Dump it Here

[Skate]

Strategy update
As "The Platinum Strategy" had a rocky start with a few members buying outside the trading plan. By group consensus, we have called the first week a false start absorbing the small losses (outside of the strategy). The ongoing weekly reports will still use the original portfolio size ($100k) with 10-positions for others to follow along.

Disclaimer
There is not one member of the group trading with the same investment level & their bet size will be reflective of this. The only issue with varying bet sizes is the calculation of the number of shares to buy as the "offer price" that is fixed & not negotiable. They have all confirmed the calculation of the number of shares to buy at the "offer price" will be simple to calculate.

Skate.

 

[Lone Wolf]

As an exploration user, it is true that you can sometimes miss a sell order...
Human error
And that sell instruction will not popup anymore
as a result i do an explore, enter the trade then check graphically that none of the remaining portfolio has a buy attached
Not ideal, time consuming

I keep a watchlist of active trades. Each week (I only trade weekly / monthly) I update the watchlist and manually step through the chart for every open position to confirm everything I hold is something I expect to still be holding. I'd do this regardless of how much faith I had in computer generated lists. Obviously this would be more time intensive if you have more systems and/or lower time-frames.

 

[qldfrog]

I keep a watchlist of active trades. Each week (I only trade weekly / monthly) I update the watchlist and manually step through the chart for every open position to confirm everything I hold is something I expect to still be holding. I'd do this regardless of how much faith I had in computer generated lists. Obviously this would be more time intensive if you have more systems and/or lower time-frames.

yes it was getting intense with many daily systems but I caught many mistakes in the past year, and better fix these asap

 

[Linus van Pelt]

As an exploration user, it is true that you can sometimes miss a sell order...
Human error
And that sell instruction will not popup anymore

I'm not sure what you meant by "And that sell instruction will not popup anymore. Is it a glitch in your explore that didn't generate a sell signal, or just you missed it, so it won't show up in following runs? I'm thinking it's the latter.

For the most part I'm using "backtest to trade" with the future bar vs. explore. Both approaches have their pros and cons.

But what I do in my explore-based trading:

* When I create a new system, I create a dedicated watchlist for that system to contain all open positions. Say @My System - Open Positions
* When my explore generates buy signals, I RMB the buy rows and choose Add selected results to watchlist
* It's much better to have symbols in the watchlist that are not in your portfolio (say it gapped above the buy price or you didn't get in for some other reason) than to have a symbol in your portfolio that you forgot to add to your watchlist. Keeping that watchlist up to date is key, but not *that* hard.
* Then, in your explore code, use the InWatchlistName function to return 1 or 0 depending on if the given symbol is in your watchlist
* You can then either highlight sells that you actually own - say red background, or an AddTextColumn that says "<<< YOU OWN THIS", or even filter the sell signals to only those in the watchlist (but this is dangerous unless you're meticulous about updating the watchlist

Anyway, using that watchlist to help flag the sells you actually own, instead of the 10's of bogus sell signals that you don't own, can help to cut down the human error. That watchlist is also useful to quickly cycle through the charts for your open positions.

It is good to delete the sells from the watchlist once they are sold, but it's better to have too many than too few in the watchlist. You can occasionally run an extract from your broker and then compare that to your watchlist(s) to clean up any stragglers.

Don't use Symbol -> Categories -> Watchlists -> Move Up/Move Down to re-order your watchlists, or you'll break every saved APX on your file system. Apparently Tomasz doesn't think this is a bug.

Instead, edit C:\Program Files\AmiBroker\Databases\NorgateData\WatchLists\index.txt to reorder your watchlists. Get an editor with a sort lines function. I like to put all my open positions watchlists at the top of the list of watchlists, another reason why I prefix them with @.

Hope this helps, and apologies if you knew this already...

Edit: D'oh, how did I miss Lone Wolf's post before composing this reply? Oh well, I put too much time into writing this to delete it now

 

[Greynomad99]

Man - there are a lot of posts here today with Skate leading the push.
Just for the record as a lot of these posts mention what sounds like algorithmic trading and indicators (like moving averages) I don't use any of these tools as most tell you what has happened, not what might happen (none tell you what will happen unfortunately). I stick to the tried and true theories like Gann and Elliott Wave which have been around for decades. If it was good enough for Mr Gann it is good enough for me but whatever works for you is the way to go - all (most) systems have some value.
My educational portfolio I set up last November and only look at once a week is up 28% for the year - which shows what a logical no frills trading plan can do for you. You don't really need more than a basic understanding of T/A to trade that system.
Anyway, I'm having a break doing some travel photography for a couple of weeks, so hope the market is kind to you.

