FPI - Fractional Product Inefficiency

 

Hi there!

Anybody heard about the FPI indicator?

Do you think it worth the time to test it? I hope that!

 

Hi CodersGuru

I think it worths. I started to test this kind of strategy last week and it works so far. However, the profits are small. Small risk = small profits.

Bye.

 

FPI works but you have to be patient. If played with large positions, you may even have income from it (which is not my case)

regards Bluefish

 

Fpi!

bluefish:
FPI works but you have to be patient. If played with large positions, you may even have income from it (which is not my case) regards Bluefish

That's right!

Just want you to update us with your results!

 
 

It is a good system, but you have to use it on an institutional feed. The problem i am having is filtering out big news spikes because it can take up to 3 seconds to process the three trades, and by that time the FPI can change drastically depending on how fast the market is moving. Don't trust FPI indicators, they are incorrect! All of them! It is a flaw in the MT platform combined with bad data. If you want to see the real fpi you have to track tick data second by second for every pair. I have attached a chrt made in excel from tick data that i collected. It matches up perfectly to the trades.

Files:
image-0001.jpg  445 kb
image-0000.jpg  129 kb
 

Nick,

What broker are you using that gives you an "institutional feed"? Might I also ask how you went about obtaining and cleaning up your tick data?

BW

 

Correlation inefficiency indicators?

Hello all,

Does anyone have a correlation inefficiency indicator (I'm not sure if this is the correct term)?

What I mean that if we have a theoretical full hedge (for example eurusd/usdchf/eurchf) and we would like to see an oscillator that shows inefficiencies between correlation at bar basis?

cheers,

Mikko

 

I added inputs so you can use whatever pairs you want to compute the FPI.

Files:
fpi.gif  19 kb
fpi.mq4  3 kb
 

Thanks!

That was exactly what I was looking for. I'm trying a few statistical methods with just visual backtesting so we'll see if this leads anywhere (I'm basically just measuring what size the inefficiencies are on different time scales)

 
TheRumpledOne:
I added inputs so you can use whatever pairs you want to compute the FPI.

Good morning,

not seeing gray midline-

if i change symbols in the code or via inputs i don't have valid fpi numbers-

thanks,

mark

Reason: