Elders SafeZone

[cpatte7372]

Gavishti,

I shall look forward to testing it out in the morning.

Thanks ever-so-much mate for taking the time.

I love this group

Good night

Carlton

[helenquenet]

Thanks so much G, seems to work a treat

[Gavishti]

Revision 1.0.1 - Dec 29, 2003
- Fixed problem with tick rounding during trend change.
- Changed display of Buy/Sell Stop text and made them persistent for reference.
- Changed computation to onNewBar to fix problem with trend direction.
- Added TrendChangeCoefficient to provide adjustment for initial stop at trend change.
- Added thin dashed line to connect the dots between stops.

Setting the right margin to 9 in chart properties works best.

Please let me know if this is placing the stops as anticipated using the information from earlier in this thread. I would also like to know if anyone finds a good set of parameters that produce profitable trades.

Thanks.

[cpatte7372]

Gavishti,

This is great. I haven't tested the formula with differing parameters as yet. However, if you want, I can email you the Elders book where I got the information from. I purchased it in .pdf format. I'm sure there is somekind of copyright law but I'll email it to you anyway if you want it.

Cheers mate.

Carlton

[cpatte7372]

Gavishti,

I'm a novice. I wonder if you could tell me what parameters would I need to change if day trading on, say, 5min or 10min chart?

Cheers

Carlton

[Gavishti]

Thanks for the offer, but I'll pass. I just picked up a copy from the library this afternoon.

I tried several different settings on 1, 2 , 5, and 15 min charts along with a few dailies. The default settings don't seem to work very well. They take you out too quick. It will take a lot of experiments to find good settings.

[cpatte7372]

Actually,

The default settings appear to work nicely when run on the IBM daily. I compared the settings with the parabolic sar. Its very tight , however it ultimately stops you out at the same time as the parabolic. I haven't had a chance to run on it 1, 2, 5 10min bar.

Out of interest, what settings work best with you?

Cheers

Carlton

[cpatte7372]

I think the best way to use this system is with a ema crossover.

Carlton

[cpatte7372]

Ouch!,

The default settings doesn't like the 1, 5, 10min chart.

Carlton

[cpatte7372]

Gavishti,

For day trading you're better off with Chandliers Exits.

Carlton

[Gavishti]

I haven't found any good setting yet. I am most interested in the minis (ES, NQ, YM) trading on 1 min bars. At casual glance an EMA of 13, offset of 1, price source of HLC/3, lookback of 8, and coefficient of 7 looked promising. But it flopped when I checked the entries and exits. When I have time I'll add some strategy code and run it through the backtester.

I also looked at some stock that I can only go long on dailies. It didn't seem to be very effective there either. On the above settings and the defaults.

After I read Elder's book maybe I will understand better how to use the indicator.

[cpatte7372]

Sounds good.

Carlton

[Gavishti]

Already tried that. I'm experimenting with different stop systems trying to find the one that suits my style of trading. I hadn't tried the SafeZone so thought I'd write the code and find out how it works.

I'm also taking a look at the Bill Williams Gator stuff. There is a trader in this forum that says this is the best system she's found for trading the YM. Just need to read the book (just picked that up tonight also) and figure out how it works.

[cpatte7372]

Gavishti

What do you mean when you say, "it flopped when I checked the entries and exits"

Thanks

Carlton

[Gavishti]

It was unprofitable. However, that was over a short period. When I get around to adding the stuff needed for backtesting I can run it against 120 days of 1 minute bars. That may give different results. It isn't whether you win or lose on a particular day, it is how it turns out over the long run.

From a quick look, it appears that settings that keep you out of the whipsaws aren't very good at catching the trends. The setttings that are good for trends kill you in flats. Again, this is from a quick look. I will have to see how it does in testing.