Forex4 System

Hello traders,

the first of five weeks (30 trading days) of my new strategy is over and I am sharing my results with

the community; some of you will remember my “Seven Majors System” thread, so will recognise similar

features to this system. The last system was on demo and was marginally profitable, but I only gave it

two months of testing; this system will be tested for one month, initially, but the aim is to keep it up

for a few months unchanged, unless losses necessitate a radical rethink.

How does the system work? It is a basic ‘straddling’ system, so it is built on two entry orders (one buy

and one sell) for each chosen pair, and the stops/limits are based on each pair’s ATR (Average True Range).

For the first week of testing, I used the following pairs, based on the fact that they all totalled 100 points

or more in daily ATR:

GBP/USD; AUD/JPY; CAD/CHF; EUR/NZD.

More importantly, on the currency correlation charts, none of these pairs at the start of the past

trading week (10th Feb. - 14th Feb. 2014) showed a strong correlation on a daily or even weekly

basis.

To sum it up, the criteria for choosing these pairs were:

  1. eight Major currencies traded in four pairs with weak correlation, meaning no doubling up on losses;
  2. ATR of 100 or more pips (on Daily time frame), as a sign of sufficient probable volatility;
  3. two orders per pair meant eight orders in total, which is not too large a number to be handled manually
    at the start of each trading day (just before my actual day job).

The aim, which was reached, was to enter all orders before 9am and, unless automatically closed, close
them manually (or remove them, if untriggered) by 5pm, GMT.

The Risk-Reward ratio was 1:2, meaning that the distance between the entry point and the stop was half that between the entry point and the limit; each stop was set as a Trailing (Dynamic) stop; the lot size for each and every order was always 4k. The latter was decided on a target of 150 pips to be gained within the trial period, thus around 30 pips per week; the amount chosen means a leverage of 8-to-1 per trade, based on the account balance; as there are floating positions (pre-dating the current system trial), the leverage calculated on the equity is closer to 10-to-1 per trade.

The system was trialled in the first week with a parallel version being served up on a demo account using the same pairs and same ATR-based targets (and same lot sizes) BUT setting the entry orders once at the beginning of the week and basing the ATR on a weekly time frame, letting them run for the entire week until
Friday at 5pm GMT.

In the real account, the system used the previous day’s (10pmGMT /5pm EST) close as the value around which the orders would be calculated; however, due to weekend price jumps, the Monday orders would be set not on the preceding Friday close but on the Sunday evening open value; in the demo account, on the other hand, the values on Monday morning would be set on the close values of the previous weekly candle.

Here are the results for the first week:

REAL account (Daily ATR targets):

Total P/L: +40.8 pips / +11.18 GBP;

DEMO account (Weekly ATR targets):

Total P/L: -79.4 pips / -27.75 USD.

The REAL account saw 13 out of the total 40 orders being triggered (that is 32% of the total),

whereas the DEMO account saw 6 out of the 8 orders being triggered (that is 75% of the total);

however, the REAL account set-up was profitable, while the DEMO one was not.

For the week ahead, I will make the following changes, and for the given reasons:

  1. the pairs chosen will be different, in that the previous ones are all showing stronger correlations
    now on a Daily chart; the new pairs are all showing weak correlation (on both Daily and Weekly
    charts), and they are:

    EUR/JPY, AUD/CAD, GBP/USD, NZD/CHF;

  2. the least volatile pairs, AUD/CAD and NZD/CHF, which score less than 100 pips on the daily ATR,
    will be traded using weekly chart ATR values, and triggered orders will be left to run through the
    entire week; if stopped sooner, they will be re-entered the following day; this may also allow any
    positive rollover to be accrued after each day’s close, depending on certain factors. By trading these
    two pairs on weekly ATR values, we may be able to more successfully hit the 100+ pips range;

  3. the times of trade for the daily orders will now be 8am to 5pm, to coincide more precisely with the
    London open and close times; trades will be entered between 7.55am and 8am and closed at 5pm
    exactly (or removed, if untriggered);

  4. the DEMO account will be used to replicate the trades of the REAL account, but they will use
    Trailing Dynamic stops, whereas the REAL account will now use unchanging stops, to allow more
    room for pairs to breathe; results will be compared at the end of the week, to see which type of
    stop allowed for better preservation of capital as well as better profit margins;

  5. entry price will no longer be based on the previous day’s close, but on the price at the time of
    the entry order being set: this is to avoid missing on moves where price has moved considerably
    since the previous day’s close.

Happy trading.

