qldfrog weekly Skate inspired system

willoneau said:

What about husband and wife running it as a business and splitting tax on profits in stead of capital gains?.

Click to expand...

It will be income indeed, same taxation , probably in a company to get the 33pc ..as an individual it would be worse, and it will not take long to reach 33pc tax rate especially if these profits are reinvested: you need some other income to buy food and pay rates
You may not split income that easily...just saying that a backtest result is 60pc lower in real world just after considering tax here.

And that any business is facing similar issues,less pronounced as you can reinvest some gain into machinery etc but whole business areas face the same problem as above
The point is: in a place like Australia, unless you can benefit from the super system..and for how long?, you are facing a very unfair battle if only with tax rates in place
So back to the only Australian game buy real estate and flip it after 5 years or bank shares..a nation moving toward a bright innovative future

Warr87 said:

there are ways to minimize tax. talk to a CPA. but i'm not sure what your point is. There are other places in the world that have higher tax than we do. Some have lower. But those who do tend to have lower do not to have the services that we have. The govt. isn't a charity, and to have the lifestyle we have things need to be paid for. if your complaint is having to pay millions of dollars in taxes, you are likely going to have that complaint fall on deafs ear...lol. if you would like a rethink of corporate tax, that sounds fine.

I would follow @willoneau train of thought and run it through a family trust with a bucket company

(also my understanding is that given the frequency you would like treat the gains as income not capital gains, unless the particular stock was kept for a particularly long time but that is something your accountant can answer)

Click to expand...

Yes indeed CG treated as income as there is low chance to keep shares more than a year...
Anyway, do not forget the compounding taxation horror in any scenario..as for the gov is not a charity, my issue is that it is for at least 50pc of the population in Australia.

a bit slack with the holiday season but time to report our weekly Christmas rally figures
Well got on kind of:
a good week:
System1 gained $773 and system2 a nice $1332
Flipper more or less static as only one stock

###########################
System I (the original)

Start 25/02/2019
Total invested $ 100,000.00
Cash remaining $22,729.49
Current portfolio value $112,394.63
Profit $12,394.63
Nb of active positions 12
Nb max position 20
Return since 25/02 12.39%
Annual return so far 14.93%
purchase value per position $5,619.73
Invested percentage 79%
dividends $ 1,387.81
Average per week: 442.6651786



no action on Monday
###########################
system II

Start 5/08/2019
Total invested $ 100,532.76
Cash remaining $834.75
Current portfolio value $102,194.55
Profit $1,661.79
Nb of active positions 20
Nb max position 20
Return since 5/08 1.65%
Annual return so far 4.22%
purchase value per position $5,109.73
Invested percentage 98%
dividends $ 189.93
Average per week: 332.358


kind of OK vs XAOA with a better week this week
3 sells, 1 buy on open monday
we will end up with 18 positions...

The flipper (systemIII) has 2 buys on Monday, one being under a takeover but as nothing is finalised yet I will take it as the chart looks decent overall

Have all a great week end and hopefully a nice holiday family break

Today, the flipper (system 3) bought MAH and VRL (vrl is under takeover offer but...)
System 2 sold TNE, IMD and PDL for a total profit of $132.5and got a buy which is not filled due to low market today and high price gap..will see tomorrow
System 1 unchanged
Surprised to see the market falling by that much today, was expecting a kind of eoy rally just to prop up profit of fund managers etc.They seem to have tried as the final fall is just 0.2% but was pretty bad earlier in the day

Another short but good week
a good week:

###########################
System I (the original)
Gain of $945

Start 25/02/2019
Total invested $ 100,000.00
Cash remaining $22,731.74
Current portfolio value $113,340.56
Profit $13,340.56
Nb of active positions 12
Nb max position 20
Return since 25/02 13.34%
Annual return so far 15.81%
purchase value per position $5,667.03
Invested percentage 79%
dividends $ 1,387.81
Average per week: $290


Caught up with xaoa
no action on Monday
###########################
system II
gain of $366
Start 5/08/2019
Total invested $ 100,532.76
Cash remaining $15,061.09
Current portfolio value $102,560.60
Profit $2,027.84
Nb of active positions 17
Nb max position 20
Return since 5/08 2.02%
Annual return so far 4.97%
purchase value per position $5,128.03
Invested percentage 84%
dividends $ -
Average per week: 88.16695652



Caught up with XAOA
Last Monday buy on MAQ could not be executed and I gave up on that one, I will need to check that my volume check is still present
But this is history
On Monday:1 sell 4 buys
we will end up with 20 positions and fully invested

The flipper (systemIII) lost $191 and buys on monday

Have all a great week end , stay safe with the fire and the preparation many do around their block or house

