Dump it Here

[ducati916]

@Rsthree you have raised some important questions & you have managed to answer one yourself. You may have also missed Duc's comment earlier today that's worth repeating.



MYR
1. "So if MYR has pulled back to or below the original auction price is there any reason not to buy it during the week"
It's not about a "negative outcome or skewing the system performance" it more to do about "unconditionally" following the trading rules. System trading is usually most effective when "implemented consistently". One problem frequently encountered by individual traders is the "difficulty" in following the rules or the system. Sticking to a system requires discipline & discipline is often difficult to maintain when their is confusion what to do next. Traders may be tempted to second-guess or modify the rules as they go along, lets not do that.

2. "I can't imagine that a buy a few days later should have any negative outcome or skew on system performance"
Yes it can, if you purchase MYR you will not be in a position to take the next signal as your portfolio will be full. Break your rules once & you will do it again & again starting one of many bad habits.

When there is "no buy" there is "no sell signal"
In saying this you can buy MYR the only issue that I can see is that the Action Strategy will never give you a signal to sell.

Skate.

Mr Skate has already given you the answer. However, here is an alternative way of thinking about it.

Some questions:

(a) When designing the system, did you have a number of ideas that you felt could be profitable;
(b) Did you backtest those ideas;
(c) Did you try to break the system (ideas) through using difficult market periods;
(d) Did you make any compromises;
(e) If yes, why;
(f) If no, you were happy with the results;
(g) Why did you go the systems route in the first place;
(h) Were your discretionary trading results disappointing;
(i) If yes, were the system results better;
(j) If no, did you seek to make improvements to the system;
(k) Did the system improve;
(l) Did the improvements made, now exceed your discretionary trading;
(m) Were you finally happy with the system: is it the 'best' that you can achieve at this point.

If you have reached (m)...why would you adjust (tamper) with it now?
If you haven't reached (m), why are you trading it?

Traders (novice/failing) will always seek a way to self-destruct. Successful traders limit the number of decisions that they make, thereby minimising the ways that they can self-destruct.

Make all of your thinking, testing, fiddling, before you enter the market. Do not start when you have money in the market: your decision making under pressure will be uniform: it will all be bad.

jog on
duc

 

[Lone Wolf]

@willoneau AmiBroker backtesting results are calculated using the opening price. To replicate your backtest results it's vital to snag the opening price & a limit order is the tool, other than that you will find it difficult to secure the opening price & replicate the backtest performance.

I don't disagree with ignoring any trade that gapped your buy limit on Monday. But one question - Will the Amibroker backtest of this week show MYR having been purchased on Monday? Or does the backtest code include something to say "Buy the open, only if the open is not more than 3% above the signal bar close"? If the backtest shows MYR as purchased, then it could be argued that picking up MYR later in the week for the original buy price more accurately follows the system as it was tested.

As a side thought, if we were trading the daily time frame I suspect most people would have their limit order in the market all day. So price has all day to drift back to your entry. But on the weekly system price doesn't get all week to come back to us since we cancel the order on Monday. Just a difference in execution depending on the timeframe which I hadn't considered before.

 

[Newt]

Always nice to feel the love!

This thread is a very valuable resource for many and has re-ignited my interest in mechanical systems.

I remember back on Reefcap days when tech/a and (I think Daryl) were developing TechTrader, more or less live on their thread. At the time, I was a purely discretionary trader (which I still am to a degree) and had a fairly cynical view towards systems. I would pop up periodically on their thread and challenge some of their ideas.

The primary issue for me then was: the system is long only. What happens when the market turns? Their exit (at the time) seemed really lagging and I just couldn't get my head around why they would potentially risk so much 'paper profits'...sound familiar? The issue was something exactly like this year's collapse.

That issue seems now to have been solved by Mr Skate. The second issue (for me) has always been buying stocks that you know absolutely nothing about (hence my preference for ETFs) and the significant risk that that can engender. The 20, 30, 40 stock portfolio and quick exits, solves that issue.

