RPI Trading

Welcome to the RPI trading system thread!

The RPI (Relative Price Index) is an indicator I have created to trade market swings. It is the core of my system. Not because it’s some holy grail magic number generator that is essential the forex market’s Eye of Sauron, but because it is a simple way to measure the relative trend of the market.

Additionally, the RPI is not designed to give us entry and/or exit signals. It is designed to essentially be a flag you raise so you can see which direction the wind is blowing.

The calculation for the indicator is very simple. It is as follows:

RPI = (C - LL) / (HH - LL)

 

1. C: Current/Close Price
2. LL: Lowest Low of the Period Being Measured
3. HH: Highest High of the Period Being Measured

The default period I use is 14. Because it’s magical? No. I use 14 because I initially started creating this indicator based off the general idea of how the RSI works but instead of measuring the average Rate of Change of the the last 14 bars, I wanted to measure the change in the time series as a whole to provide more contest.

With the RSI, the 14 period is the default because the distribution of it’s reading form a bell curve.

The distributions of returns on any one candle in the market also form a bell curve. However, the actual distribution of price over time is essentially a reverse bell curve. The distribution spends more time at the tails instead of hanging around the mean.

The goal was to create a modified version of the RSI that closer reflects the distribution of price over time.

Comparing the distributions of the RSI vs the RPI using the same period show that this model is a pretty good way of doing that. It’s not perfect, but it’s good enough.

Using RPI is simple. If the indicator is reading above 0.5, price is currently trading above the mean of the periods total range, therefore, we are looking to trade long. If the indicator is reading below 0.5, price trading below the mean of the periods total range so we are looking to trade short.

Again, the indicator is a flag you raise up so you can see which way the wind is blowing.

For all the Crucial Point acolytes out there, the idea is very similar to a rolling version of this:



There is a lot more context we can draw about the marketing using this indicator and this will be covered in later posts, but for now, I would recommend using the indicator as described, getting used to how it moves with the market and finding a period that works for you. There is no magic number in terms of what period will work for you. It's really about personal preference. I use 14.

Attached is the MQL4 source code for the indicator along with a link to the source code for the indicator on TradingView for those who use that platform.

If you have any questions, please post them below!

Trading View Indicator Source Code: https://www.tradingview.com/script/B...e-Price-Index/
MQL4 Source Code:

Following

Following

Following...looks interesting

Loaded up the TE for this thread. I trade with Oanda and I have noticed TE's on here do not process trades well unless they use standard order sizes (100,000, 10,000, 1000 units). I rarely end up using extremely round numbers like this since you can be extremely granular with your order size. MyFxBook seems to handle it a little bit better, but even they have issues so I will be adding a link to the account on there as well.

First position I plan to build into is GBPNZD.

Pair has been trending upwards for a couple months now. 3 weeks ago the RPI crossed above .50 and has remained there since in a pull back state.

The wind is long... Price is essentially 50/50 bet on moving 600 pips or more from here. If this will happen, typically can expect the move to occur within the next 4 weeks.

Putting out a feeler position to see what it does and get some skin in the game. As long as price is above my average, I will build the position on pull backs with 1:1 leverage orders. May add in drawdown with .25:1-.5:1 leverage, but I typically only like to add to winners because of the statistical significance it has on the overall trades odds.



A key decision area in the market for me is around 1.9921, how price behaves around here will determined if the wind will pick up or die.

Hi StormpikeT,

I have also played with the RPI in the past and tried to adapt it to a trading system.

My system looks like this (following the principle "keep it simple"): I look at the DAX on an hourly chart (Broker Roboforex). There are 14 H1 candles per day. My RPI is calculated with double of the number, that means a period of 28.
- Open a LONG position when the RPI at the end of an H1 candle is < 10%.
- Close it when the RPI at the end of an H1 candle is >93%.
- Open a SHORT position when RPI > 90%.
- Close it at RPI < 7%.
- If another LONG or SHORT signal comes before closing, I also take it, but only three positions in parallel.
That's it. No SL and TP considered.

For a test I loaded the last 900-1000 candles of the DAX into the Excel. The result is about 71% days in plus and an average gain of 70-80 DAX points per day. It seems to work quite well. (I have attached the Excel)


Since this can be packed well into an EA, I will unpack my programming skills.

thanks a lot and regards
Rolf

Very interesting way to use it!

You seem to be trading it the reverse of how I am using it in that you are looking for a mean reversion. Since trading the DAX, having you ever considered only trying the long side for reversion? Only ask because indexes are usually weighted to the upside so that natural upward drift will be strong baseline/foundation for any edge from the RPI to build on top of. (For example S&P500 is 3% more likely to move up than down, so trading down, you basically have a 3% + trading costs disadvantage to overcome vs trading the long side).

Looks like wind is picking up north on GBPNZD. Will be adding on dips as long as my average can stay below 2.04441.

Winds are really picking up. Currently up about 200 pips from initial entry. Added a bit more with 1:1 leverage. Currently up about 100 pips from the average.

Goal of adding is to boost what is the equivalent to an All-In EV for those who are familiar with poker. How I calculate EV with the RPI readings will be explained later. Hopefully sooner than later if I find the time.

Thursday Vibes

Some form of confidence...

This is what every trader is looking for with their system. The entire point of the system is to provide some form of confidence that you will make more money than you will lose over the long haul.

And you either fall in 2 categories:
1. You have a system, but do not attempt to validate it.
2. You have a system, and you start to validate it's edge, but you keep going deeper and deeper.

I honestly don't know which is worse. As an engineer, I think #2 is worse.

"If it were up to engineers, a car would never leave the factory"

I used to be in #2. I over engineered everything and it lead no where.

These days, I only focus on averages. Drilling down any further on any data set is a total waste of time when you are working with a continuous process like the market.

Averages are also easier to hypothesis test to see the probability of the difference being random. Use an alpha of 0.05 when testing BUT do not think for a second think there is a 95% chance that the difference is significant.

This should only be a basic test to determine if you want to look further. It's pretty easy to pass so you will find plenty of opportunity to look further. This is the signal.

Now do it again with a baseline (random entry/exit). This is the noise.

See how the t-values change between the baseline and the signal and the noise.

NOISE:


SIGNAL:


SIGNAL VS NOISE:


I'd say the probability that that the difference, if one is even present from the signal for that series, is not random is about 80%.

I repeat. This does not mean the signal produces an edge 80% of the time. God only knows how often it actually produces an edge. It just means that if an edge appears to be there, I'd give it roughly an 80% chance of it actually being there. For now.

That gives us some degree of confidence...

Keep an eye on that number though, it will like change overtime.

I'd take it one step further in the analysis/testing but would not go beyond that.

After that it starts to get really dumb.

And what is already here seems pretty dumb itself.

For clarities sake, those t-scores compare the mean drawdown vs the mean run up of each pair based on each wave based on the RPI direction.

basline is the same thing but with random start and end points of each wave.

GBPNZD trade done. Target hit and had about 2 weeks to get in on the act in/build a position. Up a cool 6%


Granted that spike could have easily just of going the other direction stopping us out... But... Since we trade on averages, we were expecting it to go either direction with 50/50 odds.


Worse that could happen is some horrible slippage doubling our loss. That is why it's important to trade small. Small enough that you could easily survive a 4x worse case scenario move.


Many don't plan like this. They think there stop loss is their max loss, and maybe some slippage will worsen it.

When market gets swept like this though you are 100% royally fucked.

Imagine if you were trading with a 30 pip stop and the market sweeps 400-600 pips like it did. You are thinking your max loss is 30-40 pips with slippage but then you take a 10-15x loss.

that will kill you.

Trade small. Trade larger time frames. Expect everything to be about 50/50 direction wise and manage your risk...

Cheers to a new week!

Here the 4 trades I made for the pair in this position:

Been eyeing gbpusd for a short. Might start a position before the Europe session.

Started a GBPNZD short. Really good RR opportunity on this one. Looking to ride it down for 1000 pips over the next couple weeks (50% chance of traveling that far during this down swing that started after last weeks close).

Shorter timeframes are pulling back but the 2 larger timeframes still point down. Daily is exhausted like no other, but weekly just turned down, and the 1 hour that has been trending is up is very exhausted and likely to turn.

Reversal level is 2.0653. If price rips through that level, I will likely close out the short.

Will not be adding to this one in drawdown since I am already entering at a pretty significant pull back. Will be adding if price moves in our favor.

Just a fair warning, markets will likely be super volatile over the next couple weeks. TRADE SMALL!

The system is simple and easy

That's the goal! A simple system/framework for making better trading decisions

Up 351 pips on first GBPNZD entry.

Added another, first entry is at BE, just incase we get a crazy reversal (very likely to happen, 50/50).

Hourly and 4 Hour waves have turned down like expected. Remember the 1hr was super extended, and that will lead the turn of the 4 hour.

In a perfect world, we will see a dump to our target, but in today's corona world, who knows what the hell happens next haha