Sunday, January 27, 2013

Week in Review: 20th January - 26th January 2013

This is a late entry, I know. But I've been suffering from a cold/flu for the last three days and it's really knocked me down.

I put in three trades this week, won one and lost two so I'm down for the week.



I saw an entry signal using my Hermes system on Monday, and used a 1.25:1 reward:risk ratio. Banked some profit on Tuesday as the AUDUSD moved north.

It's a similar story with the EURUSD. An entry signal appeared on Monday, and I put in a pending long and a pending short on the break of Monday's high and low. I set my reward:risk to 1.25:1.
Both the long and short were triggered, but in each instance price reversed and hit my stop loss.
I don't feel much at all about these losses, to be honest. If you thoroughly backtest your trading systems, you will know what to expect.
I completed my backtest and optimisation on pinbars using the weekly timeframe. I tested 11 pairs from 2001 to mid-2012 and gathered a sample size of 352 trades. It's not alot. My overall profit factor was 1.51 using a reward:risk of 0.25:1. I'll post more details in a future entry but it shouldn't take anyone long to backtest pinbars on the weekly timeframe if they don't want to wait. I'm coughing my lungs out and I want to rest for the Monday open!
I saw two interesting entry signals on the weekly chart using my Daikoku system. The signals are on the USDCHF and GBPUSD pairs.
I didn't see anything interesting on the daily charts.
I'll post more details on my weekly pinbar system, plus some thoughts in calculating the profit factor for a system across multiple pairs. Tip: your true profit factor is NOT the average profit factor across multiple pairs.

Saturday, January 19, 2013

Week in Review: 13th January - 19th January 2013

I suffered a hit to my account this week, as I'll explain.



This trade didn't go well. A low volatility candle appeared on Friday's close. My pending long was triggered on Monday before price reversed and hit my SL on Tuesday.

My position size was 3% in this particular trade, which is high, and is responsible for my account hit.
XAGUSD (silver)
This trade resulted in a win. There was a low volatility candle on the weekly TF and I took the trade using my Daikoku system.

This was a good trade. I saw low volatility on Wednesday and created pending orders on Thursday, resulting in a same-day win.

Even though I won two out of three trades, because of my R:R ratios, I still ended the week down. Plus, the trade that lost had a position size of 3% while my winners were 1%. Why the different position sizes? I felt confident with my first trade, but after the loss, I felt I needed to further backtest the Daikoku system on the daily timeframe and so reduced my trade on the USDCHF to 1%. That's trader psychology for you.
My trade on XAGUSD was only 1% because I find it difficult to accurately calculate my position size ($ per pip) on the precious metals. If anyone can point an online calculator my way, I'd appreciate it! Otherwise I tend to be conservative and use a low number of microlots.
Daikoku Conversion and Backtesting on Daily Timeframe
I only performed a little backtesting this week, using the EURUSD. I've reduced the daily range of a valid signal to 50% or less of yesterday's range. This erased a quarter of all entries that I've already gathered on the AUDUSD and USDJPY. Taking this into account, my total number of sample trades is 800.
The profit factor at the optimal R:R ratio is currently 1.29. It's tradable.
Pinbars on Weekly Timeframe
This has been the major focus of my efforts this week. I've been designing a new system on the weekly charts, using pinbars as entry signals. The backtest is nearly 40% complete. I've tested four currency pairs and have seven more to go. I hope to get one or two more pairs tested tonight. If the backtest is successful, I'll post some details next week, but so far the results are nice.

Saturday, January 12, 2013

Week in Review: 6th January 2013 - 12th January 2013

It's been a good week. I opened two trades and made some profit.



I've been working on a daily timeframe version of my Daikoku low volatility system, and found an entry signal on the AUDUSD after Monday. I used a 0.4:1 reward-to-risk ratio and won the trade.

Price came within a pip of hitting my SL before reversing. I seriously thought about closing my trade once I was halfway towards my stop loss, but I held my nerve.

It's a similar story with the NZDUSD, although price moved in a straightforward manner and hit my TP quickly.

I didn't like the high correlation between the AUDUSD and NZDUSD. Even though both entry signals occured on different days (AUDUSD on Monday, NZDUSD on Tuesday), both trades ended up staying open and trading in the same direction at the same time. But I had to trade my system according to its rules.
This has become a tedious process. I've manually backtested the USDJPY on the daily timeframe with success, although the system experience significant drawdown between 2005-2009.
So far I've backtested the AUDUSD and USDJPY. The profit factor is around 1.3 with little optimisation. It's all right. The frequency of trades is insanely high though, in my opinion. From these two pairs alone, there are over 800 trades between 2001 and mid-2012. That's about one trade per week from two currency pairs only, and I intend to focus on eleven (all the majors and Yen crosses). Very nice indeed.
With two pairs tested, I "only" have nine more pairs to go through. At this rate, I may finish by March!
I've got three entry signals that look promising.
Two signals are on the daily timeframe: the AUDJPY and EURUSD. I especially like the AUDJPY. Correlation between the two pairs is relatively low.
On the weekly timeframe, we have XAUUSD (silver), which whipsawed the previous week:

Saturday, January 5, 2013

Week in Review: 30th December 2012 - 5th January 2013

I'll try starting a new format for my blog. In the past, I'd haphazardly update my blog. While this lent to some creativity, I'd lose discipline and not bother blogging about the more "mundane" topics. I will like to update my blog on a weekly basis during each weekend, kinda like a news column, summarising events and ideas for the week.


I opened two trades this week, using my Daikoku low volatility system on the weekly timechart. Both trades closed in profit.


I'm beginning to like the NZDUSD. It has proven to be consistently profitable, whether I trade it live or during backtesting. I used a 0.25:1 reward:risk as my backtest optimisation suggested.
XAUUSD (gold)

This trade closed in profit as well when my pending long was triggered. However, I had another chance to profit with my pending short when the market reversed late in Friday. Before that happened, though, I decided to cancel the pending short since I was unhappy about my position size and left it, thinking the market had little chance to fall during the last hours of Friday. How wrong was I? Lesson: don't be lazy, keep orders open according to your system.
I was reading Bloomberg and came across this startling statistic re the performance of hedge funds:

Three of the top five funds in the Bloomberg Markets list invested in mortgage securities, and two of them are run by Minnetonka, Minnesota-based Pine River Capital Management LP. Betting on mortgage securities outpaced every other strategy, with an average return of 20.2 percent, against an industry average of just 1.3 percent, according to data compiled by Bloomberg.

I used to think that the world of hedge funds was inhabited by sharp, ruthless, analytical businessmen and women with a keen eye for profit. But an industry average return of 1.3% kinda shatters that image. That's about the same as flipping a coin. I'm not an expert in hedge funds but I'm beginning to suspect that it's an easier industry to get into than I first thought.
As a retail trader, my projected annual return will be around 10-15%, based on my performance for the last four months. This bit of news has given me confidence since I'm currently doing better than most of the pros.
I'm currently formalising my Daikoku trading system. I probably won't release the full details of this system on my blog, but for any readers, you should have enough info from my last post to perform your own backtest and see whether the system (or something similar) suits you. It's been my observation that market volatlity predictably gyrates, particularly on the longer timeframes, and you want to get in just as volatility starts to increase.
I've also started backtesting on the daily timeframe with very promising results. My frequency of trades is more than double of my Hermes low volatlity trading system(which uses Average True Range (ATR) to check for volatility). Because the potential number of trades is so high, I think my backtest may take up to a month to complete.