FPI - Fractional Product Inefficiency: The Impeccable Hedge

NeoTicker indicators

Moderator: moderators

harryporter2006
rank: <50 posts
rank: <50 posts
Posts: 1
Joined: Sun Oct 09, 2011 2:40 am
Reputation: 0
Gender: None specified

Postby harryporter2006 » Sun Oct 09, 2011 3:45 am

@ Michal


1. Firstly, (a / b) * (b / c) * (c / a) = 1 ==> You are right for normal math !

2. For forex, this formula should be: x*(a / b) * y * (b / c) * z* (c / a) = 1
For FPI, without x,y,z, FPI will fluctuate, am I right ?

3. x, y, z are necessary parameters to calculate FPI
That is why you can not trade 1/1/1 for 3 pairs

Hope I make clear my idea.
Good trade for all

Please add www.kreslik.com to your ad blocker white list.
Thank you for your support.

Relativity
rank: 150+ posts
rank: 150+ posts
Posts: 364
Joined: Mon Nov 15, 2010 4:19 pm
Reputation: 0
Gender: None specified

Postby Relativity » Mon Oct 17, 2011 11:34 am

Experimenting with FPI to find out which currency is being the major mover + how much.

Created a big FPI ring in an inversed manner, then plotting which individual values into a chart.

Here it shows how other pairs are affected on a macro scale, when SNB decided to play its nuclear option.

Interesting findings. Will keep on researching.
Attachments
FPI idea.PNG
FPI idea.PNG (42.19 KiB) Viewed 52898 times

jeuro
rank: <50 posts
rank: <50 posts
Posts: 7
Joined: Thu Oct 21, 2010 5:38 am
Reputation: 0
Location: tallinn
Gender: Male

Lets keep it alive

Postby jeuro » Wed Nov 30, 2011 12:24 pm

deleted repeat post
Last edited by jeuro on Fri Apr 27, 2012 5:04 am, edited 1 time in total.

jeuro
rank: <50 posts
rank: <50 posts
Posts: 7
Joined: Thu Oct 21, 2010 5:38 am
Reputation: 0
Location: tallinn
Gender: Male

Lets keep it alive

Postby jeuro » Wed Nov 30, 2011 12:25 pm

edited
Last edited by jeuro on Fri Apr 27, 2012 4:58 am, edited 1 time in total.

jeuro
rank: <50 posts
rank: <50 posts
Posts: 7
Joined: Thu Oct 21, 2010 5:38 am
Reputation: 0
Location: tallinn
Gender: Male

Postby jeuro » Wed Nov 30, 2011 12:28 pm

deleted repeat post
Last edited by jeuro on Fri Apr 27, 2012 5:01 am, edited 3 times in total.

Please add www.kreslik.com to your ad blocker white list.
Thank you for your support.

jeuro
rank: <50 posts
rank: <50 posts
Posts: 7
Joined: Thu Oct 21, 2010 5:38 am
Reputation: 0
Location: tallinn
Gender: Male

Postby jeuro » Wed Nov 30, 2011 12:32 pm

I read all this thread about a year ego. Very educational, probably the best found anywhere related to hedging. Thank you very much Michal.

I have been trading professionally (manually) many years, but working in automation last 2.
with main focus in hedging, and math related system.

Wondering if anyone interested to take FPI and experiment/create/design a system that would overcome the main issues with FPI, such as the spread, broker executions, etc.

For example.. True fact .. ?The frictional product almost never equals 1? but
in reality is impossible to take advantage of the 1.0003-0.9997 inefficiency with 1 set of trades, due to spreads, and that ultimately, the broker's risk management software would simple offset any possible profit by delaying execution ? but, what about if we could place several set of trades and create a form of averaging down system,.. That would accumulate and multiply the inefficiency until it was safe to trigger a CloseAll at X profit (making sure that the X profit allows the usual profit decline that occurs during the closing process).

Well, this is just one idea that could be explored. In spite that I have been in the financial area most of my life, I am not much of a pure math person, neither my partner that has a master in software engineering. Therefore, we have experience in Trading, making algorithms and coding for mt4, but lack deep knowledge in the required applied math.

