Dump it Here

[Skate]

Nick Radge shared some work he's doing on his (MR?) day strategy on twitter recently.

Let's talk about when we should exit a position
@Newt recently remarked that Nick Radge made a post on Twitter about one of his strategies. As I had a few moments I thought I would hunt that post down but I found something more interesting from Nick. Every time I mention or reference Nick's work it always creates interest.

This is "The Chartist" post that sparked my interest
When looking at Nick's chart of (LKE) the exit didn't seem right to me. I have a fair understanding of Nick's "Weekend Trend Trader" Strategy even making a few posts about it.

Over time
I've found Nick's WTT strategy was a little slow in exiting a position. Why? because of the generous 40% initial stop loss used in his strategy. Using a wide 40% stop loss always results in giving back some open profits & the chart below didn't. On the flip side having a generous 40% Stop Loss allows more time in a trend reducing the whipsaw effect.



I have changed computers twice since coding my version of Nick's WTT
There is so much written about this strategy it's not difficult to code. I've quickly set about finding the parameters once again to understand why the exit was so sharp, without giving back a chunk of open profits.

My mission was to arrive at the same signals
Now, if I can do that, I'll be able to understand the stop loss used in the chart above. Moving on, I'll throw up some charts to (a) mimic the entry & exit for (LKE) that is disp[layed in the chart (b) display an alternative stop loss & (c) determine if the exit can be sharpened. The most important part of the exercise for me is to determine what parameters Nick is using in the chart above?

Skate.

 

[Skate]

"Weekend Trend Trader Strategy"
Using a google search led me to the "Think or Swim" website below. The hyperlink has a basic description of Nick's WTT strategy. There is also some "Think or Swim" code that can be easily converter to AFL. If you're inclined to purchase Nick's WTT ebook it's also referenced below.

Weekend Trend Trader by Nick Radge Strategy for ThinkorSwim - useThinkScript Community

I stumbled upon this forum by searching for some of the momentum strategies by Nick Radge (thechartist.com.au), I found @Zachc 's brilliant post on the Bollinger Bands BReakout strategy which Nich wrote about in his book "Unholy Grails", I would like to thank @Zachc and the forum members...

usethinkscript.com

First some general parameters of the "Weekend Trend Trader Strategy"
Details are in the book Weekend trend trader Ebook by Nick Radge

This is a list of "some of the parameters" from the "Weekend Trend Trader" ebook
1. Recommended allocation is 5% of the total account capital on each position.
2. The strategy calculates & adjusts the trailing stops when the system is deemed to be "on or off" in relation to the (SMA) of the index.

Criteria for trade entry
1. Stock must be making 20 week High
2. Stock's Rate of Change must be >= 30%
3. Market Index close price must be higher than its 10 Week moving average (of the relevant index)
4. If all three conditions above are met, you may enter a trade on the Monday at market open.

Criteria for trade exit
1. If the Index is above the 10-week moving average then the recommended trailing stop is 40% below the highest weekly close.
2. If the index turns negative below the 10-week moving average the trailing stop is moved to 10% below the highest weekly close.

Skate.

 

[Skate]

Criteria for trade exit
1. If the Index is above the 10-week moving average then the recommended trailing stop is 40% below the highest weekly close.

Now I'm confused
I coded Nick's WTT Strategy using the parameters in the previous post & the exit didn't align. Maybe the parameters used by Nick are currently different from the ones posted on the "Think or Swim" forum back on the 21st of September 2019.

Admittedly all strategies tend to evolve over time
After saying this, I was able to code my version of Nick's WTT Strategy to arrive at the same entry & exit for (LKE)







What was the difference?
It was simply the stop-loss percentage that had been used.

Originally, I used the initial stop loss of 40%
As it was explained on the "Think or Swim" forum. I soon realised that Nick has used a 20% initial Stop loss in his Twitter post.

Next
Can the exit of (LKE) be improved?

Skate.

 

[Skate]

"Weekend Trend Trader Strategy"
Before I look for improvements, I might revisit the original exit as described on the "Think or Swim" forum.

When the WTT system is "on"
The WTT Strategy is deemed to be on when the Index being traded is above its 10 Week moving average. When the system is "on" the initial stop loss is set at 40% below the highest weekly close.

