Australian (ASX) Stock Market Forum

Dump it Here

OK daily vs weekly: with real money and not Backtests
https://www.aussiestockforums.com/posts/1078897/
and following post
please notice the remark about returns vs XNT index which I use as a benchmark
I just included the weekly vs daily backtests matching Skate's above period;
Mr @Skate->did not want to impose on your thread but feel free to copy/cut and paste here if you see educational value :xyxthumbs
 

I have to say that I am really starting to appreciate more the systematic approach to trading and particularly with the weekly time frame.
In the past I would have been monitoring daily with finger on the button and more prone to reacting rather than executing, which results in over trading and high costs.

Moreover, lately I've been flat out at work so placing buy/sell orders at auction on the weekend really works well to avoid more stress during the day. To be honest I had never considered this approach, so it's a valuable learning.

Lastly I want to thank Skate again for the Action strategy and the hand holding and education he's generously providing.

It struck me the other day that he is actually paying a price in providing this guidance, ie. Paying for higher trading costs than he normally would due to the small parcel sizes.

So I definitely need to buy him a beer or a soda or a herbal tea, whatever is his poison.
 
Thanks @Warr87.

Except right now I'm reading the Amibroker manual: https://www.amibroker.com/bin/UsersGuide.pdf. I'm up to page 300 so far lol. I do wish Dr. Tomasz Janeczko (https://www.amibroker.com/about.html) would invest in a tech writer where English is not his/her second language ;).

Another good place to start with Amibroker...Howard Bandy's work is excellent. This is his introductory text, first of a series of five-six books on systems trading all using AmiBroker, also some Python & machine learning content in the more advanced books.
http://blueowlpress.com/books/introduction-to-amibroker/
There a link to the free download at the bottom of the page.
 
Thanks for the system updates Skate. I remember your saying your 'old' system gave up a big chunk of profits during the corona crash, does the action strategy have a filter that protects open profits (as well as not taking new positions) and does it react quickly enough, if it exists?

Cheers
Marty
 
Another good place to start with Amibroker...Howard Bandy's work is excellent. This is his introductory text, first of a series of five-six books on systems trading all using AmiBroker, also some Python & machine learning content in the more advanced books.
http://blueowlpress.com/books/introduction-to-amibroker/
There a link to the free download at the bottom of the page.


I downloaded the book. Now I just need to find the time to read it. Possibly I could code up some of my discretionary stuff and save significant amounts of time.

jog on
duc
 
Thanks for the system updates Skate. I remember your saying your 'old' system gave up a big chunk of profits during the corona crash, does the action strategy have a filter that protects open profits (as well as not taking new positions) and does it react quickly enough, if it exists?

@martyjames, the COVID-19 crash happened so quickly (a 2 week period) 70% of my profits were lost & in hindsight, nothing would have protected me, it was a simple "Black Swan" event. (thems the breaks)

Protection
All my (weekly) systems have been designed to exit quickly with the (GTFO filter), the systems also exit when momentum stalls or slows (StaleStop filter).

What you really want to know is
"does the action strategy have a filter that protects open profits (as well as not taking new positions) and does it react quickly enough" The answer is two-fold, (1) if another "Black Swan" event was to happen again (at the same magnitude & pace), I'm sorry nothing will protect us. (2) Under normal trading conditions (if there is such a thing as normal these days) I consider we have the best protection around built into the "Action Weekly Strategy"

@ducati916 doesn't mince his words (take note)
There will be a fair bit of chop in the continuing trend, which will keep many out of the market or singing songs of doom. So for the moment: sit tight, trade the trend and do not be scared out of positions as getting back in can be tough when you have to buy back in higher than when you sold out.

Summary
1. There will be a fair bit of chop in the continuing trend
2. Those singing songs of doom will keep many out of the market
3. Trade the trend & do not be scared out of positions

Skate.
 
Thanks for the reply Skate. Im playing around with exits that use index volatility (SP500) as an additional exit for my stock system. It seems to make for a smoother equity curve and avoids some of the corona crash , still resulted in an 18% DD for 2020 but not too bad.

cheers
 
I downloaded the book. Now I just need to find the time to read it. Possibly I could code up some of my discretionary stuff and save significant amounts of time.

jog on
duc
Hi @ducati916,

I'm a noob, but have been reading up on Amibroker and Norgate, so perhaps some of the below will help? This assumes you're using Norgate Data; if not, then perhaps this may help others.