Life goes on...
flipper bought FLC and WAF
System 2 sold AQG for a loss of $39.71 after brokerage
And bought CNU, IFN,GRR and AVN
rolling on

Got an interesting message today from a most experienced trader here highlighting the fact that my system..we are talking system 1 was basically on par with the XAOA and so had no edge:
indeed to date I am basically even with xaoa if started on the same day: same performance so why bother
Obviously, there is a teaching in this and I learn a lot, I add SL oe at least SL type exits, I control my emotions, this was not a dream run as RBL was enough to loose me 4k profit or a third.
Where I find some solace is that I just confirmed one of the belief I had when questioning the results vs just adding some VAS to my portfolio;
It took my system until September to be 80% invested;
On average SystemI has been 64% invested so these 13k profits in the last 11 months were built on an average of $64k or so;
That looks much better and definitively help me to sleep better

So should I be happy? Never, I am not complacent with results like that....
My SL is better but not satisfactory, I need to see if I could review both entries and exits,
Using Peter's style display with still a few hundred dollars discrepancy but overall real

Winning rate is low, hardly better than a two up game...Where is the edge...
If I were to run that system from 01/01 to 31/12 (full 2019 calendar year), I would get

(backtests actually match my real life results so i am confident there)
I am happy with such a system

As a final check; i started the backtest 25/02(as real world system1):


and the real world results are bloody similar:13.3k gain, 52% winning rate 64% exposure

So I really did welcome that opportunity to do a full system check.
Anyone seeing something fundamentally wrong, please tell me
And it was definitively a welcome check.

The 23% return is still very respectable IMO. The limited sample may not be enough to determine your edge. I guess it also comes down to why your system wins? If it's through rising momentum and the market has been choppy, then you wont have a statistical edge outside of upwards momentum. The edge could also be its ability to withstand a drawdown or minimise exposure to undue volatility. this impacts your ability to recover from what is an inevitable downdraw. The fact that your system isn't 100% exposed but keeps pace with the XAOA shows that $ per volatility, you are more efficient.

The more I read, and listen to podcast's of other traders, the more I am realising that exits play a big part. It's been mentioned here as well. And as it was so eloquently discussed in one of the recent recent books I read, the exit determines how much of your profit you are willing to give up. I think it was Kevin Davey's podcast who also mentioned that his most successful exits is actually n-bar exits. @Skate does something similar with stale exit. I guess it comes down to: "why does your system trade", or perhaps "why does your system win".

Some ideas, as I have been thinking about this a lot.

• For a quick momentum system, you need to identify as soon as momentum has slowed. (ADX or MACD could be good candidates.)
• Consecutive down bars (or even changes in a short range RSI as used in mean-reversion type systems). If you are longer term then this may not be good as you would want to weather a PB.
  
• Tightening your stop based on volatility. To measure volatility, using the ATR is common. Perhaps a condition that has volatility increases your trailing SL tightens to preserve profit?
• Of course, a simple Close < Moving Average can be good too.

I'm still trying to figure things out for myself, and haven't even graduated to real trading yet so perhaps this should all be taken with a grain of salt.

@Warr87 , I indeed spent a long time working on exit conditions for system 1,
without going into exact code;
I now have exits triggered on:
stale , C<MA indeed,
a hard % of entry price if Signal > MAC,
conditional(market condition) C <ema and even on close below trailing ATR fct

Some of the parameters are conditional on general market conditions
So yes, getting out is basically the most complex part of system1
I hope this helps people but these are grossly the areas I have worked on.
I dismissed a simple broker trailing stop as I do not trust the outside implementation due to weird previous experience.
So I remain exposed to fast sudden major market crash

My observations on System 1: I have to start by saying that I don't know the underlying strategy. Looking at my charts of the posted system 1 buys I got the impression that the buy signals were a little late. 2 or 3 bars can make a big difference to the average win. I think @Skate was making that observation also when he posted his charts of a few system 1 buys early on in the thread.

There are ways to measure entry efficiency for a discretionary trader but they don't apply here because @qldfrog is using an algorithm. His algorithm finds the buy signal at the right time every time.

My observation that some entries are late would require a different algorithm that identifies an earlier buy signal. Without knowing the strategy setup criteria I can't offer more specific suggestions.

W% - Consistent with trend following systems in a range bound market. Once the market resumes it's trend the W% will increase to 60% while the trend lasts.

I'm as pleased as you are, to see that your getting close to what your system is designed to get.

