Dump it Here

[Skate]

Hi SKate , is your total position size 30?

Skate.

 

[willoneau]

You must have replied faster than I could edit.

 

[Skate]

I've been testing & coding @peter2 1st Blue Bar (SuperTrend Strategy) that has been discussed in his thread. I'm now a believer - his strategy performs well. Peter is onto something here & his posts on the topic are pure gold.

Discretionary verses parameters
The difference between our strategies is that Peter uses his discretion skills whereas my systematic code uses parameters resulting in selecting different positions at different times utilising the SuperTrend idea.

1st BB (Peter's Discretionary Checks)
HL - check, no resistance - check, first one in the sequence - check, no news - check, weekly trend UP - check, ASX trend UP - check, iSL below swing low, Place the buy order.
LOSING TRADE. No setup, no matter how many parameters we include, will guarantee a winning result. Only I will show you a losing trade first up.

Extra Parameters reduces false breakouts

1st blue bars are nothing special. Hundreds of them form every month. Trade them all and you'll lose very quickly. Our checklist helps us select ones that have a higher probability for success.

Nothings perfect

I post this to remind new traders that every strategy has losing trades. We cannot avoid them. However if we're persistent and manage our downside exposure we'll do quite well using this simple strategy.

Bull Market

A new trader doesn't realise the importance of a bull market.

Bull verses Bear markets
The 2 Backtests below reinforces what a Bull Market can produce.

N.B.
Optimising the strategy found using 2xATR(21) gave better results than using Peter's suggested 1.5xATR(21) for weekly systems. The best optimised system was using 5xATR(21) parameter but I wasn't comfortable with the longer hold period.

Skate's modified SuperTrend Strategy Settings (BULL Market)
Parameters 2xATR(21)
Weekly Strategy (Backtest: 1st January 2019 to 30th June 2019)
Starting Capital $600,000
40 Position Portfolio
Position sizing $15,000



Bear Market
July 2018 to December 2018 wasn't kind to most traders. Peter has posted that his 1st Blue Bar setup works better in a Bull market. Below is the backtest results confirming this.

Skate's modified SuperTrend Strategy Settings (BEAR Market)
Parameters 2xATR(21)
Weekly Strategy (Backtest: 1st July 2018 to 30th December 2018)
Starting Capital $600,000
40 Position Portfolio
Position sizing $15,000


Summary
Well done Peter, your 1st Blue Bar setup using the SuperTrend indicator with additional parameters works a treat. I encourage anyone looking for a simple strategy to play around with - check out Peters posts on the subject. Using the ASF search feature finds at least 24 separate posts.

Skate.

 

[peter2]

@Skate Wow, thank you. I've eaten a whole pack of TimTams reading through your research.

Same system tested through a bull and bear market. It was the best of times, it was the worst of times.



(1) I'm surprised that the system actually made an overall profit in the bear market. I think that your use of 40 positions helped with this. I'm assuming your universe was the AllOrds500. I'm also assuming that the portfolio held quite a few gold stocks which would have helped.

(2) Bull market comments:
Double the index performance,
W% is disappointing considering the bull market (expected >60%)
Very good AW/AL of 4.6

(3) Bear market comments:
Max DD < half the index is very good.
Ave Loss in this bear mkt is similar to that of the bull market.
Robust control of downside exposure (trade risk) in both markets is a real feature in these results.

 

[Skate]

@Skate Wow, thank you. I've eaten a whole pack of TimTams reading through your research.

Same system tested through a bull and bear market. It was the best of times, it was the worst of times.

View attachment 96854

(1) I'm surprised that the system actually made an overall profit in the bear market. I think that your use of 40 positions helped with this. I'm assuming your universe was the AllOrds500. I'm also assuming that the portfolio held quite a few gold stocks which would have helped.

(2) Bull market comments:
Double the index performance,
W% is disappointing considering the bull market (expected >60%)
Very good AW/AL of 4.6

(3) Bear market comments:
Max DD < half the index is very good.
Ave Loss in this bear mkt is similar to that of the bull market.
Robust control of downside exposure (trade risk) in both markets is a real feature in these results.

Hi @peter2

Thank you adding additional comments & I apologise in advance for the lengthy reply.

(1) I'm surprised that the system actually made an overall profit in the bear market. I think that your use of 40 positions helped with this. I'm assuming your universe was the AllOrds500. I'm also assuming that the portfolio held quite a few gold stocks which would have helped.

The SuperTrend strategy backtest was against the XAO (all ordinaries) The profit in the Bear market was mainly due to the systems Index Filter turning off in late August 2018 protecting the strategy from the pending disaster from September to December 2018. @Warr87 Index Filters are worth their weight in gold.

