Welcome to NexusFi: the best trading community on the planet, with over 150,000 members Sign Up Now for Free
Genuine reviews from real traders, not fake reviews from stealth vendors
Quality education from leading professional traders
We are a friendly, helpful, and positive community
We do not tolerate rude behavior, trolling, or vendors advertising in posts
We are here to help, just let us know what you need
You'll need to register in order to view the content of the threads and start contributing to our community. It's free for basic access, or support us by becoming an Elite Member -- see if you qualify for a discount below.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
Hi all, I am having trouble with one of the cornerstones of a good trading system: risk vs reward. So far I've been basing my Stop Loss as a function of my Profit Target. Essengially I'd risk 1 ES point to make 2. I would find a lot of times I would get shaken out of the trade only to see it take off in my origional direction. I then watched a video by Linda Raschke about Stop losses, and I started thinking perhaps I'm not thinking of risk vs reward the way I should. Meaning, instead of calculating it in X amount of dollars, I should rather be thinking in bar movement. She talks about calculating stops from Average True Range (ATR). Following that thought, then profit should also be a function of ATR as well.
This is the video I watched: http://www.youtube.com/watch?v=9rHLGJq79ZU
So some questions are:
1. Is this a sound way to approach stops?
2. Are there any downsides to this approach?
3. Is there an indicator out there which will automatically calculate the ATR of N bars back so as to easily see where the stop might be placed?
imo one of the toughest things to learn and develop are stops and targets. As you describe, many guys use atr to derive these. I'd put forward that you should include especially in stops, looking for a recent short term high or low especially one where there seems to be a confluence of orders. Again, imo , this is a discretionary item and is learned based on attributes and experience...like an art form.
If you search threads for koy, the only one that shows up is a journal that I sometimes post in. You could see many pictures there and read the trade logic of stops and targets that I use. That might be a starting point for you to work out what is best for you. DB
So I happened by the board of trade this morning and ran into a mentor with whom I lost contact in recent years. A brilliant man, a legendary trader, truly a wonder. In his mid 70's now, he was on a field trip with "at risk" high school kids …
@drunkcolonel: You asked the right questions, and I think that I can show you a number of available methods that calculate stops based on the average true range.
The idea of those stops is always the same: You take an anchor point such as a recent high, a recent low or a moving average. Then you add (for a short position) or subtract (for a long position) a multiple of the average true range and that is your stop line. You can also replace the average true range with other volatility measures, such as the simple range or the standard deviation. I will explain some of the stops for a LONG POSITION, for a short position you can apply the logic accordingly upside down.
Chandelier Stop:
The chandelier stop - also known as chandelier exit - is a classic. It was used by Charles LeBeau, one of the first system traders. The chandelier stops takes the highest high since you entered your position and deducts a multiple of the average true range to obtain the stop price. A good explanation can be found here:
The Supertrend is in fact a trailing stop based on the average true range. The anchor point of the SuperTrend is not the highest high but a moving average over the last N bars. You may then deduct a multiple of the average true range from the last price of the moving average.
I have implemented a modified version of the SuperTrend, which gives you a wide choice, these are the specifics:
-> value of MA is calculated one bar ago (saves CPU time, as recalculation is not required with each incoming tick, avoids feedback loop, stop does not move with current price)
-> value of true range is calculated one bar ago (saves CPU time, as recalculation is not required with each incoming tick, avoids feedback loop, stop does not move with current price)
-> there are 27 different MAs, a median and a mode that can be used to calculate the anchor point
-> you can use the simple range or standard deviation instead of the average true range
Exported using NT Version 7.0.1000.11
This is a new implementation of the TradeStation SuperTrend indicator for NinjaTrader 7. The SuperTrend indicator is an application of the concept of MAE (maximum adverse excursion), which was introduced by John …
BrainTrend:
The BrainTrend uses the low of the the current bar as an anchor point and deducts a multiple of the average true range from the low. In addition it has a second emergency stop which uses both the stochastics and a 2-bar momentum which is compared to the average true range. The code of the BrainTrend is not easily accessible, I think it similar to a stop indicator used by EASC Trend
Here is a small chart for the brain trend.... I have not finished the work on the indicator, it is one of my unfinished projects.