Regardless of the strategy (that I don't know) it seems strange to me that system 1 is not fully invested by now. Early 2019 was a trend followers dream market although I must temper this by remembering that the 2019 rally was a classic "V" reversal. Systems that include a market filter will lag significantly after a "V" reversal and miss many of the early buy signals that produce well above average winners. The 2019 rally did go for six months and IMO trend following systems should have had time to get fully invested especially if the manager is comfortable with the portfolio heat required to run the system.

Just some thoughts and if I can help further please ask.

qldfrog said:

system 1 at the open today bought:
LIC 610 @$8.92 for a total cost of $5451.2
MSB 2718 @$ 2.01 for a total cost of $5473.18
UOS 6292 @$ 0.88 for a total cost of $ 5546.96

Click to expand...

LIC provides an example. Your buy signal looks good but I'm wondering why your system hasn't provided a buy signal much earlier in the trend? This latest buy signal is closer to the end of the trend than the beginning. If the portfolio was fully invested when earlier signals were triggered then that's a reasonable excuse for not having it in the portfolio already.


I'm trying to help you improve system 1 performance by starting trades earlier in their trend.

Thanks Peter,
I really appreciate your input and will work on it asap
i think you highlighted actually 2 key issues on sys1
1)
Sys1 went out of full investment early November quite drastically from 100 to 80pc in a week or two, i will check the details and see what causes this.just luck..many entries turning stale or
a major problem with my index conditional entry maybe.
i designed sys1 as conservative with sys2 more dynamic but using similar algorithms
Comparing both should give me some clues.
I will share my findings should it help others.
2) late entries.LIC a good case indeed.i need to reverse engineer my code and understand bar by bar how i could improve it

You all remember how we were working hard on ensuring we were not reading the future in backtests, so i was very careful there and lately wondering if i am not actually going too far and delaying entries by one bar..will check if i do not have a pure coding issue..if i i c not detect anything wrong in the code itself


All great comments and food for thoughts..now i need to find the time to act on these

peter2 said:

LIC provides an example. Your buy signal looks good but I'm wondering why your system hasn't provided a buy signal much earlier in the trend? This latest buy signal is closer to the end of the trend than the beginning. If the portfolio was fully invested when earlier signals were triggered then that's a reasonable excuse for not having it in the portfolio already.

View attachment 99532
I'm trying to help you improve system 1 performance by starting trades earlier in their trend.

Click to expand...

Apologies
The post is in-detail, lengthy & posted in "qldfrog-weekly-skate-inspired-system thread" rather than the 'Dump it here' thread with the sole purpose of explaining my alternative entry & exits in a position currently held by 3 of my strategies. Sometimes looking at charts of other traders can stimulate thinking on a higher level to understand why those positions were taken & why.

Disclaimer - Currently I'm holding ASX:LIC in three of my strategies
I'm a current holder of (ASX:LIC) in three of my strategies & as I was mentioned in the post above it's a perfect opportunity to explain & display my actual entry points in my PANDA, CAM & HYBRID Strategies showing the timing of each strategy into the trade. Also I would like to remark that trading multiple strategies jumping in & out of trades at different times rather than buying & holding a position "over the long run" has its advantages. @peter2 has included a chart with marked entry points - my charts are actual positions taken.

Why trade the same position three times
I have canvassed this question with @peter2 here: https://www.aussiestockforums.com/posts/1046625/ & @peter2 answer is found here: https://www.aussiestockforums.com/posts/1046639/

Timing
Each of my strategies will enter & exit at different times & at different stages within a confirmed trend (confirmation is the key word) & sometimes my strategies will enter on a pullbacks within the trend.

Mechanical Strategies
Trading a combination of strategies allow me to "Dance-to-the-music-being-played"- meaning there are a multitude of entry points into the same trend as displayed by @peter2 chart of (ASX:LIC) above. Strategies have their own coded reason (subject to condition's being met) to enter & their own reason to exit a position (each of my strategy have a multitude of different & varying conditions that need to be satisfied). Entry & exits are conditional & can be different entirely to any of my other strategies within the same move, I'm aiming to take full advantage of a trend & it's very difficult for me to combine & code it all in a single strategy. In saying that my HYBRID strategy relies on three different strategies vying for the fastest confirmed entry point, only the first signal is taken & the others not displayed.

Moreover
On occasions my strategies can be in the same position at the very same time. The three charts below displays the entry points of (ASX:LIC) Footnote: all three strategies listed are currently still in the trade.

