Trading What I see !?

[Anbell]

When you do your backtesting, do you include the first 2 triggers? They are very weird unless they are there for a very specific reason. And that only leaves your 3rd trigger, which might make sense on its own, but unless it is very tightly related to the other 2, then the bot doesnt make sense (for me, for AU horses)

If the magic is in the 3rd trigger, then why not bet later?

ETA: You're betting into the market before 95% of the volume has arrived. That's fine if you have a reason for that, but otherwise?

[Anbell]

And I disagree with Shaun's advice to turn the bot off. As he has said 1000 times you never know what data you might want in the future, so keep it running but dial down the stakes.

(and you said that "a very high % of fav's have won recently." Is that true? According to what metric? Over what period?)

[decomez6]

2.0 needs 50% 10/1 needs 10% etc. Ie what bookies would call overbacked or underbacked?

i consider under backed or over backed nothing short of a fallacy.
look at the engine of a car , things work pretty much the same all the time . the only difference is when there is more impurities , you get more smoke coming out of the exhaust ....

piston price action.gif

let say you get 40% of volume at price of 2 ( under backed/over layed) OR 20% at a price of 10 (over backed / under layed ).....

-Bf algos are much smarter than we give them credit for. so in this case all what they need to do is to cleanse the impurity and keep the engine running at the same speed without changing gears...?how?

: 20% -10% = 10 % and the price of 10 shall remain stable at 10

:40% + 10% = 50% and the price of 2 shall remain stable at 2

NB: all this shall be done in the dark cover of crossmatching . this allows prices to stay the same even when the volumes % have incereased or decreased . the trader can enjoy the sight the dark smoke of stakes annihilation.

[ShaunWhite]

And I disagree with Shaun's advice to turn the bot off. As he has said 1000 times you never know what data you might want in the future, so keep it running but dial down the stakes.

I'd usually agree but I'm not sure the ground work has been done. This can be tested fairly easily.

The question is "Is there a relationship between the vwap-price gap and profitablity?"

So at a given time (let's say -2min) assume a lay bet on every runner, and log the vwap. Or in his example any horse that was the fav at that time.

Randomly select 1/2 the results and sort the cummulative PL by the price-vwap gap. The chart should be something like this if the answer is yes, (gap size on the x-axis)

Screenshot_25.jpg

Repeat with the other half of the results.

If it works then it should work over a longer time frame as more will be picked up as they go in or out of the 0.94 gap...something for the backtest maybe. This is just a snapshot.

For goat68 to just lay everything >0.94 It would have had to look like this ...

Screenshot_26.jpg

I just wanted to check that first because it doesn't even need a backtest you can do it straight from the data. The complication might be the 'reverse best'. Getting a good read on your likely fills is hard as hell, but basically if it doesn't show anything taking then offering is unlikely to help.

[decomez6]

The question is "Is there a relationship between the vwap-price gap and profitability?"

i am no data cruncher but i shall nevertheless throw my hut in the ring and say yes .

hear me out ;

1. -any price at the bottom of its trading range is by default below the vwap and vise /vasa . ( Support and resistance is the bread and butter of trading)

2.-the vwap can also be used to measure the average speed /velocity and also the intensity of a slippage e.g a fat finger . this will help especially if you dont have a price volume chart . only need to look at the distance from the vwap then build a trading decision based purely on what you see.

3. the vwap is universally the same to all observers . it is not a subjective indicator , its objectively used by most traders to determine if the price is oversold or undersold and thats why it act as a reliable support when the price is above it and resistance when the price is below it.( FOMO comes in handy)

4. i believe one can measure the profitability of vwap .. ?how? .... just keep all your other triggers identical across all the selections in question and only change the vwap triggers and see if there is a difference .(valuable information )


lets say there is more but i bet profitability is the eyes of the beholder
-

[goat68]

So it's interesting problem of conflicting views...some say keep it simple which is what ive been trying.. then above comments imply complexity...

[ShaunWhite]

3. the vwap is universally the same to all observers . it is not a subjective indicator , its objectively used by most traders to determine if the price is oversold or undersold and thats why it act as a reliable support when the price is above it and resistance when the price is below it.( FOMO comes in handy)

Not quite true. Vwap is calculated between two points in time. People are at liberty to use whatever duration they want, or more usually, a combination of durations. I'd agree that these things are objective measures but their interpretation as you describe it is subjective.

Re 4. That's what I'm suggesting he does. Plotting pl vs the price/vwap difference, but by assuming you simply back/lay everything and then you can slice and dice it anyway you want rather than testing A, then B, then C etc. That'll prove whether or not it makes a difference. Using the bare numbers is tough though. Even phrases like 'traded range' that are obvious with your eyes on it can be a mare to define in a way that's universal.

[goat68]

Im not convinced by technical variation as in 0.9 vwap vs 0.94 vwap or whatever.. i can't believe there's a magical 0.836367 vwap that is amazingly profitable!
To me 0.94vwap is just an appropriate average trigger, more to not enter at 1.06vwap, it could equally be less than vwap...
Im not sure what the magic formula is....

[goat68]

As i said it was to get an indication the horse is backed in...ie trending down at a particular volume and market time.... dont forget those parts of my trigger

[goat68]

Anyway off to chop some trees!

