Jan

3

spring is nearI found this post on Trading The Odds.

He analyses 3 set-ups on S&P 500 with respect to a holiday: assumed one would’ve bought the S&P 500 on the close two sessions before the Holiday (setup 1), on the close of the session immediately preceding the holiday (setup 2), and on the close of the session immediately following the Holiday (setup 3).

His results:

1. New Year’s Day (celebrated on January 1)

The model will take a long position on close of the first session of a new year.

2. Martin Luther King Jr. Day

The model will not take into account the Martin Luther King Jr. Day exchange holiday.

3. Washington’s Birthday/Presidents’ Day

The model will not take a long position on close of the the session(s) immediately preceding Washington’s Birthday/Presidents’ Day

4. Good Friday

The model will take a long position on the close two sessions before the Good Friday exchange holiday.

5. Memorial Day

The model will not take into account the Memorial Day exchange holiday

6. Independence Day (celebrated on July 4)

The model will not take into account the Independence Day exchange holiday

7. Labor Day (celebrated on the 1st Monday in September)

The model will not take into account the Labor Day exchange holiday

8. Thanksgiving Day (celebrated on the 4th Thursday in November)

The model will take a long position on the close two sessions before and on the close of the session immediately preceding the Thanksgiving Day exchange holiday, and no long position the session immediately following the Thanksgiving Day exchange holiday, at least if not any other criteria are met.

9. Christmas Day (celebrated on December 25)

The model will always take a long position on the close, between two session before until two session after the Christmas Day exchange holiday.

In a previous post :

1. The (calendar) Day of the Month

The model will take a long position on calendar day 30 or day 31, depending on which one represents the last business day/session of a month

2. The Day of the Week

 The model will not take into account the day of the week.

3. FOMC Meetings (scheduled) and FOMC announcement sessions

 The model will take a long position on close of a pre-FOMC announcement session as well as on close of an FOMC announcement session if the index doesn’t close up greater than +0.50%.

4. Government’s Job Report Friday

The model will take a long position on close of a session immediately preceding the Government’s Job Report Friday in the event the index closed up.

5. Option Expiration

The model will not take into account the option expiration.

This is not complete and may not add much to what you do, however, I find it interesting that counting is done by many out there. The problem is: what is the right way to count?

Victor Niederhoffer comments:

There would seem to be many problem of multiple lookback, multiple comparison, Monday morning quarterbacking, selective retrospection, and consistencies with randomness with such an approach. Most important of all, they don't take into account the diference between likelihood (which is very high) and predictability. The seasonarian left our site because of "e" and is not here anymore, but his ghost lives on in this interesting post, designed presumably to elicit this comment.


Comments

Name

Email

Website

Speak your mind

Archives

Resources & Links

Search