Caution: Math Geek at Work

Hello again! Here's a summary of what happened last week:

1. I broke even
2. My scalping needs a LOT of work
3. I had my first ever trading dream/vision


1. Longer term trades

As I mentioned in one of the previous posts, I started the week with a big negative opinion on the USD and GBP/JPY. Opened a short on GBP/JPY with too tight a stop on Monday's London session and lost a little there. Then I started building another short by scaling in and made the money back on that before I noticed around mid-week that the downward momentum seemed to be gone, so I closed that out too. Just in the nick of time, which is pretty unusual for me. Here are those trades:



I also had a short on the USD/CHF, but when all the good US news started coming out, I closed that too pretty close to B/E. I also had a long EUR/AUD position that I closed at about B/E. So with all my longer term positions closed out, I spent the rest of the week concentrating on my scalping skills with EUR/JPY.

2. Scalping

I am only using very tiny units, so I don't really make or lose any significant portion of my account on these. I might lose 10 trades in a row and not worry too much about it. This is the steep part of the learning curve and there's no reason to increase my trade size until I can consistently make a profit.

Anyway, it takes some genius to be able to go long 5 times in an uptrend and lose every time, but that's what I did on NFP Friday. I've been testing different sets of entry rules by writing them down and then trading them over and over again in a single session regardless of whether they're working or not. The idea is to build the discipline to follow the rules, and also to spot repeated price behavior that either helps or defeats the strategy.

One thing that I've been learning is that if I'm following a trend in one TF, I should not be looking for momentum in the trend direction in the LOWER TF. When one TF is trending, the lower one is oscillating around that directional bias. Say the main trend is up. If the price starts to go up in the lower TF, then what you have is WAY too much momentum at the moment. The reason is that in order to go from the bottom of an upwardly sloping channel to the top, the price slope has to be really high and unsustainable. On the other hand, it's easy to go from the top of an upward channel to the bottom; all you have to do is move sideways or slightly down.
FIRST ORDER APPROXIMATION: : When following a trend in one TF, look for tops and bottoms in the lower TFs.


3. Trading Dream

Ok, now I'm a math/science guy, not superstitious and very skeptical about any kind of paranormal stuff so I don't want to give the impression that I'm going down THAT road with this next part. There are perfectly logical explanations for why things "come to us" in dreams; subconscious mind turning the problem over, etc. But it was kind of freaky anyway!

Jack Schwager talked about this kind of thing in Market Wizards. He had a dream where Ed Seykoyta made some comment about Jack's trade and Jack closed the thing the next day I think.

Anyway, so Saturday morning I was still in bed around 9:30 or so and I was having a series of those little short dreams you have right around the time you're coming awake. A bunch of people from where I used to work were watching the Super Bowl next door, then I was explaining to someone in my kitchen that I had left that job a few months ago... and then the third dream.

The third dream didn't have any action. It was just a price chart. I could see it clearly, and there were groups of consecutive price bars color coded either green, yellow or red. On the left of the chart was a group colored green, and each of those bars had a lower low than the prior bar. On the right there was a sort of hazy group of yellow and then red bars, all of which were moving up. Right in the middle of the green group of bars was a notation that said, "This is a signal." That's when I woke up.

My first thought on waking up was "I know who killed Laura Palmer!!!" Oops, Twin Peaks flasback there... sorry.

No, actually my first thought on waking up was, "Yeah no kidding it's a signal. I mean how many lower lows can you have in an uptrend before the thing shoots up right??" My second thought was, "Wait, what? Did I just have a DREAM about a price chart??" :surprised Weird. Weird.

In my next post I'll put up my market radar and plans for the coming week.

Here are my weekly "market radar" tables from Excel. There are explanations of these and how I use them in posts #35 - 38.



There are three possible trades for which I'll be on the lookout this week.

