Saturday, May 24, 2014

Week in Review: 19th May 2014 to 25th May 2014

A relatively quiet week. Opened and won a single trade this week.

TRADES

USDCAD

I spotted a low volatility candle on the close of Monday, and went long on the break of Monday's high. I took profit at 0.5R, although I think I could've played this trade better. There appeared to be a short-term resistance level that I could've taken profit just below. 




Progress of USDHKD carry trade

Price has bounced off 7.75150, about 15 pips above the HK central bank's peg at 7.75000. It could be the HK central bank buying, or it could be traders like myself moving in. 

The position is in profit and I can already sense the temptation to close the trade and take my winnings. Unless something remarkable happens with this trade, I'll probably stop mentioning it in future posts. It's main purpose is to accumulate interest payments in the background.



SYSTEM DEVELOPMENT

My sample size of my double top / double bottom entry signal is now almost 1,000. The results continue to look very positive. I hope to enter the final stage of the backtest by next week, but otherwise there's nothing else to report. I've been struck down by a bug for the last week and felt burnt out from work, trading and backtesting, so my progress has slowed down. 

Saturday, May 17, 2014

Week in Review: 12th May 2014 to 18th May 2014

Got out at breakeven this week. Lost a trade and then won a trade that cancelled the loss. 

TRADES

GBPCAD

My live call on the GBPCAD didn't work. I went long on the formation of this large pinbar, but price didn't move up that much after I entered. Got stopped out on the third day of the trade.


USDCHF

Saw a low volatility candle on Thursday morning and went long on the break of the candle's high. I didn't know how high it would go so I stuck with my Hermes trading system's rules and took profit at 0.5R. 


Price also came within 0.2 pips of triggering my short on the same day. Thankfully it didn't, as I would've lost that trade. 

MY VERY FIRST CARRY TRADE

(warning: maths heavy)

This is probably the most important trade for this week. I opened my first long-term position trade on the USDHKD. The USDHKD has been hovering very close to 7.75000, which is a support level traditionally defended by the Hong Kong central bank (the HKD is pegged to the USD, and the lower band of this peg is 7.7500). You can find more info on the USDHKD peg here or here.

The anatomy of my trade is below (click to enlarge).


My reward is about 2.25R. Since I set my stop loss below 7.75000, there's very little chance of it being taken out. From what I read, the HK central bank is more than willing to continue the peg.

As icing on the cake, long positions on the USDHKD receive positive carry, so this qualifies as a medium to long-term carry trade. My broker currently offers a positive overnight swap of HK$2.10 per HK$100,000 (or 1 lot). 


Making HK$2.10 per day doesn't seem alot. Over a year, that's HK$766.5, or a meagre 0.76% return on HK$100,000. That's a pathetic return, wouldn't you say?

If you add leverage into the equation, it changes. I'll give my own trade as an example. 

In this case, I bought 7 lots, (HK$700,000), using leverage with a part of my account as margin. I placed my stop loss so my equity at risk is only $450. With 7 lots, I'd receive HK$2.10 * 7 = HK$14.70 per day, which would convert to around $1.90 per day. 

Over a year ($1.90 * 365), I'd earn $692 in interest just from risking $450. That's a return in excess of 150%. It would take about 236 days before I earn enough interest to cover the $450 risk, after which the carry trade becomes nearly risk-free.

SYSTEM DEVELOPMENT

Some very good news in this area. I did some analysis on my double-top / double-bottom data and filtered trades by weekday. For some reason, signals on Monday performed really well. I did more backtesting and have gone through about seven currency pairs. Sample size is 642.

This is the equity curve using $10,000 initial balance, 2% risk per trade, reward-to-risk 3:1.


This is looking very good, although the drawdown in the last few years is of some concern. Will post further updates if I find anything interesting. 

Monday, May 12, 2014

Live Trade: 12th May 2014 - GBPCAD

Live trade: GBPCAD. Went long this morning after seeing a very large pinbar just above support. It hasn't quite touched support, but the sheer size of the pinbar tells me the bulls have taken charge, and made me interested enough to buy. Reward is a little over 1R.


Saturday, May 10, 2014

Week in Review: 5th May 2014 to 11th May 2014

Just capped a profitable week. I opened and won two trades this week.

TRADES

USDCHF

I spotted this large pinbar over the weekend and went short on Monday, using my Ebisu trading system. I took profit at around 0.8733, which was pretty good. I managed to catch a little over half of the downward move before it reached support. As you can see, once price found support, it bounced back right up. Always trust your system's rules and don't get greedy. My reward was a little over 0.5R, which according to my backtest, is optimal. 



