Australian (ASX) Stock Market Forum

Dump it Here

With great Respect?
Has anybody ever tested the reality of "Computer Fraud" in Back Testing

Please show me when you were GAPPED OUT in ANY of your data trades

How does the likes Nick Radge and others handle the GAPS in either direction and in their reporting system

Do they just ignore them as if they don't exist? ( GAPS)

It reminds me of a Wolf taking a Lamb for lunch and the Lamb asks the wolf "What's for Lunch"XYZ Yacht.GIF
 
Yes and whilst you will agree to the responsibility for any losses using the code plus not to share the code, I don’t think it’s unreasonable to have a experienced third party report/audit/verify the code.

@Sir Burr as usual you have hit the nail on the head. When you are required to sign an NDA it’s not too much to expect to have verified audited results disclosed before making a purchase. Posting Backtest results is not good enough when personal finances are on the line. In my opinion, when you sell a Turnkey Strategy it should be the very best version, "but no", the master trades them with Radge Tweaks, go figure. Also, I'm with Sir Burr on this one, when issues or improvements are found they should pass on to the purchaser.

Skate.
 
The HYBRID Strategy
This strategy is a proven workhorse. I pride myself on having a respectable exit strategy but there are times a few positions will go through to the keeper.

At times "good enough" is "good enough"
Coding precise mathematical formulas on "known data" doesn't always translate well when trading against "unknown data", that's the problem with systematic trading in a nutshell.

@DrBourse touched on timing
Timing when to enter & when to exit a trade is important but how you play the 'fish' in the meantime is where the money is made. Those who buy & hold, (invest) simply ride the fluctuations. But to make serious money you need to ride the waves (trade) with precision to the best of your ability.

View attachment 147617

Drawdowns
3 times the drawdown over the last 9 months exceeded my comfort level. (marked in red)

View attachment 147618

There are times when you need to step out of the markets
Capital preservation is the number one consideration in this game. Trading is hard even in the "best of times" but if you don't know "when to hold them" & "when to fold them" it's called gambling.

View attachment 147619

Skate.

Please post the graphs (entry + exit marked) for the top 3 profitable trades of the system, over the testing period.

I'm curious......
 
With great Respect?
Has anybody ever tested the reality of "Computer Fraud" in Back Testing
Please show me when you were GAPPED OUT in ANY of your data trades
How does the likes Nick Radge and others handle the GAPS in either direction and in their reporting system
Do they just ignore them as if they don't exist? ( GAPS)

@Captain_Chaza, I understand where you are coming from with your reservations about the accuracy of backtesting results. I'm not sure if you can rely on any backtest results if you can't or don't trade the same methodology to arrive at the final figure (result). With backtesting in Amibroker, Gaps in prices are taken whereas my trading rules prohibit taking those positions. I'll explain shortly my basic trading rules.

As system traders
We have one job & one job only & that is to follow the signals religiously without exception. I put a tremendous amount of work into my trading strategies & I'm confident all the hard work has been done. Gaps, either way, are not a concern as I don't chase a price & $hit just happens with it Gaps down past an exit point.

Skate.
 
Which is better - Discretionary or rules-based trading?
The answer might depend on your understanding & definition of "discretionary" as opposed to "rule-based trading". For me, it’s really simple, because "technical analysis" can be backtested to decide if the rule-based trading has an edge. Discretionary trading is subjective trading whereas mechanical "rule-based" is objective trading. I'm not going to start a debate as to which is better as it's been done to death. Having the conversation revolve around one type of trading versus another only invokes emotions. Having a robust discussion is not my intent but it's a very interesting subject.

Skate.
 
Gut trading
The problem with overriding a trading strategy is that you'll never be sure if it's the rules or if the overriding of the rules was the driver of a losing streak. Poor trading can be attributed to poor execution. The more "inexperienced" traders will bend the rules to suit their mood & wonder why live trading fails to replicate their backtest results. Mechanical system trading is a method for trading where all trade decisions are made according to an exact set of rules (a trading system). Traders do exactly what their system tells them to do, without deviating in any way based on instinct.

