FXAnalytics - fxan.com

I've gotten the back-end of the site finished up and am finally putting some chrome on it. It's on the top of the priority list now so I would think within the next few days I'll have the basic theme finished and the code optimized.

For a while all the content will be free - I'm still unsure if what I do and have is "of interest" to the individual trader. I'll know better in a few months.

The site will be in various stages of disarray so don't expect too much right away but don't hesitate to let me know about anything that doesn't work the way you would expect.

- John

Volatility measurements show an increasing potential for a strong move. I would expect today or tomorrow we will see the pair try to bust out of this three day range.

The bearish trend appears to be a done deal as the support band has just made a higher low - indicating selling pressure has eased. That's not to say we don't take a stab lower and test 1.3100 I just think it will be very difficult to get back there (and lower) with the current levels of volatility and market participation we have recently seen. The fundamental driver would have to be something very unexpected. I guess in some way that makes a test of 1.3200 more logical or appealing to me - although I'm not sure why. At any rate gut feelings are baseless so don't put much stock in it.

If you're a short-term pattern or channel trader your train could be coming soon.

http://www.fxan.com/images/charts/eurusd-4-11-2012.png

I'm sure everyone has already spotted this... There is a nice little wedge forming in the EUR/USD on the 15m time series. Could provide an entry point on a break.

http://www.fxan.com/images/charts/eu...-11-2012-2.gif

The pair came up and touched the 1.3200 and then fell back. Right now we're seeing some initial signs that volatility is starting to diminish. The move up may not be over but it also may not have a lot left in it. If you're here on the short-term it might not hurt to start looking for the exit. I can't rule out a push higher to test 1.3200 but given the time of day I'd think we're probably looking at consolidation with a possible decline back to the 200MA or longer-term consolidation where the 200MA comes up to meet the price action. If I throw enough variables in here I'm sure to hit at least one that eventually works out - that's how we do it.

http://www.fxan.com/images/charts/eurusd-4-12-2012.png

About 3PM my time yesterday the cyclical indicator in my set put in a flat top. This is often a precursor to a move down, although I can never be sure how far. Well it turns out it was a pretty good move, which has also helped suck the bullish volatility right out of the charts (and the markets I'd assume - given the weekend and all - you're not trading tomorrow right?) So let me just tell you what you already know...

Current levels will probably hold as support and if they don't, I would expect 1.3040 to be it for today (Friday).

http://www.fxan.com/images/charts/eurusd-4-13-2012.png

I'll be back in a few days - I've got to do some coding this week to automate a few of the modeling processes I use in the trading... As I think about moving back into managed funds I know I need to manage some of the more time consuming things better than I did previously. I want to get this out of the way before I get too far entrenched in the daily writing, which is only going to increase over time. Be back in a few!

Didn't get as much done last week as I had hoped but that's the nature of the beast. Time to get back to some analysis...

Taking a peak at the EUR/USD this morning we're still caught in that range, so it would appear that cyclical trading is going to remain the better approach in the short term. Picking tops and bottoms is difficult but trying to grab a trend here could be even harder. Two things worth noting even with the current easement in the pair my cyclical indicator is showing a top... That's got me scratching my head a little and I'm not sure how much weight I'd give it - but it's there. We're also seeing the probability of a break-out or higher volatility move increase. I'd be looking for something to ramp up in the next 24 hours or so.

If the pair does move lower, factoring in the recent volatility - I'd be looking for the pair to run out of steam around 1.3058 +/- I think the fundamental driver would have to be a significant shock to get the market to participate beyond that level - and the very nature of "shock" makes it difficult to model or prepare for...

On a initial move up I'd target 1.3219 as the initial target with 1.3259 as the last leg for a sharp move. A slower grind higher with reduced momentum is of course possible but gains will take longer which always injects a certain amount of market risk in the way of a sudden reversal. - John

http://www.fxan.com/images/charts/eurusd-4-23-2012.png

Well gang - after a lot of work - I've had to swallow a pretty bitter pill. Its become apparent that the method I used to successfully manage money in the past just isn't working anymore. I've got ten years of research and work into this thing but I have to be honest with myself about the potential of it.