1. Short USD/CAD

This pair has been in a long pullback from a downtrend and now we have a statistically negative weekly bar. Could it be ready to resume the downtrend? Friday's bar was lower even in the face of all the positive U.S. news from last week. Now that that news is behind us, we could be looking at break down below unity again. Taking unity as a major S/R area, we see a descending right triangle on the hourly chart.

Plan: If the pair breaks the hourly downtrend line to the upside, bag this idea. If it breaks through unity to the downside, find a good area for the stop and start to scale into a short position.


2. Long GBP/CHF

Huh? Isn't this against the trend? It sure is and I don't really like that, so at the first sign of trouble I will be out of there. But the bullish statistical bar got my attention and so I decided to look more closely at what's happening here.

The GBP just decided to suddenly turn up last Wednesday even in the face of a surprise rate cut. I'm pretty sure all the little retail traders who play the news were selling, carry traders were selling, etc. So who the heck was buying and turning it around?? Do I detect large institutions building a position here?

Also, when I look back on the daily chart I see a classic 123 reversal formation setting up. We reached a low on 11/21 for point 1, a reversal up past 2.33 for point 2, then Wednesday's low point for point #3. We've seen higher lows for three days and Friday's high broke Thursday's as well. So does it go up to retest the 1.33+ high? Even if the downtrend is going to continue, a breach of this high could shake out a lot of shorts, allowing the big institutions to take over those shorts at better prices. Perhaps I'll hop on the back of a major money center bank this week and ride this pair up a bit...

Plan: There's a narrow up channel that was formed over the latter part of last week. Start scaling in on a break above the upper channel line with the initial stop below the lower channel line.


3. Long AUD/NZD

This is another one where the strong statistical weekly bar caught my attention. There used to be an uptrend on this pair, but the party's been over for a while. However, the price is approaching a S/R level at 1.1100 and there may be positive divergence on the RSI. However, I'm not going to anticipate a bounce. The pair needs to have a daily bar with a higher high and higher low before I would really consider going long here. So I'm really only watching this one.

Plan: Look for a higher daily bar and then re-assess.

I forgot to post my account performance in the recap from last week. I made like 0.6 percent or something like that, so essentially a B/E week.

Heyya,

Linux troll simple scalping i see!! I have been one of the thread followers from near the begining untill it closed. Just like you seem to be having - I had good days..... and the bad.

Actually, just before finding your thread I was thinking about the way im trading now, and the LTSS method. The problem with it in my eyes is that you neglect the bigger picture when trading LTSS, and also have no scope of what will/or wont be a decent move.

Maby A way to greatly increase performance would be to have a 5min chart next to the rainbow chart(5sec?) and wait for an 'Exploring bar' to enter. Well, its just an idea, ill have to try it out.

Nice thread man.

Leon

Heyya back!

Yes I've been watching the rainbow for a while now but recently cut back on the moving averages because there are really only a few that I look at. I've been reading through the LT thread and am up to about page 60 I think.


I actually do something similar. At the beginning of each scalping session I check the daily, 3-hour, and 30 min charts. I put Mouteki style (reading from right to left) trendlines on the chart from those TFs so I can see when I'm getting close to support and resistance areas. Then I check the 5 and 1 minute charts to get a main current trend.

I saw your posts about "exploring bars" over on Mick's Journal too, so I'm off to google so I can read more about the idea. Sounds interesting.


Thanks Leon!

Wow, what a great week! Great for anybody that took the exact opposite trades from the ones I took!

Yeah, so there ya go. Out of my three picks at the beginning of the week, only one went where I thought it would. My short USD/CAD opinion didn't even give me a chance to enter, luckily. Not so fortunate with my long AUD/NZD pick though as I bought near the high of week and was stopped out near the low. The really funny one though was GBP/CHF.

Last week I had this full comprehensive analysis on why I thought this pair would go up and retest the 2.33 level (I said 1.33 in my last post for some brain-dead reason but whatever). So this week, what did GBP/CHF do? It bounced around a lot, but basically it went up to test the 2.33 level, which is just what I thought it would do.