 

[qldfrog]

I'm not sure what you meant by "And that sell instruction will not popup anymore. Is it a glitch in your explore that didn't generate a sell signal, or just you missed it, so it won't show up in following runs? I'm thinking it's the latter.

For the most part I'm using "backtest to trade" with the future bar vs. explore. Both approaches have their pros and cons.

But what I do in my explore-based trading:

* When I create a new system, I create a dedicated watchlist for that system to contain all open positions. Say @My System - Open Positions
* When my explore generates buy signals, I RMB the buy rows and choose Add selected results to watchlist
* It's much better to have symbols in the watchlist that are not in your portfolio (say it gapped above the buy price or you didn't get in for some other reason) than to have a symbol in your portfolio that you forgot to add to your watchlist. Keeping that watchlist up to date is key, but not *that* hard.
* Then, in your explore code, use the InWatchlistName function to return 1 or 0 depending on if the given symbol is in your watchlist
* You can then either highlight sells that you actually own - say red background, or an AddTextColumn that says "<<< YOU OWN THIS", or even filter the sell signals to only those in the watchlist (but this is dangerous unless you're meticulous about updating the watchlist

Anyway, using that watchlist to help flag the sells you actually own, instead of the 10's of bogus sell signals that you don't own, can help to cut down the human error. That watchlist is also useful to quickly cycle through the charts for your open positions.

It is good to delete the sells from the watchlist once they are sold, but it's better to have too many than too few in the watchlist. You can occasionally run an extract from your broker and then compare that to your watchlist(s) to clean up any stragglers.

Don't use Symbol -> Categories -> Watchlists -> Move Up/Move Down to re-order your watchlists, or you'll break every saved APX on your file system. Apparently Tomasz doesn't think this is a bug.

Instead, edit C:\Program Files\AmiBroker\Databases\NorgateData\WatchLists\index.txt to reorder your watchlists. Get an editor with a sort lines function. I like to put all my open positions watchlists at the top of the list of watchlists, another reason why I prefix them with @.

Hope this helps, and apologies if you knew this already...

Edit: D'oh, how did I miss Lone Wolf's post before composing this reply? Oh well, I put too much time into writing this to delete it now

I'm not sure what you meant by "And that sell instruction will not popup anymore. Is it a glitch in your explore that didn't generate a sell signal, or just you missed it, so it won't show up in following runs? I'm thinking it's the latter.

For the most part I'm using "backtest to trade" with the future bar vs. explore. Both approaches have their pros and cons.

But what I do in my explore-based trading:

* When I create a new system, I create a dedicated watchlist for that system to contain all open positions. Say @My System - Open Positions
* When my explore generates buy signals, I RMB the buy rows and choose Add selected results to watchlist
* It's much better to have symbols in the watchlist that are not in your portfolio (say it gapped above the buy price or you didn't get in for some other reason) than to have a symbol in your portfolio that you forgot to add to your watchlist. Keeping that watchlist up to date is key, but not *that* hard.
* Then, in your explore code, use the InWatchlistName function to return 1 or 0 depending on if the given symbol is in your watchlist
* You can then either highlight sells that you actually own - say red background, or an AddTextColumn that says "<<< YOU OWN THIS", or even filter the sell signals to only those in the watchlist (but this is dangerous unless you're meticulous about updating the watchlist

Anyway, using that watchlist to help flag the sells you actually own, instead of the 10's of bogus sell signals that you don't own, can help to cut down the human error. That watchlist is also useful to quickly cycle through the charts for your open positions.

It is good to delete the sells from the watchlist once they are sold, but it's better to have too many than too few in the watchlist. You can occasionally run an extract from your broker and then compare that to your watchlist(s) to clean up any stragglers.

Don't use Symbol -> Categories -> Watchlists -> Move Up/Move Down to re-order your watchlists, or you'll break every saved APX on your file system. Apparently Tomasz doesn't think this is a bug.

Instead, edit C:\Program Files\AmiBroker\Databases\NorgateData\WatchLists\index.txt to reorder your watchlists. Get an editor with a sort lines function. I like to put all my open positions watchlists at the top of the list of watchlists, another reason why I prefix them with @.

Hope this helps, and apologies if you knew this already...

Edit: D'oh, how did I miss Lone Wolf's post before composing this reply? Oh well, I put too much time into writing this to delete it now

Thanks @Linus van Pelt ,yes, that is a nice way to double check the explore output.thanks for the hint

 

[Skate]

whatever works for you is the way to go

Nailed it.