The last issue, is the issue of actually executing and following the system. This was an issue that I also raised with tech/a back in the day: what if the market changes in some material way and the system is actually broken: how do you differentiate from a normal drawdown, to a fundamental change in conditions?

Now this may have already been addressed earlier in the thread. I only joined the thread after it was well underway and may have missed this discussion.

jog on
duc

Would have loved to have been "around" in the Reefcap days. Sounds like Captain Black was among the active members of that forum too. Would have been made that much more interesting by the 2003 - 2007 bull market.

Wondering, do you feel ASF is ahead in any way (or perhaps missing something important) that Radge's reefcap forum offered back then?

 

[rnr]

Correct - both entries & exits are always placed for Monday's pre-auction using the (+/- 3% premium). The 3% premium will either secure the opening price or it won't. With a "buy order" not being executed has no consequence or bearing to the profitability of long term strategy holding 20 positions in the portfolio. (In summary, it just misses buying that position)

Correct - If the position is still open after 10:31 am - I sell immediately "at market" as I want to be off the sucker at any price. I wait till 10:31 for the markets to settle down. (that's all)

Skate.

@Skate,

So it would be fair to conclude that the manual intervention, as highlighted above, would be the only deviation from the results of a back-test and reality.

Cheers,
Rob

 

[Skate]

@Skate, So it would be fair to conclude that the manual intervention, as highlighted above, would be the only deviation from the results of a back-test and reality

Rob, in a nutshell yes. Sometimes the (+/- 3% premium) rule won't be enough latitude for the position to be executed at the opening price when the Gap-down is in excess of 3% to the last closing price. 99% of the time the (-3% premium) isn't fully exercised (usually it isn't). To find the right exits for any strategy can be a difficult job resulting in a compromise using "manual intervention" at times. Amibroker uses the opening price for all entries & exits. Backtesting uses "DUMB" maths whereas we are not as privileged, meaning our calculations are done in the heat of the battle whereas Amibroker uses hindsight.

Skate.

 

[Skate]

@rnr has raised the issue of exits & it's only fitting that I explain a little bit about an unexciting topic.

Duc's key entry words
In previous posts I've explained how the Action Strategy enters a trend using "volatility & volume". Unfortunately I can't be more specific as private information needs to remain private.

Duc's exit strategy
I haven't checked with Duc but I'm sure he will be okay if I talk about his "Volatility Dependant Stop" in general terms instead of specifics. Using the "Volatility Dependant Stop" is a clever concept as it uses the VIX in combination with extreme parameters that I found hard to swallow, but they worked like a dream to my amazement. Bandying the words like "Volatility & Vix" is meaningless without the explanation of application "that will remain a secret".

Exits lock in profits or they avoid further loses
Everybody knows that stops are necessary, but nobody really likes them. Often you get the feeling that the stop has just thrown you out of the market before it turned in your direction & you missed the big move. To find the right exits for any strategy can be a difficult job. By now readers know that exits are more important to me than entries. There is more written about entries than about exits because each entry needs a special scenario like indicators & patterns, something unique & interesting where exits seems to be more boring until we use a different approach.

Generally
Exits are normally a trailing stop or a variable trailing stop using fixed percentages but the Action Strategy takes it a few steps further being more flexible depending on the market’s volatility. The disadvantage of just using a trailing stop or a variable trailing stop is because they can’t be adjusted to the current market conditions. The Action Strategy incorporates a volatility exit thanks to information Duc has passed on to me in private. Volatility Dependant Stop & Stales Stops works wonders when trailing stops become less reliable due to their inherent lag.

Please note
Volatility stop can change dramatically from one day to another or even more so trading a weekly trading strategy. This flexibility makes the volatility based stop superior when dealing with the reward/risk ratio compared to only using a percentage based tailing stop. The volatility based stop is better all-round. Add a StaleStop & now we're cooking with gas.

Skate.