"Wow Cap, so you made a few pips on that one huh??" I hear you ask.

Well I would have if I had created a simple trading plan like, oh I don't know... buy the pair and then sell it when it gets close to 2.33. But noooo....not me. I had this really BRILLIANT TRICKY trading plan from ACME FREAKING TRADING PLAN COMPANY THAT I BORROWED FROM WIL-E-COYOTE. :

Yeah, so my "brilliant" scaling in plan had me buying half at 2.31, then another half at 2.32, then getting stopped out of the whole position at about 2.30 which was of course the low for the entire week.

As usual, I'm still taking only very tiny positions at this point because of my inconsistent profitability (this actually IS turning out to be a smart decision) so my account balance declined by only 0.9% this week.

At this point I'm going to simplify life by eliminating this long term strategy and sticking to only the ultra short time frames. This is because I won't know if my long term strategy works or not until I have accumulated a statistically significant number of trades. That could take months or more. I don't have the confidence to wait that long, only to possibly find out that the method has a negative expectancy and needs to be refined.

I require data. A constant stream of performance data. I also require a development ==> execution ==> review ==> refinement cycle that is completed in about a week, not a year. I made a similar point in a thread on expectancy that I started here this week.

Disadvantages to scalping? You bet. Spread's a bitch and has to be defeated. I get that. Don't TAZE ME bro!! : But the ultra short term is the only TF where I won't have the urge to constantly abandon systems which may actually be profitable. With scalping, when I abandon or refine a system I'll know that I'm doing it based on adequate data, not just a few trades.

So open up a chart. Switch the time frame to 5-minutes. See that last candle way over on the right? See the little dude running around inside that last candle? Yeah... that's me.

hehe, I hear you Cap'. It was a bitch although I scraped through with a small gain. I too am changing tactics a bit, particularly with MM. I have also been in the 5-15m charts mostly and spread kills but whadyado?. I 'try' to get them profitable then move to higher TF's if price allows though hopefully to ride out the larger trend and make the spread less of a problem.

Oh, I saw a reference in your journal about "nursing the trade up into higher TFs." I meant to say something to you about it but I forgot until now. Anyway, that's an idea that I had also, and one of the scalps I did a week or two ago was a perfect illustration of how well it can work.

I had started off in the 5-sec charts (almost impossible to get a positive expectancy down there with the spread, I know), entered a position, and the momentum carried it into profit. I was trailing my stop behind the longest MA on the chart. So then I moved to the 10-sec chart and waited for THAT long MA to move past my stop, and started using that one to trail. So now basically I was moving the stop up at a slower rate.

I continued to do this, up to the 30 second chart, then the M1. So I was "handing off" the trailing stop to the next highest TF each time as the MAs on those charts started moving in the direction of the trade. Of course finally the price turned around and took me out, but by that time it was like an HOUR later (this was originally a scalp) and I had a good profit.

But here's the interesting part: when I looked back at the 5-sec chart, I of course saw all these up and down gyrations where I would have been stopped out of this trade a million times. But since I had gradually widened out my view by jumping up TFs, I held on through all this turbulence, because on those higher charts it was just a steady trend. So looking back at just the 5-sec chart made me look like some kind of genius for staying in the trade.

This concept is one that I definitely want to incorporate into my overall strategy. So I'm not entirely against the longer TFs. I just want to enter at the very low end, and as you say, "nurse the trade" up into a longer term postion. I can even imagine that every once in a while a scalp will actually turn into a position on a daily chart through this technique.

Do you know if there's a name for this? I refer to it in my own head as a "Time Frame Ladder."