40% initial Stop Loss is not used in Nick's Twitter post
Nick's WTT Strategy has used an initial stop loss of 20% below the highest weekly close (LKE)




The Original Stop Loss for Nick's WTT Strategy
On the "Think or Swim" forum back in 2019 the parameter quoted from Nick's "Weekend Trend Trader" eBook the initial stop loss was listed at 40% below the highest weekly close (LKE). There are arguments "for & against" using either stop loss percentage compared to the other. I believe this decision is better left for a trader to decide "depending" on his risk tolerance.



The next question
Can the exit be improved?

Skate.

 

[qldfrog]

Interesting reverse engineering

 

[Skate]

The next question
Can the exit be improved?

The short answer is yes
Using a "take profit stop" using a weekly (ATR) multiplier "measured bar-by-bar", is simple & effective in locking in profits when trading gets tough.

Using a variety of stops certainly helps me in my trading
A trailing stop is handy as a basic default stop & it should be the last line of defense. Relying solely on a "trailing stop" that's traditionally used in most strategies should rarely be hit in my opinion as it's guaranteed you will give back a level of open profits. Also using a unique "Stale Exit Stop" helps in exiting a position quickly which is also a nice feature. Using these simple three exits tends to work better than in isolation.




Yes, a take profit stop has helped in this isolated example
But how would each strategy compare over a 2-year or 3-year period? I'll leave that for another post.

Skate.

 

[Skate]

Before I move on
A "Trailing Stop" is nothing exotic & it's used by most traders with varying parameters with a varying degree of success. The "Take Profit Stop" is a simple (ATR) multiplier. Its primary function is to lock in profits when presented. Trading without a Take Profit Stops is setting yourself for a heartbreaking experience of "watching large open profits disappear".

The "Stale Stop Exit" is a little more complex
The Stale Stop Exit does a mighty job & is a combination of 6 conditions. It's extremely difficult for me to explain how it works without revealing the complex coding involved.

Rapid changes in price
First off the "Stale Stop Exit Strategy" gauges when momentum is shifting, slowing, or reducing but that's not all that it does as it's time-dependent as well. There are multilayers to the "StaleStop Exit" as momentum needs to be calculated bar-by-bar in relationship to previous bars. Each position is only given enough time to confirm its worth. Positions that don't keep improving are not worthy of keeping.

Skate.

 

[Skate]

Let's not look at the exit in isolation
Posting an exit of one position (LKE) shouldn't be the judge of how an exit strategy performs.

For the evaluation exercise
I'll post a backtest over the last three years from 1/7/2019 to 30/6/2022 for you to judge.

Backtest
The backtest on the left has a 20% Stop Loss Exit -- whereas the backtest on the right has a 40% Stop Loss Exit.




Blow me down
I didn't expect those backtest results. Maybe someone who trades Nick's turnkey WTT Strategy might be kind enough to post the results from their system. I'm sure the Turnkey Strategy would perform better than this.

"lipstick on the pig". Don't see any lipstick out of place here.....

Let me put some lipstick on the pig
@Newt made a remark a few posts back about how everyone thinks they have the "secret sauce". I'm the first to admit, that I fall into this category. Over the years you tend to believe your own bullsh!t when you start to see improvements with your coding.

I'll put some "Lipstick on Nick's WTT Strategy"
I'll be quickly trying to improve on his ideas. The WTT Strategy is a simple 20-period breakout, it's nothing fancy so it shouldn't take me long. Bloody hell I never prepared for this. If Nick's results were better I wouldn't have to do this.

Skate.

 

[peter2]

Next
Can the exit of (LKE) be improved?

No, absolutely, definitely not, if you stay with the weekly bars and the same trading strategy.
Sure you can use a profit target and save losing a little but then you're changing the system's trend following strategy. If you're changing the exit strategy then you're changing the system. Can the system be improved? Possibly.

Edit: Traders have to accept that $hit happens and our trade exit is triggered by a huge down bar.

 

[Skate]

No, absolutely, definitely not, if you stay with the weekly bars and the same trading strategy.

Bugger, I've just finished coding the strategy & it's ready to post
@peter2 we don't often disagree but this time I have a different opinion to the one you have expressed. All I did was code Nick's WTT Strategy more efficiently.

Now I have another challenge
In the next post, I'll try to wring a little more out of the strategy, I won't have time to change much of the code so I'll add a bit more lipstick.

