Strong and Weak signals

Someone asked me be e-mail how it was possible to follow the GDX/IWM signals if they change overnight.

I understand that it might look difficult to follow every robot's move.
I personnally do not follow every single move: I concentrate on the strong signals. I take them quickly and stay in the trade as much as possible.

Since there are two robots, I believe that traiding the strong signals of both can lead to betetr results than trading all the signals of only one robot.

I just produced the attached file, that shows all the robot short and long signals, both weak and strong, for IWM and GDX.

The idea of following a strong signal is to enter on a strong signal only, but follow the trade by looking at the weak signal: if the weak signal changes, then you close the trade. (Be aware that the signal is generated at the close of the indicated date. Hence, you enter at the open of the next day or at the close of the signal day)

You will also note that most of the buying or selling wave start with a strong signal.
Finally, I can assure you that a strong signal is built and can be seen in advance - by monitoring the evolution of the LT/ST edges), while a weak signal is almost impossible to see without the robot.

A new XLE robot will allow us even more flexibility.



Pascal


[ATTACH]8663[/ATTACH]

Pascal, I don't follow the sectors table much, but since the recent market action in the last couple of months was characterized by a strong sector rotation, could we buy with confidence those sectors, that started showing relative strength on MF?

Thanks a lot.

Pascal,

A couple of questions in regard to your post above:

The idea of following a strong signal is to enter on a strong signal only, but follow the trade by looking at the weak signal: if the weak signal changes, then you close the trade. (Be aware that the signal is generated at the close of the indicated date. Hence, you enter at the open of the next day or at the close of the signal day).

I assume you're referring to the LT/ST edges when you refer to following the strength of a Robot signal. If that is correct, then one is to watch the evolution of the LT/ST edges daily and if those edges begin to weaken, the trader is to take appropriate action, i.e., close the trade, correct? Of course, the question remains how weak the signal needs to turn before initiating a closing trade, but I'll leave that question for the time being.

Finally, I can assure you that a strong signal is built and can be seen in advance - by monitoring the evolution of the LT/ST edges), while a weak signal is almost impossible to see without the robot.

If monitoring the evolution of the LT/ST edges can provide an indication of a strong signal, why isn't the reverse true? And this seems to contradict what is said above, i.e., by following the LT/ST edges for weakness, one can gauge when to close a trade.

[QUOTE=adam ali;13137]Pascal,

A couple of questions in regard to your post above:

The idea of following a strong signal is to enter on a strong signal only, but follow the trade by looking at the weak signal: if the weak signal changes, then you close the trade. (Be aware that the signal is generated at the close of the indicated date. Hence, you enter at the open of the next day or at the close of the signal day).

I assume you're referring to the LT/ST edges when you refer to following the strength of a Robot signal. If that is correct, then one is to watch the evolution of the LT/ST edges daily and if those edges begin to weaken, the trader is to take appropriate action, i.e., close the trade, correct? Of course, the question remains how weak the signal needs to turn before initiating a closing trade, but I'll leave that question for the time being.

Finally, I can assure you that a strong signal is built and can be seen in advance - by monitoring the evolution of the LT/ST edges), while a weak signal is almost impossible to see without the robot.

If monitoring the evolution of the LT/ST edges can provide an indication of a strong signal, why isn't the reverse true? And this seems to contradict what is said above, i.e., by following the LT/ST edges for weakness, one can gauge when to close a trade.[/QUOTE]



DO not make things more complex than they are in reality: on the Robot page, you will read "STRONG BUY" for example. This means that the probability for a profitable trade is much higher than usual. So you take the trade.
The robot will also take the trade. You then stay in the trade until the robot changes the signal and closes the trade.

Your question related to exiting a trade early when the edges are weakening is entirely linked to your trading strategy. Most of the money is made by "trend following". Trading in and out is against that idea. So I'd monitor the edges to have a feeling when a strong trade is about to come. I'd let the robot take me out.

If you look at the last two trades of IWM and GDX. Both trades could have been closed at a better price, but they were not. This is OK with me. When we have a signal, we do not know if the signal will take us for a 3 days or a 30 days ride. However, if we do not take the trade, we will no know, and if we exit after a few days because "we monitor the edges" then we might miss the next big move.

So, I prefer to take the trade and exit "on a wash", but at least I was in the trade in case of a long term trend.

One example: look at the IWM edges now: they are steadily trending up. We are in oversold and the 20DMF is about to issue a buy signal (tomorrow or in two days). I believe that this will be a strong signal. We know it is coming so let's be ready!

However, anticipating is not good. You can see that today, the market tried to bounce, but this is a failure (signal is more oversold). So, we might very well go lower tomorrow. However, market makers and GS friends will all move together and will bounce this market with strength when they are ready to do so. We will catch that move!


Pascal

[QUOTE=Andrei;13134]Pascal, I don't follow the sectors table much, but since the recent market action in the last couple of months was characterized by a strong sector rotation, could we buy with confidence those sectors, that started showing relative strength on MF?