Skate.
 
Please post the graphs (entry + exit marked) for the top 3 profitable trades of the system, over the testing period.

I'm curious......

@Trendnomics I have previously posted that information. I'll show the first two & you can find the others.

The previous post can be found here

Go to the second chart in the hyperlink above (Trades sheet)
This chart displays all the trades, entry & exit points for all positions taken. The Charts below are a visual display of those positions. (CXO & AZL)

CXO Trade.jpg


Chart for CXO

CXO.jpg



Chart for AZL

AZL.jpg


Skate.
 
Which is better - Discretionary or rules-based trading?
The answer might depend on your understanding & definition of "discretionary" as opposed to "rule-based trading". For me, it’s really simple, because "technical analysis" can be backtested to decide if the rule-based trading has an edge. Discretionary trading is subjective trading whereas mechanical "rule-based" is objective trading. I'm not going to start a debate as to which is better as it's been done to death. Having the conversation revolve around one type of trading versus another only invokes emotions. Having a robust discussion is not my intent but it's a very interesting subject.

Skate.
Whatever suits your character.

imo, character should also determine time frame, number and type of instrument(s) traded, desired amount of screen time, automation (or not), etc.

No point trying to make a computer scientist trade on discretion, or a Gordon Gecko learn to code and follow pre-determined rules.

There's successes and failures in both camps, obviously.
 
Corrected repost
As the time has passed for a correction I have made the correction & reposting the slight change.

When buying "dips" turn_to_shit
When you've made a "trading or an investing mistake", cut your losses quickly as you can have another go when the market turns around & when the price starts to increase. Selling doesn't mean you "can't buy it back".

On the flip side
It pays to remember, before you sell a quality company, you're giving up the "long-term growth potential" of your buying decision. At times, deciding what to do next or what to do for the best can be the excruciating side of trading. It's for this very reason you need a solid trading plan "you can stick to".

Skate.
Good evening Skate,
rcw1 comments:

Corrected repost
As the time has passed for a correction I have made the correction & reposting the slight change.
nice

When buying "dips" turn_to_shit
When you've made a "trading or an investing mistake", cut your losses quickly as you can have another go when the market turns around & when the price starts to increase. Selling doesn't mean you "can't buy it back".
mostly

On the flip side
It pays to remember, before you sell a quality company, you're giving up the "long-term growth potential" of your buying decision. At times, deciding what to do next or what to do for the best can be the excruciating side of trading. It's for this very reason you need a solid trading plan "you can stick to"
rcw1 don't worry about 'potential growth', frankly couldn't care less, priority only to hold shares minimal amount of time and sell for a profit at some stage at rcw1 discretion, regardless of 'noise'. Hold for as minimal time to reduce risk and move quickly onto the next trade. Plan is quite simple really, target / carefully watch, a handful (not many) of stocks, to allow rcw1 100% ability to react promptly to buy and sell shares at the right time in any market. You win some and lose some, on the balance the overall challenge is win more ... ha ha ha ha

Have a very nice day today.

Kind regards
rcw1
 
priority only to hold shares minimal amount of time and sell for a profit at some stage at rcw1 discretion, regardless of 'noise'. Hold for as minimal time to reduce risk and move quickly onto the next trade. Plan is quite simple really, target / carefully watch, a handful (not many) of stocks, to allow rcw1 100% ability to react promptly to buy and sell shares at the right time in any market. You win some and lose some, on the balance the overall challenge is win more .
Being competent
@rcw1 there are some highly skilled traders on this forum who trade differently from most & it appears you are in this category as well.

Exiting late
As a system trader this is the anchor around our neck, not having the ability to exit quickly or at times not quick enough. System traders need to follow their trading rules waiting patiently for a precise setup, & only exit a position when the strategy provides that signal.

You need to be competent to trade successful
I refer to your style of trading as those who are "unconsciously competent" in their methodology to enter & exit those positions. These types of traders appear to trade using their so-called 'intuition', but are in fact, applying their vast knowledge & skill to recognise low-risk, high-profit potential, trades.

