qldfrog weekly Skate inspired system

[Newt]

So now the hard data is here,yes,my attempt at simulating the missed opportunities due to open order restrictions backfired for backtesting.
Now fixed.the system is "solid" but slippage is killing it.
I always dismissed slippage, was wrong .my initial systems were trading less volatile shares and my orders were executed.
not anymore and i pay dearly
So options are:
1)
Ensure orders are first in the queue..but much is out of my control,
2) get more traded tickers. So stricter volume ema etc
3) being a weekly system,let run the order longer during the week..but then i will probably just catch extra dogs of the selection

Very difficult to model that kind of unpredictable slippage too no doubt QF.....

 

[rnr]

To illustrate the differences between live and BT;
of the 37 trades requested at open this morning:
1 was missed:
I had an order for a parcel 8840 PNN at $0.475
PNN opened at 48c to close at 56s
so I missed $700 just on that trade today ;
The BT will capture the win.
Not that it was a bad day : my systems went up1.8% today, but yes we will always have some losses along the way

Hi @qldfrog,

I may be way off track here, but I don't understand how the BT will capture the above trade as a win based on the following price data:-
OPEN $0.485 | HIGH $0.58 | LOW $0.48 | CLOSE $0.56 with your order set at $0.475.

Cheers, Rob

 

[qldfrog]

Hi @qldfrog,

I may be way off track here, but I don't understand how the BT will capture the above trade as a win based on the following price data:-
OPEN $0.485 | HIGH $0.58 | LOW $0.48 | CLOSE $0.56 with your order set at $0.475.

Cheers, Rob

The BT buys on open at open price so will record a buy at 48c and does not check volume either .
In real world i got zero

 

[rnr]

The BT buys on open at open price so will record a buy at 48c and does not check volume either .
In real world i got zero

Sorry mate, I still don't get it!

How can the BT buy on OPEN at $0.485 when your code sets the max BuyPrice at $0.475. Even if the order was left standing all day, the LOW of the day was higher than your max BuyPrice.

Cheers, Rob

 

[qldfrog]

Sorry mate, I still don't get it!

How can the BT buy on OPEN at $0.485 when your code sets the max BuyPrice at $0.475. Even if the order was left standing all day, the LOW of the day was higher than your max BuyPrice.

Cheers, Rob

My code does not set the max price at 47.5c, my code says..buy at open
Irrespective of price
It is probably possible to put a limit to the buy at open but not easy...
The trial i did to ensure the buy was restricted to my real live 3% from last close failed lamely..so the start of the thread with wonder system
I am a taker to any ab code doing:

Buy at open if open if less than close plus 3pc , but discard and do not replace that buy if that condition is not present.
It can be done..but is it worthwhilethat would give me BT matching the real world..
And the 3pc limit has to vary based on the price range but that's not too difficult to implement.

 

[qldfrog]

Monday was very good and today very bad: systems lost 1.4% today on a flat market..outch
still above last friday close but not by that much anymore..tomorrow is a new day,,,

 

[qldfrog]

to highlight the volatility and erratic market we are facing lately especially on small caps
2 very bad days in a row, thankfully, the buy-stop was on at open and i did not buy any packet today, only offloaded on the daily systems
But lost 7k on the whole of systems or 2.3%...and as usually from half of that in the early afternoon to full scale disaster at the end of the sesssion..while the index actually caught back up during that time ...
Interesting check done:
out of the 11 packets which were not purchased due to the buy-stop trigger, none ended positive and the average loss of these 11 would have been 4.03% yet, a nice addition

 

[qldfrog]

What a week, 1 super day, 2 super losers, and today a pretty good +0.8% increase in an ASX pretty bad day
During the week, XNT..my reference, fell 1.23% hum,, not a good week;
-1.23% here go 10 years of term deposit returns..ahhhh the miracle of compounding..Anyone remember this ? We live in a weird time
but back to systems
the daily:
the intra day volatility experiment:
-$90 or -0.58% but at risk over the week end as its exit partly failed..not something I want to see again

my normal dailies:
MOA: 27% invested only -$3.2k or -6.2% BAD
DL guppy :13% invested+$0.5k or +1.17%
volatility ASX +$0.4k or +0.73% fully cash
volatility US +$0.2k or +1.36% fully invested

weekly:
QF sectors: fully invested +0.3k or +0.45%
QFDuc: fully invested +0.7k or +0.94%

overall systems: -$1.19k or -0.39%
notthat great but much better than XNT; really let down by MOA this week but this is a very agile system so volatility is expected

