Masterthegap.com has inspired me to create a program which will do just that… Master the Gap. I’ll be posting the results from 1000 days of intra-day data this week. From the preliminary results. It appears that gaps have different probabilities of filling given the day of the week (Monday, Wednesday, etc), the day of the month (1,2,3,4,etc), the month itself, and also the time at which it fills. Interesting stuff…
In my last post I talked about a program that would find a stock that would lead or lag a stock you selected. Lets take $AAPL for example. Below are the top 5 positive and negatively related intra-day stocks for Apple Computers calculated on 9/18/2010 with 5 minute bars.
However, there is a special attribute to this information. $AAPL time1 is not correlated with $DTV time1 then squared (that is how you calculate R^2. Read about it here). It is correlated with time3 of $DTV. The illistration below should help. (Please note the prices below are for examples purposes only)
On the left is the traditional stock by stock correlation. On the right is a leaders/laggers correlation. What this does is allow you to have an understanding of what $DTV will do with greater confidence than if it was correlated with $AAPL time1. The reason is this: When a stock is trending that trend can change suddenly and if you’re pairs trading or using correlated stocks it will change immediately with it. When you lead/lag the R^2 tells you that while $APPL is rising 10 mins later $DTV should be rising as well (within a degree of confidence of course). Also if $APPL is reaching a bottom $DTV should reach a bottom 10 minutes later. The chart below should illustrate this concept
As you can see as $AAPL hits a bottom at $243.62 at 11:45am $DTV hits it’s bottom at 37.70 11 minutes later at 11:56. This is the power of lead/lag. There are many strategies one can implement knowing this information.
Hope you enjoyed this post
Some time ago I began coding a program that would iterate through a list of stocks (that list being the Standard & Poors 500 (S&P500)). I use intra-day 5 minute bars going 20 days back. The information tells me how to best hedge my current intra-day positions. For example lets say $AAPL has a negative correlation with stock XYZ with an R^2 above .90. When holding $AAPL for longer periods of time I know stock XYZ will follow suit in the opposite direction. If $AAPL spikes against me and my stop lose is hit, my hedge would have covered the lose from $AAPL. I will be posting a more detailed example as well as data this up coming week.
Sayeed, Dr. Gorman and I had a successful presentation on March 5th at the Eastern Psychological Association 2010 Conference presenting “An Exploratory Qualitative Analysis of the 2008 Presidential Campaign.” You can read about my current research, including the work on Presidential Leadership on the research section of my website. Below is the short abstract along with Scribd version of our paper. In addition, all graphs are included for your viewing pleasure
An Exploratory Qualitative Analysis of the 2008 Presidential Campaign.
Short Abstract: A content analysis using Hart’s DICTON program was performed on the 2008 Obama vs. McCain presidential campaign speeches. It was found that the content of the speeches varied over time on the DICTION dimensions of certainty, activity, optimism, realism, and commonality. Obama consistently demonstrated higher levels of communality throughout the campaign. Implications for dynamic, time series content analyses are discussed.
Scribd link to our paper entitled: An Exploratory Qualitative Analysis of the 2008 Presidential Campaign.
Where are we taking this research next? We are exploring machine learning techniques to evaluate factors of leadership.
Have you ever witnessed or read about a family who repeats mistakes generation after generation? The unfortunate daughter whose father is an unemployed alcoholic who consequently finds herself in the same position later in life. The “great” buy & never EVER sell strategy that your grandfather used, then your father, now you? This is what I have defined as Micro-Evolution.
define: Micro-Evolution – Short term survival strategies that are learned unconsciously from childhood.
In typical evolution a mutation in the organism (psychological or physical) creates an attribute that will allow the organism to proliferate or suffer because of that mutation. If this mutated characteristic is more advantageous than his predecessor’s attributes, this new breed will pro-create in larger numbers rendering it the dominate one. This is different from micro-evolution in that it is not a mutation, but an unconsciously learned survival method. Regardless of how successful one can be with a different strategy, your predecessor used it and it worked. Your unconscious learned that it worked and instilled mechanisms to assure it would be used.
How do change your Micro-Evolutionary destiny in Trading?
By learning new habits through practice.
ThinkorSwim just integrated a new “on demand” feature. This feature replays the market live from a certain date forward. This would allow you to try new and different strategies. You can watch a video on it here.
Furthermore, (since my idea is now in a major platform, which I had nothing to do with) you can write a simple AutoIt Script to click the chart along allowing you to practice 24/7. I discuss this in my youtube video here: Expertise in Day Trading.
Hello! I want to welcome you to my new blog. Here I will post information regarding psychology, statistics, the stock market & their ever interchanging relation to one another. You can visit my old blog here for older posts from Andrew Ilardi. In addition, there are additional pages to peak into the other areas of my work. At the top you’ll notice 5 tabs:
- Home: The Main Blogging Area.
- About: Information regarding my background.
- Research: Current research I’m working on at the academic level.
- Services: Information on what I provide and contact.
- Capital Markets: Posts regarding daily research I do into the markets.