Reading the weekly charts
I'm a visual guy so I have code the charts to display the information that's important to me with colour coding because sometimes I'm lazy (a picture paints a thousand words - as they say). The charts display (to me) what other traders are currently thinking with a display of their emotions (the yellow ribbon) about the individual security. I won't go into a detailed explanation, other than to say it's all in the colour coding of the bar charts & ribbons. The white squares on the price bar is the "actual entry bar" & the yellow circles are the "actual exit bar". The yellow & white arrows are the signal bars.The lower ribbons display if the index is on or off (Green & Red) & the Yellow ribbon means caution (a period when the security is unloved) A yellow ribbon with two confirmations I'm off the sucker (quick smart)

Timing the 'Exit'
Season traders are all about chasing stock that has volatility & rightly so but the importance of an exit can't be overstated because it’s important to remember that the EXIT really determines the final outcome of a trade - leading to the long-term success or failure of any trading system. IMO - the timing of your exit has to be the most important part of your trading Strategy as I'm a firm believer that timing the "exit" is where the money is too be made.

Three Charts of my current trades in ASX:LIC
1. The PANDA Strategy (Variable Trailing & StaleStop)
2. The CAM Strategy (One HARD Stop)
3. The HYBRID Strategy (Variable Trailing & StaleStop)




















Also
The result of the closed trade (green dash line between entry & exit) only forms or paints when the position is closed

Skate.

Any relevant entry is most welcome @Skate , and we can all learn so much of these interactions.
As least at my humble beginner level
I have a bit of beekeeping tasks to do today but much to work on after Peter and Skate input

qldfrog said:

I have a bit of beekeeping tasks to do today but much to work on after Peter and Skate input

Click to expand...

Sorry off topic …. You are an apiarist Frog?? … Frogs and Bees … unusual bedfellows

My old Man (who is really old now) used to keep a few bees …. absolute best honey in the world when I was a younger whipper snapper ….. Nature at its best these bees are

Back on topic … Good thread

Out of topic:
@barney hobbyist only: 3 hives, a few more after swarms but keeping down to 3 in winter, bees are an amazing wonder, was just discussing with my wife about a piece of comb I got today:
admire not only the precision but also as seen with the transparency :how they are designed so that the opposite sides actually reinforce the whole structure;The walls themselves thiner than a mm

perfection..not like my trading...
Interlude over
after all our exchanges above I will investigate:
1) why am I not fully invested now: and is it an issue?
Could I have my index based conditions wrong?
2) am I actually entering too late?
Is the trigger wrong or too conservative, my code faulty?
reusing both Peter and Skate LIC example, how do I stand, why?

Pb 1:why am I not fully invested now?
the 11/11/19 and 18/11/19 saw a 100% to 64% exposure fall and only a slow ramp up after


When running Sys1 January 2020 code in backtest(with matched (I checked) explore instructions
I actually get 32entries (buys with or without sell) and have 17 active packets
But as you might have seen in my weekly reports, I have traded far more...

It is no mystery I have changed the code slightly as I was going since February, refinements and SL/exit criteria
indeed I traded 16+37=53 buys..
Initially wondering why, looking at older reports I then realised I was using aversion since discarded afterI I worked on SL and exits.
indeed stored reports with that version matched my real world action for that period.
on the 11 and 18/11 I offloaded 6 packets which when tested now would still be in
For example SHV purchased on 26/08 was sold on 11/11
but now

So with current code all along I would have been much better off with less brokerage and more exposure -> the SL exercise in December has been useful

Moral: any system changed while trading is hard to check, here the change is positive but it could have been worse and hard to detect
And if you have checked once your backtest matches your real world operations, this is only valid till your next code change..obviously but I got caught...

Now the LIC example:
indeed very late entry

I investigate

And to summarise:
I realised yesterday thanks to the seemingly unrelated: "why am i not invested fully now?" How deeply i FU initially and how i sorted it , blissfully unaware ,when working on my exit...
The market being so good last year for trends was hiding the depth of wrongdoing...but it has been expensive vs optimum total return
Even so, better than my super results...
I will also now store my explore results weekly to be able to understand the impact of gradual code changes if any
Now to a deeper issue, why am in so late? And do i want to change that as this system is supposed to be conservative over dd periods?
More work..i toyed moving into finance in the 2000's but decided for Brisbane vs Sydney...i now do it in retirement..have all a great day.stay safe with fires

It's all good @qldfrog you'll be tinkering on these systems for the rest of your days. I'm reminded of @Country Lad 's sign off, been doing this for 36 years and just getting the hang of it.

It's possible that your system algorithm doesn't know if a trade is already open in a stock and reports every buy signal when it occurs. Subsequent signals on open positions can be used to add (pyramid) to winning ones should you like to do that.

If your report lists redundant signals on open positions then I'd check the chart and look for signals that occur soon after the trend has started and buy them first.

When I look for Darvas boxes to trade I prefer to trade the 1st or 2nd ones not the later ones.