Sadly the discretionary lost $5k but overall still well ahead from an ASX ETF exposure this week

 

[qldfrog]

What a week, 1 super day, 2 super losers, and today a pretty good +0.8% increase in an ASX pretty bad day
During the week, XNT..my reference, fell 1.23% hum,, not a good week;
-1.23% here go 10 years of term deposit returns..ahhhh the miracle of compounding..Anyone remember this ? We live in a weird time
but back to systems
the daily:
the intra day volatility experiment:
-$90 or -0.58% but at risk over the week end as its exit partly failed..not something I want to see again

my normal dailies:
MOA: 27% invested only -$3.2k or -6.2% BAD
DL guppy :13% invested+$0.5k or +1.17%
volatility ASX +$0.4k or +0.73% fully cash
volatility US +$0.2k or +1.36% fully invested

weekly:
QF sectors: fully invested +0.3k or +0.45%
QFDuc: fully invested +0.7k or +0.94%

overall systems: -$1.19k or -0.39%
notthat great but much better than XNT; really let down by MOA this week but this is a very agile system so volatility is expected

Sadly the discretionary lost $5k but overall still well ahead from an ASX ETF exposure this week

Note systems tonight are 38% cash

 

[qldfrog]

As mentioned on other posts, the Frog couple is planning to go O/S for a few months .
Some of the places will have no or limited internet and with time difference, it would be too consuming to carry on daily systems.
By chance..actually due to filters...all my dailies are 100% cash tonight!;
As such, it is ideal timing for a switch /pause on daily systems or at least some of them.

While less performing, the daily volatility systems are decoupled from others and able to gain even in crashing market.I will keep them open till the new year, and will probably pause the less performing daily system.

then start moving to weekly ones on Monday:

I developed a new weekly based on breakouts so will add this one plus probably join Mr @Skate Platinum experiment as a different more cautious approach.
Will keep you posted

 

[qldfrog]

MOA will be paused and DLGuppy will carry on as well as the volatility daily systems;
MOA assets and some new funds will be moved for both platinum and a new BO based weekly

 

[qldfrog]

end of the week again
XNT went up 1.9% so good week for trading community,
overall Frog balance system and discretionary asx nyse did well this week too, but mostly outside the systems

the intra day volatility experiment:
+$117 or -0.76% fully cash before week end, was hardly played this week

my normal dailies:
MOA: -20$ and closed tonight ; was not a good experience as the bat lost -$15.5k since the start of FY;or 24%
the last code version was only run since end of august and much better but still lost just below 5k (10%) since end august

DL guppy :95% invested+$1.3k or +3.15%
volatility ASX fully reinvested +$0.2k or +0.4%
volatility US fully cash +$0.3k or +1.5%

weekly:
QF sectors: fully invested , one position too much :-( my mistake +0.2k or +0.3%
QFDuc: 95% +1.4k or +1.7%

so not flash as we are 20% cash by close tonight and only went up a bit above 1%

 

[qldfrog]

end of the week again
XNT went up 1.9% so good week for trading community,
overall Frog balance system and discretionary asx nyse did well this week too, but mostly outside the systems

the intra day volatility experiment:
+$117 or -0.76% fully cash before week end, was hardly played this week

my normal dailies:
MOA: -20$ and closed tonight ; was not a good experience as the bat lost -$15.5k since the start of FY;or 24%
the last code version was only run since end of august and much better but still lost just below 5k (10%) since end august

DL guppy :95% invested+$1.3k or +3.15%
volatility ASX fully reinvested +$0.2k or +0.4%
volatility US fully cash +$0.3k or +1.5%

weekly:
QF sectors: fully invested , one position too much :-( my mistake +0.2k or +0.3%
QFDuc: 95% +1.4k or +1.7%

so not flash as we are 20% cash by close tonight and only went up a bit above 1%

had to check:=) , if I had been trading MOA this week, I would have gained just below 1k or 2%

 

[qldfrog]

For the people following my learning path:
as explained, I closed MOA (the old bat) to switch to weekly;
If I can escape our Queensland despot, Frog will go O/S in the new year probably February.
I will need to have no daily at that stage as i will not have daily internet coverage then.
I so took the opportunity to close the fully cashed MOA and will start a new weekly, very basic BO system from Monday.
The coming months should allow me to fine tune and iron out that system before departing O/S
That is the plan.
It is not without risk as i found weekly systems frustrating during crash as you are waiting to the end of the week to offload crashing shares...