 

[ducati916]

Would have loved to have been "around" in the Reefcap days. Sounds like Captain Black was among the active members of that forum too. Would have been made that much more interesting by the 2003 - 2007 bull market.

Wondering, do you feel ASF is ahead in any way (or perhaps missing something important) that Radge's reefcap forum offered back then?

All things tend to evolve over time. In trading, that tends to be for the better, simply because if you survive and eventually thrive in trading/investing, you generally have some insights to pass on.

So on ASF currently, from the Reefcap days, there is tech/a, WayneL. and CountryBoy that I remember. With regard to systems, tech/a was the chap who really introduced me to them, but, for a number of reasons, I never really got into them.

The chap who really opened my eyes to a really useful area of trading (Options) was WayneL (although he went by some other name in those days which for the life of me, escapes me currently). That is one area of ASF that is weak: Options. They just have so many applications, but there seems to be limited interest in them.

Reefcap was a bit more US orientated than ASF. Apart from those 2 differences, it is similar.

jog on
duc

 

[rnr]

@rnr has raised the issue of exits & it's only fitting that I explain a little bit about an unexciting topic.

Duc's key entry words
In previous posts I've explained how the Action Strategy enters a trend using "volatility & volume". Unfortunately I can't be more specific as private information needs to remain private.

Duc's exit strategy
I haven't checked with Duc but I'm sure he will be okay if I talk about his "Volatility Dependant Stop" in general terms instead of specifics. Using the "Volatility Dependant Stop" is a clever concept as it uses the VIX in combination with extreme parameters that I found hard to swallow, but they worked like a dream to my amazement. Bandying the words like "Volatility & Vix" is meaningless without the explanation of application "that will remain a secret".

Exits lock in profits or they avoid further loses
Everybody knows that stops are necessary, but nobody really likes them. Often you get the feeling that the stop has just thrown you out of the market before it turned in your direction & you missed the big move. To find the right exits for any strategy can be a difficult job. By now readers know that exits are more important to me than entries. There is more written about entries than about exits because each entry needs a special scenario like indicators & patterns, something unique & interesting where exits seems to be more boring until we use a different approach.

Generally
Exits are normally a trailing stop or a variable trailing stop using fixed percentages but the Action Strategy takes it a few steps further being more flexible depending on the market’s volatility. The disadvantage of just using a trailing stop or a variable trailing stop is because they can’t be adjusted to the current market conditions. The Action Strategy incorporates a volatility exit thanks to information Duc has passed on to me in private. Volatility Dependant Stop & Stales Stops works wonders when trailing stops become less reliable due to their inherent lag.

Please note
Volatility stop can change dramatically from one day to another or even more so trading a weekly trading strategy. This flexibility makes the volatility based stop superior when dealing with the reward/risk ratio compared to only using a percentage based tailing stop. The volatility based stop is better all-round. Add a StaleStop & now we're cooking with gas.

Skate.

Hi Skate,

Based on your post regarding exits, as quoted above, I am trying to work out how you have coded (in general terms only) your exit strategies.
I'm assuming that you may have 3 separately coded (binary) stops that will send an exit signal if any of the stops have a value of 1. If that is the case and for the sake of this exercise lets call them StaleStop, VolatilityStop and TrailingStop.
With reference to the TrailingStop component, does it have multiple triggers involving say 2, 3 or 4 different outcomes for calculating a trailing stop based on the current market conditions or is there only one outcome.
Just to make this perfectly clear I am not asking you to disclose any info in relation to the VolatilityStop.

Cheers,
Rob

 

[Skate]

Hi Skate, Based on your post regarding exits, as quoted above, I am trying to work out how you have coded (in general terms only) your exit strategies. I'm assuming that you may have 3 separately coded (binary) stops that will send an exit signal if any of the stops have a value of 1. If that is the case and for the sake of this exercise lets call them StaleStop, VolatilityStop and TrailingStop. With reference to the TrailingStop component, does it have multiple triggers involving say 2, 3 or 4 different outcomes for calculating a trailing stop based on the current market conditions or is there only one outcome. Just to make this perfectly clear I am not asking you to disclose any info in relation to the VolatilityStop.

Important
This post is beyond beginner level so don't worry if it sounds double dutch as I'm answering a question.

Rob, the exits of the Action Strategy are driven by 3 inputs working separately but in concert with each other. They are all combined in a "for loop" because you have to store the bar at entry. Let me explain them one by one.

1. Trailing Stop
I use a chandelier variable trading stop driven by an index filter, this is a common alternative to an "apply stop" array. I use variables so "looping is a better alternative" for me.

2. Volatility Dependent Stop
The "Volatility Dependent Stop" is a bit more involved as it uses the "average of the VIX" of a n-period. The Volatility Stop is driven by a separate "VIX Index Filter" & no trades are allowed to be taken when the "VIX Index Filter" is off, it's a simple "Stop or Go" type filter. When the "VIX Index Filter" is on it brings into play another array which is the "average of the VIX". The "VIX array" is calculated using the average of the highest high value of the last n-period, less the average of lowest low value of the last same n-periods. The result of this calculation generates a parameter variable where the n-periods parameter changes from bar to bar. I won't say any more as the idea is confidential & the coding is complex.

3. StaleStop
The StaleStop is a momentum stop. The momentum needs to be calculated over a period from bar to bar. There has been much talk & speculation about my StaleStop but I can assure it needs to be included in a loop or it won't be efficient. If others are contemplating using a StaleStop in their strategy it needs to be in loop because it's imperative to store the bar at entry. Measuring acceleration I use a Parabolic SAR Indicator, an indicator that is not widely spoken about (SAR stands for “stop and reverse”) Like most trend-based indicators, the parabolic SAR is much more effective during trending markets than those we are experiencing at the moment.

4. Combination StaleStop, VolatilityStop & TrailingStop.
Of all the challenges we face as traders there is none more important then timing the exit by using appropriate combinations of stops to find the "sweet spot" to exit a trend when it changes direction. The placement of effective extra stops can make the difference between a quick, relatively painless loss or something more excruciating. It's also important to workout when a trend has exhausted itself. I use a combination of indicators & oscillators to respond to the challenge of stop placement & measuring trend reversal is one of them, volatility is another. As the trend turns or begins to lose momentum the exit strategy generates a signal to hop off the ride. Getting them to play nice together is another matter.

Skate.

 

[Newt]

I love it when you share this sort of detail Skate, for 2 reasons:

One, it frequently challenges established thought processes and stimulates creativity - essential for continued learning and self-evoluation.

Two, its comforting to know there are others out there pursuing systematic trading to simple but still extraordinary elegance (and complexity). If that sounds contradictory it is - as complex as it should be to improve returns without curve-fitting, but as simplified and robust as it can be to ensure reliable performance in most markets.


There were times years ago I wondered just how far down the "Alice in Wonderland" burrow I might have fallen - perhaps losing my way in the process. While I'm still not retired and living on a tropical island purchased through trading windfalls, hopefully I'm still properly grounded and continuing to improve from these ASF conversations.


p.s. Good question rnr - very much in line with Dump It Here "values".....

 

[Skate]

Skate.

 

[Skate]

Mr Skate experiment: hit quite hard:-$428 in first week (I did chase Myer as I wanted to stick to the system for compliance, unless the code is mandating purchase price can not be above 3% of last week close, which I do not believe is the case

Weekly comments
With the Action Strategy weekly updates, it's better if I let "the figures" do the talking but as @qldfrog has quoted his weekly performance trading his version of the Action Strategy it's only fitting for me to make a few comments.

How to read the Share Trader Tracker's Dashboard
Our first week resulted in a loss of (-$183) inclusive of the commission cost (-$190). In reality, excluding the commission charge the Action Strategy traded flat this week being +$7.46 up.