The below link shows the "extras"/"fundamentals" data available in Norgate Data (Gold and Platinum subscription level) (scroll the page as required):

https://norgatedata.com/data-content-tables.php#current-fundamentals

I opened the data viewer in Norgate Updater:

upload_2020-7-1_6-45-52.png

but unfortunately it is static, not dynamic, and doesn't list all the additional information. I was hoping to be able to scroll to the right (a lot) to see all the extras.

I opened the Symbol Information window in AB:

upload_2020-7-1_6-47-18.png

which does appear to list (all???) the additional information, but with data labels rather than data names. For the data names I would use the Norgate URL above.

In the AB Chart window:

upload_2020-7-1_6-50-53.png

is the Norgate Data folder, I assume created when you install the Norgate Data Updater, or perhaps the ND AB plugin.

I dropped all these on a blank chart, then selected Edit Formula..., but IMO it's easier to just view the AFL in an external editor. I like Visual Code, but Notepad works (is notepad the best Microsoft can do after 20-30 years???)

On my installation, the Norgate Data AFL is located at C:\Program Files\AmiBroker\Formulas\Norgate Data.

Viewing those files in Visual Code, the main file seems to be C:\Program Files\AmiBroker\Formulas\Norgate Data\Norgate Data Functions.afl, as this is #included in all the remaining files.

Those functions look to just be helper functions wrapped around the AFL GetExtraData function, for example GetExtraData("NorgateBusinessSummaryLastUpdate"), with the rest of the AFL being outer functions calling the helper functions (for example to Plot() something).

IMO the documentation for the extra data could be better on the Norgate website, but this AFL, plus a bit of hacking, may get you what you need.

To me, it doesn't look like Norgate Data Functions.afl is complete, compared to the extra/fundamental data available in the URL above. Plus all the variable names in the code are prefixed with "Norgate", none of which are noted in the Norgate URL as the variable names.

For example, in the URL, search Metadata, Security Information - Text, FieldName = "Business Summary", but the AFL code says "NorgateBusinessSummary". See the function NorgateBusinessSummary() in Norgate Data Functions.afl.

All of this is untested ;) - I'm making the assumption that the AFL provided by Norgate, um, works :)

If I were to augment their code, I wouldn't edit their files directly, as they might get overwritten if I re-installed Norgate Data Updater. Instead, I think I would create another folder under C:\Program Files\AmiBroker\Formulas, such as CustomNorgateData or MyNorgateData or somesuch, then put my code there. And have a good backup somewhere in case I reinstall AB and my custom location disappears! I'd then follow a similar pattern as Norgate; have one uber-AFL file with all my helper functions, then other functions that call them (such as to plot or filter). I may be wrong on this best practice approach due to my AB inexperience, but I'm sure someone will correct me if I am (and I'll learn something in the process).

I believe you are a discretionary trader, but perhaps this will help you write up some filtering rules that would save you a lot of time when hunting for your next winners?

Hope this helps,
Linus
 
Hi Linus

Thanks for the info on extras"/"fundamentals. Im with Norgate, didnt even know this was available so will check it out. I wonder if anyone combines this data with trading systems and gets enhanced results?

cheers
 
What a year 01.png
2019-2020 Trading year was looking good
I’ve spoken to a few traders recently & their trading ranges from the "worst year ever" to "the best year ever". I’m lucky to be in the latter. Trading was travelling along quite nicely, well that all changed mid-February 2020. Trading just went to mush, experiencing the fastest drop ever that I can recall. It took 2 weeks to unravel, two weeks I’ll never forget. Markets aren’t supposed to drop that quickly, don’t they know that?

Historically
New bear markets (in the days gone by) normally would begin slowly with larger declines coming much later on within the cycle but not this time. Trading systematically, “Bear markets” aren’t that scary, well they weren’t. What happened in two weeks would normally have taken months to develop & play out. It wasn’t that way this time, leaving little time to react. I say “thank god for the GTFO Filter”.

Why so fast?
The reality of the COVID-19 hit like a sledgehammer, forecasting economic chaos impacting worldwide with closures & imposed lockdowns aimed to slow the spread of coronavirus. Quick & steep selling scared most traders, shaking out the weak first. Disbelief turned to panic creating a flow-on effect, selling on a mass scale. As traders, we have the tendency to see what we want to see & act accordingly.

Skate.
 
logo Capture.JPG
Trading the bounce
Recoveries are always created haphazardly, whipsawing from one day to the next but with the COVID-19 crash, it was hourly. Even the best traders found it difficult to pick the change of market sentiment. In the next post, I'll chart up some real examples of how it was done in real-time.