Skate's Version of Nick's WTT Strategy
First off, I want to make a disclaimer, the strategy is coded as per Nick's parameters of a 20-period breakout Strategy with a few bells & whistles thrown in for a direct comparison.




Give me a sec
I'll see if I can make more improvements.

Skate.

 

[DaveTrade]

The most important part of the exercise for me is to determine what parameters Nick is using in the chart above?

The second question in the exercise is "why did he choose them?".

 

[Skate]

No, absolutely, definitely not, if you stay with the weekly bars and the same trading strategy.
Sure you can use a profit target and save losing a little but then you're changing the system's trend following strategy. If you're changing the exit strategy then you're changing the system. Can the system be improved? Possibly.

Edit: Traders have to accept that $hit happens and our trade exit is triggered by a huge down bar.

Okay, Let's compare
I've made only a few adjustments to display how settings & parameters can make a difference. @peter2 is correct when he says that no matter how good your strategy performs it's always at the mercy of the markets.




Well
I'll post what work I've done. Quick extra improvements have eluded me. So I posting two extra charts. One with a 20% Stop Loss & one with a 30% Stop Loss. Both have a "Take Profit Stop" applied.

I've managed to add a new colour
It appears all I've done is apply a different coloured lipstick. But in saying this, the results aren't too shabby.

Skate.

 

[Skate]

The second question in the exercise is "why did he choose them?".

Poker Machines
@DaveTrade when traders talk about their results they tend to do the same as those who play the "Poker Machines". They are the first to tell you about the wins but never talk about their losses. It's the same with those who push an agenda. The results displayed are normally cherry-picked to establish & add validity to the point they are trying to get across. When I post backtests I use the period each time so my results are not cherry-picked.

I'm sure Nick doesn't fall into this category
From all reports, he is upfront & honest in all his dealings. What impresses me about him is this single point - "he has passed his knowledge on to others, not only with his paid subscription but the volume of free information he supplies"

The Weekend Trend Trader
In its original form, I'm sure with additional improvements over the last few years the WTT Strategy would produce better results than I initially achieved. Also, I should say I coded Nick's WTT Strategy using the information freely available.

I was thinking
I'll make a quick post before entering my daily results only to realise the exercise turned into something that has been so time-consuming & unproductive.

Skate.

 

[Newt]

No, absolutely, definitely not, if you stay with the weekly bars and the same trading strategy.
Sure you can use a profit target and save losing a little but then you're changing the system's trend following strategy. If you're changing the exit strategy then you're changing the system. Can the system be improved? Possibly.

Edit: Traders have to accept that $hit happens and our trade exit is triggered by a huge down bar.

Peter2 said what I was thinking too have to admit. But lipstick is so damned tempting!

Actually Skate you've picked a corker of an example trade (exit) with LKE for looking at different exit approaches.

Traditional WTT and long weekly trend trading would have to trade through the strong down bar, then exit next one or two bars depending on trailing or other exits.

BUT, you could also get out after the strong bear bar (probably a significant down move against last 20 odd weeks in terms of size of bar, ATR, std dev of price bars). Peter2 has called this "strong supply coming to the market" or something similar if my grey brain recalls correctly.

AND, agree if you were going to be much more aggressive, then yes you could justify generating an exit on that very long strong bull bar before the smash down - again it would like be something like >3SD larger than previous price bars, or ATRs, or whatever measure you pick. One way of doing this might be only exit on "take profit" if market momentum is waning or terrible, rather to modify all your exits globally.

 

[Skate]

Actually Skate you've picked a corker of an example trade (exit) with LKE for looking at different exit approaches.

It was Nick who picked (LKE)
@Newt I didn't pick (LKE) for the example, I just run with it. If you read Nick's Tweet you will see he picked (LKE) in his tweet to explain why his system has an advantage when using a stop loss.

I've studied Nick's WTT long enough to spot that his chart, the one he posted on Twitter didn't look right
Why? because it didn't follow the chart pattern that I have studied. I thought it would be a quick exercise to explain that something in his methodology has changed. To my surprise, it was a slight rework of the stop loss. Instead of using a 40% Stop Loss he has changed it to 20%.

Was it intentionally done to demonstrate a point?
Well, I don't know. The strategy may have changed. Improvements could have been found. Or it could be a case where he thought no one would notice.

