Backtesting Wyckoff Method

[Triple B]

Gday

I am about to read through the Wyckoff course for the second time.
This time taking notes.
I am also finishing up master the markets for the second time.

I have also been paper trading with success on Stockwatch and ASX game.
The question:
Can I backtest These strategies . I have incredible charts only , but will get Amibroker if needed.
The intention is to hold for 2 days to 60 trading days [3 months]
mainly buying close to support after low volume test after background of high vol accumulation on markdown ,followed by accumulation then low vol test of supply.
Expected rally top will be calculated using previous rally range and my own method of calculation. Possibly confirmed with P&F

So how do I backtest a stock that is stronger than its sector, in a neutral , bullish or bear market, off support after large vol markdown followed by solid accumulation , with a low volume test for supply!!!!! oh almost forgot current price below 200 Day MA.
Not to mention my Manual trailing stop strategy that tightens at the first sign of weakness after trailing below previous days open on up bars!!

Would like to trend trade if the opportunity exists too [think A2m ]

I would like to test at least 100 trades in the 3 types of markets. that's 300 trades.
I understand 10,000 tests would be better but if I can only do this manually that might take a while

Or should I just be happy with my paper trading results in this sideways market and keep drip feeding my CFD account ? I have the minimum in there now and could take some very small positions .
However I would really like some numbers to give me an idea how I might expect to perform.
Appreciate any help
Brendan

 

[Triple B]

Forgot to mention ASX 200 only and risk management is a variable system based on 3 grades of trade.
A grade B Grade C Grade
Will be high risk[5% A Grade] with small account and reduces to "Normal " levels "if " account builds.

 

[peter2]

Hi Brendan, it's possible to back test any system with an objective set of rules. Some of the Wyckoff stuff is a pattern within a larger pattern. This is harder to clearly define, but it's possible to break anything down into smaller units.

You have to define your bull and bear market, relative strength indicators are common, a large volume markdown is easily defined.
If cond1+cond2+cond3+cond4+cond5 etc is true; you can test it.

I like idea of Wyckoff springs and upthrusts as they provide great RR setups, but they're very hard to trade if you need any confirmation. If you understand the theory behind "springs and upthrusts" then you understand the reasons for fake-outs (bullish and bearish). They're very common in the US markets and becoming more common in the ASX.

Mastering just this one pattern would provide a great edge because they seem hard to master.

B and C grade setups are for wimps. We should be trading only A grade setups.

 

[Triple B]

Hi Brendan, it's possible to back test any system with an objective set of rules. Some of the Wyckoff stuff is a pattern within a larger pattern. This is harder to clearly define, but it's possible to break anything down into smaller units.

You have to define your bull and bear market, relative strength indicators are common, a large volume markdown is easily defined.
If cond1+cond2+cond3+cond4+cond5 etc is true; you can test it.

I like idea of Wyckoff springs and upthrusts as they provide great RR setups, but they're very hard to trade if you need any confirmation. If you understand the theory behind "springs and upthrusts" then you understand the reasons for fake-outs (bullish and bearish). They're very common in the US markets and becoming more common in the ASX.

Mastering just this one pattern would provide a great edge because they seem hard to master.

B and C grade setups are for wimps. We should be trading only A grade setups.

Thanks Peter.
Yes probably should scrap B and C idea. .
Lets see how I go with Springs and up thrusts.
Spring is to take out any potential supply competition lurking below support , by triggering Stops and panicking any potential supply competition into selling ,who bought inside trading range

Upthrusts are to keep potential supply/selling competition from selling after being locked in, at or around the resistance level of Acc. ie " I cant sell this now its finally moving in the right direction"

Fake outs are Basically the same idea when above the previous Resistance level , now support Correct?
Get em excited then pull the rug out!!!

 

[Wysiwyg]

Can you post a chart of the pattern with the criteria? Stocks stronger than its Sector and below 200 MA? Charts explain a lot especially if the pattern is rare.

 

[Triple B]

Can you post a chart of the pattern with the criteria? Stocks stronger than its Sector and below 200 MA? Charts explain a lot especially if the pattern is rare.

Gday .
That should have been "Sector stronger than Market" but stock price below 200day MA .
I added the 200day MA rule. to the Wyckoff method.
If the stock filled all other criteria but not the 200 day MA rule then I would call that a B Grade Trade .
only take B grade trades if no A grade trades.
I will post all rules later tonight. The golf course calls.

 

[Triple B]

This chart is pretty much what I see .
There are a few other things , but not important here
Still have plenty to learn and am really enjoying the process.
I figure I will only improve , and maybe one day, trading will be my primary source of income.

Small arrows on bear bars in trading range represent big money buying bulk cheap stock during accumulation.
Will give trading rules soon

 

[Wysiwyg]

Do you mean the famous Wyckoffian accumulation and distribution patterns?

 

[Triple B]

Do you mean the famous Wyckoffian accumulation and distribution patterns?

