Static versus Dynamic

Bad_Badness

Bad_Badness

There is an old saying about no war plans survive the first contact with the enemy. Likewise, why would a static stop loss or target plan survive the addition of new information, i.e. data.

So not so obviously, "new information" here has to be taken into context. If you are on a large time frame, e.g. hourly, then 15 second time frame does not apply. However if 5 or 10 minute probably does.

So that is the conundrum.
- One sets up (and submit) stops or targets in a trade, and "sticks to them" despite the new information. The Static method.
- On the other hand, some don't set up (don't submit ) initial stops and targets and wait for the new data to set them, then they submit them when they become high probability of execution, based on the data. The Dynamic method.

And lastly defining what constitutes "Non Relevant versus Relevant versus Actionable new information" is where the rubber hits the road.

YMMV

PS: New traders mostly end up with the static model because it is easier to understand and implement. And if you confuse the two or get into limbo between the two, that can be "bad badness"
 
Bad_Badness said:
There is an old saying about no war plans survive the first contact with the enemy. Likewise, why would a static stop loss or target plan survive the addition of new information, i.e. data.

So not so obviously, "new information" here has to be taken into context. If you are on a large time frame, e.g. hourly, then 15 second time frame does not apply. However if 5 or 10 minute probably does.

So that is the conundrum.
- One sets up (and submit) stops or targets in a trade, and "sticks to them" despite the new information. The Static method.
- On the other hand, some don't set up (don't submit ) initial stops and targets and wait for the new data to set them, then they submit them when they become high probability of execution, based on the data. The Dynamic method.

And lastly defining what constitutes "Non Relevant versus Relevant versus Actionable new information" is where the rubber hits the road.

YMMV

PS: New traders mostly end up with the static model because it is easier to understand and implement. And if you confuse the two or get into limbo between the two, that can be "bad badness"
More...

You're dead on although its the first time I've seen this refer to as a "static method" versus "dynamic method" although I know the typical trader uses such.

I think in the beginning of developing a trade method...we as traders need to recognize these trading styles and compare the results of their trade performance when using both early in the simulator phase prior to any real money trading.

Also, its tough to backtest a dynamic method but it can be backtested manually by a trader with a strong understanding of their psychology in which they'll know how they would have adjusted their stops, trailing stops and profit targets in a trade from one trade to the next trade.

In addition, I'm not a fan of using the same position size for each trade as if the price action is the same. Reality, price action is checking due to different reasons and the position size needs to be adjusted in reaction to those changed market conditions.

wrbtrader
 
Bad_Badness said:
So that is the conundrum.
- One sets up (and submit) stops or targets in a trade, and "sticks to them" despite the new information. The Static method.
- On the other hand, some don't set up (don't submit ) initial stops and targets and wait for the new data to set them, then they submit them when they become high probability of execution, based on the data. The Dynamic method.
More...


So would you consider a target and stop based on the ATR of the last 5 bars to be static or dynamic?

Could the dynamic method also be called - Winging it?
 
Let’s visualize some scenarios and compare static versus dynamic stops:

1. Position trader with unhedged long position that gaps down below his mental stop. Most likely the viable choices are either to exit at the market or apply a statistically validated threshold around the opening price before placing a intraday stop. This could either be a static stop for the entire day, transition to a dynamic stop off VWAP or a moving average, or apply a static stop later off a intraday support level. Depending on what the statistics say, it may be worthwhile to apply a time condition to one of these stops.

2. A reversion to mean scalp is probably best served with a static stop based on time and price. For example, using a one minute chart, a reasonable price stop might be the WRB and a two or three minute time stop. Obviously, other inputs such as adverse tick action would render these stops moot as I would be looking to close the position immediately.

3. Intraday hedging off momentary edges of an overnight option position could be looked at as applying a partial or full stop, as hedging is a form of risk reduction. My hedging decisions are based on static and or static considerations.

In my opinion, static and dynamic stops each have their place in a trader’s tool kit.
 
They said:
So would you consider a target and stop based on the ATR of the last 5 bars to be static or dynamic?

Could the dynamic method also be called - Winging it?
More...

Ah ATR, good one, a little of both, I would say. Which brings up the comment from wrbtrader.

Dynamic versus Static is an arbitrary bifurcation. Sort of like weather: Sunny versus Rain. There is a continuum. Knowing the current trade setups' changing relative effectiveness in a dynamic situation, and how much to push the envelope takes skill.

As for "winging it", it often looks like a valid dynamic approach. It is not the same as experts easily adjusting in rapidly evolving situations. Experts have confidence because they have seen similar situations, thousands of time, and are familiar with the outcomes. So if you are an expert, "Yes", otherwise "No".

BeautifulStranger is correct. He gave are very good detailed examples. (Details are where the rubber hit the road). The right tool for the right situation is key. It is knowing which tools are the right one, that the skill starts. Just like a carpenter has mastered many tools and does not abuse them by using, chisels for screwdrivers, e.g. so should a skilled trader.