Skate.
 
Gut trading
The problem with overriding a trading strategy is that you'll never be sure if it's the rules or if the overriding of the rules was the driver of a losing streak. Poor trading can be attributed to poor execution. The more "inexperienced" traders will bend the rules to suit their mood & wonder why live trading fails to replicate their backtest results. Mechanical system trading is a method for trading where all trade decisions are made according to an exact set of rules (a trading system). Traders do exactly what their system tells them to do, without deviating in any way based on instinct.

Skate.

Some observations and then a question.

Observations:

From following this thread and seeing the various computer generated backtests, one thing stands out for me. That is the immediacy of the feedback. That is to say, you input your strategy, a few seconds/minutes later, the computer spits out the result.

The reality:

You implement a mechanical strategy and it progresses 1 day at a time. It is hard to equate the big picture of your backtested results comprising of 10yrs or whatever, to a 1 day plod with all the hits and misses entailed therein.

Question:

How many will go back to the periods identified as drawdown periods in the backtest and research the period in question, from both a qualitative and quantitative basis? Only by doing this will you know whether your current drawdown is subject to similar market conditions rather than simply a monitoring of price.

Non-systems traders:

Build up over time an internal market history. I have been trading now everyday, for in excess of 20yrs. Over that time period I have obviously 'experienced' the market with the qualitative inputs revealing themselves in real time. What that has led to is a two fold mechanical basis of interacting with the markets.

My point:

That trading price alone, especially in the early days of a trading career, can be very deceiving. Context is important. Lived experience is good. Second best is extensive historical research. Best is a combination of both.

jog on
duc
 
Observations:

From following this thread and seeing the various computer generated backtests, one thing stands out for me. That is the immediacy of the feedback. That is to say, you input your strategy, a few seconds/minutes later, the computer spits out the result.

Let's talk about Amibroker
What makes AmiBroker so fast? @ducati916 I can't speak for other trading software but as a long-term user of Amibroker, I can report that this software executes extremely fast. Without getting too technical it's because the AFL code is array-driven allowing maximum execution speed. The AFL language can process as much as 166 million data bars per second on a very modest CPU. The "Professional Version" runs on 64bit systems & takes advantage of running multiple cores, achieving blistering speed on steroids.

Amibroker is optimised for speed
AmiBroker uses native code optimised to gain maximum speed. In a nutshell, Amibroker is a professional technical analysis & charting tool that traders can use for (a) market analysis, & (b) chart analysis & is perfect for (c) backtesting simple to very complex trading strategies. Overall the backtest feature is extremely fast allowing you to gain information quickly to compare & contrast changes made to a strategy.

Skate.
 
The reality:

You implement a mechanical strategy and it progresses 1 day at a time. It is hard to equate the big picture of your backtested results comprising of 10yrs or whatever, to a 1 day plod with all the hits and misses entailed therein.

You missed it by this much (( -->||<-- ))
Actually, Amibroker has the ability to recalculate the (AFL) code every time new data arrives. Amibroker drives the code, whereas data is delivered by a data supplier. Data is delivered in time-chunked intervals from minutes up to "end-of-day" packets. Amibroker processes the data grouped into periodicity time frames, allowing you to trade within your risk tolerance level.

Precise coding
There are issues with the Amibroker backtesting feature is that it falls foul of my expectation, simply because of my lack of advanced coding skills. The other issue is that the strategy can generate more buys than your funds will allow, this is why getting the "PositionScore" right is so important. PositionScore is simply a ranking system as all signals are not equal in their value. Getting this part of the code wrong will alter a perfectly good strategy that can result in it being discarded.

Skate.
 
Question:

How many will go back to the periods identified as drawdown periods in the backtest and research the period in question, from both a qualitative and quantitative basis? Only by doing this will you know whether your current drawdown is subject to similar market conditions rather than simply a monitoring of price.

