Dump it Here

[Saqeeb]

View attachment 103837
Trading Mistakes (investors)
Well, it's not really a trading mistake, but rather how an investor handles a minor setback or disappointment when one of their long term positions drops temporarily out of favour. Investors buy certain shares either for long term value, dividends or perceived safety. With the recent volatility, holders can become more focused on the recent selloff, than on their long term plan. If a decision was made to sell due to emotions, a quick rebound can easily take away all the rewards built up over many years.

Trading Mistakes (traders)
Another mistake made by less experienced traders is that they "resist selling" or "never sell" unless the share price recovers above the original purchase price. Why? because they can't bring themself to sell at a loss. No one likes to lose money, but it's an inescapable part of trading. There is no method that doesn’t lose money some of the time & on occasions being prepared to accept a few losses is how we all learn to stay in the game.

Skate.

@Skate, I wish I could hit multiple likes for this post. Selling is one of the very most important aspect of trading, especially selling at the right time. This is something that I haven't been able to figure out yet.

 

[Skate]

Assumptions, taken as gospel
After reading a variety of (un-named) posts there is an insistence of some members separating what's imagined from what's real, resulting in a distinction between views. It's worth taking stock of how often our imagination, our expectations & assumptions, bleed into reality by actually changing the experience we have as readers. This isn't about denying that it's not true, it's about the cases when our perception of facts (the truth) becomes malleable.

Indecision
There’s a rational explanation why the markets behave the way that they do, it really is dependant on too many sellers or not enough buyers. Our trading decisions are based on the expectations of what others see as value. However, taking this one step further, showing how we adjust the value we see in something based on how others value it. So, it’s not just our decision that weighs into this equation, but everybody else’s as well.

Summary
We add more value based on the expectations of what others value. We also make decisions on malleable facts posted by others. We tend to place more value on their opinion rather than our own.

Skate.

 

[qldfrog]

I was going to post that in my thread but you might want to use the subject here.
as you said, too much information is not good so maybe keeping these subject for tomorrow's soapbox,

How do traders manage trading during holidays, travels
Has anyone installed AB on an amazon cloud server (aws) and run his her system explore that way,allowing work from anywhere in the world with internet without worrying carrying an expensive laptop?
Do you find this subject worthwhile?

 

[myrtie100]

No one likes to lose money, but it's an inescapable part of trading.

And to stay in the game, you must start with sufficient capital - which allows you to ride out the ups and the inevitable downs without too much pain.

This was my number one mistake learning to system trade with HomeTrader years ago. I didn't want to hear that the recommended capital for their systems were $100k - I thought I could do it with $20k!

After 15 losing trades in a row my capital base was rapidly disappearing. Needless to say, I stopped trading the system.

 

[Roller_1]

I was going to post that in my thread but you might want to use the subject here.
as you said, too much information is not good so maybe keeping these subject for tomorrow's soapbox,

How do traders manage trading during holidays, travels
Has anyone installed AB on an amazon cloud server (aws) and run his her system explore that way,allowing work from anywhere in the world with internet without worrying carrying an expensive laptop?
Do you find this subject worthwhile?

I use a VPS on Hostwinds. Have Amibroker and a API installed on it, i just log into it on my phone or ipad to run explorations and place trades etc. Does the job for relatively cheap. There are better servers though i probably wouldn't recommend them tbh.

 

[Skate]

Too much information
Making too many posts today will have a turn-off effect as some will see it as overload & too time-consuming to read (that will be their perception)

Moreover
In my previous post, I should have also mentioned that our values shift according to our situation as it alters the way we perceive & react to opinions once held, our brain tricks us to believe something different with more information & we condition ourself to accept it as gospel.

Being receptive
Some members have open minds whilst I've found most have closed minds. The point of the exercise today was to express some of my current thoughts. Those who read my series of "Soapbox" posts will relate to them while others will hold a different view to the ones I've expressed. Members who have open minds will display to a varying degree of openness accepting some of my views while rejecting others but I know the ones that resonate can make a big difference.

It's unhealthy
At the start of the posts today I said "it's unhealthy to read the views of one person only but at times it's the best we can hope for when posters are in short supply" so I'm encouraging others to post in the same spirit.

