Thursday, November 28, 2013

AUDJPY: System development on overnight carry

I didn't post my fundamental analysis on the AUDJPY as I've been busy developing and backtesting a new mechanical system. The basic backtest is done. I'm now left with some optimisation, and perhaps exploring variants of this system.

The AUDJPY is traditionally popular for carry trading, due to the interest rate differential between the Reserve Bank of Australia and the Bank of Japan. Basically you borrow JPY at a lower interest rate to buy AUD to lend at a higher interest rate, and since the interest rate on AUD has traditionally been much higher than the JPY, you pocket the interest rate difference. Most currency traders use leverage. Now suppose your leverage is 100x, a 4% interest rate differential with 100x leverage will now provide you with a 400% return on your investment. If the price of the AUD rises against the JPY, you will also profit from capital appreciation. Hence the popularity of the carry trade. 

Interest is typically collected each day at the close of the NY session. This system has you buying the AUDJPY just seconds before the NY close during a bullish trend on the daily chart. Not only will you collect the overnight carry, but bullish momentum should see you profit from a higher AUD in the next day or two.

To identify a bullish trend, all I look for is a bullish candle that appears above the 21 EMA without touching it, and the close of the candle must be within the top 50% of the candle. Here are some examples over the last few months (the dotted line is the 21 EMA. Notice no contact between the candles and the 21 EMA. If there is contact, then I assume the market is ranging and you stay out).

For purposes of the backtest, my entry price is the closing price of the candle, since we are entering mere seconds before the NY close. My stop loss is at the bottom of the candle. My sample size was 826 trades, from late 2000 to November 2013. I used the broker Alpari for data, which generally aligns with the New York close.

According to my backtest, a 0.6:1 reward:risk ratio seems best. At 2% risk and a $10,000 starting balance, we get the following equity curve.

We get a very nice R^2.

For robustness sake, this is the equity curve for 0.5:1 reward:risk.

Here is the equity curve for 0.75:1 reward:risk.

So we can see that minor changes to the R:R ratio don't break the system, which is a good indicator of robustness. Also note that these equity curves only graph the profit and loss from the trades themselves. It doesn't include the positive carry that we collect from each trade. If we include the positive carry, our equity curves will look even more healthy, but I haven't figured a way to calculate historical carry rates accurately. Anyway, it looks as if the system itself will generate profit without the need for carry.

Now having said all that, I'm not a big believer in pure mechanical trading anymore. My current philosophy is:

a) generate mechanical systems that can make money if simply traded "as is" 
b) analyse market sentiment and fundamentals
c) if your analysis lines up with the logic of your mechanical system, open a trade

So with this system, I simply wouldn't be opening longs just for the heck of it. If I feel there are good reasons for the AUD to rise, then this mechanical system should give you a good return.

Anyway, this is all preliminary. I would like to do some further optimisation and tinkering on this system so it doesn't have my seal of approval yet. I'm just posting this to hopefully generate some ideas from you guys. :)

Friday, November 22, 2013

AUD: Bulls routed

There's still a few hours left before Friday's close, but it looks like the bulls on the Aussie dollar have routed. A few news release this week looked to have negatively impacted the AUD. This includes RBA Governor Glenn Steven's speech on Thursday, who praised the AUD as a floating currency, but reiterated the RBA's right to intervene (interpreted by the market as bearish). China's flash manufacturing PMI also came out below forecast (50.4 vs 50.9). Anyone trading the AUD as a proxy for China would've been hurt. So overall, this week's fundamentals have hurt the AUD.

Technically, the AUDJPY is back at a support level around 92.500. I think this would make a good place for bulls to buy the AUDJPY at a cheap price (or for bears to position themselves for a breakout on the downside). There's a few more weeks until the RBA's next interest rate decision. However, the ascending channel that I identified earlier is very close to being broken. The high of the channel should also have been touched already (or at least, price should be close to it). Instead we are at the bottom. 

I'll post a more-detailed analysis tomorrow.

Wednesday, November 20, 2013

AUD: RBA Minutes from Nov 5

The AUD rallied yesterday after the release of the RBA's meeting minutes. I don't think the market expected much change from the RBA's statement back in Nov 5. Some relevant highlights:

+ growth of major trading partners remain around long-term average, and expect to pick up slightly in 2014
+ recent Australian growth at below-trend
+ economic headwinds from fall in mining investment, high AUD, weak public demand 
+ housing prices are rising 
+ future business investment predicted due to low level interest rates
+ next year growth expected to be below-trend but will pick up
+ RBA closely monitoring house prices
+ remains concerned about high AUD
+ lower AUD needed for balanced growth
+ no change to interest rates, RBA would like to gauge effects before making any future changes

My summary is that the RBA would like to reduce interest rates further, but is constrained by a potential housing bubble. Seems like the market interpreted this as no immediate threat of a rate cut.   