Hello traders,

this is my trading log for the second week of my new trading system, Mon. 17th to Fri. 21st February 2014:

>>> REAL account (Daily ATR targets):

Total P/L: -353.8 pips / -70.25 GBP;

>>> DEMO account (Weekly ATR targets):

Total P/L: -301.7 pips / -117.52 USD.

Both Real and Demo accounts were unprofitable, to a similar degree.

Having read pages 163 - 165 of Abe Cofnas’ “The Forex Trading Course” (The Forex Trading Course: A Self-Study Guide To Becoming a Successful … - Abe Cofnas - Google Books), which deals with ATR trading and setting stops, and having observed the way in which my trades on both live and demo accounts have been so frequently stopped out - the live account having fixed stops, and the demo account having trailing ones - I decided to revise the Reward-Risk ratio from 2:1 to 1:1, which will allow more breathing room for price action.

The new Reward-Risk ratio will be implemented from the following trading week (week 3 of 6), and traded with fixed stops on the Real account, and Trailing Dynamic stops on the Demo account.

The Currency Correlations table shows that the four trading pairs used (GBP/USD; EUR/JPY; AUD/CAD; NZD/CHF) have acquired a strong correlation on a Daily basis, but not on a weekly basis; also, the ATR seems to have dropped to below the 100 level for some of these pairs. Given these two factors, I have decided that I should allow trades to be open as weekly trades, with the aforementioned (new) 1:1 Reward-to-Risk ratio.

All trades shall be placed on Sunday (today) after 10.15pm GMT (the opening of trading on FXCM).

Positions that have not been triggered by the end of the week (Friday, 9.55pm GMT) will be left untouched; those that have
been triggered but have not hit a stop or a limit by the end of the week will be left open until the following week, and
calculated into the weekly trading stats separately as floating orders. Trades that have been opened and closed within
the week (but with some time to go before the end of the week) will not be replaced until the start of the following trading week, after 10.15pm (GMT).

Happy trading.

Hello traders!

Here are the stats for the end of my third week (out of six) of testing for this new system:

>>> REAL account (Weekly ATR targets; fixed stops):

a) Closed Positions:

Sell EUR/JPY: -124.8 (pips) / -29.36 (GBP);

b) Floating Positions:

  1. Buy GBP/USD: -2.2 (pips) / -0.53 (GBP);
  2. Buy NZD/CHF: -54.4 (pips) / -14.77 (GBP);

>>> DEMO account (Weekly ATR targets; dynamic trailing stops):

a) Closed Positions:

  1. Sell EUR/JPY: -33.8 (pips) / -13.23 (USD);
  2. Buy NZD/CHF: -39.00 (pips) / -17.71 (USD);

b) Floating Positions:

Buy GBP/USD: -2.2 (pips) / -0.88 (USD).


Conclusions:

  1. the real account lost more pips in one trade, but has two floating trades which still offer profit potential, against only one in
    the demo account;

  2. the real account trade was stopped out but soon after began reversing, only to finish the week around the same level as it
    closed; the ATR-based stop was therefore a good measure of expected price movement;

  3. there were less losing trades than in the previous week, as more breathing space was provided for these pairs; while no
    trade was triggered in the first two days of the week, this gave price a chance to really pick up directionality either way
    and only engage in the trades with some conviction;

  4. the EUR/JPY trade in the real account was initially in the money for about thirty-plus pips, and the EuroZone CPI +0.2
    reading caused a 100+ pips rally; however, there was potential for a continuation to the downside, initially, which meant
    that the fixed-stop approach was better in the beginning (whereas the trailing stop in the demo account killed the trade
    early).


STRATEGY for the following week:

on a currency correlation table (such as Oanda’s Forex Correlation | Currency Correlation Chart | OANDA fxTrade Europe) one can see that the chosen four pairs have, once again, shifted their relationship and showing signs of increasing inter-relatedness; however, only the EUR/JPY - NZD/CHF couple shows a high (negative) correlation on a weekly time-frame, which is to be expected, given the innate correlation between the Euro and the Swiss Franc.

One major problem with this system is that, unlike a trend-following system, it truly needs more volatile pairs than the ones chosen (i.e. the currency pairs, even within a bigger trend, need to be ranging significantly in order to not only trigger the entry orders but also to hit the profit targets).

I spend some time researching historic FX volatility and that for the past year, while also back testing (manually) the set-ups that I have for the current pairs, and came to the conclusion that:

  1. more manual testing will be conducted, including a view to introducing more conservative take-profit targets (75% of
    projected movement in favour of our trades, rather than 100%), and reviewing fixed stops vs. trailing stops;

  2. the pairs chosen may not necessarily be the right ones for this strategy: more volatile pairs may be required. However, depending on the time-scale consulted, the historical average will change; also different online sources give slightly different data regarding which currency pair in the medium-past has been the one with the highest volatility. Choosing the ‘right’ pairs for the system, then, will mean not only using pairs that are highly volatile but also ones that are consistently so over time;

  3. I am considering using a different Risk-Reward ratio, for which point 1) is part of the picture.

Happy trading.