When does a bear market start to become a bull market? (Do they sneak up?)
What I do know is that the markets shift unpredictably & the bounce can come in sneaky ways. Traders who obsessed over negative news didn’t enjoy the recent bounce. If you missed the bounce so far, don’t compound your mistakes by sitting on the sidelines too long, as it’s never too late to re-enter the markets. Admittedly there are to be more disappointments along the way, so be prepared.

Pain tends to focus me
Since 20th March this year, I had to endure the good with the bad. Trading the bounce was not without its challenges. I could have described my best & worst days, weeks & months in percentages but dollars adds more clarity. You can't eat or spend percentages so dollars it is. Knowing how others handle the markets may be a useful insight.

The worst of trading the bounce (from the 20th March 2020)
Worst Day (-$62,823)
Worst Week (-$33,808)
Worst Month (-$116,831)

The best of "trading the bounce" (from the 20th March 2020)
Best Day $62,414
Best Week $83,770
Best Month $146,314

Skate.
 
Ducati Two Dog Strategy Capture.PNG
"The Ducati Dalmador Two Dogs Strategy" (has been useful trading the bounce)
The next few charts hopefully prove that "The Ducati Blue Bar Strategy" is no Chihuahua when it comes to trading strategies as it's the best of both breeds (volatility & volume). As traders, we don't need a fancy strategy to make money. Personally I want to trade the price differentials along the way. The Strategy is uncannily accurate when applied to all timeframes, the signals are fast & snappy. Adding prudent "Money management" with a sharp "PositionScore" brings this strategy to life.

Picking the change (trading the bounce)
Trading the bounce has been very profitable so far but it has been a roller coaster of a ride. "The Ducati Blue Bar Strategy" when applied to the Dow & the All Ordinaries picked the turns in real-time, normally only seen in hindsight.

I'll post a chart of the two indexes showing the turn. (Both Daily Chart)

2020 COVID-19  (DJI) Crash Capture.JPG




2020 COVID-19  (XAO) Crash Capture.JPG

I prefer to trade weekly
In the next post, I'll show how I capitalised on two trades (APT & Z1P) traded weekly. Trading the same strategy, whether "Daily or Weekly" the results are similar regardless of the periodicity.

Skate.
 
In the next post, I'll show how I capitalised on two trades (APT & Z1P) traded weekly. Trading the same strategy, whether "Daily or Weekly" the results are similar regardless of the periodicity.
Ooh goody!
Happy to say that I have moved to a weekly style discretionary system... the headaches have gone!
Cheers.
 
How I capitalised - trading the bounce ("The Ducati Blue Bar Strategy")
I'll chart both weekly trades for (APT & Z1P). The "The Ducati Blue Bar Strategy" concentrates on "volatility, volume, turnover, acceleration & momentum" all the ingredients that go hand in glove with increased market sentiment.

The two weekly trades ($25k position)
1. (APT) is being traded in a Weekly timeframe & the open profits are $30,826
2. (Z1P) is being traded in a Weekly timeframe & the open profits are $30,881

Rebound APT WEEKLY Capture.JPG




Rebound Z1P WEEKLY Capture.JPG





The same two positions "if traded" in a Daily Strategy ($25k positions)
"If I traded" these using the Daily Strategy the results would have been slightly higher with additional work & commission expenses.
(APT) closed profits of $25,551
(Z1P) closed profits of $39,553

Rebound APT DAILY Capture.JPG





Rebound Z1P DAILY Capture.JPG

Summary
By "Trading weekly" both positions are currently open, whereas if I traded the positions on a daily timeframe I would have been in & out a few more times. The positions would also be closed.

Skate.
 
How I capitalised - trading the bounce ("The Ducati Blue Bar Strategy")
I'll chart both weekly trades for (APT & Z1P). The "The Ducati Blue Bar Strategy" concentrates on "volatility, volume, turnover, acceleration & momentum" all the ingredients that go hand in glove with increased market sentiment.

The two weekly trades ($25k position)
1. (APT) is being traded in a Weekly timeframe & the open profits are $30,826
2. (Z1P) is being traded in a Weekly timeframe & the open profits are $30,881

View attachment 105446




View attachment 105447





The same two positions "if traded" in a Daily Strategy ($25k positions)
"If I traded" these using the Daily Strategy the results would have been slightly higher with additional work & commission expenses.
(APT) closed profits of $25,551
(Z1P) closed profits of $39,553

View attachment 105448





View attachment 105449

Summary
By "Trading weekly" both positions are currently open, whereas if I traded the positions on a daily timeframe I would have been in & out a few more times. The positions would also be closed.

