PureBytes Links
Trading Reference Links
|
I am no statistician but can 10 occurrences in 35 years be classed as
statistically significant? I know this sounds like a rhetorical question
but it is not. I am just wondering how many occurrences would be required
for it to be useful if in fact the answer to my question is no.
regards
Philip
-----Original Message-----
From: Timothy Morge <tmorge@xxxxxxxxxxxxxxx>
To: Dtrader <dan@xxxxxxxxxxxxxxxxx>
Cc: omega-list@xxxxxxxxxx <omega-list@xxxxxxxxxx>
Date: 14 December 1998 17:51
Subject: Re: copper
|Dan:
|
|That is an astounding statistic--does it give you parameters
|for how long it takes the rebound to unfold?
|
|Best,
|
|Tim Morge
|
|Dtrader wrote:
|>
|> (i saw this on the lim site today)
|>
|> Q: What happens when copper falls more than 2% on a Friday in December?
|> A: It turns out that Copper rebounds an astounding 6% on average in 8 of
10
|> past occurrences.
|>
|> Date Day 1
|>
|> 12/10/1965 Fri 5.461
|> 12/02/1966 Fri 7.996
|> 12/14/1973 Fri 6.919
|> 12/18/1981 Fri 1.180
|> 12/21/1984 Fri 3.005
|> 12/04/1987 Fri 15.450
|> 12/02/1988 Fri 8.455
|> 12/01/1989 Fri -2.908
|> 12/07/1990 Fri 0.953
|> 12/15/1995 Fri -2.554
|> 12/11/1998 Fri NaN
|>
|> Avg 4.396
|> AvgPos 6.178
|> AvgNeg -2.731
|> PctPos 80.000
|> PctNeg 20.000
|> Maximum 15.450
|> Minimum -2.908
|> StdDev 5.626
|> ZStat 0.781
|> Variance 31.657
|>
|> 11 Occurrences
|
|