Historical probablility based exits ?

[tradingjournals]

Quote from NoDoji:


When I was reading your journal I saw you post R:R's when in fact you never posted stop loss and profit target levels at the time of the "trades". Without these at the time of the call, we have no idea what the R:R is.

We do not see things the same way, and it is a normal thing. For example, profit levels at time of entry is a random variable. If one agrees that it is a random variable, there are a number of important conclusions that can be derived from it, one of which is that one may not want to think of profit targets a priori. If you read few of my first posts, I asked if there were studies of single observation R:Rs, to which a person/persons who thought he knew, when in fact he/she did not seem to know, thought of it as a lack of knowledge. The RR is also a random variable due to others things, such as the risk part which depends among other things on the skill of the trader at the entry point.

The risk reward function is itself not fully understood, have many understandings even by the same person, and its understanding keeps evolving over time.

There is one thing that is shared by all trades: direction and price at a given time. If you were to have had my models entries' time and price, would you have been better off against them or with them (after the entry time and price) given your skill in managing risk reward? The answer depends on each person, and either answers could be valid. I am positive that you have an answer to that question after following my journal.

The edge in my view is ultimately in the mind.

 

[Sergio77]

Maybe throwing around a few good ideas would be better. The question whether systems with fixed target/stop are better or worse then systems with adaptive target/stop is not settled.

 

[NoDoji]

Quote from tradingjournals:

The RR is also a random variable due to others things, such as the risk part which depends among other things on the skill of the trader at the entry point.

Although the individual trader's management of a position that is entered based on a given model's signal, a maximum acceptable risk and minimum acceptable reward would at least provide an observer of your journal a benchmark for evaluation.

You claim that your models time entries, yet you've never called an entry price level in advance as indicated by your models (even a minute or two in advance would suffice). If the models are truly "timing" entries then it would quite simple to offer the signal a minute or two in advance and provide us with a stop loss price.

My system times entries and there's a price which tells me my timing was off by enough that the diminished R:R endangers the overall positive expectancy of my trading plan.

Quote from Sergio77:

Maybe throwing around a few good ideas would be better. The question whether systems with fixed target/stop are better or worse then systems with adaptive target/stop is not settled.

My experience based on a long-running statistical analysis of day trading has shown that in a strong trend adaptive targets produce superior results and in a wide range, a wide channeling trend, or when scalping out of counter-trend pullbacks, fixed targets produce superior results. For a day trader who trades mixed environments, not just trying to catch strong trends, my analysis has found a superior outcome from fixed targets.