Stock Market Analysis–11-12-2010

November 13, 2010 at 12:31 pm · Filed Under Automated Trader, Broader Market · Comments Off on Stock Market Analysis–11-12-2010 

S&P 500 Daily Chart

The market looks like it’s at least pausing with its first down week in a while.  This past week I‘ve been rather oblivious to the market though, since my free time has been spent working on a new auto trader.  I think I mentioned before that my day trader is hitting its limits on the amounts it can trade, so I currently have a lot of unused buying power in that account.  That means I’m just doing a lot of data mining right now to see if it’s better to create more day trading models or if it would be easier to create some swing trading models. 

A lot of people have asked how I built the automated trading stuff, and it’s actually a lot less complex than they think.  You kids nowadays have it so much easier than when I had to write everything from scratch.  With tools like Weka, it’s really easy to do the data mining and machine learning stuff.  It allows you to build latent variable models without knowing any math, and makes it easy to determine things like if the stochastics, RSI, and ADX have any meaning when you combine them.  Depending on your broker, the ease in automating the actual trading will vary, but when you first start you can use a human in the loop to execute the trades anyway.

Well, enough babbling about boring stuff.  I took a lot of profits this week and am in wait and see mode now.  I have a lot of options plays open, but am essentially delta neutral between them all.  Depending on how next week goes I’ll probably do a lot of hedging to point it in one direction or the other though.

Covered Calls

January 9, 2010 at 12:18 pm · Filed Under Automated Trader, Educational, Personal Finance · 2 Comments 

Unfortunately, all good things must come to an end.  My autotrader is still doing great, but it appears that it’s pushing up on the limits of how much capital it can trade before I move the market too much and slippage starts hurting it.  Without any changes, the return on capital will start to slowly dwindle as the capital goes up, but the return levels off.  So I need to find some new investment ideas.

I remember the first time I read about covered calls, aka a buy-write, I thought it was the stupidest strategy.  Capped upside and unlimited downside don’t seem like a good combination.  Then I learned a little more about position sizing and how to set a constant risk amount so that you don’t just view risk as how much capital you put up to establish the position.  That made the unlimited downside part a little more manageable.

Recently I’ve been reading more about real estate (related to the fact that I need new investment ideas) and getting a much better grip on controlling cash flow.  With a far better grip of expectancy analysis over when I first learned about covered calls, they’re actually starting to seem interesting to me when viewed as cash flow.  After rereading chapter 2 in the McMillan options book as a refresher, I think I may start to dabble in this a little.  I don’t know if I’ll actually follow through on this, but it may add some different types of stocks that I look at in the future.

Auto Tweeter?

April 16, 2009 at 5:24 am · Filed Under Automated Trader · Comments Off on Auto Tweeter? 

I’ve debated how much I want to publicly post from the autotrader for a while.  Since I’m not trying to sell newsletters or trading services, it really doesn’t help me at all to publicize.  If people know your exact system it can hurt you on slippage or even invalidate the system completely (kinda like giving away all your strategies and tells to someone in poker).  However, anyone reading this is probably a friend of mine, so I would be happy if having some insight into what stocks the trader is looking at can help. 

For now it will just post the stocks it triggered on after the market closes.  It will post whether they were long or short (not that it’s hard to determine), but it won’t say whether they were scalps, day trades, swing trades, or position trades, so hopefully you won’t be foolish enough to go buy anything blindly since it may already be well after the trader said to close the position.  Some days there will be nothing, and other days there will be tons of signals.  In order to not flood your google reader with these autogenerated posts I’ll put them up on my twitter feed at and on the right sidebar of my page at the bottom.  Hopefully I didn’t break anything with my 5 a.m. insomniac coding session…

Next Step for the Auto Trader

February 3, 2009 at 1:46 am · Filed Under Automated Trader · 1 Comment 

It’s been a while since I’ve written about the auto trader.  I got complacent with running the same system since last September, so that’s good and bad.  Through the holiday period I decided to just simulate the trades in the lower liquidity conditions, and it seemed to hold up fine (just with reduced activity).  The expectancy seems to be somewhat consistent now, but there are always room for improvements.

In all my eagerness to build the perfect trading system I’ve applied the machine learning to the selection of stocks and timing of trades, however, I skipped over some of the basic trading principles like scaling in and out of positions.  Scaling into positions can be viewed as 2 different trades with different entry criteria, e.g. buy on the breakout and buy on the support can be tested quite easily as 2 separate timing criteria.  I think scaling out is something that must be built into the models though, at least the way my models are organized.  We’ll have to run the numbers and see if it makes a difference.

Automated Trader – 10-30-2008

October 30, 2008 at 10:09 pm · Filed Under Automated Trader · Comments Off on Automated Trader – 10-30-2008 

Three months have passed since the last analysis of the auto trader.  After broadening the universe of stocks to screen, we’ve had a lot more action.  Unfortunately, the recent volatility has made wild swings, which give profits in my system that are nice, but unrealistic on a long term basis.  Scanning through these numbers show things that are double digits above any backtesting data, meaning that my backtesting methods suck (which would be nice) or a VIX over 50 can do funny things to the market (more likely).  For anyone else that’s testing a new system right now, automated or not, don’t trash a system that doesn’t seem to be working right now and don’t get too cocky with a system that is.  We are in unusual market conditions that are as rare as hen’s teeth, and what works today may not work for another 10 years.  We’ll see how it’s going in another 3 months…

Automated Trader – 07-29-2008

July 30, 2008 at 2:24 am · Filed Under Automated Trader · Comments Off on Automated Trader – 07-29-2008 

Over a month has passed since I last checked the exact numbers on the trader.  Maybe that’s because during the down period it basically sat on the sidelines and didn’t open a single real trade over a two week period.  Maybe if I had allowed it to trade on margin, and thus short during the incredible downturn, I would have real profits rather than paper profits, but I can wait.  So I’ve had 13 trades since mid-June and averaged 0.3 times the risk amount in profit per trade.  It was a little more realistic this time with 46% of the trades ending positive after commissions. 

The one thing I’ve learned is that the actual machine learning techniques you use don’t really seem to matter much.  Whether it’s the simplest HMM, LDA, or SVM, the choice only shifts the results small amounts.  Changing the inputs is what really gives you bigger gains.  Of course it’s obvious that garbage in equals garbage out, but I’ve had to be cut down by Occam’s razor more than a few times when I’ve gotten overly complicated with too many fancy indicators and patterns.  I guess the moral of the story is just to keep things simple and understandable and you’ll save yourself a lot of headache.

Next Page »