TDesk Whispers

[tomele]

Hi @all.

Guys, I am back to the drawing board for the Veto signal thingy (and more).

Veto Signals

Most of you seem to see Veto signals to be useful as a superior filter for signals on a higher time frame. You seem to want trades out that contradict a longer term trend. Well, I think I agree with you. Let's consider Veto signals as higher time frame signals we don't want to trade against.

Now given that, my first question is about respecting them in the calculation of Solo and Multi Signals. We would have three options to manage this:

1. Always include Veto signals in the calculations
2. Always exclude Veto signals from the calculations
3. Maintain a list of Veto signals to in/ex-clude them

I personally prefer (2). Like Bruster wrote, the entries could be quicker as they would not be diluted by the slow overall signals. That is the most striking argument. We will respect the Veto signals anyway, so why should we consider them twice and give them such an overweight? But give me your opinion, please.

Now here is another point about Veto signals: What really is a Veto signal? One that does not agree (and is FLAT or OPPOSITE) or one that is strictly contradictive (being OPPOSITE). Good question, right? I really don't know. I might let this be an user's choice (VetoSignalStrictness) like in the last version. What are you guys thinking?

Then there is the next problem with Veto Signals. What do we do if we have more than one? My opinion is to take "Veto" verbatim. That would mean that if only one of those "Must agree" signals disagrees, the overall (Super) signal is not valid. But there could also be an user input (AllVetoSignalsMustAgree) how to treat this. You could then define whether one Veto is sufficient to block the signals or all of them must agree to block. Somehow this is like ancient Greece democracy ... they never solved it, did they?

Exit Signals

All that being said about Veto signals and hoping it make sense, there is one problem left: All those Veto signal thingies can only result in one maximum outcome: To block a directional signal and force it to be FLAT. They can't force Desky out of a position. Therefore here comes the next variable into play: Exit signals.

Let's say you can define signals as Exit signals. As soon as THEY contradict, the Super signal would not only become FLAT, but Desky would be told to close all positions immediately. No matter what basket trading or other strategies you employ, all trades would be closed and stay closed until the Exit signal vanishes.

We would then have a hierachy of signals:

Solo/Multi (lowest), mostly defining entries.
Veto, allowing directional signals only in a superior trend
Exit (highest), closing Desky's trades directly

There are the same things to solve similar to Veto signals:

What really is an Exit signal? One that does not agree (and is FLAT or OPPOSITE) or one that is strictly contradictive (being OPPOSITE). I think it must be OPPOSITE, but we could make this an user input.

What do we do if we have more than one Exit signal? I think the user must be able to define whether one Exit signal is sufficient or all of them must agree to exit all positions.

Wow, here is your food for thought. I am eager to consider your views.

Cheers
Thomas

[Dragonian]

Hello Thomas,

In TDesk does the input "BlockBeforeNewsMinutes" only block signals for "x" minutes before the news, as the name suggests, or does is also block signals for "x" minutes after the news? If the former it would be useful to inlude a parameter for "x" minutes after the news.

What do you think?

Hi Bill.

Good idea. I will add an input to block signals "x" minutes after news.

Cheers
Thomas

Hello Thomas,

Thanks for that :hi:

I'm away from home myself at the moment so my input may be sporadic for a couple of weeks also.

All the best with your issues.

[bishopdotun]

Hi @all.

Guys, I am back to the drawing board for the Veto signal thingy (and more).

Veto Signals

Most of you seem to see Veto signals to be useful as a superior filter for signals on a higher time frame. You seem to want trades out that contradict a longer term trend. Well, I think I agree with you. Let's consider Veto signals as higher time frame signals we don't want to trade against.

Now given that, my first question is about respecting them in the calculation of Solo and Multi Signals. We would have three options to manage this:

1. Always include Veto signals in the calculations
2. Always exclude Veto signals from the calculations
3. Maintain a list of Veto signals to in/ex-clude them

I personally prefer (2). Like Bruster wrote, the entries could be quicker as they would not be diluted by the slow overall signals. That is the most striking argument. We will respect the Veto signals anyway, so why should we consider them twice and give them such an overweight? But give me your opinion, please.