BoJ will be releasing its monetary policy decision tomorrow. I don't expect any changes. The next RBA announcement is on December 3, a few weeks away. I believe the bulls will have another week to accumulate and drive up the AUD before taking some profit. 

Saturday, November 16, 2013

AUDJPY Forecast for 18th - 22nd November 2013

The AUDJPY seems to be obeying the ascending channel that I drew last week.

I opened a long position at the start of last week at 92.800, with my profit target at 95.400.Unfortunately, I prematurely took profit a few days ago so I'm out of this particular swing trade already. This was purely psychological. When I saw price retreating back to breakeven on Thursday, I decided to close and content myself with a tiny, nominal win. This was my ego sabotaging my trading. I'll continue to "paper trade" this particular price movement since I'm eager to see if I can successfully combine my fundamental and technical analysis.

WEEKLY FORECAST FOR 18th - 22nd November

News / Fundamentals


Last week's major headlines for the USD were bearish, with the US trade gap widening to $41.8b versus a forecast of $38.7b, and unemployment claims jumping to 339k vs forecast of 331k. The US also posted bearish industrial numbers last Thursday and Friday. Future Federal Reserve boss Janet Yellen has also indicated the continuance of QE yesterday. 

These events should contribute to bullish pressure on the AUD, especially the AUDUSD. However, with the price of USD being driven down and Yellen's dovish tone on QE and tapering, the USD retains its attractiveness as an alternative carry trade currency to the JPY. I believe any influence on the AUDJPY will be subdued.


No major headlines that I can read. 


News releases on the AUD has been mixed. Consumer sentiment and home loans are up for the month, but business confidence has slid. I don't think these events will have lasting impact come next week.


Japanese quarter-to-quarter GDP growth came out better than expected, 0.5% versus a forecast of 0.4%. However, last month's current account deficit came in at -0.13T yen vs a forecast of -0.1T yen. Other industrial indicators are also down, so overall the news is mixed. News should provide little bias in either direction next week, in my opinion. The central bank will be releasing a statement on Nov 21, but I doubt that many traders will be sitting out before then - if there are any surprises, I think it'll be further monetary loosening than anticipated. So with regards to the AUDJPY, I think the JPY will provide neutral-to-bullish pressure. 

How this fits together

News verdict: I don't expect last week's events to influence the AUDJPY this upcoming week. Neutral. I expect technicals to play out their course.


OANDA's COT report indicates that shorts on both the AUD and JPY grew last week. JPY shorts are near their highest in the last 12 months, showing extreme bearishness. AUD bearishness is not so strong.



Sentiment verdict: moderately bullish

Technical Analysis

As illustrated at the start of this entry, it looks as if the AUDJPY has established an ascending channel after respecting support at 92.500. A pinbar formed off support, lending some technical strength to a bullish outlook. The range of Friday's candle was also large, relative to the five previous candles, suggesting a lack of profit taking from the bulls. All these signs point to the AUDJPY moving up next week.

Two major events are taking place next week: release of the RBA's minutes in its last meeting on Tuesday, and BoJ releasing a monetary policy statement on Thursday. I don't think there will be any surprises, so I think traders will stay in the market before and during these events. The RBA won't be making an interest rate decision until early December so there is at least another week for bulls to drive up the AUDJPY without RBA interference. 

Overall, I expect the AUDJPY to climb, but it will most likely encounter heavy resistance at 95.600 and beyond and turn, should it arrive at these levels next week. It will take some news to push through. 

Tuesday, November 12, 2013

AUD: NAB Business Survey shows mixed signals

The NAB Business Survey came out a few hours ago. The news is mixed. The main change is a fall in business confidence, from 12 in September to 5 in October. This is still a positive number, so businesses are still optimistic, but perhaps the post-election euphoria has died. Business conditions remain mostly unchanged, although still negative. However, all numbers are an improvement / unchanged from August.

Effect on the AUD: given the mixed signals in the survey, I don't expect any major movement on the AUD. The AUDJPY fell 20 pips on the news release, but quickly recovered in less than an hour, suggesting institutional buyers moving in on the dip, a bullish sign.

Monday, November 11, 2013

AUDJPY: 11th November 2013

I lost around 13.5 pips on the AUDJPY today. The market was very choppy. There wasn't much news today, and being a Monday, I expected the market to be a little subdued, but not this much. I won two trades, but lost four. Here's a snapshot of the market I experienced today.

