Forex trading strategy - by Kevin Barry

Kevin Barry is an experienced and successful London Forex professional trader. He basically trades the major Forex pairs on 5min charts, using MACD histogram divergence signals and Pivot levels.

The following explains Kevin's strategy for applying/combining MACDH divergence signals and intraday pivot levels into his own Forex trading strategy.

Although this article was written by Kevin prior to his subsequent purchase of the MACDH divergence kit, MACDH divergence signals were at the time available to him on 5min charts.

Pivot level calculations

Kevin uses the traditional calculations for my intraday pivot levels taken from the previous day's High/Low/Close. However, when trading FOREX, there are extra levels of Support/Resistance halfway between these pivot levels.

(LastHigh/LastLow/LastClose are the High/Low/Close of the previous day)

Pivot:= (LastHigh + LastLow + LastClose) /3; Displayed in Green bold

R1 := 2*Pivot - LastLow; Blue
S1 := 2*Pivot - LastHigh; Blue
R2 := Pivot + (LastHigh-LastLow); Red
S2 := Pivot - (LastHigh-LastLow); Red
R3 := 2*Pivot + (LastHigh-2*LastLow); n/a
S3 := 2*Pivot - (2*LastHigh-LastLow); n/a

M5:=(R3+R2)/2; All middle pivot levels displayed in light grey

MetaStock Intraday Pivots indicator
The MACDH Divergence kit includes Intraday, EOD and Weekly pivot indicators

MetaStock -> Tools -> Indicator Builder -> New
-> copy & paste complete formula between "---8<---" lines.

Pivots - intraday

{ Plots daily pivot values on intraday charts.

Apply only to weekday data:
MetaStock -> Tools -> Options -> Chart Options
-> untick "Ignore weekend data".
With thanks to Kevin Barry for pivot logic.

Copyright © 2004~2007 Jose Silva.
The grant of this license is for personal use
only - no resale or repackaging allowed.
All code remains the property of Jose Silva. }

{ User input }
pds:=Input("Use Highest/Lowest of past x days",1,260,1);

{ Message }
message:=Input("(Plot on intraday charts)",0,0,0);

{ Day's start }
dStart:=DayOfMonth()<>Ref(DayOfMonth(),-1) OR Cum(1)=2;

{ Previous x Days' High }

{ Previous x Days' Low }

{ Previous Day's Close }

{ Pivot }
Pivot:=(Hd+Ld+Cd)/3; {Plot in Green bold}

{ Support }
S1:=2*Pivot-Hd; {Blue}
S2:=Pivot-(Hd-Ld); {Red}
S3:=2*Pivot-(2*Hd-Ld); {dark Green}

{ Resistance }
R1:=2*Pivot-Ld; {Blue}
R2:=Pivot+(Hd-Ld); {Red}
R3:=2*Pivot+(Hd-2*Ld); {dark Green}

{All mid-pivot levels light Grey}

{ Plot on intraday price charts }


The Forex trading strategy explained

For illustration purposes, I have attached the chart from August 27th 2004, showing the intraday pivot levels, together with the MACDH divergence signals numbered 1-6.
I only trade from the London open at 07.00 to lunchtime New York around 18.00 GMT so they're the only signals in which I'm interested.
Please excuse the dodgy artwork (manual MACDH divergence signals)!

Click on charts to enlarge

click on chart to enlarge

The 27th is pretty much a random choice of day. As I've mentioned earlier, the MACDH divergence signals generally work well in isolation but, perhaps, even better when combined with the pivot levels.

In the main, I use a binary BB indicator and pivot levels, together with specific candlestick formations and MACD divergence for confirmation, for my trade entry. I look to enter trades just above/below the pivot levels as they are very often reversal points.

Once a pivot level is broken, I expect the price to reach the next level. My profit objective is always two pivot levels away from my point of entry. Of course, I often miss out on very large price moves by capping my profits but, hey, my tastes are pretty simple.

So, let's take a look at the MACDH divergence kit signals:

1. Long at 10.10. Very good signal. Support at M3.
Also worth noting that with the Cable, there is frequently S/R at 1.xx00, 1.xx20 and 1.xx80.