Yes the patterns combined with vertical bar charts with volume. Taking into consideration the background and history

The problem I see with an automatic type backtesting system with Wyckoff method is that the automatic system may not be able to take in consideration the background information.
For example: a low volume test of supply near support after a period of accumulation.
how can you program that?
That signal would only be valid as a sign of strength with the background of accumulation.
Sure you could program for a low volume bar at support level but if the accumulation is not there you would get too many false false backtesting results. I assume!?
This is just one example I can think of. There are many others.
All the parts of the wycoffian charts need to have the background information taken into consideration as the chart is analysed.
In other words the Method is not mechanical like many TA strategies.
One must treat each and every chart like it is a new Story. the end of the story does not not make sense if the beginning and middle are not read first.
I know nothing about Automatic backtesting , and am a novice at Wyckoff and VSA.
Perhaps there is a way , but I imagine it would be very complicated and not something a newb like me could tackle in a reasonable amount of time.

If you look at the chart I posted above, all the information in the text boxes needs to be somehow programmed into the backtesting software along with the spread and volume. Then the program would need to identify the support and Resistance levels at the same areas as I do , As this is important for calculating possible price targets for rallies..
I'm starting to think manual backtesting is the only way to get decent results with Wyckoff.

 

[tech/a]

All can be coded.
You’ll need to do it with python.

 

[Triple B]

All can be coded.
You’ll need to do it with python.

Thanks Duck.
Spent a few hours on this rainy Sunday Reading up on python.
Very interesting.
Will definitely be using that in the future.

 

[Triple B]

Was watching this video and a question very similar to mine was asked on the webinar.
about AFL whatever that is rather than backtesting. I imagine problems would be similar coding for backtests
The answer starts about 1:06 One hour 6 minutes in

 

[Wysiwyg]

I'm far from knowlegeable on Wyckoff but we could learn things maybe.

Yes the patterns combined with vertical bar charts with volume. Taking into consideration the background and history.

What you posted is the way I see it too. The pattern is a template but the duration is quite variable. These things can bottom bounce for a long time. The duration isn't too important as we only want to buy on an uptrend resumption with a volume breakout above previous pattern resistance, yes? To code the template would be hard because the range, intensity and duration of each phase is infinitely varaible. Therefore identification of each pivot point (SC, AR, ST, LPS, SOS) would be difficult as they happen but more obvious after. A lot of hindsight goes into the pattern.

I'm starting to think manual backtesting is the only way to get decent results with Wyckoff.

What about creating a watch list of stocks that fit the general pattern and wait for them to match that last phase. It isn't a common pattern.

 

[Triple B]

The duration isn't too important as we only want to buy on an uptrend resumption with a volume breakout above previous pattern resistance, yes

That's one way.
I am buying after the first low volume test of supply as close to support as possible.
This I believe increases the RR ratio. the trading range may be 10% wide so you pick up the 10% plus what ever you get above
resistance , usually seems to be 10%min depending on where accumulation started.

One thing Wyckoff says in his course, is that the stock should rise soon after the low volume test.within a couple days. If it does not ,that can be a sign of weakness.
My charted example above has shown that sign of weakness.
One of my rules is a maximum hold of 20trading days for a low volume test that does not rise straight away. this is to give the trade some Time to shake out the impatient money that are looking for the same trade as me!!
I have a "Ready to go Long "Watchlist made up of mostly these types of trades.
There is also trend trades mixed in that list.
I will separate them and use that list as a kind of backtest. Might as well ,use it as I have the hours of work in it.
The video I posted along with the others in that series are very good.
I have also looked at Bruce Frasers Wyckoff power charting series on Stockcharts.
Looks excellent so far.
Better go do some work
Bye
Brendan

 

[tech/a]

After 20 yrs with Wyckoff and VSA
I have a lot to put up here on the topic of testing and various
hybrids developed by many.

Will need sometime which I don't have right now
But Ill be back.

 

[Triple B]

After 20 yrs with Wyckoff and VSA
I have a lot to put up here on the topic of testing and various
hybrids developed by many.

Will need sometime which I don't have right now
But Ill be back.

You are a generous bird.
appreciate the help

 

[tech/a]

I think those attempting to develop a trading system using Wyckoff and or VSA
are attempting to use the analysis in a way that it isn't intended.

Both methods use Patterns both multiple and in individually read in context of a chart to indicate
momentum---the beginning or the end or the continuation of.

The reason you don't see systems is that there aren't and wont be any profitable systems based upon
either method using currently available systems testing software. If there were Gavin Holmes would have
it up on every website he could get his hands on.

Todd Krueger would have launched his software after 5 yrs of its coming-------soon.

I've seen so many posts around the place on how disappointed people are that the Arrows (They call them signals)
on a Tradeguider chart that often show a negative result.
They are INDICATIONS of possible price action which in the past has shown the same characteristics in the future.
They are NOT signals.---BUY or SELL.

----and-----They forget VOLUME!

