Currency strength (like Hanover)

[garyfritz]

OK, try this. I dropped my modified code back into your shell. See the Comment() in start().

The spin-up initialization isn't perfect, but I think it's Good Enough. It has to call CurrencyStrength() on the last 2 * (SmoothLen + SpeedLen + AccelLen) bars.

And that's Enough for today. Yardwork awaits.

[astral77]

Further to our posts in the NB thread I thought it is better to not divert the discussion in that thread. I have put your indicator on charts and compared them side by side with HANOVER and they seem to be more or less in tune (i.e. the relationship between rising and falling currencies are the same). Understandably the values are not exactly the same but that is to be expected.

Kind regards

[garyfritz]

That's good to know, astral. I hadn't actually done that comparison. I know the indicator's results make sense compared to the actual movements of the appropriate pairs.

You can tweak the inputs to modify the results a bit. Larger numbers for smooth/speed/accel tend to produce smoother lines but more delay. Running it on different TFs produces different results, similar I think to Hanover's.

I suspect I could probably use one "length" input, and just scale the smooth/speed/accel by that value. I don't think there's a lot of value in setting those 3 params separately and that would simplify it conceptually. Currently all 3 phases are optional -- you don't have to smooth, you don't have to calculate speed or accel (but you have to calculate speed to get accel) -- but I'm not sure that's useful. I suspect the full model as I showed in the recent screen shots, with length ratios of about 2:2:1 or so, if probably Good Enough.

[stimpy]

i mentioned this one before when you guys started looking at hanover if anyone is interested - might have been mentioned here but here it is if not..

http://www.trade2win.com/boards/forex-s ... post993638


i have played and played with corrie stuff just cannot get it going....

[astral77]

Hi
Please see attached, on the left is David`s (HANOVER) Recent Strength and on the right is your indicator. The relative position of the currencies is the same, IMHO if anything yours is more accurate at this time of night (23:30 GMT).
Kind Regards

[garyfritz]

Thanks, Astral. Many things look similar between the two, but there are very definite differences. One that I think is significant is EUR and CHF. In all my tests the green and yellow lines (EUR and CHF) have been exactly parallel, on top of each other. Which is what you would expect, given how EURCHF has been nailed to 1.2000 for quite a while.

I noticed in your PDF that green and yellow diverged. ??? So I looked at a EURCHF chart, and sure enough last Thursday it jumped, and currently it's up around 1.2017. So my indicator showed very clearly when the EF pair moved. David's indicator didn't show them in lockstep before so it wasn't noticeable when the pair moved. Mine seems to be more sensitive with less noise.

Stimpy, I've got my hands full with this one. I'll look at it if I get a chance but I think I've got a good one here.

[aacd_1]

hi,gary,
Thanks for you indicator. I have a question about it.

From 2012.6.8 15:19 AUDJPY went up slowly，but from the indicator , tha AUD is decreasing and JPY is increasing, AUDJPY should go down.

How to explain this Phenomenon？

[garyfritz]

It's a little tough to say since I can't tell if you have them on the same timeframe. But likely it's a result of the inputs. Unfortunately this indicator is a bit sensitive to its inputs. What did you use?

FYI I've basically abandoned development on this indicator. Baluda's Currency Slope Strength seems to fill the need nicely.

[aacd_1]

It's a little tough to say since I can't tell if you have them on the same timeframe. But likely it's a result of the inputs. Unfortunately this indicator is a bit sensitive to its inputs. What did you use?

FYI I've basically abandoned development on this indicator. Baluda's Currency Slope Strength seems to fill the need nicely.

I use a 33/33/22, BarsBack = 500, in H4 timeframe as you described somewhere. 50/50/10 seems also.

thanks for your advice, Baluda's looks more sensitive to price action and yours looks more smooth.

[frankie]

hi,gary,
Thanks for you indicator. I have a question about it.

From 2012.6.8 15:19 AUDJPY went up slowly，but from the indicator , tha AUD is decreasing and JPY is increasing, AUDJPY should go down.

How to explain this Phenomenon？

Ok i will jump in here with my thought on this.. Sorry but i havent read the whole thread, but I guess that this indicator, as the others out there ccpf.. are based on moving averages. This gives this kind of indicator a downside, as the indicator is calculated on the MAs and what happens when a price stalls after bigger moves? the MAs flatlines and the indicator will then travel towards the zero line. Simple explained, when pricemovement becomes lesser over time the indicator will show less movement, even though there still is movement, but the waves is smaller in sizes, therefor.. as an average of all pairs involved, the same trend will happen and, lets take AUDJPY.. both currencies decrease in volatility and both currencies will gravitate towards the zeroline.. giving you a wrong image.. or right as they both become weak and non moving.. But the indicator, as you read it, indicates that AUD should get weaker (sell) and JPY stronger (buy). But it's only flatlining..

Thats why i am looking for an indicator like this, but based on something other than MAs. But what should it be based on? I don't know, but when i get an good idea i will get back here..

The only way you can use this kind of indicator as they are build right now is, use currencies that moves away from the zeroline! Why? becasue volatility is growing and if AUD is going down away from zero and JPY going up away from zero you have a indication that shorting AUDJPY would be a better choice at that moment... until the indicator indicates weaknes (turning or slowdown).