I really thought based on some of the recent results things were going in the right direction - but, like is so often the case with trading, the returns did not persist into the future. If I can't make money as a trader I have no business in the Business... So with that I'm hanging it all up for good.

I wish you all the best in your efforts. - John

This sounds very drastic. I have the impression you have years of experience and were very successful. Don't you have the drive left to take things from another approach?

I appreciate the spirit of that post - nothing wrong with "you should hang in there!".

Probably a little background is in order.

When I entered Forex it was after having spent several years on the equities side in a professional capacity. In a lot of ways I was ahead of the game in terms of my research and back-testing ability and I came into it from a position of strength. From 2000 to 2003 I spent over $100k on R&D. During that period I tested every approach I could find. In my mind I left no stone unturned. From simple to incredibly complex algorithms, I ran it through it's paces. After three years I settled on Fast Genetic Programming in a Distributive Environment. It was cost prohibitive at the time but showed the most promise. I didn't stop there though, even while that process was grinding along making money I continued to research and test.

After that decade of work I've come to see my successful track-record as little more than luck at the time. Even if it wasn't, the life span of the approach versus the cost to maintain it could never be justified unless there was massive capital to allocate early on, which rarely happens.

I gave up a long time ago trying to convince anyone of something that they don't believe - I understand how belief biases work and most of us are programmed from infancy to have them and hang on to them. Once someone "believes" they can trade, no matter what information they have to the contrary, they will continue to hold the hope they can over-come any shortfall they have whether it's know to them or not.

All I'll say at this point is - for me - after having allowed the equivalent of 2400 computers evaluate 1.2 billion trading programs over the last several month - without a one being able to generate consistent positive returns - the fat lady has sang.

I'm no longer willing to allocate time, capital and so on to an endeavor that appears to be directly tied to chance. That's just me. I know there will be a lot of nay-sayers, but the truth is - few have spent the time or resources I have in this field and fewer still will be hear ten years after they have made their claims of success. That is of course the greatest measure of success - Time. It's not hard to be a hero for a week or a month. But very, very few people will survive a decade with their Holy Grail. They of course don't know that yet and would never consider it to even be an option, but time will have her way with them. She is the mother of reason (one way or another).

Did you try trading systems lab GP system?

- guandi

No but I have extensive experience with Discipulus the backbone of their system. In 2007 when the Patent Expired on Disc I took a hard long look at the guts of thing... Personally I wouldn't touch Trading Labs... here's why.

First, the last time I checked the licensing fees were double of that of discipulus. At the time Disc was $5000 a year and TL was $10,000 - that may be different now. But if you're savvy enough to use something like Disk you don't need all their bells and whistles to properly structure your model or the testing. You can easily do it with something like Excel.

Secondly, I think TL injects to many variables into their system. It runs the risk of putting users into the endless loop of research, trying this and that until there "appears" to be a fit.

Thirdly, Fast Genetic Programming (FGP)- IMHO - requires massive distribution to be effective. The program is written in low level code and is much faster than many high level applications but this still doesn't overcome the complexity of answering the trading riddle. Between software and cpu's you can easily (I mean EASILY) drop $50k into what I would consider to be an entry level system to make this thing work.

Finally, If TL was actually able to generate profitable models with disc, it wouldn't make sense for them to sell the product - it would make sense for them to manage $100m in assets. That's where the real money is at (or should be). It's no mistake that Retail Investing is such a big business... It's so big because at the end of the day IT'S THE ONLY MODEL THAT IS PROFITABLE. That's the bitter truth.

I wish everyone well in their trading, if nothing else I hope you keep your risk levels in check until you have traded long enough to make an educated decision regarding the task. This will be my final post. Good luck to you all - John

Respect. All the best.

Well gang, I'm back. I guess you should never say never... After a lot of research I finally forced myself into new thinking about the old approach. It helps when your field (Genetic Programming in my case) is still actively being developed and expanded by others. There is a lot to be said for fresh eyes on a problem. I've been back to trading now for almost a year, just on a much smaller basis then in the 2005 when managing money.