It's not important
Whether my views are right or wrong isn't important, others who hold an alternative view should express them without taking a firm stance that their opinion holds more value. The "Dump it here" thread is for others to express their ideas as it's not a contest of their ideas.

Quality posts
When members make quality posts it really does influence how others think to some extent, because it shapes our reality from what we see & read. New members are joining our community everyday & if their expectations are not met after reading a few posts they will not hang around.

Skate.

 

[Skate]

I was going to post that in my thread but you might want to use the subject here. as you said, too much information is not good so maybe keeping these subject for tomorrow's soapbox, How do traders manage trading during holidays, travels Has anyone installed AB on an amazon cloud server (aws) and run his her system explore that way,allowing work from anywhere in the world with internet without worrying carrying an expensive laptop? Do you find this subject worthwhile?

@qldfrog, I'm off my soapbox for now.

Good topic
I admit you have raised some good points in your post & as I have experienced all that you have suggest I might take the time to lightly touch on how I handle it.

QUESTION
"How do traders manage trading during holidays, travels? Has anyone installed AB on an Amazon cloud server (aws) and run his her system explore that way, allowing work from anywhere in the world with internet without worrying carrying an expensive laptop?"

How I handle Amibroker
The winters of the last 5 years I've been in the Northern hemisphere. It was a delight not to experience an Australian winter. Last year I was working away for 6 months & traded without an issue. To trade using Amibroker & Norgate data you don't need an expensive computer & today you can pick up a laptop that will do the job for a few hundred dollars.

1. I wouldn't go anywhere without my laptop, nope "not ever" even if I wasn't trading.
2. I have a variety of computers with Amibroker loaded on each one & they are synced through Microsoft "One Drive"
3. I have 3 Norgate Data licences

Now here's the rub
Forget all that, with a cheap laptop, internet access & "Team viewer" you are set to go as you can control your trading computer from anywhere in the world, hassle-free. If the laptop is lost or stolen it holds no personal or sensitive information.

Skate.

 

[qldfrog]

Now here's the rub
Forget all that, with a cheap laptop, internet access & "Team viewer" you are set to go as you can control your trading computer from anywhere in the world, hassle-free. If the laptop is lost or stolen it holds no personal or sensitive information.

I would try to avoid carrying the actual hardware around if I could, so the question: thanks for your explanations on how you manage your own setup.
Accessing my own computer would mean having it running 24/7 here in Oz while away , and hope no storm power outage, bushfire or burglar spoil the dream so my thought of AWS

 

[Roller_1]

I would try to avoid carrying the actual hardware around if I could, so the question: thanks for your explanations on how you manage your own setup.
Accessing my own computer would mean having it running 24/7 here in Oz while away , and hope no storm power outage, bushfire or burglar spoil the dream so my thought of AWS

That's why i use a VPS just access from remote desktop app no problems. I have Amibroker that runs a batch > produces the CSV easy. But like i said Hostwinds isn't the best has crashed a few times but that just means opening Amibroker again basically but still annoying.

 

[Warr87]

I would try to avoid carrying the actual hardware around if I could, so the question: thanks for your explanations on how you manage your own setup.
Accessing my own computer would mean having it running 24/7 here in Oz while away , and hope no storm power outage, bushfire or burglar spoil the dream so my thought of AWS

That's why i use a VPS just access from remote desktop app no problems. I have Amibroker that runs a batch > produces the CSV easy. But like i said Hostwinds isn't the best has crashed a few times but that just means opening Amibroker again basically but still annoying.

This is something I think about too. Occasionally with work I have little to no access to my usual setup. Good to know that a VPS can be used. I may not have access to my laptop or desktop for about a month or so soon, so having the VPS could be a option. The alternative was setting it up on my siblings computer as a backup and show them how to use it. Not ideal.

 

[qldfrog]

Unless someone has done it before, i might try to go the AWS way .
Keep own laptop when doing heavy research but otherwise just run the afl scripts in a batch and get csv results.
That means 2 licences for the data unless i can get some free data

 

[Warr87]

Premium date allowed you to have your data on more than 1 computer (such as a desktop and laptop). I'm unsure if Norgate allow the same thing.

 

[Roller_1]

Premium date allowed you to have your data on more than 1 computer (such as a desktop and laptop). I'm unsure if Norgate allow the same thing.