While I'm here in my journal, and since it's Pick Your Pocket Sunday (PYPS), : I'll take the opportunity to not only have myself a damn fine cup of crème brulé flavored coffee but also to list out the success factors which I would like to include in a complete strategy. Right off the top of my head, some possible "edge producing" tools are:

1. Positive carry
2. Scalping = immediate feedback on expectancy
3. Time Frame Laddering to turn short term wins into long wins
4. Varying strategy based on current price action condition (see diagram)
5. The power of confluence (MMAs is a good example)
6. Macroeconomic modeling to anticipate economic release surprises
7. Market intelligence system based on a database of pending events
8. The creation of "free positions" without stops (only possible in my account where USD is the countercurrency).


Hopefully I'll remember to create a post on each of these ideas at some point. I won't make the mistake of trying to incorporate them all at once though, because this will lead to immediate priority lock! So this is more of a "strategic interests" list for me to keep in mind as I go forward.

I'll make a brief comment on item #4 now though. The "Price Action Map" below is something I completed today after thinking about the essential structure of price behavior and how it really is composed of only a few types of conditions and events. If anyone's interested in more detail about this, let me know and I'll post more about it.



Anyway, my coffee is done so I'm off for now...

A few things from fti in the TA fallacy thread are starting to sink into my subconscious and are now starting to come forward the more I analyze the charts, I'll try to explain.
One method I exercised the other day was buying in in reverse to rescue a trade, it looks dangerous (and it can be if you don't have a strategy to manage it) but if executed properly can save you and even enhance your profits, this also takes full advantage of the 'swings' in all time frames but this method also needs a good understanding of why stops are not as important as you might think - particularly in short term trading.

It also has to be taken within the context of the larger time frames, within larger time frame trends are the minor trends which also have their own swings etc. Once you picture this you can capitalise on them if using proper MM to size your postions as to not over expose yourself on any trade if it 'really' goes against you and you can't 'rescue' it in a larger TF.

Does any of that make any sense??

I do set a stop but it is in no way anywhere near the market I'm trading at the time of entry. I'm experimenting with breaking my initial risk into 20% chunks for my entry 'units', this give me a lot of room for loss for a start. For example, if my account is 100k and my risk is 1%, that gives me a SL of 1k!, broken down to a unit of 20%, this can take quite a bit of room before it's hit. It also allows me to snowball in reverse (up to my risk) so when the swing comes back in my favour I already have good exposure to break even on the swing or continue if it goes in my favour.
If the market goes in my favour from the start, I'll put more units in to take advantage of the trend and maximise my exposure, once well into profit I'll keep adding units beyond my initial risk until the market tells me otherwise.

You may want to play with different %'s of risk but the play is to get into the market with plenty of room and capital if needed to dig yourself out of a bad trade. It 'may' not seem as though the profit is there but you are less likely to be stopped out of trades, you can stay and fight for it using the swings of the different TF's to your advantage.
Once you're trading in higher TF's the principal is the same, the difference is you use your 'new' capital for risk management and exposure

Sorry for the long winded post, I should have put it in my journal but it fit's well here to in regards to SL's and not being hit.

Yes, it makes sense and I'll respond with a Zen inspired quote of my own. The ZB say that there is an entire universe inside every blade of grass. I say...

Inside every candle is an entire price chart. And your price chart is inside someone else's single candle.

I saw the trade that you're talking about on your journal, but I didn't check back to see the explanation yet. So I'm not sure of the specifics of what you mean, but I understand the general idea behind it. It's a matter of keeping enough capital in reserve so that you can shift perspectives, see if the "losing trade" is actually something that can work as an initial entry in another TF and go from there. Or am I way off?

Nope, you understood pretty well considering I typed it out quickly
Love the quote too.

Another analogy would be to not keep crossing the road to change busses, while the bus you're on may not be heading your way at this moment (it may have to pick up more passengers ) doesn't mean to say it's not heading towards your destination.
If you're on the wrong bus though...you're on the wrong bus

OK, I'm opening up the week by taking two (hopefully) very long term positions. This move incorporates factors #1, #4 and #8 on my list.

