I’m recently came across some Python Code snippets for Machine Learning maintained by Stephen Marsland. I’m not a programmer or coder but I know several of my readers are.  Follow the link and enjoy.

I consider myself a Bullish guy.  I believe in the markets and that they’ll recover one day into another fabulous Bull Market.  Heck, I’m still long in my retirement accounts and still plowing money into the markets as part of my long term investing strategy.  However, there is a short term reality out there that any sane person can’t ignore.  The markets are still sucking bad and this rally is probably going to fall apart now or around the 1000 level in the S&P500.

How did I come to this? I did it through three ways: Technical Analysis, my recently updated Rapidminer neural net classification model, and my Monte Carlo simulation model. You can check out my example classification  model in the tutorial section to get an idea how to build one.

Bottom line: We don’t have the right ingredients in place for a new Bull Market and we have more lipstick to put on this pig before we build a new Bull Market.

My whole AUDUSD neural net model is shot to hell because of the massive selloff last year, so I’m forced to rebuild it from scratch.  As I ponder the inputs for the model this time, I’m spending time familiarizing myself with that market again.

So what do I do?  Well I pull up a chart of the AUDUSD currency pair and slap a Fibonacci Retracement tool on it.  Its amazing how many traders use this tool because prices tend to “retrace” to certain price levels BECAUSE so many traders use Fibonacci.  Self full-filling prophecy I guess.

Anyway, I’d be playing a breakout of 0.7461 in this currency pair, if it can ever get there.

(click to enlarge)

I’m so looking forward to start trading my favorite market of all time again, FOREX!  I have my many neural net models to update for sure but this time I’m going to test out some automated systems, aka robots, as well.

I’ll probably go back to trading the London market open using a breakout strategy if its still viable.  I like that strategy alot because I’m usually asleep when the trade is entered! LOL!

As I wrote in my Feb 2009 post,  “Where’s the Market Bottom,” I believe the real estate and equity markets have more downside coming.  Now someone’s IYR model , is setting up for a possible test of its March lows. Hehehe.

I built a fast IYR model myself using Stock Neuromaster to see if I can match his “box” model.  My model show a short signal generated on 4/1/09 for 4/2′s open, which remains short this morning.

Overall I’m hoping that we see a test of the equity market bottom and maybe even a break of it, I want to go balls to wall long with the remaining 401k monies have I have. Only time will tell what happens.

I’ve been working on updating my various stock and forex models, both in Rapidminer and other platforms, and just built a test model for EWS using Stock Neuromaster. It seemed that it had some false signals at the most recent bottom but managed to catch a small leg up. I wonder if it’ll go long again or stay out of the market as shown!

Remember, the signals tell you to enter or exit at the next day’s open from where they are generated.

I recently downloaded the new version of TraderXL and was surprised to see a major update to the ClassifierXL module (as part of the NeuroXL suite). I’ve used this module before to classify like groups of stocks and identify (per my requirements) the right stock to buy out of a group of many. Major updates to the module include a better GUI interface and the inclusion of five neural net functions, namely the Threshold, Hyperbolic Tangent, Zero-based Log-sigmoid, Log-sigmoid and Bipolar Sigmoid functions. To see what it can do, I’m attaching a recently classified ADR stock scan spreadsheet from www.aaii.com.I downloaded this scan from AAII, used the zero-based log-sigmoid scan, and classified the stocks into 5 similar groupings.After it crunched the data it created two charts and a color coded spreadsheet from your data.If you flip to the charts in the spreadsheet, you’ll notice that cluster 1 and 5 have large groupings of similar stocks.These clusters represent the most interesting of the stock groups and should clue in the data modeler to some possible opportunities in the data. Let’s say you are interested in investing in a China based company and you have lots of data from a stock scan to go through. How can you identify a good candidate for more due diligence? First open the spreadsheet and then using the pull down data sorting menus to select China as your country of choice. The data in the spreadsheet will sort and show 7 China based stocks, with 5 being in Cluster 1 and 2 being in Cluster 5. Now this is interesting data revelation to me because not all of these 7 China based stocks are being classified as the same. If you further drill down the data by selecting the Top 10 EPS Growth Estimate, then you are left with 4 China based stocks in Cluster 1: LFC, JOBS, BIDU, and MR. These 4 companies should give you a good smaller list of stocks for further review. Granted, this example was a fast way of doing a complex data analysis but the ClassifierXL module helped simplify the process. The neat thing about this module is that it does all the heavy lifting for you and organizes the data in an easy to use spreadsheet!

I’ve used EasyNN plus in the past and I still use it today. I think it’s a great product for a lot of general neural net applications and the occasional stock fundamental analysis model. I use EasyNN plus primarily for Yahoo Stock Fundamental Data analysis and building data models that help me select the “right stock poised for growth”. I know I can do that in Rapidminer but I find it quicker to use EasyNN plus because of the screen outputs that let me identify stock sensitivity in a blink of an eye (more on this later).