2&3. Short at 11.20 and 11.35. Again, good signals.
Strong overhead resistance at M4 and 1.8020.

4&5. Long at 1.25 and 1.45. These signals are best ignored.
Resistance at R1 and 1.8000 and the last high is lower than the high of the day just after the London open.

6. Long at 17.50. A decent signal for a quick few pips maybe.
However, IMO, we are too far away from a pivot level to consider entering a trade and the trend for the day is down.


Below is the same chart showing just the binary Bollinger Band indicator, which provides me with many of the trade setups.
It uses the classic 20/2 parameters; I haven't found anything that works any better.

click on chart to enlarge

The best BB signal of the day was around 15.00. As there was no significant confirmation, I would still have taken a long trade once the price broke back above M2, but with a reduced exposure.

My target would have been M3, which had been providing support for most of the day, but watching closely for a reversal at the pivot level, which is, in fact, what happened. Time to exit stage left at breakeven.

The long MACDH divergence signal at 17.50 coincided with both a BB setup and an Engulfing Bull CS signal. As I mentioned, it just might have been worth a small wager despite the lack of pivot level support.
The break back down below S1 would most certainly have been the exit signal.


Below is a print of my MS screen that I've just taken.

click on chart to enlarge

There is nothing that unusual happening here, but you will get an idea of how a day often pans out. I use an hourly chart on the left, in order to get a feel for the general direction in which things are going, with the five minute chart on the right for the trade setup. I actually use my broker's one-minute chart in order to finesse the trade entry.

As well as the manually plotted pivot levels, you can see the modified CS expert, a stochastic, the MACD (spot divergence using that plot if you can!), the 20/2 binary Bollinger Band indicator, a volatility indicator based upon the Kase oscillator that comes as an eSignal addin, and a binary 5/20 MA crossover as a trend indicator.

It's also worth mentioning that the FOREX market reopens on Sunday.
Therefore, which HLC prices should be used for Monday's pivot levels?
IMO, the Sunday session is best ignored and Friday's prices used for Monday's pivots.

Postscript, 15/Dec/2004

"Hello Jose,
Things are going nicely with your MACD Kit. A great signal yesterday, which I acted upon, pretty much paid for the package so I'm already ahead of the game! Thanks. As I mentioned, I'll send you some proper results as time goes on."


Click on charts to enlarge

click on chart to enlarge


click on chart to enlarge

The red indicator is a 20-period volatility and the magenta plot is a momentum indicator.

Postscript, 24/Feb/2005

Kevin's reply to EquisMetaStock usergroup query on detecting early trends.

I only trade reversals. My two main reasons for doing so are that (if you're correct) you can catch most of a move rather than just part of it and, more importantly, every stop can be very tight because there is an obvious and logical place to put it; i.e. once the low/high that triggered your trade is breached, you're wrong and it's time to exit stage left immediately.

A few suggestions for you:

1. Bollinger Bands - watch for a run up/down outside the band and a cross of the price back inside the bands. Stop goes above/below the high/low of the trigger bar plus one point.
If the trend resumes and hits your stop, just watch for a repeat of the set up and do it again.

2. If you're trading stocks, watch out for a volume climax on the high or low of the reversal.
If you see one, double your bet.

3. Watch for a reversal at fib levels. Everybody else is watching them so you should too.

4. Use pivot levels. Same things applies. The beauty of fibs and pivots is that they are amongst the few truly predictive indicators.

5. Divergence between the price and the MACD. One of the very best ways of spotting an upcoming reversal. If you're too lazy (like me) to keep a watch for divergence, take a look at Jose's MACDH divergence kit. All the hard work is done for you.

Trend following is only for the timid ;-).

Last thought

Another factor that I've omitted to mention that might be useful, could be the choice of the trading day on which the strategy would be executed.

For example, a day on which several major economic events were due to occur could be a good candidate for trading.

Also, after two or three days on which the market had been range-bound, is when a trending day is more likely to occur. Conversely, perhaps the day after a big trending day where there were no economic events of note due, would be a good day to avoid.

Kevin Barry

(email address supplied on request)