-------------------
I'll throw out this idea about "time" and a trade:
After a trade is entered, the premise certainty decays over "time".

I put "time" in quotes to indicate "time" means passing of relevant events, as well as minutes, hours or days.

Here is an example for clarity, using swing trade LONG in ES Sept 3rd-11th . (This example is a gross simplification, of course).

1) Assume premise on the 3rd was "Buy the dip when it nears the 50MA, Exit on momentum near the 75% of the all time high."
2) Entry was near the 50 MA and target was set.
3) Now it has bounced (event 1), and now retesting the 50 MA (event 2)

The Exit premise has changed A LOT with those two event: The certainty of the exit decayed because the momentum weaken and now there is a full retest.

The point is, "time" changed the Exit premise certainty. Expect it and be ready to adjust. Or better yet anticipate it and incorporate it into your premise-plan. Reevaluating and adjusting (if needed), after "time" has passed is a skill. Advice: "Stay on top of your trades and unbiased but don't overmanage your trades either".

The BOYD loops of Observer, Orient, Decide, Act is illustrative. The faster and more accurate your loop is, the better, imo.
------------------
 
SunTrader said:
I would call it ... price action. Which is absolutely dynamic.
More...

The market is always dynamic. However its dynamics ARE measurable in terms of velocity and volatility.

I think @Bad_Badness is making the case that a fixed target and stop is a 'static' approach and is somehow inferior to what he his calling a 'dynamic' approach. Because the dynamic approach by his definition uses time and events after trade entrance as input data for adjusting stops/targets vs a static approach which only observes pre -trade data for all decisions.

I'll throw out this idea about "time" and a trade:
After a trade is entered, the premise certainty decays over "time".
More...
This is what MFE/MAE testing will do for you, it will give you all the validity of the trigger event over X time or X volume traded.

If your trigger event is price bouncing on the 50MA and exiting at 75% of the ATH, one can easily test that. One can easily test if the desired result has not occurred over x bars/time/volume then exit the trade as well.

An automated trading system can be built based on more data than can be viewed manually.

* Question - What happens when price trades into the 50 MA?
* Observance - Seems like when price touches the 50 MA it always goes up
* Hypothesis - Price that trades down to the 50 MA bounces up to 75% of the ATH
* Analyze - Backtest the hypotheses derived from the observance
* Conclusion - After price trades to the 50 MA it moves up to 75% of the ATH a statistically significant amount within X amount volume transacted or X time expired.

System rules derived from the above scientific method have a static stop/target but if a time threshold or volume traded threshold has been hit and the static stop/target has not been hit, exit the trade.

This is static approach by @Bad_Badness definition because all decisions were made before the trade entrance, yet it includes post traded price action, or lack there of, in exiting with a contingency.

Applying ATR into a system's rule-set and allowing for early exits based on time expiry or volume transacted are static as they are predefined, but dynamic in their behavior. They require no human observation or overriding once the trade has been initiated and they are respective of price action.
 
They said:
The market is always dynamic. However its dynamics ARE measurable in terms of velocity and volatility.

I think @Bad_Badness is making the case that a fixed target and stop is a 'static' approach and is somehow inferior to what he his calling a 'dynamic' approach. Because the dynamic approach by his definition uses time and events after trade entrance as input data for adjusting stops/targets vs a static approach which only observes pre -trade data for all decisions.


This is what MFE/MAE testing will do for you, it will give you all the validity of the trigger event over X time or X volume traded.

If your trigger event is price bouncing on the 50MA and exiting at 75% of the ATH, one can easily test that. One can easily test if the desired result has not occurred over x bars/time/volume then exit the trade as well.

An automated trading system can be built based on more data than can be viewed manually.

* Question - What happens when price trades into the 50 MA?
* Observance - Seems like when price touches the 50 MA it always goes up
* Hypothesis - Price that trades down to the 50 MA bounces up to 75% of the ATH
* Analyze - Backtest the hypotheses derived from the observance
* Conclusion - After price trades to the 50 MA it moves up to 75% of the ATH a statistically significant amount within X amount volume transacted or X time expired.

System rules derived from the above scientific method have a static stop/target but if a time threshold or volume traded threshold has been hit and the static stop/target has not been hit, exit the trade.

This is static approach by @Bad_Badness definition because all decisions were made before the trade entrance, yet it includes post traded price action, or lack there of, in exiting with a contingency.

Applying ATR into a system's rule-set and allowing for early exits based on time expiry or volume transacted are static as they are predefined, but dynamic in their behavior. They require no human observation or overriding once the trade has been initiated and they are respective of price action.
More...

Nicely stated. Had I tried to explain what you just said...it'll be 10 pages long.

wrbtrader
 
You must log in or register to reply here.
Back
Top