That is correct. From Norgates Website FAQ:

"How many installations are allowed?

The Norgate Data Updater ("NDU") may be installed on 2 machines, as long as both are for the subscriber's personal use."

I'd be interested to hear your thoughts on aws if you use it.

 

[martyjames]

Hi Skate

Question, if your system produces several buy signals on one day and your maximum open positions allows that many new buys, will you take those signals on the same day, or do you 'spread them' (new buys) over a period of time? Is there any thought about preferring to take buy signals on a 'down day' when the market is down , rather than buying into strength on up market days, or do you simply take buys as they come up regardless of above factors?

Cheers
Marty

 

[Skate]

simply take buys as they come up regarless of above factors

Hi Marty, I simply take the signals as given. (in PostionScore order until the Strategy is full)

Skate.

 

[martyjames]

ok -thanks!

 

[Skate]

I purchased Amibroker 2 weeks ago and know it has excellent backtesting capabilities, hence the newly formed interest in 'systems'.(bit of a learning curve though!)

Learning curve
It may have taken you 20 years but purchasing Amibroker is a great first step in the right direction, learning how to drive Amibroker will be your second. Trading is still the hardest game in the world because you're playing with some of the sharpest, most intelligent, well informed & unethical minds in the world. The professional traders you are competing against, the fund managers, the quants with their "Machine Learning Supercomputers" are out there just waiting to eat new traders for lunch.

Two goals in building a system
In building a trading system there are two goals you should have in your sights, (1) the ability to detect a trend as early as possible & (2) your system must also be able to distinguish "false signals", the wheat from the chaff so you can avoid taking a wrong position. It's a slow process of learning Amibroker, even when you get there, every system requires time to be able to produce results, something lost on those just starting out.

Get out quick
Timing of the exit is critical to any strategy. Cutting trades early is the the secret of being a profitable trader.

Skate.

 

[ducati916]

View attachment 103829


How we think anchors how we act
What's happening with the markets at the moment is anyone's guess. I'm amused that the media can express with conviction exactly what is causing each & every fluctuation not only in our markets but markets around the world. It's a pity they aren't clairvoyants, it would make my trading much easier.



Skate.

The media submerge the signal in pure noise. History on the other hand provides signal. The issue with history (backtesting) is one of interpretation.



Choosing a timeframe in which to observe returns is an important decision, as opposed to simply the variability. With a 15% and 10% volatility the variability between a 1 minute chart through to a 1 day chart is marginally higher than a probability of 50%. A coin toss. This will play havoc with your mind. You are receiving almost pure noise. Hour to hour that translates into 30 parts noise to 1 part signal.

Over 1 month this improves to a ratio that is 2 parts noise to 1 part signal. This translates into a 67% probability of success. Over 1 year a 93% probability of success.

Mr Skates system:



Sitting far higher than my hypothetical 15%. The signal to noise ratio will be far higher. Yet by decreasing the time of your observations to shorter time periods, you run the risk of lowering that signal to noise ratio: you increase your chances of (discretionary) fiddling based on nothing more than noise.

There is a new thread (just popped up) re. 'Why is the market rising'? On my own thread I purposely post a mixture of noise/signal, mostly out of curiosity as to whether (either version) creates comment or opinion.

There are occasions (bounces) where the signal to noise ratio can and will compress into shorter timeframes. This is in part to the big Quant driven strategies (the non-day-trading strategies) in the US fiddling with their parametres and shortening some of their inputs, the result being (seemingly) greater sensitivity to the news. Even the big chaps fall prey to this phenomena. You see it in the data which demonstrates the higher churn rate, with increased choppiness and higher than average volatility. My hypothetical 10% annual volatility is currently far below average volatility over the past 1 month, which increases the noise, further hiding the signal.

Now I don't trade mechanically, so Mr Skate will need to confirm: when backtesting (history) the test is over significant periods of time. The 'interpretation' of that history is manipulated through the choice of indicators etc. Clearly value can be added through that 'interpretation'.

When I look at history (which I do a significant amount of) I am comparing stocks not to their indicators, which are after all derivatives, but, fundamental factors. Here (again Mr Skate can confirm/deny) I am not sure whether the 'systems' do that. Is it necessary? Arguable I guess.

