To help giving me a good idea of a profit...I programmed a simple console program to show me a general increment of profit and final value on a weekly basis.
It is interesting, that if I maintained a 20% profit weekly with a 500USD bid, I could make 6 million in a year. But, then again, 20% may not be as easy as I think.
I have put in 500USD last week, buying EUR/USD and selling CAD/JPY...EUR/USD had realized $25 profit. Unrealized P/L has gone up and down from $25 to $49.
I think selling CAD/JPY while buying AUD/JPY is a mistake. I need to do more analysis on this. Now the very profiting Canadian pair is paying for my Australian pair.
So I'm only up 5% in a week.
I figure, 10% is more constant for me. One year projection value would be 71k. However, 5% gain would yield only 6k. But that is good enough, as FX trading is my small pet project.
Soon, I would program more complex calculations.
Quick tips on Strategy Pattern.
sites for very up-to-date currency rates:
http://www.x-rates.com/
Timing is important. Currency pairs usually rise or drop at certain time:
ie. EUR/USD rises from 3am to 10am EST.
AUD/JPY:
rises in the morning, 2am, 5am.
drops in the late morning, 9am
At minimum or maximum...of course.
Determination of such, is a study.
2:40AM 9/7/2012
rising a good deal to 7am.
That rise yesterday made me pay more attention to improving technically and be less ignorant.
That rise at 22:45pm EST 9/5/2012, really kicked me, because it made me lose around 100USD. Taking away all the profit I've made and back to 500USD the original deposit.
This lesson forced me to study closer and learned to use a useful feature called "Trailing stop". This way, I cannot lose big.
It was however because of some complaints against OANDA regarding this stop-loss/trailing-stop feature not working well, that made me lack interesting in acknowledging the term in the first place.
There's much to learn. But I am more careful now.
As of now, I am climbing back up, slowly.
Interesting drastic changes starts at around 2am EST.
at 4:20am/4:30am EST, all rates are going crazy.
start around noon, most rates become negatively boring or risky, but mostly boring. Must be watchful of it even at 9am. This is very obvious in CHF/JPY, CAD/JPY cases.
I made a little program using C#.
When I set high or low limit watcher, at night, for possible direction jump, it rather accurately gives me the time when the rise or drop starts.
When I study the 1 minute regular and 2 minute R/D freq. rate, it usually indicates the opposite direction for the incoming rates, especially when the 7 consecutive points I collected indicate increment instead of decrement.
Both high/low watcher and R/D Freq. data should be used together to determine buy or sell.
I am watching 9 pairs:
AUD/JPY
AUD/CAD
AUD/NZD
AUD/SGD
CAD/JPY
CHF/JPY
USD/JPY
EUR/USD
EUR/NZD
of all, EUR/NZD is the only more active player during the noon-9pm EST.
So it is better to play for this pair during day time for drastic changes.
For analysis? http://www.myfxbook.com/forex-economic-calendar
Suggested news read: http://priceonomics.com/the-trade-of-the-century-when-george-soros-broke/
Population growth rate should also be a factor.
Initial phase:
Download EUR/USD hourly data (as spread sheet) from http://www.quantshare.com/sa-421-6-places-to-download-historical-intraday-forex-quotes-data-for-free
Get table of economic calendar from myfxbook.
Plot graphs on Excel to compare the events between the calendar and the exchange rate.
Regarding setting FX rates, it seems that there's no centralized entity. Rates are set by buyers and sellers in the market (Thanks to econbizer from Forexfactory - Damn I wish I had gotten the $100k PHP developer job there). For further info - Two important keywords:
algo-trading & arbitrage.
Key events to watch on Calendar:
GBPUSD - BRC Retail Sales Monitor y/y (monthly). If falls, then GBP falls.
EURUSD - US Unemployment Rate - the higher, the higher the rate.
Consumer Inflation Expectations: if higher than previous, than wait for peak (20 mins?) then expect drop in rate.
GBPUSD - German Factory Orders m/m - short effect.