Part of the process in getting GP to work again (although differently) was deconstructing the old data-set. A lot of the proprietary indicators that I used in the set or as an overlay/trigger were no longer helping the algorithm. Their impact on the modeling process had dropped to zero. That required a lot of new work, something that will have to be an ongoing process to keep the approach viable. I've also changed how I use GP in the decision making process, but that I'll keep to myself for a while

Some of the old data-set has application on the institutional/corporate level and I've launched a new company to slowly push that work in their direction, it's a pretty long sales cycle though, plus I need to establish some track-record there as well.

Some of the indicators also have value for the individual trader, so I'll also eventually be relaunching Fxan so traders can add those to their decision making tools. Right now though I just want to get back to writing and analyzing and thinking about the markets and the next trade. Looking forward to posting again. - JP

Just a quick update on a fairly slow trading day for me. The 5m time frame in the EUR/USD is signaling a high probability for a break. My volatility model is peaking. If you're sitting on your thumbs today there may be a few pip on the table when this moves.
http://fxan.com/content/wp-content/u...m-10-21-13.png

95% traders go broke, only 5% survive
I saw this thread in the Rookie Talk forum and thought I would chime in with my own experience - albeit from the restrictions of my own commercial prison.

I have no idea where that statistic came from or what the basis for it is. Over the years I've run across a couple of research pieces that try to address it, but they always leave a few important (IMHO) components out, which I'll leave for a different day.

My gut feeling is, 95% might be too low. Almost all of the traders that came in with me in the Class of 2003 are gone. 2004, 2005, 2006... Gone. I joined this forum in 2007 and almost never see anyone from that period. I think if these researchers looked at a ten year window they would find a much smaller percentage than 5% have survived without blowing up an account.

So is my advice not to trade? No. Why? Because it's your right to. If you have an interest in the markets you have a right to pursue that interest. That doesn't mean it won't cost you though.

So what is my advice? Trade small. Really, really small for a very long time. Trading may be a hectic light-speed, action packed process. But learning the nuances of it is the equivalent of growing good grapes for great wine. Every year is different. No challenge or combination of fundamental drivers is exactly the same as before. You have to trade small and if your goals don't allow for it, you need to readjust and rethink why you're doing it. - JP

I'm watching the conversation in the thread mentioned in the last post. A few of our members are discussing monthly returns. what they are targeting, what they need to make for this to be worth it etc. I understand the drivers behind that thinking, my opinion is, you can't do that. If you need to make x% a month to achieve whatever goal or reason you have for trading it will never happen. Maybe 1 person in 100,000 will be the white swan in the distribution and get so lucky they generate 60% a year with no considerable draw drown. If you do believe 5% a month is a reasonable return I would encourage you to sit down at your computer and make a personal list of the reasons why. Then research those out and see if they hold up. By research I mean real research. Look for papers written at universities, stuff that is unbiased. Unfortunately when you ask a question like, is five percent a month reasonable on a forum, everyone will say yes because 99% of the participants are close to your personal experience level and goal bias.

To really learn you have to move yourself into unbiased research and build opinions from people who have been at it much longer than you and have no financial interest in getting you to trade (that's hard to do btw).

My point here isn't to discourage you from trading, if you want to do that you should. But, the capital should be true risk capital and nothing else until you have proven to yourself over many months (years?) that what you believe is based on fact and not just hope. - JP

Thought I'd blow the dust off this old thread. Can't believe it's been so long since I've posted - time flies. I've been under the hood of a massive project for some time, not out of the weeds yet but getting closer. Hoping in the next year or so to be able to pull the trigger on managed funds again. Right now working on an institutional product that may have a trickle down retail offshoot. Too early to tell. Maybe I'll stop by in two years and update... haha.

Well, I wrapped that project up this summer and started trading again in September (very small account). I'm going to grind along with this through 2015 and let the metrics shake out then make a decision about what follows. I was going to setup a Trade Explorer but for some reason FXCM is truncating my return data, I have to get it from their prop platform... so for now I'll just post up the chart. I still may do the TE for verification. Not in any rush to prove anything - need to slow grind this for a while.

Quick look at the decision matrix this morning. The only thing that is really standing out here is the CHF & EUR/JPY. They are about a .99 correlation right now so it would be one or the other for me. Honestly at this point I'm probably done for the week but will revisit it Sunday night and see where things are.