How Day purchased and Week of Month affects option trading sucess

The Bayes Analytic engine is designed to predict probability of success for a given offer to meet a goal.   In some instances the goal could be will a particular customer buy from a specific offer.   I originally designed the stock prediction Engine to answer the question of which recent trades have the highest probability of meeting a specific goal such as increasing in value by 23% over a 7 day period.   The idea was that if we treat every option trade like an offer if we could isolate the set of offers which provided the highest probability of reaching the goal then we could buy those and skip the trades which offer a lower probability of delivering a profitable trade.

To answer these questions the stock prediction engine builds thousands of statistical measurements which it compares against historical trades and then based on a fundamental assertion that winning trades which have the greatest statistical correlation to the current trade can be used to help filter the existing trades to identify the best trades.

As part of the statistical learning system it stores millions of individual metrics some of which can be useful even without the rest of the engine.   These statistical metrics change over time and part of the logic built into the engine in the ability to tune contribution from more recent data versus historical data and to compare how the results change across time to identify patterns.

Statistical Excerpts Option Purchase Day

Data snapshot as of 6/4/2013.   We had about 6 months of option trading data for SPY at the time of this snapshot.  

Goal: 30% gain in 7 days

  • Spy PUT option –
    • Purchased on Monday has a 66% chance of meeting the goal  but if purchased on a Thursday only had a 52% chance of success.
    • If Purchased on the first week of the month is much less likely to succeed than those purchased in the 3rd or 5th week of the month.   Week-1=14%,  Week-2=38%,  Week-3=73%,  Week-4=65.64%, Week-5=86%.   Based on this if you want to play the SPY short then you may be best served working the 3rd and 5th week of the month and going fishing for the first two weeks.
  • Spy CALL option
    • Purchased on Thursday had a 80% probability of success while a purchase on Wednesday has a 66% chance of success.
    • A call option purchased in 5th week of the month have a a very low success rate.     Week-1 = 74%,  Week-2=87%,  Week-3=81%,  Week-4=74%, Week-5=33%.     You  will obviously maximize your chances of success if you purchase in the 2nd week of the month.  You are better off going golfing than buying a call option on the 29th, 30th, 31st of any month.

    These are only a small fraction of the total set of computed statistics.  when they are combined they can help identify specific trades that are likely to be profitable.  For example a call trade in the 5th week which happens to be a Thursday on either the 29th, 30th or 31st is a very good statistical bet.  The stock prediction engine combines these automatically but they can also be used manually to help traders tune their activities or they can be used to tune the output of the stock trading engine.

 

Day of Week used for Customer Targeting in Marketing

The same feature would be useful in a marketing or sales context.   A good example of this is the shopper visiting premium retail stores there may be a woman who tends to visit on both Tuesdays and Saturdays but on Tuesdays she is alone and on Saturdays she had her kids.   If looking at this customer the store could determine that an offer for child cloths on Saturday is more likely to be successful while an offer for perfume would be successful.   An interesting facet of the statistical engine is that can make indirect inferences many humans would miss.  For example even if the woman has never purchased a mans wallet, if she has purchased child clothing she has a 62% greater probability of buying a mans wallet than a woman who has never purchased child clothing.

Sophisticated Targeting needed for Wireless communications

In a wireless communication marketing the same kind of metrics can be gathered but the raw data set is broader and more rich.  For example we know when the customer purchased their last phone and when they purchased their last high value accessory.  We could use those to determine if those who are between 1.3 and 2.0 years since their last phone are much more likely to respond favorably to a new phone advertisement than those who purchased their last phone a 3 months ago.    This can be combined to see if they just purchased an expensive battery or charger for their existing phone would respond favorably to an early upgrade when buying the underlying part.    An even more advanced use is that if we had access to quality of service (QOS) metrics from the hand sets the cellular vendor could tell me ahead of time if the device I am looking at ordering will give yield poor performance near my home.    If they have marginal coverage they may have a better chance of keeping me as a loyal customer if they sell me a higher powered 3G device than upgrading me to new 4G device that will not hold a call.  A few moves like this where the vendor may risk a small amount of revenue to ensure I make an informed decision would gain a lot of loyalty and would be a lot less expensive than replacing me with another premium high margin customer.

 

Required Disclaimer

Forex, futures, stock, and options trading is not appropriate for everyone. There is a substantial risk of loss associated with trading these markets. Losses can and will occur. No system or methodology has ever been developed that can guarantee profits or ensure freedom from losses. No representation or implication is being made that using this methodology or system or the information in this letter will generate profits or ensure freedom from losses.  Forex and Option trading can result in losses that exceed the original principal balance.

Hypothetical or simulated performance results have certain limitations. Unlike an actual performance record, simulated results do not represent actual trading. Also, since the trades have not been executed, the results may have under-or-over compensated for the impact, if any, of certain market factors, such as lack of liquidity. Simulated trading programs in general are also subject to the fact that they are designed with the benefit of hindsight. No representation is being made that any account will or is likely to achieve profit or losses similar to those shown.

Even results from Live cash trading can be subject to specific market conditions that may not repeat in the future and as such, duplicate results from future trading is unlikely to duplicate past results.    Changing the dollar amount traded can cause different behavior in live trading markets especially when trading large positons that can exceed the liquidity available in the market and cause changes in pricing behavior.

Bayes Analytic LLC provides software that can produce trading signals.  The customer is responsible for choosing a configuration and parameters for the software that meets their own goals.  The customer is responsible for conducting their own tests and only the customer can activate the software to start trading.   The software runs in an account the customer has logged into and then activated the software.   Bayes Analytic has no control of,  influence over or visibility to the signals specific to given user because we have no visibility into configuration parameters the user has chosen to operate with.    The Bayes Analytic software is provided without Warranty on a As Is, Where is basis.  It is the customers responsibility to test the software to ensure it meets their trading requirements.   Every time Bayes Analytic releases a new version of the software the customer should conduct new tests to validate the new version continues to meet their requirements because every software change could have unexpected side effects that may not be obvious until the customer has tested it in their environment with their configurations.   The Bayes Analytic software may run as a script inside of other software packages or talking to API that Bayes Analytic has no control of or Influence over so the customer should test entire ecosystem to ensure it meets their trading requirements.    Bayes Analytic may provide the software in source form since that is required by some trading systems but it remains the exclusive copyrighted property of Bayes Analytic and may not be reverse engineered or redistributed.    The customer is responsible  for choosing their own broker and installing the Bayes Analytic software so it can trade using the desired account.  Bayes Analytic has no control over or influence of the broker and many brokers have different ways of quoting spreads,  charging commissions,  flow of orders and latency of information.  As such a strategy and software that performs well at one broker may and probably will require changes to perform well at other brokers.  It is the customers responsibility to test the software with their selected broker to ensure it meets their trading requirements.

Leave a Reply