The formation of a strategy

[Human Pirana]

Hi folks,

So as I am sure you can tell this is my first post, lets hope its a goodun.

In short: I am not (yet) doing this for real. I have been "paper trading" on and off for a couple of years (and studied FM's at uni), but have never established any criteria for trade entry and exit that a monkey and a dartboard couldn't compete with. I endevour to find one, and it is my intention to fill you guys in on how I get along; For those of you who are in a similar position, perhaps we can pool our ideas together and speed up the process? Experience has shown me that mistakes are easier to learn from if your peers are there to witness them.

For those who have established rules that work, I hope that the inevitable errors and frustration that follows is some source of entertainment for you all.

Anyway down to business.

Market traded:

I will begin trading USD/JPY. My reasoning is that while the US and UK markets are open, and when I will make my trades, the Japanese markets are closed so I have less to keep an eye on.

Platform:

I'll be using the FXgame platform that OANDA offer from their site. I have found it is the most intuitive to use - and of course doesnt expire, in case this takes much longer than expected. Also I think they offer the lowest spread of around 2 pips for USD/JPY, although I dont know if this is the same spread that Cash account customers get (I doubt it).

Targets:

I will set daily targets of +15 pips, and cease trading for the day if any positions fall below -5 pips. I dont know if this is a high or low target (methinks a little low), but it is my target nonetheless. I do not expect to find some magic rule that will make me rich, but if I can make enough profit for "beer vouchers" I will be satisfied (I will make my proper job responsible for any bail requirements that may result).

Strategy stage 1:

My (albeit small) experience from paper trading so far tells me that I should watch the following Indicators (is this the right term? I get confused with all the terminoligy! In particular what is a "whipsaw"?????):-

EMA - I use 15 and 5 periods.
there is rarely a significant movement up or down without these crossing beforehand. However, every incident of crossing EMA does not imply a movement will follow. Furthermore, I havent got a clue which way the movement is likely to be - to begin with I will assume that the crossing of EMA's indicates a move back to equilibrium - that is, the direction of the movements oscillate.

RSI - the default in FX game is 14 and thats what I stick to.
I'm new to these, but it seems that a breach of the 75% barrier indicates a "buy", and a breach of the 25% level a "sell". NB the default levels in FXGame are 70/30, but I've erred on the conservative side.

Fibonacci:
Fans are the only Fib. tool I can use guesswork to be any use whatsoever. However I know from elsewhere that this magical series pops its pretty little head up EVERYWHERE!!! [incidentally, for those closet geeks amongst us a book by H.E. Huntley on Dover is a great little read about golden ratio's etc...].


Sentiment:
Today the USD/JPY broke below the 119 levels, with support around 118.60.The sub-prime fears over in the US and the extent to which they will affect the economy as a whole is (as yet) a little uncertain. Also I think poor earnings announced today both in the UK and the US pushed indices a little lower (didnt the NYSE pause trading because of the size of the move??). Today yields on the 10yr note fell to around 4.75 ish, which *i think* means that the "real money" is moving from equity markets into something a little less volatile. Obviously any fall in yields will reduce the USD/JPY rate.

What do I think this means? I reckon... err...

well I havent got a clue. Clearly the USD is under pressure; but i struggle to find much info on the JPY economy to consisder the other side of the equation. For the sake of discussion I will suppose that the weakening of the USD is a little exagerated; perhaps some better info from the UK will cause a rise in GBP/USD - and Ceteris Paribus - the USD/JPY will rise on the back of that.

OK so tomorrow I will take a punt using only the indicators and discretion I've shared here. I'll let you know how I get on. Any comments and/or thoughts are much appreciated. As I mentioned earlier, If anyone is in the same position I'd be happy to share ideas - two heads are better than one and all that.

Good luck to everyone else!

Oliver

 

[Human Pirana]

First Step

Morning all

CORRECTIONS:

wrt previous post: I mixed up the "buy" and "sell" signals from the RSI barriers - 25% => BUY; 75%=> SELL

Yes I know I've spelt Piranha wrong, Truth be told my first choice was "Dash Riprock" but it had already been taken.


As promised I will start to document my trading positions today...

Today's Sentiment:

Overnight USD/JPY fell further below the 118.60 support levels mentioned yesterday to settle above the 118.00 resistance level (would this be to do with integer numbers?). Figures released early this morning for the JPY economy fell in line with expectations, with CPI figures around -0.1% Y/Y - if I understand this correctly, this is to be interpreted that the Yen will buy you less now than in will in the future (-ve inflation?) => the Yen will strengthen in the long run (?).

Today GDP figures are released for the US and I will trundle along with the assumption that they will come in worse than expected - the only reason for this is that my general feeling of the US economy is that it is weak, and GDP figures lower than anticipated will follow as a result (yes I know this a very simple and "weighty" assumption to make, but it's all I've got to go on!).

Also I will *guess* that the US 10yr note yields will continue to fall as "real money" continues it's move out of equity markets on the back of credit risk [NB but wont this be good for the USD in the long run? If investors move their cash out of domestic and foreign equities into Treasuries shouldn't this do the dollar good?]