Kase DevStop:
Cynthia Kase has tried to refine the concept by replacing the average true range with an average true range plus a confidence interval. The main problem of her concept is that she did not define well the stop and reverse points. I do not like this concept, because calculating the volatility of the volatility for the purpose of defining a stop line is mathematical overkill, it is simply not needed.
I am not posting a chart here on purpose, but wanted to have the concept included in this list.
Stops based on Average True Range, explained by Sylvain Vervoort
I really recommend to have a look at a few articles that were published by Sylvain Vervoort in Technical Analysis of Stocks & Commodities in 2009. It is a series of 3 articles which can be found in the May, June and July issues. In particular the second article deals with stops based on the average true range. Sylvain Vervoort uses the close as an anchor point and then deducts the average true range. This is basically a SuperTrend with the period of the moving average set to 1.
A lot of good recommended techniques and possibly the only thing I can add is to stress (in my experience) stop and target placement (and trailing stop behaviour) ought to stem from price pattern rather than from attempting to inflict an arbitrary risk reward ratio on the market, except to the extent the ratio may perhaps be derived from rigorous back- and forward testing as one cornerstone of a complete system. In this case if the ratio inferred from price action doesn't meet either profit expectations (assuming the setup itself qualifies as high probability as defined by the system) or any ratio required by the system then the trade is passed over.
Indeed I would never "prescribe" a reward-to-risk multiple to be attained. Probably the best method to define the trailing stop for a given entry method is to look at the maximum adverse excursion (MAE) that the system generates without a trailing stop.
You can then plot the P&L of each trade against its MAE. Introducing a trailing stop would in fact eliminate all trades, for which the MAE has exceeded the trailing amount. This eliminates both profitable and losing trades and replaces them with losing trades that produce the trailing amount as a loss.
Example: Below is a scatter plot of 2,600 trades backtested over a period of 18 months. If you draw a vertical line at $ 550 to eliminate all trades to the right side, this should be beneficial as
-> most of those trades have been losing trades
-> the overall dispersion of the results has been reduced allowing for higher leverage
However you have to keep in mind that all those trades, also the profitable ones, will become losing trades with a loss of $ 550. The trailing stop will therefore not contribute a lot to overall profits, but will definitely contribute to reduce the largest drawdown.
The optimal trailing amount depends on the trade distribution and not on any benchmark value ofr the R-Multiple.
Another method you can try is simply crafting your own custom trails and stops.
A lot of traders like to "lock in" profits when the market moves your way. This is easily illustrated with a simple breakeven that you set so that what is potentially a nice profit doesn't turn into a total loss if the market retraces. (or you can move your stops UP closer upon a certain amount of runup).
I've found pretty good success with parabolic trails, which allow the trail to be large to start and then shrink, either in a constant rate as the trade goes in your favor, or at a variable/accelerated rate.
Again, this can be achieved using a series of discrete "levels" or thresholds which modify the stop, or you can use a simple parabolic equation to give you a more refined adjustement.
I've personally found that a parabolic combined with a final minimum trail amount can significantly enhance the profitability of many types of systems.
"A dumb man never learns. A smart man learns from his own failure and success. But a wise man learns from the failure and success of others."
Thanks so much everyone for your responses, and sorry I could not get back to contribute sooner with my thoughts as I ended up losing a lot of work when a nasty lightning storm took out my other computer. Thanks a lot Florida...
Anyways I wanted to take some information presented here and start to run with it... so I spent the last few days trying to put together an example. I figured not only would this help me understand the concepts better, but also help me think about how I can best introduce the concepts into my trading. I hope you find the thoughts well thought out and organized. Any comments are most welcome.
Getting Started:
In an attempt to organize my brain as well as have an easy to follow structure to the posts, I'm going to try and organize my writings into Observations, Questions, Answers, and finally Decisions. Most of the questions I write are for the community to hopefully answer, however I will make some attempts at answering my own questions through my own analysis. My hope is that if I post some thinking that is "way off the mark" or somewhat flawed, those more knowledgeable than I will chime in with suggestions or their answers to the various questions.