SYSTEM DEVELOPMENT


There's been a promising development on this front. My Hermes low volatility system has been my bread and butter for the last year and I revisited it to see if I could expand the number of entry signals without sacrificing (too much) profitability. So far the results are looking interesting. I added an ADX(14) indicator to filter out ranging vs trending conditions and found the system performs better during ranging periods, which surprised me. But it's still early days and my research might lead to a dead end (as it usually happens).




Tuesday, May 6, 2014

What is a quant trader? - Youtube interview



A very educational interview with a quant trader. Some notes that I took:

- a return of 15%-20% pa for a retail trader is good (depending on risk)
- quant traders mainly deal in the micro-second / nano-second timeframe
- retail traders should NOT compete with high-frequency quant traders
- quant traders are more interested in exploiting the physical characteristics of the exchange itself such as the exchange's physical hardware. They are less interested in traditional fundamental or chart analysis.
- institutional quant traders are more interested in managing risk than maximising profit (mainly due to investors hating volatile returns)
- a weakness of quant trading are black swan events. Quant traders may become so data-driven that they forget the possibility of extreme events that can destroy their models.

Monday, May 5, 2014

Business plan changes over the last year

I last revised my business plan about a year ago. I spent this afternoon going over my numbers and changes in beliefs and here are the changes I discovered:

1) My risk tolerance has become much, much more conservative, probably because I realised just how badly drawdown can kick you in the guts. I initially planned position sizes ranging from 2% to 12% equity, depending on each trading system's expectancy. These days I only risk an absolute maximum of 2% equity per trade, dropping to 1% when drawdown exceeds 10%.

2) I now use an anti-Martingale money management strategy to limit my losses during drawdown. A normal Martingale strategy would be to double your bet (or position size) after a loss. Suppose you bet $5 and lose, your next bet will be $10. If you lose $10, your next bet will be $20. If you lose again, you will bet $40 etc. The idea is that eventually you'll win a bet and make up your losses. But as you can see, every time you lose, the size of your bet gets big pretty quickly. Lose enough bets in a row, and the size of your next bet will be more than the money you have left. When this happens, you are ruined.

An anti-Martingale strategy does the opposite, in that you reduce your bet or position size after a loss. So if I bet $5 and lose, my next bet will be $2.50. If I lose again, I will bet $1.25, and so on. This is what I use during a bout of drawdown (10%+). This is especially useful for traders since it will save your skin when you have a negative edge. As the old saying goes, a dollar saved is a dollar earned. The downside to an anti-Martingale strategy is that it will take longer to crawl back from your losses. If I lose $5 and start betting $2.50, I will need to win two times to make up for my loss. But you will protect yourself from ruin. 

3) Conversely, I use a Martingale money management strategy when I'm winning. My strategy is to increase my position size to 2% once I've grown my initial equity by 10%. I'm still deciding if or when I should increase my position to 3%, perhaps once I've grown equity by 25%. But the logic is that if you're winning and have a good read on the market (or if your trading systems are performing optimally), you should jam the pot and bet more while these conditions last. 

4) I had a very optimistic goal of becoming a full-time trader by 2015 (next year). This is now unrealistic, especially given my reduced risk tolerance. 2017 might be a more-accurate possibility. 

5) My original plan envisioned trading 11 currency pairs. I trade alot more pairs now, including gold and silver and more obscure pairs like the USDHKD or USDSGD. Perhaps next year I'll expand to trading other markets like oil or stock indices. 

6) Probably the biggest change is the shelving of my weekly timeframe systems. I know myself alot more now, and I don't have the kind of personality that can hold trades for weeks without tampering with them. 

Sunday, May 4, 2014

Week in Review: 28th April 2014 to 4th May 2014

TRADES

I entered a bit of drawdown this week and lost three out of five trades, with another trade left open over the weekend. 

EURJPY

Normally I use a pre-determined R:R ratio to place my profit target. For this trade, I decided to use support / resistance levels for my profit target, and took profit just before the resistance level at 142.47. Price then reversed and would've taken out my stop loss if I had kept my trade open. Using discretion, I was able to catch the meat of this trade.



There's not much else to talk about. At the moment, I've shelved my backtesting on the double-top / double-bottom entry signal. I've tested a few more pairs and while the results are still positive, I found the amount of optimisation to be too high.

UPCOMING SIGNALS

I really like this pinbar that formed on the USDCHF daily timeframe. It's range is over 1.5 ATR(14) (meaning it's relatively big and thus significant). This is how I plan to trade it on Monday. I'm using discretion here and will be putting my profit target just above the support level shown on the graph (click to enlarge).



Reward would be about 0.9R.

There is a similar bearish pinbar on the USDJPY, but I already have a short open on the EURJPY. I don't feel like opening two simultaneous positions on JPY pairs.