MYER - Chasing the price
One of the Action Strategy rules is that you aren't allowed to chase a price. It's not about a "negative outcome or skewing the system performance" it more to do about "unconditionally" following the trading rules. System trading is usually most effective when "implemented consistently". One problem frequently encountered by individual traders is the "difficulty" in following the rules or the system. Sticking to a system requires discipline which is often difficult to maintain when there is confusion about what to do next. Traders may be tempted to second-guess or modify the rules as they go along, let's not do that. The Action Strategy has not logged Myer as a buy so you will never get a signal to sell as a result.

The markets have not been kind to us
I've been reading the recent comments with interest. It's a perfect time to remember that trading is not always kind to us, losing weeks are inevitable & frankly there is nothing we can do to prevent it. To a great degree, our success or failure in the market is a function of our luck & timing. We would like to think that our results are a direct consequence of our insight & efforts, but in reality, luck plays a big part in how we do.

Signal quality
As with most indicators, the signal quality depends on the settings & the characteristics of the underlying security. The right settings combined with decent trends can produce a great trading system. The wrong settings will result in losses & frustration as markets don’t always trend they bounce around from one day to another. Sometimes price fluctuations are hard to handle knowing that the tools most suitable for analysis in these types of markets differ from those that I'm currently using to take advantage of trending market. Comparing the current volatility to before the time of COVID19 you start to get a sense of how the markets are affecting "not only us" but the "markets in general".

Skate.

 

[qldfrog]

View attachment 103494
View attachment 103495




View attachment 103496




View attachment 103497


View attachment 103499


Skate.

You mean Monday....
Have a great weekend.so you did not chase any of the buy.how will you compare to backtests?
Do you code your 3pc margin above the last close in the code?

 

[qldfrog]

You mean Monday....
Have a great weekend.so you did not chase any of the buy.how will you compare to backtests?
Do you code your 3pc margin above the last close in the code?

Sorry you answered my question as i i was typing it....

 

[willoneau]

MYER - Chasing the price
One of the Action Strategy rules is that you aren't allowed to chase a price. It's not about a "negative outcome or skewing the system performance" it more to do about "unconditionally" following the trading rules. System trading is usually most effective when "implemented consistently". One problem frequently encountered by individual traders is the "difficulty" in following the rules or the system. Sticking to a system requires discipline which is often difficult to maintain when there is confusion about what to do next. Traders may be tempted to second-guess or modify the rules as they go along, let's not do that. The Action Strategy has not logged Myer as a buy so you will never get a signal to sell as a result.

Hi Skate just curious you have only listed 15 companies when 19 were taken?

 

[Skate]

Hi Skate just curious you have only listed 15 companies when 19 were taken?

@willoneau well spotted, I'm in the process of replacing it with the correct capture as the worksheet was not in the correct position when I did the screen capture.

If I'm unable to rectify or replace the graphic I'll repost it again.

Sorry about that...

Skate.

 

[willoneau]

Curious why AST, BGL, BFC are missing?
I know why MYR isn't there.

 

[willoneau]

Thanx Skate

 

[Skate]

Curious why AST, BGL, BFC are missing?
I know why MYR isn't there.

@willoneau the post found here https://www.aussiestockforums.com/posts/1072660/ has been corrected to include the missing securities.

Skate.

 

[qldfrog]

OK, interesting take for a different broker, same orders:
At open, this is what I got:
MYR in red is different and this has been discussed
you will notice that in yellow AST was purchased at 1.97 vs Mr Skate $1.95
But I got in green cheaper GRR at 0.2354 and RBL at $1
This is a kind of mystery for me


you will also notice the number of shares is slightly different based on the broker software establishing the amount of shares for a given requested amount and the max price I used. Still roughly similar but do not expect exact same figures
If discarding the Myer purchase which was a kind of misunderstanding in the execution, we would end the week with a $186.44 loss (vs $183) and $1095 cash left from our initial 20k
I will resync on Monday and swallow a $250 loss, at this stage, for the mistake.