, with the majority ending with positive returns.
I would personally take the average of 100 runs and use that as the expected profit/loss value for the test period.
Sounds like you are trying to do a de-facto Monte Carlo simulation to me IronHeart
Monte Carlo the whole way!
good point…
is it me or is lewis hamilton overrated?
Well, I tried to explain, but perhaps I didn’t do that good of a job at it. It’s difficult to explain something that’s mostly instinct. Here’s some rules to guide you.
-
Never take a trade against the trend of the TF immediately higher than your home chart, that is the chart you are getting the signal from.
-
If you take a trade against the trend of your home chart (but always with the trend of the time frame one above), like most acceptable Tymen CBL entries will be, it is considered counter trend. These CT CBL entries have a very high probability of turning positive, but a lower probability of of trending for a long time. In most cases, close to 9 out of 10 by the math, you will not have time to use the 5 lot strategy because the trade will run out of steam around the 2nd lot. So you can employ one of three winning strategies, which ever works best for you.
a) Use the Tymen 2 lot TP1 & TP2 strategy. Everyone who has used this in conjunction with only taking trades that are in agreement with the trend one above the home chart (where the cbl occurred) has made pips to the best of my knowledge.
b) Put your 1 or 2 lots (which ever you chose, be consistent) on at the beginning, like in the Tymen strategy, but let them run to they turn against you or reach 2X your stop and take profit. Your goal is to walk away with maximum pips. You will never let a trade that goes positive lose pips. If it goes positive, always exit before it turns back negative.
In this case you are treating the CT trade like a scalp, you won’t allow the trade to go the full distance to your SL. The SL is only for emergencies. You will exit much faster, as soon as you are convinced that the trade is going against you. You will cap your maximum loss at around 1/2 your normal stop loss plus spread.
c) In the rare case, about 1 out of 10, that the trade goes into a BB walk after you have already locked in TP2 profit, you can risk a small portion of that profit and convert to the 5 lot in mid trade. Since the odds aren’t great that your home chart trend will actually change in any one trade, you should be slow to put on extra lots and fast to close all lots and walk away with most of your pips if you try this conversion in mid trade. Do not try it unless you see a O-BB price action on your home chart.
That summarizes how I trade CT trades. There are many other factors I consider such as fundamental news, higher TF charts like the daily and weekly charts, optimization of entries and exits using lower TF charts, trade filters and such that we have discussed before, but these are the essentials.
Trading with trend on home chart and the TF above is a different case. I always use the 5 lot for trades with trend. If you can get a CBL with trend on your home chart (and always with trend on the chart one TF above) then take it. These are the best trades to take, but they are rare.
If you want to get into a long sustained trend that is already in progress, this is a very risky proposition. Your Win/Loss ratio will be less than the CT trades above, but if properly managed your risk reward ratio can be many times greater. Always use a CBL off an outer BB to enter these trades, just as Tymen taught. But as I said, you will rarely get a large enough retracement to an established trend on your home chart to get a valid CBL signal.
In this case look down one TF and see how many valid CBL’s were offered in the direction you want to trade over what you consider a reasonable time to wait to see and entry signal. Say around 4 hours for a day trader. If you see none, and that will most often be the case if it is truly a consistent and established trend, then drop down another TF and see how often CBL entry signals are offered. Say that is the 15M chart and you see one was offered three hours ago, you may chose to wait another three hours for the next one. Patience is a virtue when waiting for retracements. But say there were none in the last 4 hours, you can drop to the 5M chart, there should be some there, so you wait for another. They tend to come around often there.
When it looks like you are getting a 5M CBL entry, make sure to check and see if you are also getting a higher TF entry developing, if you are take that because it’s better. If not take the 5M entry CBL. Your stop will be smaller and you may want to consider other factors than just the bottom of the extreme candle when setting the stop, like time of day, volatility of the pair, ATR, etc.
This will either move with you to turn into a multi lot trade, or against you. If it turns against you, it was not a good entry and it’s best to cut your losses, which should be small, and try again for a better entry. This is a tricky trade to master, but it’s not unusual to make 200 to 300 pips off it, so it’s well worth learning. I suggest you practice this many times in demo before trying it live.
Is that better? Do you have any specific questions?
This is consistent with my experience. I didn’t want to state these results as it might cause too great of expectations in newer traders. But now that it’s out there, yes, it is consistent and quite possible based on my experience.
TA, the technical definition of a trend is described by Tymen as a series of higher lows for an uptrend and a series of lower highs for a downtrend. I like to see three in a row to call it a trend, but I suppose technically, two in a row qualifies as a series.
Drawing a trend line does help to visualize these trends. It isn’t required, but I do it when trading is slow to give me something constructive to do and keep me from overtrading. Many other traders do too, so you will often see price using these old trend lines, some going back weeks or months, as support and Resistance.