Please post here if want to keep this alive and find new new angles of FPI or wants to contribute with your math skills.

Relativity
rank: 150+ posts
rank: 150+ posts
Posts: 364
Joined: Mon Nov 15, 2010 4:19 pm
Reputation: 0
Gender: None specified

Postby Relativity » Wed Nov 30, 2011 1:06 pm

jeuro wrote:I read all this thread about a year ego. Very educational, probably the best found anywhere related to hedging. Thank you very much Michal.

I have been trading professionally (manually) many years, but working in automation last 2.
with main focus in hedging, and math related system.

Wondering if anyone interested to take FPI and experiment/create/design a system that would overcome the main issues with FPI, such as the spread, broker executions, etc.

For example.. True fact .. ?The frictional product almost never equals 1? but
in reality is impossible to take advantage of the 1.0003-0.9997 inefficiency with 1 set of trades, due to spreads, and that ultimately, the broker's risk management software would simple offset any possible profit by delaying execution ? but, what about if we could place several set of trades and create a form of averaging down system,.. That would accumulate and multiply the inefficiency until it was safe to trigger a CloseAll at X profit (making sure that the X profit allows the usual profit decline that occurs during the closing process).

Well, this is just one idea that could be explored. In spite that I have been in the financial area most of my life, I am not much of a pure math person, neither my partner that has a master in software engineering. Therefore, we have experience in Trading, making algorithms and coding for mt4, but lack deep knowledge in the required applied math.

Please post here if want to keep this alive and find new new angles of FPI or wants to contribute with your math skills.


I welcome you to our forum Jeuro =>

I am no math wizard neither, just know some general practical basics like e.g. bell curves, distribution plots, standard devs ..etc. I don't think its much, but it does help in my trading.

I had been working on an offshoot version of FPI, to assist me in making decisions for carry trade strategies. I might want to make synthetic currency indexes/charts in future out of this thou.

Here's the thread : TrendTrader

quantdeanj
rank: <50 posts
rank: <50 posts
Posts: 1
Joined: Mon Oct 31, 2011 12:07 pm
Reputation: 0
Gender: Male

Postby quantdeanj » Sat Dec 10, 2011 10:04 am

I have developed the FPI technology which includes aggregation for 12 liquidity hubs (direct bank and or ECN FIX connections, all non institutional brokers) including EBS, SolidFX, Hotspot, Integral, Rocket etc etc, automation algos for FPI as well as STAT ARB, risk management, position tracking etc in addition its possible to now to take advantage of linear ARB and Tri ARB opportunities. A key part of implementing this strategy is realtime netting which we have as well via a couple of destinations (which I wont name here). Now we need scaling capital so any one with deep pockets to provide working capital and act as a partner in this project, contact me at arb @ tentops . com. Only serious investors / partners apply please.

SteTrader
rank: <50 posts
rank: <50 posts
Posts: 1
Joined: Thu Dec 08, 2011 9:25 pm
Reputation: 0
Gender: None specified

Postby SteTrader » Sun Dec 11, 2011 5:15 pm

michal.kreslik wrote:I got a question about how does such an FPI trade look like.

Let's use a commonplace FPI ring:[list][*]SellShort EUR/JPY[*]Buy EUR/USD[*]Buy USD/JPY[/list:u]I picked this one because it features low spread. To make the example a real-world one, let's consider the Oanda Bid/Ask spreads. They are fixed to:[list][*]EUR/JPY: 2 points[*]EUR/USD: 1.5 points[*]USD/JPY: 2 points[/list:u]Now let's find some FPI extremes for this ring:



The FPI extremes of this magnitude happen every day. There are many times bigger differences sometimes. But let's settle for this "everyday" sample magnitude for now.