The AUDJPY remained in a 5-15 pip range for most of the Asian and European session. My losses mainly stemmed from trying to trade upside breakouts from this range, especially after some good news on the AUD came out today (new month-to-month home loans increased by 4.4% vs forecast of 3.6%, suggesting increasing confidence among Australian borrowers). 

After four losses, I felt emotionally unbalanced and halted trading. 13 pips isn't alot, but it's the whole concept of being intellectually wrong that tilted me. 

I still think there's good upside potential on the daily chart. Volatility has frozen after a steep fall, and the ascending channel is still in place. It looks as if the 92.000 - 93.000 is behaving as a significant support / resistance level in the last few months.

Two news releases are coming out tomorrow: Japanese M/M Tertiary Industry Activity, and the National Australia Bank's Business Confidence survey. It's the NAB business survey that I'm most interested in. The Reserve Bank has been busy talking the AUD down for the last week. If good news comes out, this should hopefully spur some upside movement and shake out the range over the last two days.  

Sunday, November 10, 2013

AUDJPY Forecast for 11th - 15th November 2013

I discovered a neat COT / sentiment chart at  TimingCharts. If you're looking for free COT / market sentiment information, this is a useful resource.

WEEKLY FORECAST FOR 11th - 15th November

News / Fundamentals


US NFP beat expectations last Friday, so short-term USD momentum should be bullish. This increases the probability of the Fed tapering in the next few months, especially if good news continue. The debt ceiling deadlock between Obama and the Republicans hasn't been resolved though, but I think the market will ignore this until January 2014. 

If tapering is on the cards, what does this mean for the AUDJPY? With USD and EUR interest rates now very low, investors have been using them as alternatives to the JPY for carry trades. Carry traders may start switching their borrowings from the USD to JPY, so this is good news for JPY-paired carry trades. Specifically, this should benefit the AUDJPY and NZDJPY. If the US is improving, this should increase appetite for risk, which is also beneficial for carry trades. 


The major EUR headline from last week was the ECB's reduction of interest rates to 0.25%. This is on par with the Federal Reserve's range of 0.0 - 0.25%. This should benefit EUR-based carry trades, especially the EURAUD and EURNZD, but harm competing carry pairs like the AUDJPY and NZDJPY.  

However, the stimulus should be beneficial for the world economy and provide a boost for commodity prices, which especially benefits Australia. I think this is how the market saw it last week.


The RBA chose to maintain its current interest rate at 2.5% last week, although the RBA has indicated its willingness to reduce rates if needed. The next meeting is scheduled on December 3. This means that bulls should have a window for the next few weeks to drive the AUD up without RBA intervention. Bears will be staying out. 


I haven't observed any important JPY events for the last week.

How this fits together

News verdict: moderately bullish


OANDA's COT report indicate that the large traders are more bearish on the JPY than the AUD. 

Sentiment verdict: moderately bullish

Technical Analysis

Volatility on Friday fell significantly, as we probably saw some short profit-taking before the weekend. This means many of the bears on Thursday are now out.

In light of bullish news and sentiment analysis, we may see an ascending channel forming. If this is the case, I would look for longs on Monday and possibly ride it for the next week or two.  92.500 would make a good support level since it's near Thursday's and Friday's lows and would make the bottom of the ascending channel that I've graphed below. I'd place my SL just below 92.000. TP would be towards the top 25% of the channel. 

Intra-day, I'd be looking for longs on Monday.

That wraps up my plan for next week. I may be completely wrong as this is the first time that I've used all three forms of analysis (fundamental, sentiment and technical).

Friday, November 8, 2013

AUDJPY - 8th November 2013 - Day trade #2

Lost my second trade for today. My forecast isn't playing out so I'm out for the day.

This trade was cheap, with a stop loss of only 2.4 pips. I filled my short at a good price, 92.891. This took place at a good resistance area, where we have confluence with a short-term round number (92.900) and the 50% fib retracement level.

My reward was around 7R, so this was a long-shot trade.

Net pips lost = -2.4 - 0.55 = -2.95 pips

AUDJPY - 8th November 2013 - Day trade #1

This trade failed. Went short at 92.777 but got taken out at 92.825.

I went short when I saw a high-volume, low-volatility candle appaer off 92.800. The AUDJPY generally likes to behave in 10 pip moves, so I thought this was a good area to go short after the bounce from 92.700. 

Net pips lost = -4.8 pips - 0.55 pips commission = -5.35 pips

AUDJPY - 8th November Intra-day Forecast

Bearish. The last swing low on the daily chart got taken out yesterday, and it looks like a lower high fractal is forming today, so I'm looking for shorts.