Skate.


From the number of posts and backtest results posted, I'm pretty sure most accept that the weekly is the way to go. Of course weekly doesn't mean that you exit at the end of the week, the trade could run for months. The problem is that it is somewhat counter intuitive. Traders tend to think that they can catch the daily fluctuations and that by doing so they will never suffer temporary drawdown, converting that all to profit. The truth is that that is incredibly difficult to do and even if you could, often you need the top and bottom prices of the day to enter/exit. Watch a few day-trading videos and see the incredible consistency of traders missing the trend for dollars and trading counter-trend for pennies. This is the exact trap that awaits for those that trade daily systems.

It is also incredibly difficult to exert the mental discipline to re-enter or re-exit a stock that you have just entered/exited, almost immediately, which in practical terms means that you lose your position. This can be incredibly costly if that stock then goes on a run.

Further, when trading daily fluctuations, you will find yourself increasingly paying attention to the headlines. This almost invariably leads to the poor-house. Trading the headlines is a monumental waste of time/effort and most importantly money.

If you can't sleep at night, have to watch the market, etc: your risk is too high or you have no idea what you're doing.

jog on
duc
 
Hi @ducati916,

I'm a noob, but have been reading up on Amibroker and Norgate, so perhaps some of the below will help? This assumes you're using Norgate Data; if not, then perhaps this may help others.

The below link shows the "extras"/"fundamentals" data available in Norgate Data (Gold and Platinum subscription level) (scroll the page as required):

https://norgatedata.com/data-content-tables.php#current-fundamentals

I opened the data viewer in Norgate Updater:

View attachment 105427

but unfortunately it is static, not dynamic, and doesn't list all the additional information. I was hoping to be able to scroll to the right (a lot) to see all the extras.

I opened the Symbol Information window in AB:

View attachment 105428

which does appear to list (all???) the additional information, but with data labels rather than data names. For the data names I would use the Norgate URL above.

In the AB Chart window:

View attachment 105429

is the Norgate Data folder, I assume created when you install the Norgate Data Updater, or perhaps the ND AB plugin.

I dropped all these on a blank chart, then selected Edit Formula..., but IMO it's easier to just view the AFL in an external editor. I like Visual Code, but Notepad works (is notepad the best Microsoft can do after 20-30 years???)

On my installation, the Norgate Data AFL is located at C:\Program Files\AmiBroker\Formulas\Norgate Data.

Viewing those files in Visual Code, the main file seems to be C:\Program Files\AmiBroker\Formulas\Norgate Data\Norgate Data Functions.afl, as this is #included in all the remaining files.

Those functions look to just be helper functions wrapped around the AFL GetExtraData function, for example GetExtraData("NorgateBusinessSummaryLastUpdate"), with the rest of the AFL being outer functions calling the helper functions (for example to Plot() something).

IMO the documentation for the extra data could be better on the Norgate website, but this AFL, plus a bit of hacking, may get you what you need.

To me, it doesn't look like Norgate Data Functions.afl is complete, compared to the extra/fundamental data available in the URL above. Plus all the variable names in the code are prefixed with "Norgate", none of which are noted in the Norgate URL as the variable names.

For example, in the URL, search Metadata, Security Information - Text, FieldName = "Business Summary", but the AFL code says "NorgateBusinessSummary". See the function NorgateBusinessSummary() in Norgate Data Functions.afl.

All of this is untested ;) - I'm making the assumption that the AFL provided by Norgate, um, works :)

If I were to augment their code, I wouldn't edit their files directly, as they might get overwritten if I re-installed Norgate Data Updater. Instead, I think I would create another folder under C:\Program Files\AmiBroker\Formulas, such as CustomNorgateData or MyNorgateData or somesuch, then put my code there. And have a good backup somewhere in case I reinstall AB and my custom location disappears! I'd then follow a similar pattern as Norgate; have one uber-AFL file with all my helper functions, then other functions that call them (such as to plot or filter). I may be wrong on this best practice approach due to my AB inexperience, but I'm sure someone will correct me if I am (and I'll learn something in the process).

I believe you are a discretionary trader, but perhaps this will help you write up some filtering rules that would save you a lot of time when hunting for your next winners?

Hope this helps,
Linus

While I certainly appreciate the effort, I haven't even had time to print the book, never mind read it. I'll get round to it eventually...

jog on
duc
 
Top