So the outlook today is weak for the pair. Even my basic understanding of charts tells me the current trend on 5min charts is downwards (see attached), so I have jumped along for the ride. I will close my position when either RSI breaches 25% or if the EMA's cross - from the charts we can see the last significant trend was upwards - and I'm on the downward movement as we speak - so given my "oscillation" guesswork the next trend after the crossing of EMA's will be upwards - upon which I will have exited the trade (we hope!).

I will not yet set stop/limit orders as I dont want the trade to close; this is only play money and I dont want to limit the number of opportunities I have to develop the system (also why I'm using 5min charts; I guess no-one with the power to influence the market gives a hell about 5min charts so I hope to increase the time frame in the future) . We will get to these in time.

Anyway I went short at 118.47

will keep you posted

Oliver

 

[Human Pirana]

Chart

Please see attached.

 

[Human Pirana]

Lesson #1

Morning again,

well this is going great! Already I have learnt something new to include in my system: DONT BE WRITING POSTS WHILE A TRADE IS OPEN!

closed position with a loss of 56 pips. OUCH!

For my own purposes I will take the view that had I not been writing to you good people I'd have closed the trade at the crosshairs after the EMA signal, for a loss of about 20 pips.

Sentiment still negative; methinks the uptrend witnessed was something to do with Fib. retracements.

Oliver

 

[Human Pirana]

Trade # 2

Right, Given lesson #1 I have opened and closed a position since last post.

What follows is a commentary on what's going on and my thoughts.

Sold at 119.28

The RSI is just about breaking the 75% level, and the EMA's haven't crossed since last trade. I infer that the next EMA cross will signify a move downwards - confirming the direction of the pair.

looking to close either when RSI < 25% or EMA's cross.

Oliver

 

[Human Pirana]

Trade # 1 cont.

The EMA's have crossed and so I expect the pair to drift downwards...

next EMA cross => close position

 

[Human Pirana]

Trade #1 cont.

well the EMA's have crossed so strictly speaking I should close the position now. However the EMA's actually have been flirting with each other for a while while the pair move sideways (as I have noticed they do sometimes...).

Im gonna stay in this for a little longer... once the EMA's seperate (either for a move up or down) I'll review the situation.

Oliver

 

[Human Pirana]

Trade # 1 cont.

The EMA's have consumated their relationship and have gone their seperate ways.

Will stay in the trade until the next EMA cross or RSI < 25%

Oliver

 

[Human Pirana]

Trade # 1 cont.

Positioned Closed at 118.54

Profit of +74 pips

with the RSI hanging around the 27% level, I closed the trade.

 

[Human Pirana]

Trade # 1 cont.

looks like I made the right call and closed at the right time.

Given that the RSI didnt make a clear breach of the 25% level, I would have definately closed the trade once the EMA's crossed - which still would have been a profitable trade.

Oliver

 

[Human Pirana]

Lesson #2

OK, time for a debrief over the last trade and a review of our system so far.

The last trade was great for the psyche... especially given the first one came in at a whopping -56 pips! (for audit purposes I'm sticking to the loss of 20 that I would have incurred if I hadn't been messing about here!) - so our 2 trade old system shows a profit of +54 pips.

well what IS our system?!:

Using the last trade as a model (we wont use the first because it was a LOSS!)

ENTRY SIGNALS:

went short once RSI climbed above 75%, and the EMA's hadn't yet crossed to signify an end to the most recent trend upwards

So for the next trade, we'll look for a breach of one of the RSI levels and EMA's that still show a trend in the opposite direction we want the pair to move.


EXIT SIGNALS:

Closed the position when the RSI was around 27%.

For the next trade I'm going to move the RSI levels to 70 - 30. I reckon it'll be a clearer system if there are definate breaches for entry and exit signs, instead of me going on about the RSI "hanging around" any particular level - what counts as "around"?????

So our next Primary exit signal is a breach of either the 70 - 30 level, and our Secondary exit signal is the crossing of the EMA's to suggest an end to the current trend (I seem to noticed that an EMA cross tends to lag behind an RSI breach... anyone else find this?). If either of these two happen its time to close.


I noticed that the exit of my trade would have coincided nicely with an entry into a long position. For now we'll stick to the conditions we've got, but perhaps we can include parameters for continuously moving in and out of long and short positions regularly? This will require the utmost concentration as we'll always be in a position, but avoiding any times when the pair just trundle along sideways it could prove quite profitable. Will investigate.



Moving on.... has anyone done a regression or stat. analysis of the "proper" interbank quote (for any of the major pairs) against that offered by a bucket shop? Methinks it might be useful to know if the quote OANDA give me is any more volatile that the real one, its Beta value etc...

anyway enough for now

Oliver

 

[Depth Trade]

Hello HP

You seem to be doing good considering your trading soley off indicators. There are many reasons for currency moves. Interest rates, commodity positioning, technical positioning.
A bias in direction along with consistent play of your strategy could provide pips.

DT

 

[fibonelli]

Hi HP,

Good to see your progress.

