But will it live up to it in forward testing ? - page 2
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
It is on a live Demo account already and has been so for about a week.
Thanks for your advise.
I posted on this forum a few weeks ago that I had taken a couple of years and tested in excess of easily a thousand strategies and run what will have been tens of millions of backtests on those strategies to have finally arrived at one where I was finally confident enough in it to commence forward testing on a Demo account as a prelude to running it live - It is a stage very few of my strategies reach.
I have been a little more diligent this time around as my first forex trading experience saw me throw $5K down the drain as an early EA failed me - or rather I enthusiastically let it lose too early in its proving.
I had attached to the original post the Backtest graph and stats.
The great thing about this EA is its very low DD% and it still had a great return on the initial investment.
At that time and so as not to contaminate the results with money management issues, The EA traded all trades as one lot in size.
My original $5000 demo start bank is now $ 8,652.75.
I will continue to forward test and while I have no idea of how long it will take me to commit to backing the EA with real $ I probably need to see a good couple to a few months of results to throw a bank of about $5K at it.
I have now added money management to the EA - I am prepared to risk the [optimised] 16% of my Account Equity as it keeps the DD reasonably low.
I have attached to this post the backtest graph and stats having added this money management. It is the exact same period and data used on this backtest as was the last where I posted the graph and stats.
In these stats the DD is larger than the original 0.67% - it is now 1.46% but given the profit lift from the original (1 lot each trade) of $132,154 over the roughly 27 Month period between 2014 and mid 2016 we now see that the profit in the same time has jumped to $4,775,151 for the same period. Simply the Lots for each open got much bigger as the equity grew. The relative return comparisons to the original 1 Lot test has convinced me to turn on the money Management in the Demo.
My losses as a proportion of the gains in this second backtest results are still acceptable both in numbers and value.
The EA was developed from an observation that the next bar will nearly always be either Higher or Lower than the current bar - this happens with a certainty of upward of 90% of the time on all time periods and on some it is as high as 99%. I note that I have found little else with as much predictability in the 1 Min Bar charts.
You can create a little EA that counts higher or lower next bars as a percentage of all bars and cycle thru bar time periods from 1 Min and then add 1 Min increments right up to - well I tested up to 20,000 minute bars to find that as the Bars (time) increased so did the percentage of truth that the next bar will be higher or lower.
Anyway from this observation I developed this strategy and built the EA.
I simply use Pending orders generated at a price distance above and below the current close with a regular constant time frequency.
These Pending Order [bombs] lay in wait to try to catch and then ride a future Bar that has a significant enough rise or fall in it to generate a profit.
You need to be patient or have a Robot to run this system.
It sounds and, on a chart where the open prices are shown as a horizontal lines, looks very Grid trading like - but I use only Pending orders and convert to live positions only once the entry criteria have been fulfilled.
When at first I started testing I found that the EA triggered a lot of orders at the wrong end of a bars price movement - that is I was opening orders at the end of a rise or a drop within the Bar and so I incurred a lot more losses and a greater DD. I then developed an order open or entry condition to highlight when the price is at an extreme to its recent average thereby suggesting a higher potential for it to retract from its rise or fall [price] trajectory and if so I don't open the trade and instead delete that Pending order to ensure it will not be triggered.
I do still catch a few bars that don't have a decent enough price rise or fall left in them to make a profit or where the intra bar price direction is reversed quickly after opening an order, enough to trigger my very tight SL conditions and so do I lose out - mainly commissions lost- but it is these very tight SL rules that do minimise the individual order losses and the overall Investment DD.
The profit happens when the Pending Orders laying in wait, like mines waiting for a target price to cross their path, are triggered and so convert to open position(s) and the Bar, during the positions short life, makes a steady and significant up or downswing. The [trailing] SL follows very closely behind the price fall or rise, recalibrated at every tick, locking in profit gains. The moment there is even a slight backward price tick or two in the price this triggers the SL and the position(s) close to lock in the profit.
Pending Orders that go way out of the money (their open price is too far away from current price) are just deleted by the system.
It is really a fairly simple EA and strategy.