[ShaunWhite]

So it's interesting problem of conflicting views...some say keep it simple which is what ive been trying.. then above comments imply complexity...

Std scale vs selectivity decision really. And your choice depends on your personal idiology. Are you a filter type guy (small scale higher return ) or are you a large scale guy (big turnover lower retunes) Both are quite different and both work, but being good at one is easier than trying to do both or keep flip flopping. What kind of trader are you planning to be? Figure that out then put all your energy into that one thing. Just like the manual guys the auto guys tend to specialise in one area too.

[ShaunWhite]

Im not convinced by technical variation as in 0.9 vwap vs 0.94 vwap or whatever.. i can't believe there's a magical 0.836367 vwap that is amazingly profitable!
To me 0.94vwap is just an appropriate average trigger, more to not enter at 1.06vwap, it could equally be less than vwap...
Im not sure what the magic formula is....

That's what your analysis and test tells you. The peak value and a tailing off either side. Unless your trigger value shows that gradual pl change as you change it (normalised distribution) then it's not a meaningful indicator. With all that data and testing setup there shouldn't be any 'could equally be' because the analysis is all about finding those optimal values and establishing whether or not they make a difference.

Post a pic of your cumulative PL plotted vs signal strength (just bet everything initially so you get the full spectrum from - 100% to - 100%.). Then we can see what it's telling you. Ideally 2 random sets of data if you have enough to double check the dreaded back fitting.

[decomez6]

3. the vwap is universally the same to all observers . it is not a subjective indicator , its objectively used by most traders to determine if the price is oversold or undersold and thats why it act as a reliable support when the price is above it and resistance when the price is below it.( FOMO comes in handy)

Not quite true. Vwap is calculated between two points in time. People are at liberty to use whatever duration they want, or more usually, a combination of durations. I'd agree that these things are objective measures but their interpretation as you describe it is subjective.

Re 4. That's what I'm suggesting he does. Plotting pl vs the price/vwap difference, but by assuming you simply back/lay everything and then you can slice and dice it anyway you want rather than testing A, then B, then C etc. That'll prove whether or not it makes a difference. Using the bare numbers is tough though. Even phrases like 'traded range' that are obvious with your eyes on it can be a mare to define in a way that's universal.

+1
Very true , the trading window is the key to the objectivity of the vwap .
the interpretation of what is being observed is also very subjective .

BA s ‘ vwap matchpoint’ comes to bare.

And there-in lies the answer to the question if vwap is profitable?
Taking
A. trading window &
B. the observer ( trader)
Into consideration the answer should be a yes.

[goat68]

3. the vwap is universally the same to all observers . it is not a subjective indicator , its objectively used by most traders to determine if the price is oversold or undersold and thats why it act as a reliable support when the price is above it and resistance when the price is below it.( FOMO comes in handy)

Not quite true. Vwap is calculated between two points in time. People are at liberty to use whatever duration they want, or more usually, a combination of durations. I'd agree that these things are objective measures but their interpretation as you describe it is subjective.

Re 4. That's what I'm suggesting he does. Plotting pl vs the price/vwap difference, but by assuming you simply back/lay everything and then you can slice and dice it anyway you want rather than testing A, then B, then C etc. That'll prove whether or not it makes a difference. Using the bare numbers is tough though. Even phrases like 'traded range' that are obvious with your eyes on it can be a mare to define in a way that's universal.

+1
Very true , the trading window is the key to the objectivity of the vwap .
the interpretation of what is being observed is also very subjective .

BA s ‘ vwap matchpoint’ comes to bare.

And there-in lies the answer to the question if vwap is profitable?
Taking
A. trading window &
B. the observer ( trader)
Into consideration the answer should be a yes.

Im sure there's very key in what you said there, but i don't understand?

[decomez6]

Not quite true. Vwap is calculated between two points in time. People are at liberty to use whatever duration they want, or more usually, a combination of durations. I'd agree that these things are objective measures but their interpretation as you describe it is subjective.

Re 4. That's what I'm suggesting he does. Plotting pl vs the price/vwap difference, but by assuming you simply back/lay everything and then you can slice and dice it anyway you want rather than testing A, then B, then C etc. That'll prove whether or not it makes a difference. Using the bare numbers is tough though. Even phrases like 'traded range' that are obvious with your eyes on it can be a mare to define in a way that's universal.

+1
Very true , the trading window is the key to the objectivity of the vwap .
the interpretation of what is being observed is also very subjective .

BA s ‘ vwap matchpoint’ comes to bare.

And there-in lies the answer to the question if vwap is profitable?
Taking
A. trading window &
B. the observer ( trader)
Into consideration the answer should be a yes.

Im sure there's very key in what you said there, but i don't understand?

As Shaun said . the point at which you start calculating the vwap will determine the value which is true to you . This value will also be true to all others that start calculating the vwap within the same trading window similar to yours .
Now assuming all of you are chasing the same goal and using the same strategies a trend should come to bare .
You can then log in that trend tagged under the vwap specifications ( trading window ) and save the observations so as to use it for further interrogation .
Then use BA. vwap matchpoint to start a new calculation and see if the participants of the new window have a different trend or the same as the previous observed .
NB . You should only compare oranges with oranges and the only difference should be the trading window and the vwap is your pivot point.