EURUSD - 2012-04-28

Not many updates lately, because, meh.

My breakout trading strategies have not been working lately. I like to pretend it's because something strange is happening in the market, and point to the following chart as an example:

Daily EURUSD chart with trend line

Nothing much will happen until the EURUSD either breaks below 1.30 or above the bearish trend line. Until then, chop city. 

The following chart shows price has barely poked it's head out above the Ichimoku Kinko Hyo chart, but looks to be stalling again, which is consistent with the above chart and the bearish trend line:

Daily EURUSD chart with Ichimoku Kinko Hyo

In reality, other currency pairs have been moving, it's just that my breakout strategies are most probably flawed and I should quit using them. 

Looking at this chart, we can see that despite the contained nature of the EURUSD, there was in fact a trend showing on the 4 hour chart:

4 hour EURUSD chart

It's pretty choppy, the candle wicks are long and indecisive-y, there's really no way of fitting in a nice breakout there, but despite all that, there's a bullish trend line.

I haven't had any decent trades on EURUSD for a long time. I've had a little better luck with AUDUSD, and some with GBPUSD. Last week I placed a nice bearish trade on EURAUD, and this has restored my faith in trading a little. 

I think it may be time to ditch trading EURUSD and concentrate on something else for a while. 

Robot Results April 23rd - 27th

Let's do this!


Inside Bar

Last week lost $275, this week up $217, with one order open for another +$304. Two AUDUSD wins, one big AUDUSD loss. Balance is $4839.

These are the results are "optimizing" the settings with the back tester. They're OK results, I suppose. Over all, since switching to the H4 time frame, this robot has been steadily making small amounts of money. Also, it's not using true inside bar detection, which adds a small amount of impressiveness.



Pending Reversal Candle

Last week lost $366, this week up $83 with another order open for +$63. Balance is $5279.

When you consider this robot has been running for quite a few weeks now, a +$279 return isn't that impressive. But this is actually one of my better performing robots. It's done well in weeks when others have failed miserably, and it's relatively consistent, so far.

While ultimately it may not be a winning strategy, this robot along with "Inside Bar" have shown that the H4 time frame is perhaps the one to be using, and has contributed to an idea about momentum I've been slowly formulating the last couple of weeks.


Daily Close Breakout

I didn't have much hope for this robot, but it's surprised me for a few weeks in a row by pulling out win after win. Last week it gained +$289, but this week gives up a few weeks profit with a $706 loss. Balance is $4209.

Looking at the chart, we can see the price on the AUDUSD was close to reaching a support, and then rallied unexpectedly. This robot lost there.

While optimizing failed, I think I should adjust the stop loss to be half the range of the breakout zone.



Weekly Breakout 2

Lost $467 last week. This week says "screw it" and loses $1089. Balance is $3627. One order is open for +$161.

During back tests, this robot performed really well across several currency pairs. So perhaps it's just having a run of bad luck with the market lately. Although at this rate, it will have no money left to trade with when the market "comes right".

I should adjust the stop loss, perhaps even do an optimization of the settings.

I have another robot based closely on this one, but uses a dynamic breakout zone (still based on the weekly closing price). It's also done poorly lately.

To think that this was at one stage my primary hope! Well, we'll see I guess.




Open Close Bot H4

$413 down last week. Down $196 this week. Balance is $4339, after having to reset the demo account last week.

I'm tempted to mess with this one and "improve it", but I don't think much can be done with it. While you can clearly see a sort of up-trend in EURUSD this week, look at the length of all those candle stalks! Especially to the upside. You'd never think this trend would continue, yet it does.

The primary use of this robot is educational. Given that it used to do quite well, I now use it to gauge market conditions, and so use it as a reference bot.

--

So that's it! Two robots up slightly, the other three down.

The breakout robots I had so much hope in have let me down, badly. Many of the other robots I am demoing at the moment are breakout based, and I don't think any of them have returned a profit in quite a while.

The market changes and goes through various phases, and this is obviously not a breakout system phase. It seemed like it was for a while, because it seemed I couldn't lose with a breakout system a few months ago.

I need a system which works regardless of market phase, and breakouts isn't it.