Skate.

 

[othmana86]

Spot on @Skate.

Chart below is of one of my live systems. Top green chart is the system's P&L, which includes open and closed positions.

Started trading the system in early 2015. It basically did nothing for the first 2.5 years, but really found it's momentum around mid 2017 and has really got some good traction since then. It was a real test of my patience to keep trading that system for 2.5 years and make little progress.

View attachment 132081

This makes me feel a whole lot better. I started my system back in feb, the same week we had a serious market crash. I dropped about 15 percent of my starting capital. I had the confidence to keep running knowing that the drawdown was within my systems limits. I tested the system quite significantly. Multiple MC runs and all. While it did recover some of it it’s been super choppy and not yet profitable. If you stayed strong for two years… I think I can do the same . In saying that it’s so hard not to doubt your system and constantly think “WTF is wrong with this thing”. So thank you for sharing

@Skate! I concur firing up your system at a particular time could have significant effect on the system. Will just have to wait so it smoothens out

 

[DaveTrade]

I'm finding all these posts about systems very interesting, it's got me thinking that a purely mechanical system would be the way to go for me as my main income and then my discretionary trading can be an interest on which I can vary the work load to fit in with my other interests in life. I'd like to try and build a 100% mechanical and robust system of minimums. Minimum time to operate, minimum drawdown, minimum consecutive losses and minimum stress. Can it be done? can I do it? I don't know. I think I can but I'm not so sure if the system I end up with will be as perfect as I now hope. The markets are like that, they can be perfect for a short while but then everything changes. This is the big advantage of discretionary trading, you can vary how you trade depending on the market conditions.
Anyway thanks for listening and keep talking systems.

 

[cynic]

A minimum drawdown, minimum consecutive losses, minimum time to operate, and minimum stress, system, is very easy to envisage, but not at all desirable to those seeking to profit from participation in the markets.

Perhaps one might derive more benefit from entertaining softer, less minimal, parameters.

 

[DaveTrade]

A minimum drawdown, minimum consecutive losses, minimum time to operate, and minimum stress, system, is very easy to envisage, but not at all desirable to those seeking to profit from participation in the markets.

Perhaps one might derive more benefit from entertaining softer, less minimal, parameters.

Thanks for your response Cynic. As has been mentioned in a previous post by Skate, your brief message has a much broader and wide spread impact than it may seem on the surface to some readers. From my readings in this thread I think that my biggest challenge may be with the stop-loss and drawdown. I knew that a trade off between return and risk would have to happen, what I don't know at this point in time is if I can find the balance that will give me my peace of mind in both. Obviously my leaning with this system would be towards lower risk and I'd be willing to pay for that. Each person should own and drive the car that makes them happy, a young guy goes for the sporty car and the older guy may choose the the more practical or more comfortable one.

 

[MovingAverage]

In saying that it’s so hard not to doubt your system and constantly think “WTF is wrong with this thing”. So thank you for sharing

Agree...it is one thing to run backtests and think "results look good I can trade this" but it is a whole other thing when it comes to live trading that system.

One of the things that helps me with live trading my systems on a day-to-day basis (particularly when it is not performing) is to also have a good understanding of specific trade characteristics that can be obtained from your backtests. By that I mean many people are very focused on the big picture performance metrics of their systems such as CAGR, DD etc (which is all well and good and very important) but to understand the profile of specific trades is also important and certainly helps me on a day-to-day basis. More specifically, what I like to have a good handle on is the distribution of individual trade profit ($ or %) and the distribution of individual trade loss ($ or %). By this I mean I know that backtests show the average profit per trade might be 23% + or - 10% and the average loss per trade might be 15% + or - 5%. This is what you tend to focus on a day to day basis (how much did that trade make or lose) when live trading so to ensure the profit or loss of each live trade is within system backtest expectations gives me a level of comfort my live trades are in line with system backtests.

The other thing I find very useful in terms of tracking my live trades is to do some general short, medium and long terms backtests. For example short might be 100 trades, medium might be 2000 trades and long might be 4000 plus trades. When starting to live trading you might focus on your short term backtests to see how your live trading is going and as you get more live trades under your belt start focusing on the medium term backtest and over time switch to the longer term backtests. Problem is many people will run backtests over an extended period of time (many thousands of trades) and when they start to live trade wonder why their live trades aren't meeting the longer term trends and behavior. I think this is a very important point that a lot of folks overlook. Systems behave very differently over different time frames so having some insight into that also re-enforces my confidence in the system.

