qldfrog weekly Skate inspired system

[divs4ever]

once CBA is on again, I can complete the total investment result, another screaming night in the NYSE and total investment up 0.9% last week nearly equal to the XJT but after living expenses, shed, patio, renos, holidays..etc
View attachment 209002

Very conservative now, even with these screaming results:
overall investment as follow:
View attachment 209001

CBA might be cheaper on Monday after a massive snarl up after an unexpectedly long downtime caused by planned maintenance , seems all sorts of people ( but not me ) got caught when the payment systems ( including pensions ) weere not available until mid-morning

( CBA is on my AVOID list , but my exposure is via various ETFs and LICs , so maybe i will get an opportunity with one of them )

 

[qldfrog]

Well another glorious week. Let the good times roll.
The dogs of 2024 went really good again, even laggards are now favourites
+2.75% on the week nearly at break even:

The system not so good with a -1% loss on the week:

But the super overall did very well again:+1.4% and 17% cash allocation
Another ATH

Getting all the dividends also helped the overall investment and we should breach again our ATH.
We will see tomorrow after the US session.
Hopefully will be able to do a debrief on the RRG system behaviour in the last 6 months or so

 

[qldfrog]

RRG system analysis
If I run my system in backtest mode between its actual start date and today, I get a great result:


So a 25% profit vs current real world system not even positive;
So question is why?
I so listed the backtest trades vs the real world trades
and highlighted in green the properly matching trades

The difference was on the selection of 2 entries on the 24/03 following 2 exits (in both cases);
I entered into NUF and ISML for the system whereas the backtest entered DRO with a $10k ultimate profit and SPR
Both ISML and NUF exited the next week with a minor win and created 2 new entries vs the backtest etc etc..cascading effects
And to be very honest, I suspect these 2 entries are a manual error;
I was in Sri Lanka the 2 previous weeks, Singapore on the 21st of march and landed back in Australia on Sunday the 23/03 when I had to enter the trades for the Monday open.
It is so very likely that I missed the Amibroker output or score ranking.
That would be the most likely reason;
There is still one worry, on the week of the 20/06, the backtest selected ZIP for a big profit whereas my real world selection went for LTR, so do we have a forward looking issue in the score setting?
It demonstrate 2 things:
1) however rigorous you want to be, if human is involved , **** will happens.
Life ,death or accidents, holidays..Sure , all is setup on an AWS server for remote access but your phone screen is small, internet not so great..not ideal and prone to error;
It also demonstrate how relatively fragile results are.
There ois nothing really wrong in the implementation run vs the backtest, but that single error missed some good entries and so increase the turn over..and the real system has a WAF parcel which will cost a bucket again vs the backtest which is WAF free..