I presume you're familiar with the concept of momentum divergence. Check out the "IBEX" thread if you're not.

Good luck

Fibonelli

 

[Human Pirana]

30/7/07 record

Hey folks, hope everyone had a good weekend etc...

anyway here's the P/L of todays trading:

Trade 1 short 118.74 close 118.79 (-5)

Trade 2 short 118.89 close 118 89 (0)

Trade 3 long 118.89 close 118.76 (-13)

Trade 4 long 118.16 close 118.42 (26)

Trade 5 short 118.83 close 119.08 (-25)

Total P/L for the day -17 pips

So not that great today!

The things to pick up from today's trading is enter and exit ONLY when the system recommends - of the trades I made, only the last two "fully qualified" under the conditions I had set out previously and I missed one out completely. Looking back (retrospect is a wonderful thing!), If I had done as I was told, todays activity would have been:

Trade 1 short 118.89 close 118.35 (54)
- this is the same entry as trade 2 above, but I got an itchy finger and closed it early to break even. I should have exited after the EMA's had crossed and once the RSI fell below 30% at around 11:00 am.

Trade 2 long 118.16 close 118.52 (35)
- I got the entry to this right, but again exited too early. It looked like the EMA's were about to cross for the second time, signalling the end of the current trend cycle. Anyway they didn't, and there was another 9 pips to be made before the EMA's DID cross for the second time at 16:00 ish.

Trade 3 short 118.83 close 119.08 (-25)
- well I got this trade spot on according my system, but it came in for a loss.

- total P/L for the hypothetical trades = +64 pips.

OK so what to learn from today...

A new addition will be to wait until there is a DEFINATE signal - that is to say, trade on the data from the last set of candlestick, EMA and RSI instead of trying to get in/out earlier. E.G. Actual trade 1 looked like RSI > 75% at the time, but looking back there was no def. breach - this trade came in at a loss.

That'll do for now, however two points to note for the future; Trade 3 was recommended but was a loss... the RSI > 75% but the trend continued quite strongly... I either need to find a method to determine where to put a stop loss to prevent -25 pip trades. As it stands I closed when the RSI breached the 75% level again.

Better still find an indicator that'll show whether the trend is likely to continue at an accelerated rate as it seems to have done in this case, instead of reversing.

Also, the EMA's crossed at about 17:40 to show the start of an up-trend, which it did until the EMA's crossed again at 22:20 - rising by around 60 pips. The RSI's showed no signals and I missed out big time. However, in between these two EMA events I went in and out of a Losing RSI indicated trade... so I dont quite know how to go forward

Anyway tomorrow I'll stick to my system as it is with delayed entry and DILIGENCE!
Losing trades like #3 today i can manage as long as i dont miss out on trades #1 and #2.

DT: Thanks for the encouragement... without any info on Commodities (I guess Oil and Gold are the ones to keep an eye on?), interest rates (US 10yr yield?) or technical positioning (Option position strikes?) I used the US indices as a guide for a "top down" approach... and got it wrong. hehe! If I can't get around the losing trades my system indicates, maybe I will apply some PIP weighting for "with" and "against" economic sentiment (i.e. If sentiment is +ve for the (intra)day, place more $'s per pip on upward trends than downward trends as indicated).....?

Fib: No not come across momentum Divergence until you mentioned it... did a bit of research over the W/E and got confused... MACD, stochastics etc... they all seem to be different measures of the same thing (and if I'm honest, alot of the technical analysis stuff I have looked at seem to over compicate matters...there are HUNDREDS of the things!) In an FX situation aren't there only Price and Time to consider?. Stochastics I am familiar with (not in a FM situation)... a really useful source would be some accurate description of what these indicators are actually a measure of, and how they are calculated (i.e. the calculus)- not what they may indicate - any ideas?

Many thanks to you both for your guidance. I am running a S'sheet of my performance trade by trade and will put this up once I have enough trades to be significant.

Oliver

p.s. I was wondering if Fourier analysis could be done on trade cycles - dump the data where there is no position, and stick all the tick data during open positions together and look for a cycle... given the same "boundary conditions" (i.e. trade entry and exit) might there be a price/time pattern? anybody heard of this being done?

p.p.s. One last thing.... what about historical INTRA-Day FX rates... I can find daily rates no prob., but I want to investigate the relationship between USD/JPY and GBP/USD at different times of the day. In particular I'm looking for any correlation (-ve correlation) between USD/JPY and GBP/USD while both the USD and JPY markets are closed (from around 8am - 2pm ish). If the "value" of USD and JPY are largely static due to closed markets (other than Gold and Oil pressure on the USD, but I could include this as well from the LIFFE or wherever), maybe a factor of USD/JPY will be the appreciation (depreciation) of the USD against the GBP? I got at this earlier on but 10 mins with PcGive and I'd had enough! The intra-day prices would make this much easier to do as I'd simply dump the data from 2pm - 8am and work with the rest.

p.p.p.s. I actually have no idea if anyone is reading my posts, but writing all this stuff down is certainly helping me get my thoughts together