I think it is also important to maintain a healthy level of doubt about your systems. IMHO systems need to be constantly watched to ensure things are going as expected so you can react if something is not right. Systems can and do go haywire so being on top of things will add to your level of confidence to stick with it during its tougher times, which happens with all systems.

 

[MovingAverage]

I'd like to try and build a 100% mechanical and robust system of minimums. Minimum time to operate, minimum drawdown, minimum consecutive losses and minimum stress. Can it be done?

All those minimums will equal minimum profit. That is a very de-risked system (which is fine) but with minimal risk there is minimal profit.

 

[Skate]

@Skate! I concur firing up your system at a particular time could have significant effect on the system. Will just have to wait so it smoothens out

Wishing has no effect when it comes to trading
When it comes to system trading we all start out thinking we have something special. At times we believe we may just have the "Ducks Guts" of strategies believing the returns will start flowing in from the get-go. But live trading has a way of grounding you in reality. As @othmana86 said, "timing can have a significant effect". The market doesn't care what you believe, or what you think, the market will decide what you deserve.

Agree...it is one thing to run backtests and think "results look good I can trade this" but it is a whole other thing when it comes to live trading that system.

The Platinum Strategy
With each new strategy, you code it in such a way trying to garner an edge. Markets are constantly changing & boy have they changed in the last few years. I'm just saying before Covid, trading was like shooting "Wood Ducks" in a barrel some of the time & wondering what's going wrong the other times. Being "sucker-punched" always hurts but now it feels like a common occurrence & most of us are sick of it.

Skate.

 

[Skate]

I think it is also important to maintain a healthy level of doubt about your systems.

How good is "The Platinum Strategy"
As Scott Morrison famously said in his address winning the last Federal Election, "I believe in miracles". So do I, that's why I pray a lot, or at least I do when everything goes pear shape.

It takes time for a strategy to develop
If you think you are going to hit the jackpot trading a new strategy straight away you will be sadly disappointed.

Let's have a look
"The Platinum Strategy" might turn out to be my best strategy so far. When I first started trading back in 2015, I waited to start off at the start of the financial year making the assessment must easier having a full year under my belt. I'm now wondering how my new Platinum Strategy would have gone if I started trading it at the start of this financial year?

Backtest period (July 2021)
Well, that's disappointing & not what I was expecting.




Skate.

 

[Skate]

The Platinum Strategy
July trading results didn't live up to my expectations at all. But hey, the strategy has backtested & paper traded with pleasing results so let's give it another month. The strategy can't perform any worse, can it?

Backtest period (July & August 2021)
WTF, is going on? Well, that's embarrassing - I certainly wasn't expecting this. This is the point where "indecision" starts to creep in. Questioning if the results you achieved in backtesing & paper trading might have been down to luck. Let's persevere, give it another month, it's only money.



Skate.

 

[Skate]

The Platinum Strategy
I'm now wishing I didn't start posting accurate results - I should have "cherry-picked" the date range to show the strategy in a better light. I'm sure you would need "balls-of-steel" to keep trading it after two months of dismal performance. Well, I've gone this far, I'll keep going.

Backtest period (July & August & September 2021)
Now, that's more like it. For some strange & unexplainable reason, the system seems to be back on track. Let's give it another month as it's starting to pick up a bit of steam as expected.



Skate.

 

[Skate]

The Platinum Strategy
After 3 months we are making headways, I'm now in two minds if the momentum can keep going, or is it going to fizzle & burn?

Backtest period (July & August & September & October 2021)
There's been a slight knockback, similar to what is being reported in other threads, so there is no need for concern.



What is concerning
I have to start it all again on Monday. Accepting early setbacks when starting a new strategy is a hard pill to swallow.

Skate.

 

[qldfrog]

All those minimums will equal minimum profit. That is a very de-risked system (which is fine) but with minimal risk there is minimal profit.

Indeed a term deposit would fit the bill?and not too bad long term...

 

[Newt]

Not suprised at the lumpiness this financial year TBH Skate. And sideways is better than down

Would be interested to see the odd weekly equity curve on some of your longer backtests. Fantastic extra info in how these strategies hold through difficult periods, defend in market pullbacks, surge in the goodtimes. We're all tempted to go for the nice smooth 45 degree equity curve but of course its not realistic when trading weekly.

Know what you mean about sucker punches. Plenty of opportunities to learn from school of hard knocks in recent years!