The solution for intelligent exit condition?

New comment
 
It's possible to have a good entry system, but we have to think about the hard part which is knowing when a trade should close.

- Fixed take profit is a random number of points. If it is large, the market might go up and go down again causing you to lose the profit. If it is small, you might miss a trend continuation.

- Standard trailing stop can go wrong if the market is very noisy and volatile

- Exit signal indicators are not ideal if the signal is lagging 

The only thing I saw so far which seems reasonable is a PSAR trailing stop. The stop loss is wide at first which helps avoid institutional stop hunts, and then it "accelerates" if your position moves favourably. 
There has to be more tools which are beyond trailing stops. PSAR has nothing to do with economics. It is based on principles in mechanical engineering.


 
Conor Mcnamara:
It's possible to have a good entry system, but we have to think about the hard part which is knowing when a trade should close.

- Fixed take profit is a random number of points. If it is large, the market might go up and go down again causing you to lose the profit. If it is small, you might miss a trend continuation.

- Standard trailing stop can go wrong if the market is very noisy and volatile

- Exit signal indicators are not ideal if the signal is lagging 

The only thing I saw so far which seems reasonable is a PSAR trailing stop. The stop loss is wide at first which helps avoid institutional stop hunts, and then it "accelerates" if your position moves favourably. 
There has to be more tools which are beyond trailing stops. PSAR has nothing to do with economics. It is based on principles in mechanical engineering.


The first similar indicator that comes to mind is the ATR-based Chandelier Exits indicator. There are also Supertrend indicators and Renko line break (virtual Renko) indicators that chase price in a similar way. There are various implementations of those posted around the web for free.
 
Conor Mcnamara:
It's possible to have a good entry system, but we have to think about the hard part which is knowing when a trade should close.

- Fixed take profit is a random number of points. If it is large, the market might go up and go down again causing you to lose the profit. If it is small, you might miss a trend continuation.

- Standard trailing stop can go wrong if the market is very noisy and volatile

- Exit signal indicators are not ideal if the signal is lagging 

The only thing I saw so far which seems reasonable is a PSAR trailing stop. The stop loss is wide at first which helps avoid institutional stop hunts, and then it "accelerates" if your position moves favourably. 
There has to be more tools which are beyond trailing stops. PSAR has nothing to do with economics. It is based on principles in mechanical engineering.


I use 4 layers:

(0) hard 500-pip SL for rescue only, never moved.

(1) virtual ATR initial stop.

(2) USD-based lock once profit hits target.

(3) ATR chandelier trail.

(4) close on 20% give-back from peak. No fixed TP, no indicator lag.

 
As trail stop you can use the Donchian channels which are based purely on price, an atr multiple trail stop which takes in consideration the volatility or a percentage you allow the price to reverse(e.g. 15-20%). Also you can use both donchian channels plus keltner channels if you want to get the best from both.
 
Conor Mcnamara:
It's possible to have a good entry system, but we have to think about the hard part which is knowing when a trade should close.

- Fixed take profit is a random number of points. If it is large, the market might go up and go down again causing you to lose the profit. If it is small, you might miss a trend continuation.

- Standard trailing stop can go wrong if the market is very noisy and volatile

- Exit signal indicators are not ideal if the signal is lagging 

The only thing I saw so far which seems reasonable is a PSAR trailing stop. The stop loss is wide at first which helps avoid institutional stop hunts, and then it "accelerates" if your position moves favourably. 
There has to be more tools which are beyond trailing stops. PSAR has nothing to do with economics. It is based on principles in mechanical engineering.


In my opinion, there is no universal “intelligent exit” that works for every market condition. A fixed TP is not always wrong, but it should not be just a random number of points. It should be based on something measurable, for example ATR, ADR, market structure, volatility, or risk/reward.

For me, a better exit system would combine several conditions:
1.initial SL based on market structure or volatility.
2. TP based on ADR/ATR, previous highs/lows, or risk/reward.
3. exit if momentum disappears or an opposite signal appears.
4. optional time-based exit if the trade does not develop.
 

Previously I always used a fixed number of take profit points as a placeholder. I only tested my EAs on demo accounts. My thinking was to use this fixed number of points for take profit, and then let an intelligent trailing stop adjust the position dynamically. But as markets shift unexpectedly sometimes, the take profit definitely needs to be adaptive. I was looking at the Chandelier Exit, although it seems to weaken its signals when the market is choppy. I'm trying another tool called "Profit Maximiser" which is also built upon the ATR, I'll see how this performs.

 
why not using the same entry indicator as exits ? "if it's good for entries it's good for exits" is my motto
 
Jean Francois Le Bas #:
why not using the same entry indicator as exits ? "if it's good for entries it's good for exits" is my motto

I tried, and it is not really great. For example the MACD (and impulse version) was designed for entry signals. Exits can be too soon with an entry indicator.  It can say "I think this is the end of the trend, but I have to calculate more ticks first to know for sure". Then it is too late for automated systems, because the bot says "ok I don't care, I'm closing the position now".

You need something which doesn't bend to whipsaws easily. The PSAR is one of those things. ATR based indicators as well with a multiplier (which chandelier exit is built from). You want the trend to be maximized for exit signals. The exit indicator can be used in two ways...you can use it to calculate a sharp exit, or else you can use it to maximize the trend for swing trading.

If you're only concerned with scalping then you could use a good momentum indicator for both entries and exits.
 
Jean Francois Le Bas #:
why not using the same entry indicator as exits ? "if it's good for entries it's good for exits" is my motto
My motto is more like... The exit must be faster than the entry─hence, the different logic applied to each.
New comment