1. LONG NZD/USD in an amount equal to my account balance (1:1 margin) with no stop.

2. SHORT an equivalent amount of silver (XAG/USD) at about 14 with a stop at 15.10

The price action on both looks favorable as silver is in a downtrend on the daily chart and is just breaking below its prior low now. NZD/USD has made three higher highs and two higher lows on the daily chart since mid-October, and my buy was at the bottom of a big angry red bar that was still above the prior lows.

That's my long term play of the week I think. I will be adding to the NZD/USD position in the future from any accumulated short term profits. The only time I plan on moving out of that position is if another currency's rates surpass the NZD rate.

Well I just hopped on the carry trade bus. I hope it's going my way!

Am I some kind of crackhead or what? :surprised This short position in silver is not only kicking my butt right now, but the underlying reason that I took the trade actually doesn't make sense. I put it on as the "other half" of my long NZD/USD position so that I would be able to take advantage of the full 8.25% NZD rate instead of just the 4% differential. So I figured I would short a metal (silver because I have more flexibility in postion size) to get the "other half" of the carry.

This is nonsense. I already AM getting the "other half" of the interest risk free on my margin balance. There was no need to enter another position to do this.

Sigh. Ok, so now what do I do with this? (I'm not really asking readers this, so don't post suggestions, LOL! Just talking...errr, typing my way through it for myself.)



Those are TD trendlines, and the MAs are there to just provide some longer range context. Anyway, this "downtrend" that I thought I was selling into hasn't materialized and it's looking more like a pullback in the larger uptrend now. The stop on there is an emergency "system down and my shoes are on fire" stop, not an exit strategy.

So what I'm going to do is see how price reacts to that upper TD supply line. If it breaches with some strength then I'm out of there. If it breaches but then re-enters the consolidation immediately then we'll see. If it fails before reaching it, or bounces off of it with gumption (by gumption!) then I'll start thinking about possible profits out of this craziness.

At any rate, this position is too large and I really would rather be out of it.

Mainly because of the short silver trade that I discussed in the last post, I managed to dump 3.9% of the account this week.



1. Long Term

Even when I'm out of that trade though, the account will still fluctuate a bit more week to week because of the long NZD/USD position that I have at 1:1 leverage. That's an income producing position with no stop. I don't consider myself to be "short the dollar" on that position because my account balance covers the short dollar side. Thus, this is a straight absolute value asset position like a stock or a bond.

So my plan is to use any weekly profits (assuming I ever have any ) to dollar cost average into this position over time and take advantage of the rate. The only time I will move out of this position is if another currency overtakes the NZD in interest rates. For example if the AUD and NZD rates are ever equal, I would move half of the position into AUD/USD.

2. Scalping

OK, so much for the long term. On the scalping front, I said last week that I require a constant influx of data and I've got that now. The chart below shows the progression of my Actual Win Rate vs. my Required Win Rate as the week went by.

RWR = average loss / (average loss + average gain). I consider my "edge", if any, to be the gap between the two values. As you can see, the market has a huge edge at this point. Prepare to hurl...



Well as we can see, my win rate is below 30% while my RWR is around 75% (mainly due to a large loss that I took on Thursday after letting a profitable trade turn around on me while I was trying to be tricky and outsmart it). The 75% RWR means that my risk/reward is such that I would need to win 3 out of 4 trades to break even, which translates to having losses that are three times as large as my wins.

Anyway, this simply won't do. The spread is of course a big factor, and another is the constant "ruleset hopping" that I'm still doing. Now I had a fairly well defined set of rules in post #33 supported by some statistical analysis in post #34. This is another example of why having a trading log is valuable because I have no idea why I got away from that ruleset, but it still looks good.

3. Next Week

A. If that silver trade is not profitable soon it needs to go.

B. I will go back to my old ruleset for scalping and see what the stats look like for that. Then I'll adjust the method accordingly.
Plan ==> Execute ==> Review ==> Repeat!

