Boris and Kathy's FX Blog www.bktraderfx.com

8/18/2007

 

Direction and Amplitude

Novice traders often think that finding proper direction is the most difficult aspect of trading. No doubt getting direction right is a challenge, but with experience that task can be mastered. After some time in the markets many traders can accurately forecast direction as much 60% to 70% of the time. However in order to succeed, a trader must no only predict direction but also the amplitude of the move. Put simply we need to accurately forecast not only if the instrument will go up or down but by how much it will do so. And amplitude can be maddeningly hard to estimate. Every time we see a moving average crossover, a Ballinger band breakout or an event risk reaction we can never be sure if the move will last for 10 pips or for 100.

This week was a good case in point. We had two long term trades that initially moved our way but ultimately reversed and hit the stops. Our short EURGBP stayed very quiet while the rest of the market was caught in a whirlwind of activity. The trade even slowly moved 10 pips our way as our forecast of stronger UK Retail Sales and therefore more bullish cable posture proved accurate. Bu the move was short lived. The market this week wasn’t paying attention to economic news – it was strictly trading off risk aversion theme as carry trades were liquidated wholesale. The EURGBP got caught in the cross fire of volatility as the 1000 point declines in GBPJPY ultimately pushed the pound lower.

In the GBPUSD trade we had a different dynamic but similar results. We bet on the fact that after Thursday’s massive sell off the high yielders would be due for a bounce. Going long GBPUSD into the Asian open we were right from the start and the trade moved 40 points our way. But when the Nikkei opened for Friday trade, panicked Japanese investors sold stocks at a frantic pace pushing the index –800 points. The pressure was just too much for the currency market and the pound buckled hitting our stop and falling for another 100 points lower before finding some support.

Some of you took us to task for letting a winner turn into a loser and rightly so. Generally we trade with two units so that when 1 unit moves into profit we take a small profit and move the rest of the position to break even. This, in fact, is the only practical solution to the direction/amplitude problem. Since you never know whether the move is good for 10 or 100 points you take some money off the table and look to ride the second unit to a much larger profit. However, we decided that using two units on our long term trades exposes subs to too much risk. A typical 100 point stop could generate a 200 point loss on two units of trade. Therefore we prefer using only one unit. The trade off for assuming less risk is that we will sometimes miss banking small profits. Of course some of you chose to use 2 units even on long term trade recommendations and wrote to us that you’ve been able to bank profits that way. You are more than welcome to do so. In fact the most successful BK subs are those that use our ideas but adapt them to their own trading style. For our part we will try to manage these trades tighter and will trail our stops to or near breakeven once they move in the money in order to minimize risk.

Overall the market continues to be dominated by themes rather than event risk and therefore we will maintain more of our focus on longer term trades rather than news ideas. We expect volatility to continue for at least another week as investors try to assess the magnitude of last weeks damage, We also have Category 5 hurricane Dean barreling into the Gulf of Mexico which may wreck havoc with oil markets and impact financial markets as well. All in all it promises to be another exciting week in the markets and we will be back to you with fresh trading ideas.


Comments: Post a Comment



<< Home

Archives

09/2006   10/2006   11/2006   12/2006   01/2007   02/2007   03/2007   04/2007   05/2007   06/2007   07/2007   08/2007   09/2007  

This page is powered by Blogger. Isn't yours?