Some of the very best signals come from a combination of understanding Very High AND Very low
Volume!
The question I think all should ask is what is the LIFE of an Indication or Signal.
Until its proven or disproved
(1) A Value stock--possibly months or years.
(2) VSA,Wyckoff Can be seconds or minutes to weeks dependent on time frame.

Combining a Very short signal in a trading system with a very long (In conjunction with chart context) Condition is in my view
pointless.
The best use of both methods is in identifying opportunities in the NOW.
In and out in the NOW.
Discretionary.

Systems
Sure you can use VSA but make sure your not looking for a daily signal and con texting it to 6 mths of trading.
The only relevance is where your signal should be related to.
(1) Bull
(2) Bear
(3) Ranging

IS IT RELEVANT?

I think there is a sad lack of understanding of the implementation of VSA/Wyckoff.
Hence some really poor code and copies and crap systems. I've only seen a few.

Software for systems testing is limiting when looking at patterns but with language like Python and more using it
we may see a new wave of advanced systems coming on to the domestic scene.

 

[Triple B]

Thanks Duck.
So basically learn and practice, learn some more, practice, learn ...............
I guess I read about how important backtesting is and figure I need to do that with an automated system to get enough tests to give valid results.
Probably best to do manual testing ie practice ..practice ......
As Ben Hogan would say.. "The secret is in the dirt" practice .........
At least I enjoy it!

 

[tech/a]

You need how to implement and trade VSA
What to do when it works out right and doesn’t.

Post some charts and your analysis
Keeps you honest and you’ll get some comments
The charts I post on Charts of interest are VSA based
I don’t go into a lot of detail due to time but anyone is welcome to comment

 

[Triple B]

You need how to implement and trade VSA
What to do when it works out right and doesn’t.

Post some charts and your analysis
Keeps you honest and you’ll get some comments
The charts I post on Charts of interest are VSA based
I don’t go into a lot of detail due to time but anyone is welcome to comment

Ok
Here is one
I could not [paper ] Trade this as I had no more capital . all tied up in profitable trades!!


So red down arrow Indicates medium effort to rise [med vol] met with strong selling pressure.
Prior strong distribution on bullish bars is also a sign of pending weakness.
Weakness confirmed after red arrow 2 bear days with gap down.
If I was in this stock I would have been out the day after the red arrow.

Then comes the Large vol Bear bar
The big money sold the balance of stock the day before . weak holders cant find buyers , there may have also been bad news.

The big money, lets call them Composite Operators or CO pick up the stock at the bottom of the bear bar[Big Volume]. Contrarian CO perhaps. Who else is going to buy all that crappy falling stock!
The next day there is still plenty of selling pressure as shown by medium vol and bear bar.
2 Days of selling pressure ends with medium vol bull bar. the buyers were fairly strong as can be seen by the top pin fairly long.

Then 4 Days of selling pressure as shown by the 4 bear bars, however low volume suggests this is more a lack of demand than than strong selling pressure.

Then we have the lowish vol Doji?

This is where the lowish selling pressure is almost overcome with demand. supply is starting to dry up!
Then the med-large spread bull bar on low-med volume indicates selling pressure has decreased and the bulls may be starting to buy up. possibly a lot of retail trader in this bar as doji is a classic buy signal.
The bull bar is followed by med to low selling pressure . for a few days.
Then we have a long pin bear bar. A push of supply to drop prices to collect as many stops as possible.
Probably got many of the buyers stops on the previous big bull bar, easy cheap Stock!!

The big pin bar is followed by lowish volume mainly selling pressure for a few days.
Then as the price drops the spreads narrow and the volume drops dramatically.[Three green arrows, my buy signal]
there is not supply left to buy. It was almost all taken out by the big pin bar.It looks as though everybody who was going to sell has sold.

So Now the CO thinks they have most of the supply, So they Stop any selling pressure to see if there is any demand.
There is buyers! Indicated by the Smallish bull bars after the green arrows. The low -med volume with slightly larger spread indicates little selling on those days.
After a few days letting the buyers get a bit of cheap stock the CO applies the selling pressure again shown by the very small spread bars.
A large bear bar with med vol indicates firm selling pressure to relieve the buyers in the last few days of their cheap Stocks.

CO Is now happy they have as much floating supply as they are going to get.
They simply hold on to the supply and let the demand do the work.
Maybe some good news , or a well timed company report to help demand along!?
They sit back and watch the retail traders and weak money do the work for them.
its all demand for a few days with no large volume distribution.
The large push across the resistance is mostly from traders bidding 'at Market 'orders .
But the ask is way up high . I wonder who is asking such high prices for this Stock?
If you can get any at all!!
Then the red Bear bar back across resistance.
No Real vol there which indicates selling from profit takers[Lots of stops hit] with maybe a small helping hand from CO
to help the price lower , and collect more stock for the distribution.
The last little red Doji is probably selling pressure from profit takers. low vol little supply

So Where will it stop ?
$4.60 will probably be a strong resistance.
I will do a P&F and see what that comes up with.

Sorry was so long , but that's what I see.
Any more chartsI do will be abbreviated, takes me 5min to analyse and 1 hour to type!