The Day Of Week effect, does it still work

Discussion in 'Strategy Building' started by Murray Ruggiero, Aug 21, 2007.

Do you believe in day of week bias in the S&P500 ?

  1. No it's a myth

    42 vote(s)
    61.8%
  2. Yes, but only Monday being a up day

    4 vote(s)
    5.9%
  3. Yes but only Monday and Friday are reliable

    10 vote(s)
    14.7%
  4. Totally, I use each days effects in all my S&P500 patterns

    12 vote(s)
    17.6%
  1. does your work adjust for Tuesdays often behaving like a Monday, after holidays?

    I would suggest stripping out entire weeks where there is a Monday holiday, and seeing how M-F behave without this effect.

    Do these T, W, R, F behave the same as T, W, R, F on 5-day weeks???
     
    #31     Aug 24, 2007
  2. Good and hopefully Murray finds answers to his questions he presented to ET.

    Also, I'm glad to see you do have an interest in a different aspect of technical analysis besides indicators. :cool:

    This is one of my favorite type of TA research and application.

    Re-reading not necessary because I don't have a problem with that statement as long as you understand that Murray is asking questions...

    I'm sure answers will be welcomed without the I know all and you know little type attitude.

    I'll post a more direct question as the thread evolves concerning trade management of the tendency and possible alternatives.

    Trade management is where my extensive research occurs at.

    Once again, glad to have you aboard a different area of TA.

    Mark
     
    #32     Aug 24, 2007

  3. Well this Friday was up. Will see what Monday will bring:confused: :confused:
     
    #33     Aug 25, 2007
  4. I'm strongly interested into what key market events occurred in late 2001 to change things for Wednesday.

    I'm also curious about 1998 key market events that changed things for Tuesday.

    Knowing such can be useful.

    Mark
     
    #34     Aug 25, 2007
  5. Murray Ruggiero

    Murray Ruggiero Sponsor

    These are good question but if I did that , you would tell me I am just crunching numbers and curve fitting. You are correct these are good points. I did these studies in the 1990's and some Tuesdays after holiday mondays act like Mondays and other did not. The number of cases was too small for me to think it was reliable.
     
    #35     Aug 25, 2007
  6. Murray Ruggiero

    Murray Ruggiero Sponsor

    I will get to it and post it next week.
     
    #36     Aug 25, 2007
  7. Murray Ruggiero

    Murray Ruggiero Sponsor

    I did not say it was the only problem but it was a major one.
    Another includes, what if Friday is a holiday ?. Also , the markets being close because of 9/11 create a problem. The market was closed on Tuesday and then reopened the monday after 9/11 the 17th.
    Finally , Your last problem was Ronald Reagan Death which closed the market on the Friday June 11th. So technically it was not a normal holiday so, that why this might be a bit of a trick question.

    Here is another problem, I don't remember System Writer which was what Joe originally coded it in. Did it close that last trade if that last bar was a monday, or did closing it require another bar of data ?.
     
    #37     Aug 25, 2007
  8. Once again, glad to have you aboard a different area of TA.

    Mark


    To me, seasonals are not automatically qualified as "TA".

    I consider this as fundamental analysis. I dont care for unsupported seasonals that say "hey, the price for the DAX has moved higher from August X to September Y for 13 of the last 15 years." That is what I discussed in that post. Many times, these are probability flukes, based on testing large numbers of instruments over every permutation of date ranges. Many of these will look good, but not hold up in eal trading.

    However, it is a known that energy prices move at certain times of year, based on temperature changes, summer driving habits, etc. Grain prices move seasonally based on harvests, etc. etc.

    Of course, the experts know this, and options, contracts, etc. tend to reflect a lot of this knowledge, rendering it difficult to capture profits many years...




    [/B][/QUOTE]
     
    #38     Aug 25, 2007
  9. I am simply asking whether T-F on short weeks look like T-F on long weeks... If T looks like M, that is informative...

    Personally, I think backtesting with smaller samples require to move towards a median model. Large days can throw the numbers off. Or at least, consider throwing out a couple of highs and lows for each day...
     
    #39     Aug 25, 2007
  10. Yeah, it really depends on the type of market seasonal tendency (I use both fundamental and technical).

    Also, even the fundamental tendencies can traverse into technical tendencies once you get into alternative methods of trading the tendency.

    Simply, in my opinion, after all the research and analysis...

    It comes down to how to trade it.

    If not, why bother with studying it in the first place.

    Anyways, that's why I want to research some of the changes that occurred on the prior chart that was posted.

    I want to see if the tendency change due to fundamentals or technical reasons...possibly both. :cool:

    Good night all.

    Mark
     
    #40     Aug 25, 2007