Empirically, the secular (fundamental) trends tend towards 18yr. periods. Not exact by any stretch, but close enough. Macro market views (if you catch big chunks of the trend) can earn significant returns. Even higher if, you can avoid the cyclical downturns in the secular trend.

The signal: design methodologies that utilise the longest time periods possible. Try and avoid the noise when designing a system.

jog on
duc

 

[Skate]

Now I don't trade mechanically, so Mr Skate will need to confirm: (A) when backtesting (history) the test is over significant periods of time. The 'interpretation' of that history is manipulated through the choice of indicators etc. Clearly value can be added through that 'interpretation'. (B) When I look at history (which I do a significant amount of) I am comparing stocks not to their indicators, which are after all derivatives, but, fundamental factors. Here (again Mr Skate can confirm/deny) I am not sure whether the 'systems' do that. Is it necessary? Arguable I guess. Empirically, the secular (fundamental) trends tend towards 18yr. periods. Not exact by any stretch, but close enough. (C) Macro market views (if you catch big chunks of the trend) can earn significant returns. Even higher if, you can avoid the cyclical downturns in the secular trend. The signal: design methodologies that utilise the longest time periods possible. Try and avoid the noise when designing a system.

Duc, I've condensed your post & bolded the questions that require answering. It's important I make a few comments first. My technical responses can be a bit dry sometimes & hard to understand, that's why I've "dumbed-down" them for beginners to comprehend. The Action Strategy is simple in concept & design but the coding to generate the signals is very complex. There are two parts to the conditional entry being "volatility & volume" in relation to the mean that's also dependant on turnover, acceleration & momentum. The exit strategy uses three independent indicators being: (1) Variable Trailing Stop, (2) Stale Stop & (3) Volatility Dependant Stop which is clever in design.

(A) "Mr Skate will need to confirm: when backtesting (history) the test is over significant periods of time. The 'interpretation' of that history is manipulated through the choice of indicators etc"
Backtesting of the Action Strategy is limited by my subscription level of Norgate data to a 10 year period. I've not found this to be an issue as the data of the last 10 years is relatively current & it's more reflective to the current performance of the strategy. In saying this, one of the main requirements of any good strategy depends on the accuracy of signals being consistent over many different time frames. Indicators are just indicators & they "indicate" when a condition has been met. Indicators don't manipulate the data what so ever. I should also say that "backtest results" are not that important but they need to be consistent over multiple time frames. I use a multitude of indicators because as they say: "two heads are better than one" & it no different when it comes to indicators, choosing & getting them to work together is another story. Indicators have an inherent lag so it's critical the parameters match the time frame being traded. Backtesting allows for optimisation of the parameters & these parameters need to align with the strategy being tested.

(B) "When I look at history (which I do a significant amount of) I am comparing stocks not to their indicators, which are after all derivatives, but, fundamental factors. Here (again Mr Skate can confirm/deny) I am not sure whether the 'systems' do that. Is it necessary?"
As you say: "I am comparing stocks, not to their indicators but, fundamental factors". The Action Strategy is a mechanical trend trading system & it's "Index-based", the All Ordinaries index to be precise. The All Ordinaries has a mix of volatility that's required for long term growth of the portfolio. The "momentum indicator" is pivotal in the process of selecting stock with the required "volatility & volume". I touched on it earlier about the "lag" of the indicators used to measure "volatility & volume". For me to reduce the "lag" of these indicators required some juggling so they would quickly adapt to the volatility of the price movement. The responsiveness was optimised to the values of the input period & it works a treat.

(C) "Macro market views (if you catch big chunks of the trend) can earn significant returns. Even higher if, you can avoid the cyclical downturns in the secular trend"
Duc, If I could sum up the Action Strategy it would be your words above. The Action Strategy has been developed to catch "the majority of the trend", not all the trend. The Action Strategy is very reliant on multiple confirmations, momentum & strength being two of them sometimes delaying the entry, but it's "better to be safe than sorry". The index Filter working "hand in glove" with the "Stop & Go" indicator goes somewhat in alleviating "cyclical downturns" & "market sentiment".

Skate.

 

[Skate]

Skate.