I have some luck with reversals with large buffer zones on the higher time frame charts, but will need to keep investigating and developing these ideas.

Robot Results April 16th - 20th

They're a little late this week, but here they are - robot results from the 5 we've been regularly reviewing.

Inside Bar

Last week it was up $362.  This week down $275.

Pending Reversal Candle

Last week, this robot surprised and was up $384. This week down $366, but one order is still open, currently up $133.

Daily Close Breakout

Another surprising robot - last week up $323. This week pulls in another $289. Amazing. I ran it through the back test optimizer, but the results weren't great. I think there is something wrong with the history data. One of these days, I'll get around to trying to fix it and re-optimize. 

Weekly Breakout 2

Last week down $354. This week down another $467. This system was described in a Forex Factory thread, and seemed promising. But after several weeks of losses, I'm starting to have my doubts. While it is easy enough to see where the entry's should be, it's not always easy to get them, and the robot doesn't always enter when it should either. It's frustrating.

Also, sometimes when I trade other people's systems, I get caught up in their hype for it, and continue it well past the point where I should give up.

With this system, I don't know what to think. I can see potential in it, but I just can't seem to make it work. When other systems, such as the inside bar, seem to work better, I sort of think I should give up on this one.

We'll try it a few more weeks, and see what happens.

Open Close Bot H4

Last week down $413. This week the demo account expired and I had to reset it, but this old robot posts another loss of $464. Starting to forget that robot ever succeeded at all in the past. 

--

So there we have it. Another pessimistic week. Only one robot which returned positive results, the "Daily Close Breakout" bot. This one is dubious at best.

However, it along with some other robots not shown here gave me an idea - time vs distance (momentum) is important in these breakout bots. If the price can move a certain amount in a given time, this can be used to judge the momentum and possible continuance of a move. Thus, by adjusting the buffer zone around the consolidation patterns which the breakout robots use to judge when to place trades, and the order expiry time, we might have a bit more success.

EURUSD - 2012-04-25

Trading has been difficult lately. Probably because I'm learning and don't know what I'm doing. It's nice to think it's just because the market is being a pain, but then again, the market will always be doing strange stuff, so I just have to deal with it and hopefully learn something.

I've taken to watching 5 min charts for long periods of time, and then missing any moves or being too scared to trade when I think I spot something interesting. Frustrating.

I've nearly given up on my breakout strategies. They work ... sometimes. It's easy to see where the breakouts have occurred after the fact. It's even easy to see them as they occur. But getting stop loss and take profit levels right is damn near impossible. When you can see a trade happening right in front of you and still can't get it right, then what do you do?

Anyway, pessimism over for a few minutes, let's look at a few charts. While no excuse for failing trading, they may explain a little some of the troubles lately. 

Daily EURUSD chart

The EURUSD is getting itself into a triangle wedge shape. It refuses to go below 1.30, but seems unable to break higher, either. No movement means no trading opportunities. At least, not many good ones that I'm able to capture.

Daily EURUSD chart with Ichimoku Kinko Hyo

On the Ichimoku Kinko Hyo chart, we see some hot cloud based action. Price is almost poking it's nose out above the top of the cloud, but not quite.

Other currency pairs aren't giving much clue as to direction either. GBPUSD has been rallying, AUDUSD diving, and USDCAD consolidating. Bah.

One good thing out of all of this is that I've been learning about range and swing trading. I'll keep practicing. Perhaps I'll have a good strategy at some stage.

EURUSD - 2012-04-14

A difficult week to trade EURUSD, but there were a few good moves - we just had to be patient and watchful to catch them.

This week the EURUSD sort of  moved choppily upwards, and ended the week with a big move down to end up pretty much where it started. Behold the weekly candle:

(Weekly EURUSD chart)

The long upwards wick on this candle and close below last weeks closes suggests continued bearish movement.

So I may not be that great at predicting price movement from charts, but it is a lot of fun to review them later
and see where price fits into the various technical indicators.

(Daily EURUSD chart, 100 & 200 period moving averages set to CLOSE)

See how last weeks price action (the last 5 bars of this chart) flirted with the 100 period simple moving average?Amazing!

(Daily EURUSD chart, 100 period moving averages set to HIGH & LOW)

The HIGH and LOW of the 100 period simple moving averages are similarly interesting, and I'm impressed with how often price flirts with these lines on various time frames. Here we can last weeks price action was contained between both of these lines! The move up Thursday reversed from the 100 HIGH and move down Friday stopped at the 100 LOW. Astounding!

(Daily EURUSD chart, Ichimoku Kinko Hyo)

Ichimoku Kinko Hyo is fast becoming one of my favorite charts to look at because price seems to respect the various levels presented nicely. Here we can see price hit the blue line and bounced down, but not quite able to exit the "cloud".


Ok, so all of this is fascinating, but it doesn't tell us where the price may go next. Many technical indicators are pointing towards a bearish move (moving averages, weekly trend line, etc). However several other currency pairs have been trying to break higher, and if they succeed, may drag the EURUSD higher as well.

Also, while technically it looks like EURUSD should move lower, if we expect the unexpected, it will move higher. Gah!

Let's look at another chart which might help explain EURUSD indecisiveness:

(Daily EURUSD chart, 100 period Bollinger Bands, 1 & 2 standard deviations plotted)

If you excuse my poorly drawn yellow trend line, we can see a triangle forming in conjunction with the strong support level of 1.30 which price has so far 3 times failed to break through. Also, price is consolidating between the 1 standard deviation lines of the 100 period Bollinger Bands. 

So EURUSD price has been erratic lately, but generally it is moving lower, but stalling above 1.30. It's not sure where to go, but doesn't want to move higher, but also doesn't want to break lower. 

It's winding itself up for a big move. I'm guessing down, but we'll have to see if the 1.30 level holds, and also see what the other currency pairs are up to. 

Robot Results April 9th - 14th

Half way through April already. Time flies when you're testing robots. Let's see how the 5 demo test robots did this week.


Inside Bar

On Wednesday I ran this through the back test optimizer, and adjusted the settings for EURUSD and AUDUSD. After a few losses, it caught a big move on AUDUSD and ended the week up $362. Last week it lost $289. Balance is $4898.


Pending Reversal Candle

Last week made $185, this week $384. Interestingly, no losses. Balance is $5562, which makes this one of the more successful robots I'm running - so far. I ran this robot through the back test optimizer too, so we'll see how it performs next week. The adjusted settings increased the stop loss, so if it loses, it'll lose big.


Daily Close Breakout

Last week made $243, this week up another $326 with one order open for +$67. This robot is surprising. It did so poorly on the H1 timeframe, but on H4 has had two good weeks. No losses this week either, despite a choppy market. Balance is $4626.


Weekly Breakout 2

Last week I left this robot alone completely, and it made $906 profit. This week it lost badly in the beginning of the week, as the market chopped around. I had almost lost faith in it (again). Then it captured a couple of nice moves, recouping a little, but not enough to make profit. It's down $354, with one order open for +$97. Balance is $5184.

I'm impressed that it does recoup some lost profit when the market decides to make a move.


Open-Close Bot H4

Lastly, my trend detecting benchmark robot. This poor thing hasn't had a winning week in quite a while. When I first started testing it, it did really well, with several straight weeks of profit and was up a few thousand. So I was surprised when it started failing, and for so many weeks in a row. But now, with the exception of 2 good weeks about a month ago, this thing has lost week after week. So it just goes to show that sometimes your positive results are more from luck and timing than because you have a good robot.

This thing needs several H4 candles to go in the same direction, and this hasn't been happening much lately.

Down $184 last week, it loses another $413 this week, balance is $3659. Note it scales it's lot size up and down depending upon the account equity, so to lose more money this week than last is really something, considering it was using smaller lot sizes.

I may run this through the optimizer just for fun next week, but I suspect it's strategy is just doomed.

--

So there we have it. A fairly directionless market this last week, yet 3 out of the 5 robots posting gains. I am surprised.

EURUSD - 2012-04-13

After a steep drop last week, the EURUSD has barely budged this week. Some other currencies rallied a little Thursday, which may have helped pull EURUSD up slightly last night. 