Changing the Stop Loss
You can see from my charts you can get a better exit using a 20% Stop Loss, whereas in reality it may not be physically traded with that setting.

I guess Nick is the only one who can explain why the Stop Loss was changed
The change might be temporary or after further evaluation over the years, he has elected to change it permanently. Than again, the markets at the moment are not conducive of a wide 40% Stop loss



Skate.

 

[Skate]

The Weekend Trend Trader
In its original form, I'm sure with additional improvements over the last few years the WTT Strategy would produce better results than I initially achieved. Also, I should say I coded Nick's WTT Strategy using the information freely available.

Hopefully, someone will post
There are those (forum members) who would trade Nick's WTT Turnkey Strategy. I'm hoping that a backtest will be posted for a direct comparison.

Backtest Details
$100k portfolio with a backtest period from 1/7/2019 to 30th June 2022.

My previous best effort
The previous best effort of backtesting the WTT Strategy was acceptable. To be truthful, it was a struggle to improve on the previous results.

With a little bit more time
There were areas for improvement. I'll post my previous best effort then I'll post the backtest for my final version of Nick's WTT Strategy.




Now my "Final Version" of Nick's WTT Strategy
Improvements were hard to find in the final version. The next capture below this one is a side-by-side comparison between the two strategies with red highlights.




Side-by-side comparison
The upgraded version of Nick's WTT Strategy with a direct comparison to my previous best version. The upload unfortunately is a little smaller making it hard to read.

There are improvements
Not only with the Net Profit, but also increased winners. reduction in the number of trades, lower drawdown, & better Car/MDD metric. The improvements came from adding my "Ulcer Index Indicator" to the buy condition, & passing on those riskier trades.




What is the Ulcer Index Indicator?
The Ulcer Index Indicator attempts to estimate the “stress” of a position by estimating price retracements. The indicator is based on the notion that downward volatility is bad, but upward volatility is quite good. It increases in value as the price moves farther away from a recent high price & also falls as the price rises to new highs. I hope my explanation of how I use the "Ulcer Indicator" helps you better understand why some traders are luckier than others. Indicators play a big part in my trading & they have certainly helped me. The Ulcer Indicator's sole purpose is to control the drawdown risk (not eliminate the risk) without reducing the profit potential of a strategy.

Metrics
Two areas of a backtest that carry more weight for me in deciding whether to keep developing a system, a system that you could trust trading live. (1) "Maximum System Drawdown percentage" & (2) the "Ulcer Index". Both should be low. A low drawdown & low Ulcer index should go hand in glove to give you the confidence to trade the strategy. I've previously explained how I take advantage of the Ulcer Index Indicator even supplying the parameters that I elect to use.

Finally
# I can now put this exercise to bed.

Skate.

 

[Willzy]

Hopefully, someone will post
There are those (forum members) who would trade Nick's WTT Turnkey Strategy. I'm hoping that a backtest will be posted for a direct comparison.

Backtest Details
$100k portfolio with a backtest period from 1/7/2019 to 30th June 2022.

My previous best effort
The previous best effort of backtesting the WTT Strategy was acceptable. To be truthful, it was a struggle to improve on the previous results.

With a little bit more time
There were areas for improvement. I'll post my previous best effort then I'll post the backtest for my final version of Nick's WTT Strategy.

View attachment 143663


Now my "Final Version" of Nick's WTT Strategy
Improvements were hard to find in the final version. The next capture below this one is a side-by-side comparison between the two strategies with red highlights.

View attachment 143664


Side-by-side comparison
The upgraded version of Nick's WTT Strategy with a direct comparison to my previous best version. The upload unfortunately is a little smaller making it hard to read.

There are improvements
Not only with the Net Profit, but also increased winners. reduction in the number of trades, lower drawdown, & better Car/MDD metric. The improvements came from adding my "Ulcer Index Indicator" to the buy condition, & passing on those riskier trades.

View attachment 143665


What is the Ulcer Index Indicator?
The Ulcer Index Indicator attempts to estimate the “stress” of a position by estimating price retracements. The indicator is based on the notion that downward volatility is bad, but upward volatility is quite good. It increases in value as the price moves farther away from a recent high price & also falls as the price rises to new highs. I hope my explanation of how I use the "Ulcer Indicator" helps you better understand why some traders are luckier than others. Indicators play a big part in my trading & they have certainly helped me. The Ulcer Indicator's sole purpose is to control the drawdown risk (not eliminate the risk) without reducing the profit potential of a strategy.