@ducati916 what a great question
I can only speak for myself. After developing a strategy, backtesting can decide if you are on the right path, taking it to the next level. First off, I should make a disclaimer in saying that I don't trade using the same methodology as Amibroker software uses. It is so easy for Amibroker to do all its calculations after the fact "at the end of the day" (EOD) whereas I place my trades in the pre-auction "before the market opens". This will make a big difference in the overall results you will achieve.

Doing it my way
I take my backtesting a little further requiring additional coding. Why? so I don't overtrade or be caught short not having enough money when the opening price "GAPS" as @Captain_Chaza has previously mentioned.

Skate,
 
Knowing the opening price to do your backtest calculations
What does this matter? Amibroker fudges the backtest results as it knows the opening price because it uses (EOD) opening prices to do the calculations of the number of shares to buy with the strategy "PositionSizing".

Not knowing the opening price
I need to handle these calculations quite differently to ensure (a) that I do not overspend in relation to my "PositionSize" & (b) to limit my exposure to price gapping at the open. A minor point of difference but to me it's massive. Using the same code using both methods the individual results achieved can be like chalk & cheese.

Skate.
 
After backtesting, the hard work starts
The generated backtest signals are then traded theoretically using my trading methodology to evaluate the actual returns. It's an important step in my system development & evaluation before comparing any metric of the testing. Explaining my method further would become a dry boring subject.

I'm amazed
There are some who would invest funds without thoroughly evaluating a strategy within an inch of its life. I place a high value on my money where others simply don't, trading the alternative way to me IMHO is called gambling.

Skate.
 
Knowing the opening price to do your backtest calculations
A big difference with backtest.
It is very interesting, at least for me, to backtest a strategy currently being run against the actual performance.
That allows me to detect:
Error in code or data
Highlight the importance of 2 other differences between backtest- real world:
1
i do place , as Mr Skate, my orders before the open.and am limited by my platform in my order price range which needs to stay quite near the last closing value.
Bad gaps either way make me miss both sell and buy actions, usually negatively impacting the portfolio
2
that order total amount, as specified before open will at most be reached.more often than not,will not and you buy a smaller parcel often
3
trading halts are not bothering the backtest which will buy another candidate on open, no such chance in the real world where you will act manually after the open and at best buy another parcel during session
Item 1 especially can affect results significantly,especially for very dynamic systems and this is enough to change a backtest or paper system from positive to loss maker.
Experienced in the past year.
These backtests checks usually trigger me to add a few tweaks to strategies ..and so restart the clock for BT actuals checking .which is not good
Hope it helps.
 
After backtesting, the hard work starts
The generated backtest signals are then traded theoretically using my trading methodology to evaluate the actual returns. It's an important step in my system development & evaluation before comparing any metric of the testing. Explaining my method further would become a dry boring subject.

I'm amazed
There are some who would invest funds without thoroughly evaluating a strategy within an inch of its life. I place a high value on my money where others simply don't, trading the alternative way to me IMHO is called gambling.

Skate.
I think everyone values money, since it's essential to life. However not everyone is technically adept enough to go through all the steps that you do. I often think I'm not smart enough for this game, but I also think of what a family friend said. He was a very wealthy guy, and when I was young and stupid I asked him 'secret'. He said "You don't have to do everything right. You just have to do the few right things right". Meaning - figure out what is essential and do those steps as well as you can.

In relation to trading, I think this is true. Of course everyone would love to have a straight-line 45 degree walk-forward equity curve over 10,000 bars. But it's not within everyone's reach. I don't think that means 'game over'. There's hundreds of ways to skin a cat.
 
Last edited by a moderator:
Lifted from Twitter
1. Bad risk management destroys a great trading edge.
2. Remaining calm in market situations that once used to stress you out is a sign of maturity as a trader.
3. Without self-discipline, your success is forbidden.
4. In the beginning, you have to work hard to learn. In the end, you have to learn to work smart.
5. There is a vast chasm between being smart and being wise. May we be smart enough to seek Wisdom.
6. It's so difficult to walk away from a winning trade, and even more challenging to leave a position when you're on a losing one.
7. Every pro trader was once an amateur.
8. Trading is a mental game.

Skate.
 
Top