(60 minute EURUSD chart)

In the chart above, the dark orange line shows the weekly open, with the light orange showing approximate weekly open. We can see that price has barely moved away from this weekly line.

Other currencies have rallied, and several such as GBPUSD and AUDUSD show some possible good reasons to do so. We won't discuss them here, we'll just keep looking at EURUSD and see if it perhaps could keep going up.

(Daily EURUSD chart with 100 period moving averages plotted)

Generally the EURUSD still shows a bearish trend on the weekly chart (not shown). On this daily chart shown above, we can see price is stuck between the two 100 period moving averages (set to the HIGH and LOW prices of each candle respectively). These lines act as barriers the price needs to push through to get some direction going. Momentum has been sort of worn down through several attempts at breaking away from these lines. Direction still appears to be bearish, but price looks tired!

(Daily EURUSD chart with Ichimoku Kinko Hyo plotted)

Lastly, the daily Ichimoku Kinko Hyo chart, which I'm becoming more fond of, shows the price stalled at the lower cloud level, rested there, and now it's rally has stalled beneath the red and blue lines. Plus the top of the cloud is looming above that. This suggests strong resistance to overcome to the top side.

However the stubborn 1.30 level is not far below either. And with see-sawing news from Europe and the USA, we'll probably continue to see the price do not much for a while.

Although technically things look set up for a continued move lower, if other currency pairs continue to rise, in particular GBPUSD and AUDUSD, I'd not be surprised to see a break higher.

Crap History Data

Here's an example of a problem with back-test history data.

Here are two sets of back-test results from the danw2 - 34 ema wave + cross + pending robot. The idea of this robot is to wait for a candle to cross several EMA's, in this case 3 x 34 period EMA's, set to HIGH, CLOSE, and LOW, and then set a pending stop order above or below those prices.

In back-tests on my usual machine, I get these results:

Back-tests on 60 minute EURUSD chart, 1-1-2011 to 1-1-2012

61% win rate, $2334 profit over a year. Apart from the equity curve flattening out around the $12k mark halfway, a pretty good result.

From these results, I expected the demo tests to return some positive results, even if, like most robots, it will fail over all. But on a demo account, it's barely won anything - all I see is loss after loss, after loss.

I was curious, so downloaded the history data for EURUSD on another computer running another copy of MetaTrader 4, and re-ran the back-tests. This is the result I got:

Back-tests on 60 minute EURUSD chart, 1-1-2011 to 1-1-2012

Vastly different. A mere 30% win rate, and $1506 loss. 

Why so different? The settings are the same. I have downloaded history data in the same fashion in both instances of MetaTrader 4. I have deleted the history data from the folders and re-downloaded several times, yet I keep having the same problem. 

The clue may be in the "mis-matched chart errors". The first result shows a lot. The second, not as many. 

Searching the web for this problem, I'm told to go to the History Center in MetaTrader 4, select EURUSD, highlight the 1 minute bars in the history data, and then click download. 

I don't know why this is different that simply selecting EURUSD and clicking download as I usually do, but it did seem to fix the problem. I'll just have to keep an eye on the mis-matched chart errors from now on. 

Robot Results April 2nd - 7th

A better week for the 5 robots we're demo testing.

Inside Bar

Lost $133 last week. This week down another $289. It's kind of funny, since this robot has done so well when the others were performing poorly.

I suspect I need to revise the stop-loss and take-profit settings for this robot, perhaps even make them ATR based. If I get around to it, I shall run some back-tests with optimization. Still have a reasonable amount of hope for this robot to perform somewhat adequately.

Balance is $4535.




Pending Reversal Candle

This robot seems to make money in the weeks when the Inside Bar loses. It's almost comical. After several losing weeks, I was about to give up hope, but last week it made $166. This week it's up another $185. I'm reasonably satisfied with the stop-loss and take-profit settings for this robot, but may run it through the back-test optimizer anyway. Balance is $5177.



Weekly Breakout 2

$465 down last week. A promising system which I kept meddling with, closing trades early and adjust stop-losses too tightly. This week I left it alone. It was an exceptional week for the Forex market, and thankfully this robot captured the breakouts perfectly, and made $906 profit in 4 trades, after losing $220 earlier in the week. Balance is $5538.