The legend will more or less be:
Observations = O
Questions = Q
Answers = A
Decisions = D
Supplemental = S
Not Applicable = NA
First in order to speak about ATR and understand how this concept can lead to more successful trading with regard to Stops and Profit Targets, I want to make sure that I've got the correct understanding of it in the first place. I thought it best to present my learning process through an example trade that walks through step by step. AFTER THE FACT NOTE: when I actually began this analysis I had been using Heikin-Ashi bars, however when I lost the analysis on the other computer via lightning strike, I desided to redo it with OHLC bars believing this is a more "back to basics" price action approach. I wanted to keep the analysis as "clean" as possible. In doing so, one interesting thing I noticed is that when looking at one chart you might see a signal that tells you to trade, and when looking on another chart your perspective on entering the market changes. Moving on.
Chart One:
Observations:
O1. We see that the market is trending down.
O2. We see that the market starts going sideways for a bit.
O3. Looks like a potential breakout to the downside.
O4. Market turns right back around but fails to continue higher.
O5. Continuation of the down trend.
O6. Pullback touches the EMA15
O7. Next bar goes red and closes on the low: Trend Continuation Bar.
Questions:
Q1. Am I missing any key observations in this analysis?
Answers:
A1. I do not know if I've missed anything essential, I think I have the pertinent information for a decision.
Decisions:
D1. We're setup based off of observations, and with [ O7 ] we decide to pull the trigger and go short.
O1. We see that we have entered the market short, our order being filled at 1338.75.
O2. With our decision [ D1 ] on the previous post, we can now look at our previous bars counting backward from the Filled Bar.
Bar One: Range between 1339.50 and 1338.75 or 3 Ticks.
Bar Two: Range between 1340 and 1339 or 4 Ticks.
Bar Three: Range between 1339.75 and 1339.25 or 2 Ticks.
Total Ticks under consideration: 9, Total Bars under consideration: 3, thus 9 divided by 3 gives us an Average True Range of 3.
Questions:
Q1. Where do your Stop Loss and Profit Targets go as a function of ATR?
Q2. Are some bars more relevant to that calculation than others?
Q3. What multiple qualifies as reasonable given considerations such as: our instrument ( ES in our case ), our trading chart ( 3 minute ) and the current trading time ( 9:06AM )?
Answers:
A1. It depends on how many bars (periods) are used in the calculation, which leads to [ Q2 ], as well as the multiple which leads to [ Q3 ].
A2. What do you get when you cross an elephant with a rhino? Eliphino. Bad jokes aside, my thought is that if one goes back too far and encounters a bar of disproportionate range (say a couple of strong breakout bars) and factors that in, say compared to more recent bars (e.g. a couple of weaker pullback bars), that might skew the calculation into a wider stop than desired.
A3. Again I don't know as I'm relatively new to futures trading, however as per the PDF provided by Harry, under the Exit Priorities - Initial Stop Loss section: "Wide stops are usually best because they will tend to provide a higher percentage of winning trades. Tight stops can be bad because they often cause us to exit trades that would eventually be profitable." The presenter also makes a note that: "My favorite initial stop is placed 2 or 3 ATRs below my entry point".
Decisions:
D1. Without having a real answer to [ Q2 ], I'm just going to throw out a number and guess with three.
D2. Without having a well qualified answer for [ Q3 ], I'm just going to go with the conclusion presented in [ A3 ], and thus in our case go with two ATRs for our trade.
D3. With our calculation of ATR as 3 Ticks from [ O2 ] coupled with our decision [ D2 ] which stated our chosen multiple as 2, we now decide to implement a Stop Loss of 6 Ticks, and a Profit Target of of 12 Ticks (assuming our system implements a Reward Risk Ratio of 2:1. This makes our Stop Loss a value of 1340.25 and our Profit Target a value of 1335.75.
Supplementary Thoughts, Information, and Resources:
S1. Given this resource: Average True [AUTOLINK]Range[/AUTOLINK] ([AUTOLINK]ATR[/AUTOLINK]) - ChartSchool - StockCharts.com I believe the calculations in [ O2 ] to be correct. Essentially using Method One we take just the difference between the High and Low. I believe my understanding is that Method Two and Three are for bars in which there is a gap. Since we have no gap with our current example, I believe Method One is the correct method to use in this instance.