Since I’m often making quick decisions when trading 1H and below, I use short cuts to help visualize trends, like which way is the BB mid band (20 SMA) pointing, is price above or below the 20 SMA. I will also look at the price at the far left of the chart and the current price at the far right to see if it has moved up or down over that time frame. I will also use the default Stoch with 8,3,3 settings to help visualize the trend. If I need more visual aids, I may use the RSI, MACD, ADX and Zig-Zag indicators. I keep all this other junk on my junk chart and only refer to it as needed, just about every day once or twice, but not on every trade. Well, I do look at the 15M Stoch filter on every trade as previously described.
As we move to longer time frames, for those wanting to make twice the pips with half the work, we will have much more time to make decisions, so short cuts won’t be as valuable, but the lessons learned on 30M and 1H TF’s will be invaluable when making decisions at the longer time frame levels. Some may chose to stay at shorter TF’s because they have a system that works great for them there and they enjoy more active trading. That’s fine with me. You should trade the TF you are most comfortable with.
Did I answer your question and do you have any others?
Good suggestion IH. Many thanks to you and NB for the fine work you both have done over the course of this thread.
I have considered doing that - but at a few minutes per run, or longer if I go back a couple of years - it is just prohibitive.
One thing I have discovered is that 2009 was a bad year for Tymen’s strategy. Really bad. 2010 on the other hand has been a great year.
NB,
The alternative is to have it running on multiple computers so the load is shared (parallel processing). There are a number of senior dedicated thread participants who could run these simulations using mt4’s strategy tester.
Even if you only got 10 of us to help, I am sure that we could run these backtests over the weekend; say 10 times each on a 2 year timeframe.
Just an idea…
Cheers,
hachi
Thanks Graviton,
Your answers are excellent and comprehensive as always. The problem is me, not your explanations. Maybe I’m getting senile! 
I have become confused as to what is a trend and counter trend now, which is puzzling, since it used to be relatively easy to tell the two apart using just one TF.
I have attached two charts, a home (M30) and home + 1 TF (H1). We have what I would say is a CT entry on the home chart.
Can you see my dilemma here? Which trend do we refer to on the Home+1? The same CT as the M30, or the SMA20, or the bottom left to top right uptrend, or the 3 point touch uptrend?
Seems like the more I learn, the more I discover I don’t know. No wonder they say, “ignorance is bliss” (but neither is it enriching)
I’m in the same boat, I think. Am I right in thinking that the main trend is the direction the PA is heading in on the higher TF chart, whilst the CT is the direction of your home chart (if against the higher TF trend).
ie,
1H shows an upward trend
30m shows a CBL forming for a short entry downward trending trade.
The trade you are entering is a CT trade? This is the method I’ve been using to avoid trading CT, as here I would leave the market be until the PA again reverse on the 30m chart then enter long if a CBL presents itself, going with the longer TF trend. Now I’m slightly confused I’ve been avoiding CT correctly…
On a side note, I’ve had some good success with setting my stops at ATR x 2 + spread. It’s noticably improved my chances of avoiding an early whiplish and keeping me in good trades, whilst still protecting me from mistaken entries. Since I’ve employed this method, my account has started a healthy increase, which is nice. I’m still experimenting with the idea of changing this formula for times of high volatility, such as market crosses.
Also, I’ve been experimenting with moving down to 5m charts. I only do this if the 15m chart is showing a bounc of the mid BB in the direction of the main trend. I’m uncomfortable taking a CBL here (maybe others do take a CBL and have success?), so I move down to the 5m chart and look for a squeeze CBL. More often than not, or so it seems so far, there is a CBL. I then enter using the 5m as my new home chart, using the ATR from this TF to set stops and managing trades here.
The idea is to eventually become competent enough to move up the TF when the SL have increased sufficently to meet the higher TF ATR SL requirements (using 2 pips profit = 1 pip moved up SL).
Finally, I’m willing to help out with some backtesting if required. I’m working all weekend anyway so won’t be using the home computer - can leave it running and doing it’s thing for as long as needed.
Cheers all,
Cord
I’m currently looking at expanding upon the idea of not just using ATR to measure the intial SL, but also as the means by which the trailing stop is calculated.
There are some interesting sites I’ve found through google detailing using ATR in this manner, and I’m currently looking for a good indicator that does it for MT4.
Instead of using my current formula of (SL = 1/2 max profit), I might potentially start using the price set by the ATR trailing stop as soon as BE is met. So I mean, once I have reached BE by the normal method, I will ATR trailing stop to calculate the exit for each lot.
Might sotp me getting stopped out all the time on my lots (and missing big pip moves). Conversely of course, it could spell the start of losing many pips by not exiting soon enough.
I’ll do some testing and if anything useful comes out of it I’ll let you guys know.
Cord