C. I'll also be trying out a very simplified method of intermediate term trades based on a discussion in this thread: http://www.forexfactory.com/showthread.php?t=61664

D. I won't be doing much of any of these things actually because it's a holiday week. So HAPPY HOLIDAYS EVERYONE!!

I'm putting this in my own log so as not to broadcast it all over the place. I noticed a couple of days ago that I have a 1KT voucher (I had to check to see what this was) from Merlin! :surprised I must have said something halfway intelligent somewhere but I don't know what, LOL.

Anyway, I don't know if he goes around reading people's logs, but if you see this Merlin, thank you for the sentiment! Thank you also for this great site, and for the excellent article you wrote on the methodology you use for system developement and testing.

I'm actually only at the "consciously incompetent / aha moment" stage in my trading life where I'm just struggling to break even most of the time, but I'll get there!

Hello there Journal : and Happy Holidays to all!

I didn't do any trading this week because of the slowness, so I concentrated on my latest research project instead. I'm still in the short silver position which went against me a little more this week. However, it may be stalling below 14.80 so I'm still hanging in there until next week.

My research project on price action and structure has three planned parts, a study at the macro level (patterns of many bars), micro level (patterns of a few bars), and a combination of the two (macro behavior inside a single bar). I am finished with the first part which brought my notes up to 24 pages so far. They won't be up on my site's Research Archives until I finish the whole thing, so below is the summary page from the first section.

NOTE: All of the data I'm using is on EUR/JPY only. I don't know if the findings are universal across pairs or not but they do seem to be universal across time frames.


-----------------------------------------------------------------------------
MACRO PRICE ACTION - SUMMARY OF FINDINGS
-----------------------------------------------------------------------------

12/28/07 – 0415GMT:

1. Setting a zig-zag threshold of 250 pips for daily bars, 50 pips for 10-minute bars and 15 pips for 10-second bars I found remarkable similarities in behavior among all three time frames.

2. There are only four different 2nd order types of price points; HH, HL, LH, and LL. There are only eight possible types of price leg between any two of these points which consist of either rising or declining uptrend, downtrend, range contraction or range expansion legs. This is an analytical conclusion, not an empirical one.

3. Pullback legs in trends retrace the prior leg by a median of about 80% while trend extensions have a median length of about 110% of the prior pullback leg. Range breakout legs have a median length of 140% of the prior range contraction leg. Range expansion legs (HHLL or LLHH) are generally followed by pullbacks with a median of 60% of their length. I didn’t see any particular affinity for pullbacks to classic Fibonacci ratios, but I didn’t do an extensive search for this.

4. More than 65% of price legs are trending type legs (HHHL, HLHH, LHLL, LLLH). About 60% of price bars are inside such trending legs, with 25% being part of range expansions and the remaining 15% being part of range contractions.

5. Trends generally last for only 3 or 4 legs before being interrupted by a range contraction or range expansion, which tend to last for only one or two legs.

6. In EUR/JPY there was cyclic behavior by which trend directions tended to reverse after interruptions as opposed to continuing in the same direction. The types of interruptions tended to alternate too, so that if a trend was interrupted by a range expansion, the next trend would most likely be interrupted by a range contraction.

7. The alternating trend & alternating expansion/contraction cyclic behavior was present more than 70% of the time in all three time frames. A single cycle tended to go through an average of eight such alternating conditions before breaking the pattern, i.e.
Uptrend, contraction, downtrend, expansion, uptrend, contraction, downtrend, expansion. Each of these conditions might consist in turn of several individual alternating price legs.

That's it for now. I'll be studing micro level reversal and continuation patterns next. Also, I'll soon be getting more formal with trading system development testing and implementation in order to stop my tendency to "system hop."

My sister sent cookies, so I'm off to make coffee! C\_/ mmmmmm...

Whoops, I did it again. Forgot the account update. Even though I didn't do any trading the silver position brought the account down by another 1.2% last week. The position actually lost more than that but the NZD/USD long made up for part of it. Here's the update.

HAPPY NEW YEAR!!!



~~~2008~~~~