Now here is another point about Veto signals: What really is a Veto signal? One that does not agree (and is FLAT or OPPOSITE) or one that is strictly contradictive (being OPPOSITE). Good question, right? I really don't know. I might let this be an user's choice (VetoSignalStrictness) like in the last version. What are you guys thinking?

Then there is the next problem with Veto Signals. What do we do if we have more than one? My opinion is to take "Veto" verbatim. That would mean that if only one of those "Must agree" signals disagrees, the overall (Super) signal is not valid. But there could also be an user input (AllVetoSignalsMustAgree) how to treat this. You could then define whether one Veto is sufficient to block the signals or all of them must agree to block. Somehow this is like ancient Greece democracy ... they never solved it, did they?

Exit Signals

All that being said about Veto signals and hoping it make sense, there is one problem left: All those Veto signal thingies can only result in one maximum outcome: To block a directional signal and force it to be FLAT. They can't force Desky out of a position. Therefore here comes the next variable into play: Exit signals.

Let's say you can define signals as Exit signals. As soon as THEY contradict, the Super signal would not only become FLAT, but Desky would be told to close all positions immediately. No matter what basket trading or other strategies you employ, all trades would be closed and stay closed until the Exit signal vanishes.

We would then have a hierachy of signals:

Solo/Multi (lowest), mostly defining entries.
Veto, allowing directional signals only in a superior trend
Exit (highest), closing Desky's trades directly

There are the same things to solve similar to Veto signals:

What really is an Exit signal? One that does not agree (and is FLAT or OPPOSITE) or one that is strictly contradictive (being OPPOSITE). I think it must be OPPOSITE, but we could make this an user input.

What do we do if we have more than one Exit signal? I think the user must be able to define whether one Exit signal is sufficient or all of them must agree to exit all positions.

Wow, here is your food for thought. I am eager to consider your views.

Cheers
Thomas

Thomas, thank you for the deep thought you gave to the Veto signals. I am commenting based on my experience with the indicators. There is no statistical evidence. Just what I see in the short time I have used TDesk.

- On the calculation of Solo and Multi Signals. I prefer option 2 (Always exclude Veto signals from the calculations). There is no need to treat the veto signal twice. We will respect it anyway. If the veto signal is disabled, then all signals will be used in the calculation of the solo and multisignal.

- I consider veto signal as one that does not agree (and is FLAT or OPPOSITE). So long the direction of the veto signal is either undefined (Flat) or Opposite, then ignore the solo and/or multisignal.

- If using multiple veto signals, one of them should be sufficient to block the signals.

- On exit signal, I would exit a trade if the veto signal is opposite. Flat veto signal should not trigger exit.

- If there are more than one veto signal, the user should be able to define whether one Exit signal is sufficient or all of them must agree to exit all positions. I would exit if any of my veto signals goes against the trend. Waiting for all to agree might result in late exits.

Regards.

[rapple]

Hi Thomas. Here are my thoughts

-I think, for a start at least, there should only be ONE veto signal (on a higher timeframe) to avoid confusion

- User selects if veto is used for ENTRY, EXIT or BOTH

- On the calculation of Solo and Multi Signals. I also prefer option 2 (Always exclude Veto signals from the calculations). If veto-ENTRY selected, ENTRY must be in the direction of veto, along with a percentage of other indicators.

- If veto-Exit selected, EXIT trade when veto signal goes FLAT or OPPOSITE (user selected). Trade exit is not based on Solo or Multi

Kind regards Richard

[cozybooks]

Hi Thomas,

I agree with Richard on the one veto on a higher timeframe for entry and/or exit. I agree with others that excluding it from the calculations would be best.

My idea originally was to have a separate lower timeframe veto for entering trades, but I think with a little creativity on the signals I decide to use, I could accomplish whatever I need. Keeping a veto to just one signal would keep things a lot more simple to start. My experience has been that as soon as I try to add extra levels of entry or exit to keep losses smaller, my losing trades grow and my winning trades shrink.