UPDATE:

three out of my four pairs from last year are now open, at the start of the trading week (Sunday evening), so I will leave
this set-up to unfold; the Buy EUR/JPY order may open later in the week, and if not, I will wait for the other three orders
to close first, to try to start with a clean slate the following week.

As an alternative to the current system, I am considering trading one currency pair at the time, over a long time, as it may lead to a better understanding of each pair concerned. The rule for trading a pair would be that it should have a volatility of at least 100 pips on the ATR (Average True Range) reading.

UPDATE for my fourth week…

The system seems a little ineffective…

The same problems seemed to occur with this system as with the Seven Majors: it needs a lot of volatility to work… On quiet days, nothing happens, OR, you get false triggers and then you are stopped out…

When the big swings happen (e.g. during news releases or other big triggers), then you will bag those pips… However, where my initial Risk-Reward ration was 1:2, it had a negative edge; when I changed it to 1:1, it performed better, but the wider trigger points meant that you had to wait sometimes a lot longer for something to happen (and even then, you may not have been guaranteed a profitable outcome).

One difficulty is that I started leaving orders running from one week to the next and, in the meantime, I have been placing other (shorter-term, smaller) orders beyond the chosen system’s pairs… Keeping track of this rather fluid system has become rather complex, especially as the closed orders disappear from the trading platform at the New York close, so if you do not
make a note of them you then have to run a report and extrapolate that manually…I have been in a busy week and so I do not have a note of those orders without getting them from a report (NB: the report does not give you the number of pips gained or lost in a trade: you have to work it out yourself, which adds more time to the calculations).

More to the point, I know that most of my transactions in four weeks of this system have ended losing me pips/money, so I would have to really have a thick skin not to take notice! I am taking notice and I am taking a break from the system until I know what better alternative could be found.

Still, I am not saying that it was a waste of time, and I think that some of the questions (e.g. currency pairs correlation and Risk/Reward ratios) are still going to be keeping me busy for a while!

Happy trading.

Hello traders…

I think that it is always good to look back on previous threads one wrote, and to see how one’s mind has been shaped by events, or what impact a decision has had on our future…

In the case on my last post here, nearly two months ago, I was saying that the system was losing me money, and that I had decided to cut my losses by abandoning it…

Well, the equity and balance of my account have been steadily rising since I dropped the ‘Forex4’ system, and the question is: does conforming to self-imposed rules work in itself? The message sent out by most experienced traders is that we must have a trading plan… Often times, this can and will be misinterpreted by tech-focussed newbies, who will devise (like I did) a trading plan in the form of a system with clear rules and fixed parameters… The good thing about this is that it removes the emotions from a newbie’s trading, and gives him/her a sense of impartiality, of not becoming too involved with their trading decisions… However, having a mechanised or semi-automated (or manually run) ‘system’ is not always the same as having a strategy or trading plan… In other words, the kinds of systems that a newbie (and even a more experienced trader, I imagine) could devise and run may still need fine-tuning, and, in some cases, they actually require more and more sub-parameters, in order to cope with the varying conditions that a chosen pair or pairs may present the trader with as time goes by… A simple system may then become increasingly more complex, and requiring a lot of maintenance, meaning that their time-saving potential would only really be realised by full automation (Expert Advisor or other form of coding). . . However, having a fully automated system is not, per se, the same as having a strategy or trading plan… A trading plan or strategy could incorporate a system (like the ‘Forex4’ one that I had devised) but the reverse is not necessarily true: a fixed system may be devised and running, but in itself it does not constitute a strategy, as my mentor-ess PipNRoll had once highlighted to me, in her high wisdom… Unless a system is more than a set of mathematical rules from a chart, and incorporates FLEXIBILITY WITH VARYING MARKET CONDITIONS, for example, it will only work in some conditions, but not in others …A human trader, on the other hand, will be much quicker at assessing whether a pair, for example, had gone from trending to ranging, and could adapt his/her trading style to the changing conditions or even decide to wait for conditions to return to what they would consider more in tune to their trading strategy/style…

An interesting debate, no doubt…

Regards

Having read pages 163 - 165 of Abe Cofnas’ “The Forex Trading Course” (The Forex Trading Course: A Self-Study Guide To Becoming a Successful … - Abe Cofnas - Google Books)

Well, I finally bought that book, eight months later…it arrived today…it looks great…

:slight_smile: