Moving Average as an Exit Criteria

[TheBramble]

Moving Avergae as an Exit Criteria

Stop me if this has been discussed before, but I was idly thinking of how to use current market characteristics (whatever timeframe 'current' might mean to you) to determine when to trigger my stop.

Using a trailing stop in a long position, I don't move the distance. So, if I've got a 10c trailing stop, if the price retraces 10c from the highest point since opening the position - I close the position.

I use a different trailing distance for each stock, somewhat percentage based, but not totally.

Then I wondered. If the MA as a lagging indicator would be useful in telling me when to get out?

MAs are pretty good at factoring in most recent market action (by definition that's precisely what they do).

Therefore, using a trailing stop of (for a long position) the highest high value of the 5-period simple MA (exponential, vol-adjusted etc. don't make that much difference in the smaller periods).

So when the closing price on any bar drops below its 5-period MA - I exit.

Does this make sense as a viable exit criteria?

(Obviously you'd use an exit of the closing price rising above the lowest low MA value since opening on a short position.)

Has anyone done any work in this area?

I'm keen to do more analysis, but thought I might save a bit of time asking here first.

 

[adrianallen99]

Strange you should post this as I have been looking at something very similar, for an exit to a long position would require the position to drop 10 points below the moving average (for example 10 points below a short term trend on a 200 period moving average on a 10 minute chart).

 

[adrianallen99]

I have to admit though, it seems to work better for an initial stoploss when you enter the position as if you get a decent uptrend and it moves away from the MA, you end up losing too much of your profit waiting for it to hit the stop before exiting - maybe it would work with using a different MA for the trailing stop.

 

[TheBramble]

Well, with a 200-period MA yes, you would be exposed to a large move against your position before being stopped.

That's why I was proposing a very short 5-period MA.

 

[rog1111]

FWIW I have found that a stoploss based on x (1,2,3 etc) times the average true range (ATR) seems to produce better results during backtesting than MA based exits. I don't have any logic to offer as to why this should be.

rog1111

 

[adrianallen99]

I have in the past used 2*ATR as my stoploss.

Out of curiosity, is the 5 period MA on a daily chart?

 

[TheBramble]

Rog - I'll have a look at ATRs as well.

AA - the timeframe is immaterial. I've looked at it over 1 min and daily. Seems to work well on that range.

 

[TheBramble]

Rog (I don't hang about!) - just checked the ATR stoploss scenario and I'm not sure I understand how you'd apply it.

Plotted 1, 2 and 3* ATR and it gives the varying ATR values as a relative price range. So on (for instance) AAL the stock price was in the 1170-1300 range and the ATRs were in the 10-55 range.

Are you saying you use the ATR to determine the size of your stoploss at any given point on an open trade?

Doing that would have you increasing the size of your stoploss as volatility/price-range increased. Surely that could be bad news?

 

[adrianallen99]

The way I have used atr, was 2*ATR on the day I entered the position, I then trailed this on an EOD close. So if the ATR was 9 on the day I opened the position I would trail a stop at 18 points.

 

[adrianallen99]

I have used MA as I stop while trading the Nasdaq 100 before, basically I found a MA that it seemed to be following/bouncing off. The position stayed open for just over a week and made over 70points. I sold near the top of the chart and when it broke through the MA it carried on down.

 

[oatman]

if you get a decent uptrend and it moves away from the MA, you end up losing too much of your profit waiting for it to hit the stop before exiting

How about a crossover with weighted aves?

 

[adrianallen99]

Perhaps a 10 and 40 Period MA with an exit point when the 10MA crosses below the 40 MA

 

[oatman]

21 & 55 weighted was my preferred.
Can't resist a Fib number :cheesy:

 

[rog1111]

TheBramble

Sorry I should have clarified. For example as my initial stop I might use 1,2,3 times the 10 day ATR ie not just the latest 1 day ATR, but the average of it over the last 10 days, to get a better idea of the average daily price range. I would never widen my stops once in the trade, only possibly narrow them as profit increases.

rog1111

 

[chump]

Back shifted moving average crossover ?

 

[BBB]

Chumps on the right lines here I think. A displaced MA is excellent if you want to fit an MA to the current trend.

ATR is another excellent tool in risk management.

Sometimes I simple place my stop at the recent swing low/high. If its taken, then somethings up - change of trend or consolidation/sideways price = get out.

Different tools for different jobs. x MA may work well for one trade, but fail you on the next. Only your own personal experience will teach you which, when and why. Let the market tell you, not a solid rigid rule.

 

[Simon]

Let the market tell you, not a solid rigid rule.

Very easy to say, but until you have the experience & discipline...