- Joined
- 12 January 2008
- Posts
- 7,397
- Reactions
- 18,492
Mate, R U OK?
Is it possible to re-run you code and report the backtest results for say the approx five years only 2018 to 2022?
I am interested to see how the more recent data pans out.
Hum, i should look back at my old bb strategy, got a bit more knowledge under the belly, and the previous version was run/tested on corrupt data.Published on the 20th July 2021
"One of the most popular strategies was the Bollinger Band Breakout (BBO) strategy named “One of the Top 10 Trading Systems of All Time” by Futures Truth and remains a popular trend following strategy".
Disclaimer
The backtest results are from my version of the "Weekly Bollinger Band Breakout Strategy". The results aren't too shabby & the drawdown is lower than most trend trading strategies as I incorporate the Bollinger Bands as an additional exit strategy. We all tend to have our own unique version of John Bollinger's idea & it's at the "very heart" of my "BBO Strategy". Many will trade a variation of the BBO strategy because it's a trusted workhorse that works across all time periods. The BBO is a brilliant idea of John Bollinger.
So there is no cherry-picking
All backtest going forward will be on the dates suggested by @entropy. The backtest period is from 1/1/2018 to today.
View attachment 145182
Skate.
I mentioned earlier in the thread that I was starting new strategies at the start of the new financial year. I have already started two of the three strategies & both are traveling along nicely "at the moment".
Yep, totally agree that this lag is huge, especially after a panic dip in the market. The only purpose of a off/on market filter is to reduce the max DD. Once this has been activated and reduced the DD compared to the market, the portfolio manager needs to get back into the market quickly in order to take advantage of any rally. Moving averages and even volatility based indicators have too much lag for me after panic selloffs.The issue I have using the garden variety (SMA) index filter is the tremendous amount of lag. This lag guarantees you will get in & out of a position late.
For those interested
Time Stamp 1:55 pm Today 10th August 2022
WTT% Strategy (10 x $10k positions)
The WTT% uses the number of advancing positions in the relationship of decliners converted to a percentage as an alternative to using a simple moving average index filter. The issue I have using the garden variety (SMA) index filter is the tremendous amount of lag. This lag guarantees you will get in & out of a position late. I believe my alternative overcomes this issue.
"Skate's Lipstick on a Pig Strategy" (10 x $10k positions)
First off, I should say @Newt's accidentally named this strategy "Lipstick on a pig". In a previous post, I was saying, "vanilla" strategies can be improved. These improvements, using additional buy conditions, parameters, & indicators, are the "Lipstick on the pig" he was referring to.
View attachment 145183
View attachment 145184
View attachment 145186
Skate.
Do you consider it useful to pay attention to indices as an overall filter?
By that I mean if a candidate stock is in a sector whose index is flat or falling do you avoid it?
Say the index for all industrials in the All Ordinaries is declining would you automatically avoid these stocks until they move to an uptrend?
Fwiw, i tried such a system looking at trend per sector something initially qfsec in my journal.Nice set of trades there, Skate, well done!
Especially since you are adapting some long-established methods: I would have thought that the juice would be all squeezed out of those oranges so it is encouraging to me as a newby that the old ideas are worth re-investigating.
Amongst the noise you have detected some strong signals!
Your point about looking for signals other than ones like SMA's, which can be by their nature quite lagging, is a good one.
Do you consider it useful to pay attention to indices as an overall filter?
By that I mean if a candidate stock is in a sector whose index is flat or falling do you avoid it?
Say the index for all industrials in the All Ordinaries is declining would you automatically avoid these stocks until they move to an uptrend?
Fwiw, i tried such a system looking at trend per sector something initially qfsec in my journal.
(a) I'm a basic trader (b) I trade the ASX All Ordinaries (c) I trade multiple Mechanical Trend Trading Strategies (d) I trade the ideas of others, as there is no reason to reinvent the wheel (e) I spend most of my time trying to improve on those ideas.
A simple moving average to keep you on the right side of the market is better than having none at all. The "WTT% Strategy" incorporates a simple buy filter. If 50% of the companies in the "All Ordinaries" is up over a week, it indicates "to me" it's safe to buy.
something which appears profitable and relatively straightforward holds instant appeal.
I do find the turnkey thing a little strange - selling a code but then using a modified version of this seems a little off particularly those who are buying it but without the necessary coding knowledge to make their own modifications
Anyway, back to the filters - was just intrigued how to referred to it as a 'Buy' signal. Is that to say that the red periods are not necessarily looked upon as 'sell' ones (apologies if I have missed this over the past few pages). I know with Nick's he starts with a farily large stop loss which moves up to 10% as soon as the filter turns on.
can see that yours is 'off' a lot more of the time in comparision but with the benefit of reducing the off/on lag from what I can tell.
Certainly there is no free lunch with these things and always more to consider.
Skate, thank you for this detailed and instructive post!Trade anything going up
That statement should be with a proviso. Buy only positions that are going up that meet precise conditions & sell them when precise conditions are met. The strategy will find those positions no matter the sector.
Improving on an original idea
(LKE) has been mentioned on Twitter & in this thread so I'll use that as the example to demonstrate an improvement IMHO. The WTT strategy is a simple 20-period breakout. The original WTT strategy produces the buy & sell signals shown on the chart. The next chart is my recoding to take advantage of the exit. To reiterate, the exit is the money maker.
View attachment 145219
Being more selective
Strategy signals are generated when precise conditions are met. Being more selective makes sense when stronger signals are coming along at a great rate of knots.
View attachment 145220
I use a variety of buy filters
Earlier I discussed that a garden variety "Index Filter" using a simple moving average is robust but the lag can play havoc with your strategy returns. It's worth remembering that there is always someone making money even if you are not.
Index Filter versus a Buy Filter (the coloured ribbon)
A simple moving average to keep you on the right side of the market is better than having none at all. The "WTT% Strategy" incorporates a simple buy filter. If 50% of the companies in the "All Ordinaries" is up over a week, it indicates "to me" it's safe to buy. Using the percentage method treats all companies "as equals" rather than being weighted by the top few. The "green" colour of the ribbon at the bottom of the chart denotes when it's safe to take a position.
The upper & lower chart
In the upper chart, the original WTT strategy depends on a simple moving average to determine when a position can be taken, whereas the lower chart depends on the percentage value of all companies in the "All Ordinaries" to make that decision.
View attachment 145221
Summary
There is no reason to reinvent the wheel. I spend most of my time trying to improve the wheel, otherwise, we would still be using the wheels that were used on wagon trains all those years ago.
Skate.
Skate, thank you for this detailed and instructive post!
I have collected "rises versus falls" data for a while in the belief that "winners keep winning, losers keep losing" but have not been sure how to implement a sensible strategy. Your examples are quite motivating!
We use cookies and similar technologies for the following purposes:
Do you accept cookies and these technologies?
We use cookies and similar technologies for the following purposes:
Do you accept cookies and these technologies?