If you’re scared of Rapidminer, want to do more with all kinds of data, and don’t want to spend a fortune, EasyNN plus is the product for you!

What I like about it:

EasyNN plus is a lot easier to use than Rapidminer but a little bit harder to use than Stock Neuromaster. While Stock Neuromaster is primarily geared toward stocks, EasyNN Plus can be applied just about anything. Overall its best features are its output screens. My favorites have to be the Sensitivity, Relative Importance, and Training Error screens. From those screens I can figure out what variables in my input data are very sensitive to change (Sensitivity), which variables are driving the outputs (Relative Importance), and which rows of data might contain bad information (Training Error).

The product lets you load in CSV and Excel files, checks your data for any errors, allows you build macros and scripts, allows you to forecast, and lets you do a lot more tweaking with your data than other programs.

What I don’t like about it:

As I mentioned above, it’s harder to use than other programs so you’ll have to have a basic understanding about neural networks and how they work. You’ll have to know the basics such as validation, momentum, lag, and how to set up your data into input and output variables. Loading in the data is easy but prepping it for your model takes more time than I originally expected, and there is a slight learning curve.

Bottomline:

Despite these minor shortcomings, I use EasyNN plus in my routine stock data modeling and will continue to do so.  It’s a great little contained neural net building system for a fraction of the price of other products. Granted it’s not free but they have a 30 day trail with plenty of samples for you to check out and make that determination yourself.  I highly recommend this product overall and its perfect for the user who has a basic understanding of neural nets but wants to dabble more in the addicting world of financial modeling.

You can download the 30 day trial executable file here: EasyNN Plus

My weekly S&p500 timing model is still showing elevated volatility levels but a major easing off the insane levels from last year.  While this is a major improvement, we are still not out of the woods.  The real bottom isn’t here yet and I expect a little more pain before everyone throws up their hands and capitulates.

Overall the model remains in BUY mode from 11/14/08 and I’ve been adding more money to my 401k and IRA’s as a result.  I have one last slug of $$$ left to drop into the markets when this last “pain event” occurs.  I also added some shares of GE, JNJ, and TX to my long term holdings.  I will looking to add more stocks (domestic and foreign) across the board this year and I’m busy using my AAII stock scans to find those canditates!

I entered a small position in EWH yesterday after my newly updated neural net model said “go short young man.”  So I went short @ $10.95 and currently sit on a small loss.  Based on how the index futures are behaving this morning, my stop might just get triggered. We’ll see.

Update: Sure enough I was stopped out at the open.  Look at EWH go. :)

Now that the new year is here and we went through the greatest meltdown since the Great Depression, what shall I do?  Well, max out my 401k for sure and update my neural net models!!!  I’ll be busy working on them but I’ll probably starting moving toward more genetic algorithmic models this year as they tend to adapt better to the shifting market place.

My market timing model seems to be holding up but I’m going to review it to make sure its still giving me the right signals.  The volatility indicator is showing a major pull back in market volatilty and if it continues to fall, I’ll say we’ve made a bottom.

On the flipside, Obama’s Infrastructure Stimulus plan seems to be going ahead as all my clients are scrambling (compressing schedules) to make the projects I’m working on meet the stimulus schedule.  That means I’ll be super busy from now till April.  If the stimulus really comes on full force, then I’ll be busy for years.

I believe that we are fast approaching a bottom based on my market timing model and simulations.  I know I’ve said that before but these levels of panic can’t sustain themselves for long without some sort of capitulation.  Below is a superimposed weekly chart (from 1990 through last week) of the my market timing indicator relative to the S&P500.

It’s a pretty good indication that some sort of top (not necessarily thee top) was made if a rapid rate of change in the indicator occurs while simultaneously making a new high.  Usually at these times I like to skim some profits and put them in bonds or cash.

Likewise, if the indicator experiences a massive rate of change while simultaneously making a new low, you have a pretty good idea that you’ve made a new bottom of sorts.  The timing model can’t tell me if I’ve made the BIG top or bottom, so I have to rely on other models to tell what that probability is and gauge how much money to take out or put into the markets.

Granted, we are seeing things in this market that are unprecedented but my model seems to be coping farily well.  The scary thing is I just watch my indicator blow through the 3.0, then 4.0, and now approach the 5.0 level without blinking.  These markets are in a free fall and I’m feeling the pain!

So how far will she go?

A hastily drawn yellow trend line on the chart shows the S&P500 finding support around the 1000 level first and the model issued a major BUY signal at the level.  Unfortunately we’ve already blown through 1000 so the next level of support, from inspecting the previous BIG bottom in 2003 is around the 800 level.  Then using my Genetic Algorithm model and Monte Carlo simulations, I come up with my most probable BIG bottom.  That’s the 700 level, more than 50% off the all time market high.

Could it go lower than 700, you bet!  But I’m willing to dump more money into my 401k at that level and then hold on tight.