Metrics
Two areas of a backtest that carry more weight for me in deciding whether to keep developing a system, a system that you could trust trading live. (1) "Maximum System Drawdown percentage" & (2) the "Ulcer Index". Both should be low. A low drawdown & low Ulcer index should go hand in glove to give you the confidence to trade the strategy. I've previously explained how I take advantage of the Ulcer Index Indicator even supplying the parameters that I elect to use.

Finally
# I can now put this exercise to bed.

Skate.

Hi Skate,

I've noticed that with your backtests your only going back to 2019 to compare results, I'm curious as to how your systems compare since say 2000 to today with delisted stocks and index constituants considered.

Kind Regards
Matt

 

[Skate]

Hi Skate,

I've noticed that with your backtests your only going back to 2019 to compare results, I'm curious as to how your systems compare since say 2000 to today with delisted stocks and index constituants considered.

Kind Regards
Matt

@Willzy thank you for taking interest in my last series of articles concerning Nick's WTT TurnKey Strategy. The point of these posts was to understand & explain to others why "The Chartist" Twitter chart for (LKE) didn't look quite right to me. The exit displayed by Nick in his Twitter feed didn't display the normal tendency of a system that incorporates a wide 40% initial Stop Loss.

Backtesting
Comparing a backtest from many moons ago isn't productive. It's even less productive & misleading when it comes to optimizing the parameters & settings. In my opinion, you can include too much historical data resulting in a series of false & misleading results. I've made multiple posts explaining the reasons why I prefer a shorter backtest period to a larger time frame. One of the major reasons is that the results are more reflective of changed trading conditions.

I haven't the time to do a search
@qldfrog articulated why he also preferred a shorter backtest period where he explained the reasons succinctly & much better than I ever managed to. In saying this there are those who have prosecuted the alternative view with vigor. I'm not stirring this conversation up again.

In saying this
I started trading on the 1st of July 2015, over that time I've traded like a mad man & traded more strategies than I care to remember. Coding my version of Nick's WTT Strategy is a recent endeavour but if the WTT was available back in 2015 I would have most likely traded it with all the others. Now I'm curious to see if it would have performed back then to now. For me to revisit a backtest any further back than 2015 is not on the cards,



Skate.

 

[Skate]

I'm curious as to how your systems compare

How do they compare?
Trading this calendar year has been a tough gig. I've done a backtest comparison of Nick's "Original WTT Strategy" from the information that is free off the web. Then I compared it to my Updated Version of the same WTT Strategy.

Don't take it as gospel
I'm sure someone who owns Nick's TurnKey WTT Strategy will post a more accurate backtest results to set the record straight. The backtest below is from the 1st of January 2022 to today.

Going from what I know
The concept of trading a 20-period breakout system is a solid idea. I've managed to code the idea IMHO more efficiently with increases in the performance of the overall strategy.

Net profit is not the metric you should be chasing
We all tend to code a strategy to exhibit an increase in the Net Profit, but that's not the only metric that should be of interest. Coding is always a trade-off. When optimizing parameters it's so easy to fall into the trap of "curve fitting a system" if you don't have your wits about you.

There are sometimes more telling metrics that need to be considered
From the holdings period to increased winners with a reduction in the number of trades also needs to be considered. Most of all, I'm always seeking to achieve a lower drawdown that results in a better Car/MDD metric.

The backtest comparison
Between Nick's Original Parameters from the "Think or Swim" forum versus my "Upgraded Version" of the same strategy



In Summary
Neither version of the WTT strategy performed all that well but in my defense, my upgraded version of the same strategy didn't lose any money at all -- but rather snaffled a few dollars along the way.

Skate.

 

[Sid23]

I started Nick's WTT (default parameters) beginning this year, down 2.3% (a lot better than my BBO which started 12 months ago & down 22%).

Below are 3 x backtests for WTT (default parameters as per Nick's code) of different periods that Skate used, . Even though stop loss starts at 40%, all my trades were stopped out at 10% once the index moved down (ribbon turns from green to red). As Nick has said, it is very rare to be stopped out at the initial 40% stop.



Regards,
Sid