(2) Bull market comments:
Double the index performance,
W% is disappointing considering the bull market (expected >60%)
Very good AW/AL of 4.6

The Full Bull Market Trade list - The SuperTrend Strategy (January 2019 to 30th June 2019)


3 Best Trades (Bull Market) as at 30th June 2019 (the Backtest period)


# This is Important - These are the ACTUAL Trade EXITS for clarity


3 Worst Trades (Bull Market) as at 30th June 2019 (the Backtest period)


The Best Bull Trade CHART - ASX:ISX
Peter colours the price bars BLUE - whereas my charts tell me a story by the colours I use, so they are retained. For explanation & clarity I have included a "BLUE plot line" when the SuperTrend Indicator in on.




(3) Bear market comments:
Max DD < half the index is very good.
Ave Loss in this bear mkt is similar to that of the bull market.
Robust control of downside exposure (trade risk) in both markets is a real feature in these results.

The Full Bear Market Trade list - The SuperTrend Strategy (July 2018 to 30th December 2018)


3 Best Trades (Bear Market) as at 30th June 2019 (the Backtest period)


3 Worst Trades (Bear Market) as at 30th June 2019 (the Backtest period)


Best Bear Trade CHART - SFG
Peter colours the price bars BLUE - For clarity I have included a BLUE plot line when the SuperTrend Indicator in on.


Confirmation
The SuperTrend Indicator is not too shabby as a simple strategy, well done Peter for highlighting a strategy that has the potential to help others.

Full Year financial results - 2018/2019 strategy results


Skate.

 

[Lone Wolf]

Hi Skate,

Just something to be aware of is that not sure all supertrend indicators are the same. I was looking yesterday and I can't get my supertrend indicator to match Pete's for the AIA chart in his weekly portfolio thread.

Supertend can either use median or closing prices for the calculation. The version I had uses median by default. Also, there are multiple ways of doing the ATR calculation. I believe BullCharts lets you select the method and it looks like Pete chose Exponential MA. If you look up the ATR function in amibroker there's a comment below it from the developer stating that it uses the Wilders calculation.
"Note that original formulation of ATR (the one that AmiBroker implements)
uses WILDERS smoothing (not simple moving average)"

I don't mean to detract from your findings. Even if your version isn't exactly like Pete's, it doesn't take away from the points you make. You still took the supertrend indicator, made a system from it and your results are valid for your version. But the stats may or may not be applicable to Pete's version.

 

[Skate]

Hi Skate,

Just something to be aware of is that not sure all supertrend indicators are the same. I was looking yesterday and I can't get my supertrend indicator to match Pete's for the AIA chart in his weekly portfolio thread.

Supertend can either use median or closing prices for the calculation. The version I had uses median by default. Also, there are multiple ways of doing the ATR calculation. I believe BullCharts lets you select the method and it looks like Pete chose Exponential MA. If you look up the ATR function in amibroker there's a comment below it from the developer stating that it uses the Wilders calculation.
"Note that original formulation of ATR (the one that AmiBroker implements)
uses WILDERS smoothing (not simple moving average)"

I don't mean to detract from your findings. Even if your version isn't exactly like Pete's, it doesn't take away from the points you make. You still took the supertrend indicator, made a system from it and your results are valid for your version. But the stats may or may not be applicable to Pete's version.

Hi @Lone Wolf

Thank you for your post in the 'Dump it here' thread, I appreciate when knowledgeable & highly respected member decide to add to the discussion. Your remark clarifying that all SuperTrend indicators are not the same is 100% correct. May I also add, how the indicator is implemented varies from trader to trader. As a point in case - yourself, Peter, as-well-as myself have all used the indicator differently. In my case I've used the indicator to formulate a complete SuperTrend trading strategy.

It's a simple indicator
The Supertrend is a simple trend following indicator & it works well in trending market. As you say there are multiple ways to calculate the ATR. Amibroker uses the Wilders smoothing to calculate theirs. @peter2 has demonstrated the SuperTrend is an excellent indicator of trend direction, the heart of any trend follower.

SuperTrend has two parameters
1. Time Period
2. ATR Multiplier

The two parameters Peter has used
The time period of (21) & the ATR multiplier of (2) was the foundation of my trading system. The different ATR multipliers signal the degree of price volatility. It's also important to note that using the old method of buying when the Supertrend closes above the price and a sell when it closes below the closing price is a sure way of losing your money quick smart, so I'm advising others not to use the indicator this way - just don’t do it.

The Blue Line on my chart
I’ve overlaid the Supertrend indicator on a weekly price chart in my previous post (the Blue line) displaying when the trend is ascending. When there is a change in trend the indicator line turns "blue" which is simple to use & understand.