 

[qldfrog]

and while in a talkative mood:
daily Guppy had 40000 EUR and a sell order at open this morning;
sold at 15c the share ended up jumping 42% today closing at 18.5c...arrrrrg 1.4k missed
I made a 43% gain (bought at 10.5c) but could have done so much better
Maybe my exits are too tight...

 

[KevinBB]

It is not without risk as i found weekly systems frustrating during crash as you are waiting to the end of the week to offload crashing shares...

You could always introduce the rule that position entry only happens on a fixed day of the week (say, Monday), but position exit is whenever the selection criteria fails. I don't know your system, but it could be as simple as a MA crossover turning for the worse.
KH

 

[qldfrog]

You could always introduce the rule that position entry only happens on a fixed day of the week (say, Monday), but position exit is whenever the selection criteria fails. I don't know your system, but it could be as simple as a MA crossover turning for the worse.
KH

Yes, i tried to backtest using a daily SL on a weekly system, just before the 2020 crash.it was not very good as BT result.i disabled it...and the crash came.
Since, i have had very tight exits on weekly systems, and took the aggressive systems to daily

 

[qldfrog]

This morning at open, started the new weekly QFBO:
6 positions purchased but could only get $2k of PNN vs $5k requested..we will carry on this way
It was not a good start as the 6 positions lost 1k on day one with AVZ falling 17%
Pretty eventful
QFsec and Platinum missed TZL which is doing a SPP at 12.5c vs last price of 17.5c..we probably dodge a bullet here
QFDuc missed KRM which went into trading halt
and QFSec missed MQR which jumped 30% today.....
my systems went from +$5k around 11Am to +$400 at close:-( , as per the usual daily grind, except that today it was a 1.5% overall fall between morning and close, not usual 1%
But had a great day on discretionary....so mood is still sunnies and lollipops

 

[qldfrog]

Sharing what I think is valuable experience;
For quite a while, I experienced significant difference between my backtests results and the same code and realm trialed on some other members setup;
I run amibroker and premium data feed with the plugin
we had endless back and forth between the merit of historical data (from Northgate Data) or not (my view being it has limited value as we had a great set of market samples in the last 3 y and a relatively stable set of tickers.
)
I also remember getting slightly different results between systems explore on my main platform and a cloud setup..with the same code..
Anyway, a few in depth tests in the last few days really pointed to deep differences and when narrowed -> showed actual differences in the watchlists used in Amibroker, these preset lists handle the Realm filter: belongs or not to All ord, ASX 200, etc;with my list definitively outdated
After a full reinstallation yes, my list were up to scratch again and my backtests were matching parallel runs as expected.
When you do an update as i do at least daily, these lists are not updated;
You need to do, in Amibroker, a "Tools->ASX-Premium " script run at least weekly;
yes it is somewhere in the manual
RTFM..so my mistake
but genuinely the first time I amaware of this;
As your realm definition decays, it slowly as days go, make your testing and exploration/run suboptimal to the point of irrelevance after a 3y old installation.
Pretty nasty IMHO;
That explains why I was getting better outcomes on wider universe..less influenced by this issue..

Worse, in Premium Data there is something called

which I used, thinking it was actually refreshing these lists..well it seems it does NOT.
So lesson learnt now, that has probably cost me the price of a few German cars in the last 2y;
All my systems optimisations and basically code feasibility have been affected;
i have fixed the QFBO version so will run a QFBOv2 from next week.
But I have serious job ahead for QFDuc, QFSec and DLGuppy; both volatility systems remains relatively unaffected

 

[DaveDaGr8]

Is there a reason you are still running Premium data instead of upgrading to NDU ?