I'm pretty sure one of Steve's sayings ties into this...

Thanks for taking everyone's input. We will trust in your mastery to sort it all out.

Steve

[Dragonian]

Hi Thomas,

I agree with Richard on the one veto on a higher timeframe for entry and/or exit. I agree with others that excluding it from the calculations would be best.

My idea originally was to have a separate lower timeframe veto for entering trades, but I think with a little creativity on the signals I decide to use, I could accomplish whatever I need. Keeping a veto to just one signal would keep things a lot more simple to start. My experience has been that as soon as I try to add extra levels of entry or exit to keep losses smaller, my losing trades grow and my winning trades shrink.

I'm pretty sure one of Steve's sayings ties into this...

Thanks for taking everyone's input. We will trust in your mastery to sort it all out.

Steve

Hello Thomas,

Add me to the above list of agreers

[Ehrenmat]

Solo/Multi (lowest), mostly defining entries.
Veto, allowing directional signals only in a superior trend

Hey Thomas,

this excerpt sums it up pretty well, excluding the Exit signals. I only have concerns of TDesk becoming too complex with all the described options / variations. But they are indeed needed to find any trading setup's sweet spot.

In detail:

1. Solo/Multi must all agree for an adequate entry

2. Veto must at least not be in the opposite direction of Solo/Multi signals
- this is because of the nature of higher TF signals since they need more time to confirm a trend

3. To this point, I don't clearly see the advantage of the Exit signal category. In my opinion, this can be simulated with the first two signal categories and the exit threshold.

Whatever it will look like... I'm looking forward to it



Cheers,

Mat

:hi:

[cozybooks]

I think Mat's post has led me change my vote. I feel like superslope changing color back and forth with every 5 minute signal.

His post reminded me of something. Back when i was doing a lot of testing with with the Superslope EA that combined Superslope with Flying Buddha, I found that I did marginally better when I didn't automatically close trades when the higher timeframe superslope changed to the opposite color, unless the lower TF agreed. I had a lot of months testing and live trading on that EA, so I had a fair amount of data to draw upon.

Differnt scenarios:
HTF Super Slope changes to buy, LTF signals are saying buy - you close all sell trades.
HTF Super Slope changes to buy, LTF signals have changed back to sell- I did better when I left my sell trades open until the LTF signals changed back to buy as well.

More than half the time I would lose a little bit more by having it set this way, but inevitably, I'd get a trade in which the LTF sell signal would run for quite a while and more than make up for all of those small extra losses, and once in a great while, it would be that start of really big run and by the time the LTF signal changed back to Buy, the HTF signal would have changed back to Sell, which would leave the trades open to hopefully go into profit rather than having closed for a loss or continue to run for more profits than they would have made otherwise.

Now whether this would really hold up over months and years of data, who knows, but it would argue for the option to have the veto signal just to be used simply to agree or disagree with the other signals, and not be used in the calculation otherwise.

Ideally, if you used a veto signal in its most simple form, this would mean:
Not open any new sell trade unless the veto signal is sell
Option to not close any sell trade unless the veto signal is buy (or optionally flat)

Steve

[tomele]

Hi @all.

Thank you for all your input. I have put on my thinking cap. It will take a while to get all the strings together.

Cheers
Thomas

[SteveHopwood]

For what it is worth, and all remember that I am an idiot, veto signals merely confuse me.

I have my charts loaded with my SS, Peaky etc. They generate a LONG signal, then Desky closes all sells and starts buying. Opposite for an original SHORT signal.

I tell Desky what to do when they generate a FLAT signal.

I am applying the same principles to manually managing those long-term SPB trades that got caught out last summer. Some I have been able to close at a profit after several months. Others I have hedged and closed the hedge at considerable profit when the signal changed again.

This is working well for me so far, so I do not use veto signals.

Doesn't mean I am right not to, mind.