Daily Commentary - Posted on Sunday, March 28, 2010, 10:05 PM GMT +1
Setups, Probabilities and Odds
I’m just playing around with / testing a mechanism on how to quantify the next session’s probabilities and odds for a higher / lower close on the then following session not only based on and triggered by a single setup on the previous session’s close, but by a combination of (all of) those setups which – from a historical and statistical perspective – show probabilities and odds significantly above / below the respective at-any-time probabilities for a higher / lower close on the next session, in the first step based on price (open|high|low|close|first hour|last hour) alone (not taking into account market breadth).
The improvement is that in the past I manually checked for all potential setups triggered on close of a session, and analyzed up to a maximum of 5, but was not able to automatically identify and check for all (combinations) of those setups being triggered on close of a session.
Table I below not only shows the list of setups triggered on close of Friday, March 26, but the current number of consecutive sessions with a higher (+) / lower (-) price / daily return as well.
How to read Table I:
On Friday, March 26, the SPY
Index 2: opened higher (above the previous session’s close) on the second (+2) consecutive session,
Index 13: closed below the open on the third (-3) consecutive session,
Index 29: was trading below the previous session’s close at the start of the final hour of the session now for a first time (-1),
Index 30: the intraday range contracted for the second consecutive session (-2), and
Index 31: closed belowthe midpoint of the session on the second consecutive session (-2).
I then checked for the respective historical probabilities and odds (since 01/01/1990) whenever the SPY showed an identical combination of any two of those setups listed above, e.g. the probabilites and odds after posting a higher close a first time (Index 5, +1) in combination with trading below the previous session’s close at the start of the final hour of the session for the first time (Index 29, -1).
Table II below shows the SPY‘s historical (since 01/01/1990) performance (number of occurrences, the combination of setups and streaks, percentage of winning trades, profit factor, median return, geometric growth rate per trade, maximum gain and maximum loss) on the then following session (next day returns) immediately following a session where the respective combination of setups and streaks had been triggered in the past (Table II itemizes only those setups where the probability of a winning trade exceeds 66.67% or undercuts 33.34%, and the sample size (number of occurrences) is at least a double-digit number (>= 10)):
It is at least interesting to note that with respect to those setups being triggered on close of Friday, March 26,
- the raw number of positive setups (16) exceeds the raw number of negative setups (8) triggered by a factor of 2,
- the simpliest combination of setups – 5 (1) | 29 (-1) the SPY closed higher for a first time while the SPY was still trading below the previous session’s close at the start of the final hour for a first time as well – is indicating a negative outcome on Monday’s session: 10 higher and 22 lower closes, with a (negative) profit factor of 0.30 and a median return of -0.55%,
- almost all other negative setups triggered are only based on the fact that the SPY hasn’t left an unfilled opening gap down (Index 2: a high above the previous session’s close for 21 sessions) for a month now, while the list of positive setups (showing a higher sample size as well) triggered is widespread, so from a statistical point of view the overall bias for Monday’s session might be positive.
I’m quite sure that running the stats listed above day by day in the future, we’ll get some very interesting outliers, means where historical probabilites and odds either with respect to a single combination of setups (I manually checked for up to 5 potential setups triggered in the past) and/or with respect to the overall sum of combination of setups triggered are heavily skewed in one or the other direction.
The next step will be to incorporate market breadth and check for combinations of price / breadth based setups as well.
to be continued …
Disclaimer: No position in the securities mentioned in this post at time of writing.
The information on this site is provided for statistical and informational purposes only. Nothing herein should be interpreted or regarded as personalized investment advice or to state or imply that past results are an indication of future performance. The author of this website is not a licensed financial advisor and will not accept liability for any loss or damage, including without limitation to, any loss of profit, which may arise directly or indirectly from use of or reliance on the content of this website(s). Under no circumstances does this information represent an advice or recommendation to buy, sell or hold any security.