The basic idea
The SuperTrend indicator is to give buy & sell signal to the traders but I don’t use it that way. I use the indicator in-conjunction with parameters & a momentum indicator to enter the move. The SuperTrend indicator by itself is not enough giving a degree of false signals & that's why @peter2 uses his discretionary skills to enter a position.

It's not the Holy Grail
The Supertrend indicator fails in a "Bear or sideways-moving market" & that’s the very reason why I've added an Index Filter to my strategy design. Like any other indicator, the Supertrend indicator works best when used in conjunction with other indicators. Meaning the indicator is best used in combination with other indicators & parameter or else it will end up generating the wrong entry signals. The SuperTrend indicator is not appropriate for all markets & situations.

It helps
The indicator works when the market is trending & is a wonderful tool to know the current market trends. It clearly depicts the distinction of the downtrends and uptrends helping you to make smarter decisions. There are times when you get false signals which are yet again not as high as compared to other indicators. Hence, it is best to use the right combination of multiple indicators, especially a momentum indicator.

False signals
There’s no technical indicator which can be 100% accurate including the Supertrend indicator. The indicator also generates false signals in the sideways market, though it gives a lot less false signals as compared to other indicators I’ve used in the past.

I don't mean to detract from your findings. Even if your version isn't exactly like Pete's, it doesn't take away from the points you make. You still took the supertrend indicator, made a system from it and your results are valid for your version. But the stats may or may not be applicable to Pete's version.

No worries, I wasn't trying to emulate @peter2 results but my post was merely an exercise explaining to others how a simple strategy Peter applies in his trading can be very profitable. I've posted details & trading results of my MAP Strategy, my BOX strategy even my CAM strategy - the SuperTrend strategy is just added to the list, hoping to help others.

Skate.

 

[peter2]

@Lone Wolf and @Skate Thank you both for the additional discussion on the supertrend indicator. It's a little more complicated than a moving average and can be calculated in many ways. Like all indicators there are charts where the supertrend nails it and we think it's the best, then there's price action that defies all indicators.

I could select a moving average and colour the bars blue when price closes above the MA (red when price is below) and I would trade it in a similar manner to the 1st BBs.

It's interesting that skate mentioned that the 5xATR produced the best back test results in his recent 1st BB optimisation. A long time ago, another knowledgeable ASF member posted that a 6xATR produced the best results in his back-testing. He was a longer term trader and wasn't concerned by the higher portfolio heat and loss of open profits. Profitable investors aren't worried by these temporary deviations that the market produces.

Back on the supertrend indicator, I use it as it represents a significant price move from a recent low. We know that all up moves off a low don't turn into trends but a few of them do. If we start with a larger time frame, say a weekly and we know the weekly trend is up, then a price movement off a low (1st BB) in a daily chart has a higher probability that the prior up trend is resuming.

The 1stBB can get me into smaller price swings that I can trade as well as longer price trends.

 

[Skate]

Skate.

 

[Skate]

I've had a quick 'PM' alerting me to the fact that the Portfolio Dashboard didn't quite add up this week & if others are wondering it all to do with a Dividend payment that is included in the P&L Total.

Just a quick note to suggest there is a problem with this weeks figures.
Gross Profits = 29330
Gross Losses = (26172)
Profit = 3158

P&L is showing as $3581 which is a difference of $ 423.







Apologies for the confusion..

Skate.

 

[Skate]

Rough week ahead
The US President Donald Trump launched a sensational tirade on Twitter, announcing increased taxes on products from China as the stock market plummeted.



Adages
Never a truer word spoken...
Little strokes fell great oaks...
The pen is mightier than the sword...

Confucius once said
“Before you embark on a journey of revenge, dig two graves"

Skate.

 

[Sir Burr]

ASX codes
I want to right upfront & say "I don't even know what companies I'm buying" let alone the sector they belong to, I'm just buying ASX codes. But what I do know about the ASX codes I buy is that they meet my buy conditions & a predetermined list of parameters.

I'm interested in this because I'm the same except there are things that Amibroker can't tell me. Maybe takeovers and mergers or other reasons for bypassing a trade on the list.

I plug the ASX code into here: https://www.asx.com.au/prices/company-information.htm to see if there is something "funny" going on before continuing. I don't read all the pdfs (unless something seems odd), just a quick scan of the headlines.

Note that I don't like overriding my system and it's rare I exclude a trade.

Are there any circumstances where you would cross a trade off?
Apart from takeovers or mergers what could other reasons be? Maybe share dilution?