RBA Statement - November 2013

+ growth with Asian trade partners remain average
+ little change in forecast for next year
+ Australian growth remains slightly below average, but good signs are emerging (rising consumer sentiment & sales)
+ house prices now above 2010 peak, optimism in housing market
+ labour remains soft, unemployment expected to rise next year
+ inflation at or below target
+ expect mining investment to decline

RBA policy:
+ desires lower AUD
+ content with monetary stimulus already in place
+ open to cutting rates if needed while meeting inflation target (current inflation = 2.2%, target range = 2-3%)

My thoughts

Housing bubble is emerging, which the RBA may be wary of pumping further. Mixed signals and rising consumer sentiment suggest no drastic need to cut rates. Inflation is in the lower target range, so the RBA still has some bullets left. RBA will try to talk down the AUD and stir uncertainty regarding interest rate cuts. 

RBA probably won't cut rates until the end of the year, if at all.

Thursday, November 7, 2013

AUDJPY - 7th November 2013 - Day trade #1

I only took one trade today.

I took a look on the daily chart and felt that the AUDJPY would behave bullish for the next week or so, based on technicals.

This meant that I'd be looking for long opportunities today.

Earlier today, weakish Australian unemployment figures came up, with the unemployment rate nudging up 0.1% and 30k full-time jobs lost. While this was bad news, I didn't feel it was that bad, and could be a statistical blip. I felt the plunge had been overplayed, and buyers would be looking to get in at some point, which would be around 93.400. At this point, price had also moved beyond 50% of the daily ATR(14), suggesting there isn't much downside left for the day.

Once the range formed around 93.400, I looked for tell-tale signs of an upturn, mainly high-volume, low-volatility candles. I timed it a little too early, though, as you can see. Price dove a bit before finally breaking out on the upside. I took profit just below 93.500.

Net pips won = 6.1 pips - 0.55 pips = 5.55 pips

I was aiming for 93.600 but when price stalled at 93.500 for a couple of candles, I decided to close here. I got the gut feeling that the market will range for most of today. There isn't any strong reason to rally after the earlier plunge.

I got work so that's all my trading for today.


I just took a look at my chart after finishing work and noticed that the AUDJPY shot up and completely recovered from the AUD-related unemployment plunge that took place earlier today.

The rallies were triggered by non-AUDJPY news events. The European Central Bank reduced rates to 0.25%, and US jobless claims for this week came well within expectations. How did these events influence the AUDJPY?

The AUDJPY reacted positively to the EUR interest rate drop. This meant good news for Australia, and bad news for Japan. Why? Some thoughts:

- yield on the EUR fell lower, making the AUD relatively more attractive for yield seekers (the AUD still retains one of the highest interest rates amongst developed countries. With the 2nd most traded currency now near 0 percent, the AUD shines even brighter).
- the Eurozone is a manufacturing zone. Stimulus should help commodity prices, thus help the AUD. In contrast, it doesn't help the JPY, which competes against Europe. 

Now, the AUDJPY's reacton to US jobless claims is quite interesting. US jobless claims came well within expectations (336k vs 335k) so why the highly volatile reaction? I'm quite puzzled. US jobless claims did fall, but as it was within the overall forecast, this should have been priced in already. I guess this was good news, even if expected, and the USD rallied anyway. I'm guessing the AUDJPY moved due to some correlation with the USDJPY, taking out stops on the upside and triggering its own rally.

Chris Lori's AUDJPY course

I just bought Chris Lori's AUDJPY course (found here). The course comes in two components: a DVD set and an online component. The DVD set will take a few weeks to arrive, but I have access to the online material and found it interesting so far. 

This also means that I'm shifting my focus from the EURJPY to the AUDJPY. As an Australian-based trader, I think this will make a better fit than the EURJPY, especially for daytrading where AUD and JPY news releases will occur during the Sydney / Tokyo sessions. Also, living in Australia provides an information advantage on the AUD side. 

Sunday, November 3, 2013

Still plugging away on EURJPY daytrading system

The EURJPY has become my obsession over the last month. My system development has moved lower from the 5M timeframe to the 1M timeframe.

Many people say that the 1M chart is just noise, but it really depends on what you're looking at. If you're trying to trade candlestick patterns like pinbars or dojis, then I'd agree. The candlestick patterns produced are probably meaningless. However, information that is "smoothed" out tend to be more clean, such as moving averages or average true ranges. 

The following is an equity curve of my "range breakout" system on the 1M chart. The thing with the 1M chart is that you can catch some REALLY big breakouts. A 20 pip reward breakout may not seem much, until it's measured against a 2 pip stop loss, which is what you're dealing with on the 1M chart. The equity curve is based on a 4:1 reward:risk ratio and 1% risk per trade. The sample size is only 100. 

These trades took place over six days in April 2011. Achieving a 50% return in such a short span seems a little too 'holy grail'.