Input information for the Impeccable Hedge open (considering Oanda spreads as described above):[list][*]FPI: 0.9999[*]bar close date/time: 6/29/2006 3:15 PM[*]price: close[*]EUR/JPY: Bid 145.61 / Ask 145.63[*]EUR/USD: Bid 1.2647 / Ask 1.26485[*]USD/JPY: Bid 115.15 / Ask 115.17[/list:u]Input information for the Impeccable Hedge close:[list][*]FPI: 1.0004[*]bar close date/time: 6/30/2006 9:45 AM[*]price: close[*]EUR/JPY: Bid 145.88 / Ask 145.90[*]EUR/USD: Bid 1.2753 / Ask 1.27545[*]USD/JPY: Bid 114.34 / Ask 114.36[/list:u]Trade analysis:[list][*]SellShort EUR/JPY[*]open: SellShort 100,000 EUR/JPY @ 145.61; we pay 100,000 EUR, we get 14,561,000 JPY[*]close: BuyToCover 100,000 EUR/JPY @ 145.90; we get 100,000 EUR, we pay 14,590,000 JPY[*]profit/loss = -29,000 JPY[/list:u][list][*]Buy EUR/USD[*]open: Buy 100,000 EUR/USD @ 1.26485; we get 100,000 EUR, we pay 126,485 USD[*]close: Sell 100,000 EUR/USD @ 1.2753; we pay 100,000 EUR, we get 127,530 USD[*]profit/loss = +1,045 USD[/list:u][list][*]Buy USD/JPY[*]open: Buy 100,000 USD/JPY @ 115.17; we get 100,000 USD, we pay 11,517,000 JPY[*]close: Sell 100,000 USD/JPY @ 114.34; we pay 100,000 USD, we get 11,434,000 JPY[*]profit/loss = -83,000 JPY[/list:u][list][*]Overall profit/loss at the time of trade close in USD:[*]EUR/JPY trade: -29,000 JPY / 114.36 (USD/JPY) = -254 USD[*]EUR/USD trade: + 1,045 USD[*]USD/JPY trade: -83,000 JPY / 114.36 (USD/JPY) = -726 USD[*]Sum: 1,045 USD - 254 USD - 726 USD = 65 USD netto[/list:u]So our Impeccable hedge trade did earn us net 65 USD after spreads. That's 6.5 pips net if you translate it to EUR/USD trade. You may say it's nothing to write home about, but it's sure 6.5 pips in your pocket without risking anything. You see, we were

FLAT

all the time and still were able to make a profit. A fairy-tale magic :D

You may be curious what happens if you miss the FPI = 1.0004 exit point. Take a deep breath before I tell you!

.... ready?

.... stop cheating, you didn't take a deep breath, man...

.... so

why, absolutely nothing happens :) As I said in the article, your Impeccable Hedge position may be open for as long as you wish without exposing you to any risk. It's like having EUR/USD and EUR/USD "futile hedge" open. You don't gain, you don't lose. You're just flat. So you may sit back, relax and watch the ripples on Miami Beach and wait until FPI is say > 1.0003.

The wonderful thing is FPI always wiggles up and down, so sooner or later, you'll be able to exit with a small profit :D Now isn't life beautiful? :)

To give you an even better idea of how FPI works, let's have a look at this common-or-garden sine wave I created in excel:



You see, the sine has got its extreme points at the crests and then always returns back. It cannot get alarmed and go out of the boundaries. The same is with FPI. It's like a somewhat crippled sine wave. Sooner or later, the crest will appear.

Michal
Hi, Michal, do you have any more news about this issue five years after this beautiful article?

brunobiondo
rank: <50 posts
rank: <50 posts
Posts: 1
Joined: Sat Dec 17, 2011 4:14 pm
Reputation: 0
Location: Rome
Gender: Male
Contact:

Your code in Metatrader4 Language (MQL4)

Postby brunobiondo » Sun Dec 18, 2011 12:47 am

Hi michal,
I'm a software programmer trying to become a honest trader for living,
and I'm wondering if, before attempting to port your FPI code written in C# to metatrader4 language MQL4, that's substantially C language,somebody already has done the job for MT4.

Thank you for assistance
bruno

Please add www.kreslik.com to your ad blocker white list.
Thank you for your support.


Return to “NeoTicker indicators”