[You'll find me at http://tradingtheodds.wordpress.com/ as well]
On Friday's session the market recovered from a (modest) morning sell off to finish with solid gains for a 4th up-day in a row. But it is not only the fact that the market closed up 4 consecutive sessions, Friday's session was remarkable with respect to several other events and setups triggered as well:
- The ratio of NYSE Advancing Issues / Declining Issues closed above 2 the 4thday in a row which is the second occurrence only since 02/01/1990 (last occurrence on 11/28/2008). Even 3 consecutive sessions with a ratio of NYSE Advancing Issues / Declining Issues above 2 are rare events (9 occurrences since 02/01/1990).
- The running 4-day summation of the Arms Index (TRIN, ratio of [Advancing Issues/Declining Issues]/[Advancing Volume/Declining Volume], adding up the last 4 sessions) closed below 3, declining the 4th days in a row. Since 02/01/1990 there were only 208 occurrences when the 4-day summation closed below 3 (on average 10 occurrences only per year). If more volume goes into advancing issues than into declining issues , the Arms Index (OTCQX:TRIN) closes below 1.0, and if more volume goes into declining stocks than advancing stocks the Arms Index (OTCQX:TRIN) closes above 1.0. It is therefore a proxy for supply and demand and the health (confirmation or non-confirmation) of a market advance. So a 4-day summation of the Arms Index below 3, declining 4 consecutive days, with a NYSE Advancing Issues / Declining Issues ratio above 2 4 days in a row means that there was a more than usual strong underlying bid in the market (to say the least).
- Wilder's True Range closed at 1.89% on Friday's session, for the lowest True Range (and 'quietest', least volatile session) of the last 7 weeks (since 02/13/2009).
- Although the S&P 500 closed higher the 4th day in a row and posted three higher highs on the previous 3 sessions, it did not post a higher high than Thursday's high on the 4th consecutive up day.
- SPY volume was especially light and came in at the lowest level of the last 8 weeks (since 02/09/2007).
- The S&P 500 underperformed the S&P 500 Equal Weighted Index the 4th day in a row (highest capitalized stocks, regularly those being part of core portfolios of institutional investors, underperformed the less capitalized stocks).
- The VIX (markets expectation of 30-day -calendar days- forward volatility) closed more than -7.00% below the 21-day (trading days, equals 30 calendar days) realized (historical) volatility, reflecting some kind of complacency among traders which regularly had short-term negative implications in the past.
The following table shows the SPY‘ behavior and the respective performance of those 17 occurrences/trades since 10/01/2007 on the session immediately following 4 consecutive up-days in the SPY (day 5):
|No.||Date||open||high||low||close||close - open|
(’open’, high’, ‘low’ and ‘close’: percentage change in comparison to the last session’s close; ‘close - open' speaks for itself: any positive percentage change means a close above the open and vice versa)
The following table shows -percentage wise- the historical odds (which do not represent the true chances that the event will occur, but the amounts that -on average- will be payed out on winning bets/trades in order to determine the expectancy and favorable opportunities) for the next session's outcome concerning the magnitude of change of the next session's close compared to the previous session's close (assumed one would have taken the /trade on the last session's close) as well as the magnitude of change concerning the next sessions close compared to the respective opening quotation (if one would have taken the trade on the next session's open, in the event the direction of the opening provides any tradable edge).
- 1st column: at-any-time odds (taking into account every single trading day),
- 2nd column: odds for the following session after the S&P 500 posted a higher close one session only (’w/Survey I‘),
- 3rd column: odds for the following session after the S&P 500 closed higher on 4 consecutive sessions (’w/SurveyAll‘).
All setups are broken down by 'up' (odds if the following session opens higher), 'down' (odds if the following session opens lower) and 'on average' which doesn't take into account the direction of the opening at all (average of all at-any-time sessions and sessions after the setup has been triggered respectively). Odds significantly above or significantly below their respective at-any-time odds (in this case +/-50.00% -which is a really challenging bar- to overcome, but this percentage is up to everyone’s decision what may be regarded as ’significant above’ or ‘below’) are marked by a green (for a probable bullish outcome) and red (for a probable bearish outcome) background color. This should make it possible to catch on a glimpse if (any), where (e.g. on the open, for a higher/lower close or intraday strength/weakness) and to what extent (compared to historical odds) the following session (Monday, April 6, 2009) possibly provides a tradable edge.
(click on link)
('EOD': percentage wise end-of-day change, close x days later compared to the close when the signal was triggered; ’C-O’: close - open, percentage wise intraday change)
Bottom line (short-term only):
- concerning w/Survey I (the session following an any up-day in the market), and due to the recently short-term mean reversion character of the market, the odds are favoring the short side, but only to a minor extent. The SPY closes -on average- down -0.38%, and -on average- -0.26% below the opening, while the at-any-time odds would be -0.13% and -0.09% respectively. And in the event the market opens lower, the odds concerning any potential down side potential after the open are -on average- with a +1.18% gain on profitable trades (short on open) and -1.02% loss on loosing trades higher than the respective at-any-time odds, and
- concerning w/SurveyAll (the session following 4 consecutive up-days in the market) all odds are favoring the short side of the market (but remember: not the true chances, but the payout only). Going short on the close of the 4th consecutive higher close and/or the next session's open would be much more profitable than going long respectively, and to a significant extent (and the "significance" level is set to at least 50% above the respective at-any-time odds).
What what about the potential consequences of the other setups (the ratio of NYSE Advancing Issues / Declining Issue, the TRIN, among others).
Among others, Wilder's True Range and the current gap between implied (VIX) versus realized volatility show negative implications concerning Monday's session as well, but breadth tells -not concerning Monday's session but for the medium term- a different story.
From the breadth related setups triggered on Friday's session, for further investigations I choose the TRIN because it has the best possible (highest) sample size in order to increase it's statistical validity.
Since 02/01/2000, there were 62 occurrences when the running 4-day summation of the Arms Index (TRIN, ratio of [Advancing Issues/Declining Issues]/[Advancing Volume/Declining Volume], adding up the last 4 sessions) closed below 3. The following table shows -percentage wise- the historical odds for the outcome over the following two weeks (10 sessions) concerning the magnitude of change of the close 10 session's later compared to the previous session's close (assumed one would have taken the wager/trade on the last session's close).
- 1st column: at-any-time odds (taking into account every single trading day over the course of the following 10 sessions),
- 2nd column: odds over the course of the following 10 sessions after the running 4-day summation of the Arms Index closed below 3 (’w/Survey All‘).
(click on link)
Bottom line (medium-term):
- Over the course of the following 2 weeks, and after the running 4-day summation of the Arms Index (TRIN) has closed below 3, the market showed a significant edge on the bullish side (significance defined at 25% above the respective at-any-time odds). Average magnitudes of changes and the sum of all gains of profitable long trades are -over the course of the following 2 weeks- both significantly higher than the respective at-any-time odds, and the sum of all losses (of unprofitable long trades) are lower than the respective at-any-time odds (for a significantly above-average profit factor).
So -from a historical and statistical perspective- probabilities and odds regarding those setups listed at top of my post are indicating a probable short-term pullback early next week, but we may see another (or a continuation of the) run-up in the markets after the short-term pullback has run it's course, especially with respect to the strong breadth we've experienced during the last couple of sessions (and which might not disappear without showing it's strong hand again any time soon).
Disclosure: Long BGZ (Direxion Large Cap Bear 3x Shares, ETF) at time of writing.