Daily Close Breakout

Down $497 last week before I switched it to the 4 hour time-frame (but then forgot to restart it until the end of the week). It performed OK, but then again this was a difficult week not to make money in as pips were raining from the sky. Closed the week up $243. Balance is $4229.


Open-Close Bot H4 

Several weeks of losses, last week giving up $563, and this week despite the strong trend and the trend following nature of this robot, still lost another $184.

This robot is not able to capture breakouts, and gets messed up in consolidation. Still, it's interesting to see how it performed. Almost all of my older robots are based on trend following.

--

So there we have it. A lot of movement in the market this week, the break-out based robots performing as expected and capturing profit, a couple with mixed results, and the trend-following system still taking a beating despite a trend developing.

EURUSD - 2012-04-07

Excitement! Action! Movement! After the FOMC meeting minutes were released at 4am Wednesday morning Queensland time, EURUSD broke the lovely channel I had drawn for it and plummet like a plummeting thing. 

(4 hour EURUSD chart)

While excited I was to discover the linear regression channel tool in Metatrader 4 and apply it to a chart, for the last two weeks EURUSD has failed to make any higher highs which should have indicated something was about to happen. Well, it did, and the move after FOMC was violent and sudden. 

Luckily I had several (mostly demo) systems in place to capture breakouts, and they all did very well this last week. 

While it's exciting to see some movement in EURUSD and capture some pips, the daily chart shows the move down may be limited already. Price has really only just broken back through the daily 100 period simple moving average. Also, from the above 4 hour chart, the horizontal yellow line indicates the 1.30 support area which may be tough to break through. 

(Daily EURUSD chart with 100 & 200 simple moving averages plotted)

Also, price action has penetrated the cloud on the Ichimoku Kinko Hyo chart, but not escaped it. The bottom of the cloud may act as support.

(Daily EURUSD chart with Ichimoku Kinko Hyo)

We'll see what happens next week after the Easter holidays. 

EURUSD - 2012-04-03

9.54pm on a Tuesday night here in Australia, and EURUSD starts making a move downwards. Exciting. It's tempting to sell, but wait, what have we here?

(60 min EURUSD chart with 100 & 200 period simple moving averages and channel)

We have the 200 period simple moving average on the hourly chart, as well as the bottom line of a nicely fitting linear regression channel. The bottom line is a supportive trend line. 

EURUSD failed to make any new highs yet today, so it's possible it will keep dropping. Despite this, however, recent behavior indicates this move is only going to be small. 

EURUSD - 2012-04-02

The last week of March saw the EURUSD make a big leap on Monday, then stagger around in a seemly drunkenly fashion for the rest of the week, stalling slightly below the high it made on Monday. 

I was confounded and confused until I realized I could fit a nice bullish trend-line on my chart, and EURUSD had bounced off it on Thursday. 

(60 minute EURUSD chart with resistance line in blue and trend line)

So much for the bearish interpretation of various indicators I made a couple of weeks ago, this thing is still heading north, assuming it can push through the blue resistance line. 

Assuming it is still going up, where is it going? A while ago I theorized that EURUSD might be retracting from a big bearish move from Oct 2011 to Jan 2012. 

 (Daily EURUSD chart with fibonaci retracement lines)

Price hit the 50% retracement of that move, stalled, moved back to the 23.% line, and looks to be trying to move up again. But it's currently stalled just below the 50.0%.

So, without spending more hours trying fibonacci lines all over the place trying to fit, I have no idea what's going on, at least on the daily time frame.

Yesterday when playing around with the Metatrader 4 tools, I found that the linear regression channel tool snaps to various highs and lows. I didn't know that before - I assumed it was similar to the equidistant channel tool.

Anyway, this is very useful, as it formed a pretty channel with very little effort:

(60 minute EURUSD chart with resistance line in blue and linear regression channel)

Here we can see the latest short term bullish EURUSD move, fitting beautifully into this channel. 

Since it took me so long to find, this pattern may be almost finished. But that's OK, as now I can look for channel breaks, selling below the lower channel line or buying above the upper channel line. We'll see what happens.