Thanks a lot.[/QUOTE]

My backtests have shown that when you buy a sector that just issues a buy signal and if you do so against the market, you will end up on average with a neutral position (the market direction is strong, but your sector's opposite direction is helping you). The best is to buy a sector when yo have a sector buy signal. And the best sectors to buy are those with the worst price RS... until shorts have finished to cover. Then, you need to move to the sectors with strong thrust or best ratings


Pascal

[QUOTE=Pascal;13140]My backtests have shown that when you buy a sector that just issues a buy signal and if you do so against the market, you will end up on average with a neutral position (the market direction is strong, but your sector's opposite direction is helping you). The best is to buy a sector when yo have a sector buy signal. And the best sectors to buy are those with the worst price RS... until shorts have finished to cover. Then, you need to move to the sectors with strong thrust or best ratings


Pascal[/QUOTE]

Where on the website do we find the sector's relative strength and what constitutes a buy signal or a sell signal for that matter? I know I should already know this. Also, concerning using strong robot signals to buy and using a signal change to sell, if I want to protect a good or satisfactory profit from the original strong buy signal should I then sell when the next buy signal change to a weak buy?
Steve

[QUOTE=slgerritz;13143]Where on the website do we find the sector's relative strength and what constitutes a buy signal or a sell signal for that matter? I know I should already know this. Also, concerning using strong robot signals to buy and using a signal change to sell, if I want to protect a good or satisfactory profit from the original strong buy signal should I then sell when the next buy signal change to a weak buy?
Steve[/QUOTE]

You first vist the filters section and click on the first link that leads to the repository file (more explantions are on the FAQ). Then, when the repository file is open, you click on the sectors sheet.


[ATTACH=CONFIG]8673[/ATTACH][ATTACH=CONFIG]8674[/ATTACH]


Pascal

[QUOTE=slgerritz;13143] Also, concerning using strong robot signals to buy and using a signal change to sell, if I want to protect a good or satisfactory profit from the original strong buy signal should I then sell when the next buy signal change to a weak buy?
Steve[/QUOTE]

Steve,

This is part of a discretionary decision that the robot will not make with its current algorithms (second column).
Pascal has backtested such strategies and you can see that there is not much performance difference, but you increase the number of trades (and costs, commission, stress) by about 20%.
Based on these results, we decided it was not worthwhile to complicate further the robot signals.
Again, the robot signals and statistics can be used as a discretionary guide for experienced full-time traders, but for most people, they are the optimal compromise, especially for novice or part-time traders and for people with a full-time job who only trade EOD signals.
Billy

[ATTACH=CONFIG]8675[/ATTACH]

[QUOTE=Pascal;13148]You first vist the filters section and click on the first link that leads to the repository file (more explantions are on the FAQ). Then, when the repository file is open, you click on the sectors sheet.


[ATTACH=CONFIG]8673[/ATTACH][ATTACH=CONFIG]8674[/ATTACH]


Pascal[/QUOTE]

I found the sectors spreadsheet. There is no column labeled RS. Is term relative strength or RS synonymous with "Strength Aver Tot EV" ?
Steve

Thanks for that profit vs trailing stop overview Billy. Interesting results.

Is this true for the IWM bot as well as the GDX bot ? I'm asking because if I remember correctly the GDX bot was described as being more of a reversion to mean system then a trend system.

[QUOTE=slgerritz;13150]I found the sectors spreadsheet. There is no column labeled RS. Is term relative strength or RS synonymous with "Strength Aver Tot EV" ?
Steve[/QUOTE]

The "worst stocks" sheet carries a list of sectors ranked by price RS (Col F & G).


Pascal

[QUOTE=Rembert;13151]Thanks for that profit vs trailing stop overview Billy. Interesting results.

Is this true for the IWM bot as well as the GDX bot ? I'm asking because if I remember correctly the GDX bot was described as being more of a reversion to mean system then a trend system.[/QUOTE]

Good morning, Rembert.

These results were for the IWM robot only. If Pascal has also backtested for the GDX robot, I'm sure he will post the results here (he's out of office for a short while). Billy

[QUOTE=Billy;13153]Good morning, Rembert.

These results were for the IWM robot only. If Pascal has also backtested for the GDX robot, I'm sure he will post the results here (he's out of office for a short while). Billy[/QUOTE]

I just returned. The GDX robot only uses the LT signals. So there was no point of testing the same trailing stop strategy.


Pascal

Pascal,

Just a few days ago I was thinking to modulate the robot signals in three ways, according to the neutral, long/short, strong long/short labels. But I see that in your Long_Weak signals table you only have strong signals and standard ones. That is, two types. My real question is: is it worth having three types of weights or two make more sense?

Many thanks.

[QUOTE=Adriano;13156]Pascal,

Just a few days ago I was thinking to modulate the robot signals in three ways, according to the neutral, long/short, strong long/short labels. But I see that in your Long_Weak signals table you only have strong signals and standard ones. That is, two types. My real question is: is it worth having three types of weights or two make more sense?

Many thanks.[/QUOTE]

The best is to back-test a portfolio that trades both GDX/IWM using the two levels and then using three levels. The idea is to be 100% invested whenever you have at least one strong signal. If you have two striong signals, you are 50/50 invested in both.

I'd like to do that, but I am working on some urgent specifications right now for the real-time system.

Pascal