 

[Skate]

I'm interested in this. Note that I don't like overriding my system and it's rare I exclude a trade. Are there any circumstances where you would cross a trade off? Apart from takeovers or mergers what could other reasons be? Maybe share dilution?

Hi @Sir Burr

Thank you for your question & it’s a good one.

I take every trade as they come, I’ve even taken trades that I think will be dead set losers but what keeps me stepping up to the plate can be summed up in two reasons.

1. It’s not my job to think & what I think is irrelevant to my trading style.
2. My systems are throughly tested over many years & would have included something similar (barring Trump that is)

So in short, I don’t override my strategy not even for one trade. I’ve planned what to do & I do what I’ve planned.

Inclosing
I’m not a cafeteria trader, meaning I don’t pick & choose which trades I will take & which trades I’ll pass on.

Skate.

 

[willoneau]

Hi Skate, did you look at pyramiding at all to reduce draw down?

 

[Skate]

Hi Skate, did you look at pyramiding at all to reduce draw down?

Hi @willoneau, thank you for your question. Your question may be the catalyst for others to seek more information about pyramiding into positions..

My drawdowns (posted backtest results in the 'Dump it here' thread) are well within acceptable limits for my risk tolerance. When formulating a strategy, pyramiding was always found to be detrimental to the systems performance. Using a "Take Profit Stop" did much the same. Being a trend follower it's essential to get into the move (the trend) as soon as possible & pyramiding defeats this purpose.

For Simplicity
Formulation of a strategy (for me) is not about seeking the best or ultimate returns but seeking to find the "Goldilocks parameters" to suit the strategy & it's always a trade off to some degree. Strategy development is always a constant balancing act between risk & reward.

Confidence
If pyramiding into to a position gives you the confidence to take the trade I see no reason not to use it. The good thing about traders, we are all different.

Skate.

 

[Gringotts Bank]

Scaling in is ok with certain MR set ups, but it does alter my target price. if I'm at 3x my usual size (due to continued scaling in), I either look to exit at a much lower or much higher target price than for a single lot. The much lower price will usually be a break-even trade. The much higher target will be about twice the usual target. This is another way of saying "if a stock is going to revert after several successive buy signals, it will revert big time. If it's not going to revert, there will be a small window of opportunity to exit before it continues lower".

 

[Sir Burr]

2. My systems are throughly tested over many years & would have included something similar (barring Trump that is)

Thoroughly tested here but when the system snags a shareprice spiking to a takeover price, that's my Trump.

I use rank based on turnover but fine without.

 

[Sir Burr]

I feel like a dump this morning ...about takeovers.
Note my goal as a system trader is to simulate as close to actual trading as possible. Also, I’m not a short term trader.

There are two situations for takeovers, firstly holding a stock for a period then taken over (can’t avoid). Secondly, a buysignal is triggered by a takeover bid (can avoid).

This Amibroker code is relevant for both situations in backtesting delisted stocks. Trades that have been entered yet have no sell signal and these will "probably" have an exit value at or near takeover price.

NonTradedPeriod = 30;
SymbolDT = DateTimeAdd(DateTime(), NonTradedPeriod, inDaily);
Delisted = BarIndex() == (LastValue(BarIndex()) -1) AND DateTimeDiff(Now(5), SymbolDT) >= 0;
Sell = Sell or Delisted;

Better still, if you use Norgate data: https://norgatedata.com/amibroker-faq.php#exitpriortodelisting

If you don't use this for backtesting, you could be left with a bunch of stocks blocking buysignals. These stocks will withhold your capital and effect backtest results.

In real time trading, would you have exited on the announcement of a takeover, waited until the takeover is completed or accept a new ASX code?

I sell before a takeover is completed but should you choose to accept your mission of a new ASX code, what happens to exits? There would be no exitsignal for your new ASX code when running an Amibroker exploration as there would be no buysignal. Perhaps it's possible to manually add a trade but you could save the ASX code to favourites in Amibroker and check stops. Depending how complex exits are, this could be tricky. i.e. if your exit uses the buyprice or has some timed stop from the entry signal.

For the second situation where a takeover triggers a buysignal you could use one of the above choices (except “exited on the announcement”) but using the Amibroker code above simulates not buying, except if in real time the stock doesn’t end up being taken over!

Mr Monte Carlo is helpful guy

 

[Sir Burr]

Damn!
Scan this week and top of the list is EPM with a big spike and a takeover announcement.

Fridays close $2.45.

"ERM's founder and top shareholder, Trevor St Baker, who speaks for 27 per cent of the company's shares, said in a statement he would accept the offer of $2.465 a share if no higher bid emerges".

 

[qldfrog]

So do you get in hoping for future higher bidder , or run away