$SPY and 20 day high/low cycles
this is a similar study done $IWM , few years ago
this study is on $SPY sequences of 20 day low closings and 20 day high closings.
If a current close is a 20 day low coming after a 20 day high closing, calling it as the 1st 20 day low closing (marked as 1, under #), and if another 20 day low closing is printed, then call it as a second 20 day low (marked as 2 under #) after a 20 day is already printed, and so on. The sequence is counted till a new 20 day high is printed.
Currently we printed a 1st 20 day closing as on Mon close on $SPY.
since 1994 , there were 491 , 20 day low closes , of which only 21 , stopped at the first 20 day low close , before regaining a 20 day high close.
since 1994 , there were 102 such cycles of 20 day low close to 20 day high close and then to 20 day low close , of which , one cycle continued as long as 29 consecutive 20 day low closes ( from 3 Apr 2002 , with $SPY at 82.9 , till the final 20 day low close of 58.78 as on 23 jul 2002 , before regaining to do 20 day high close of 67.81 on 14 Aug 2002 )
below the last two 20-day high /low cycles on $SPY , for your understanding of the concept
# is the 1st 20 day high or 20 day low cycle , and as you can the last 20 day high cycle of consecutive 20 day highs , that started on 11th Sep 2017 till 26th Jan 2018 printed 46 , 20 day high’s before hitting a 20 day low close . a prior 20 day low close cycle that started on 10 Aug 2017 printed 3 consecutive ( not consecutive days ! ) 20 day low closes , till 18th Aug 2017 , before regaining a 20 day high close .
caution no one knows how many consecutive 20 day low’s or for that matter how many 20 day high’s will be printed in a cycle till it lasts . for ex: if you think we are over done in printing 46, 20 day closing high’s , no !! we have over done even more than that on these , 3 prior times
05-Feb-18 ( 46) , 10-Mar-11 ( 50) , 27-Feb-07 ( 51) , 10-Oct-95 ( 74)
below is the distribution of the 20 day low close sequences ( not consecutive days , it is a sequence of 20 day low close’s before we start numbering 20 day high close’s again )
|sequence||#||% age stopped||% age not stopped|
sequence : how many 20 day low closes were hit in this cycle before reclaiming a 20 day high close
# : no such 20 day low cycles out of 102 , where the sequence is stopped at
% age stopped : % age of 20 day low close sequences that stopped at that sequence number , for ex : there were 21 of 102 , 20.6% on whole stopped just by printing 1 20 day low close , before reclaiming a 20 day high close ,
% age not stopped : it is the reverse of the % age stopped
conclusion : 79% of the 20 day low cycles , continue hitting at least one more 20 day low close , before reclaiming 20 day high close !! which basically warrants caution for all the bottom ( on close basis ) callers ! may be at a 3rd 20 day low close , where i’d have 50% confidence in calling the bottom close , better at 5th 20 day low close , where i have 70% confidence !!
write to us or comment to use , if you didn’t understand the analysis !!
you might want to consider subscribing to our daily quant rants email !!