Thursday, March 14, 2013

There's More Than a 90% Probability of a SPX Correction Before April

That's a Business Insider-type title. Have to grab you.

We have been anticipating strong resistance between 1555 and 1575. SPX is now trading 1% off the top of this range. As Ryan Detrick points out, the past 10-day-in-a-row-advance in the Dow is the weakest since 1990. We think the SPX resistance zone is likely a main reason. 

Another reason is timing. SPX has had at least one 5% correction by May every year since 1996. 

The same can be said about every year since 1980, with only 3 exceptions: 1985 (it corrected in June), 1989 (September) and 1995 (none). Those exceptions bear no resemblance to today as the prior year in each case was either flat or strongly down and demand was therefore pent up. Last year, in comparison, SPX rose 12%.  In the first chart below (zig zag 5%), those exceptions are shaded yellow. 

The key is this: excluding those years, SPX has had at least one 5% by the end of March 93% of the time since 1980. If you include the first week in April, the probability rises to 97%. Even if you include those other years (1985, 1989, 1995), the probability is 88%. The current trend is running headlong into an exceptional bias. 

In only one case did SPX correct later than April and that was in 2006, when SPX rose until the first week in May. 2005 had a 5% gain and by May, SPX was up another 6% (vs 9% today). Overall, the set-up is not hugely dissimilar from 2013. Over the next 3 months, SPX fell 7% and more than erased all gains for the year. 

These probabilities are even more noteworthy when you consider that SPX is sitting right within resistance. Over time, SPX (like other indices), reacts very close to where it has reacted previously. See the next chart below - the red lines show where SPX reacted off prior levels during the 2003-07 rally, the orange lines show where SPX reacted during the rally from 2009 onwards. They line up well. The next level is 1555-75. 

There are other reasons to expect a strong reaction near current levels: read about strong consecutive price trends at the end of a long uptrend here and the strong similarity of 2013 to 2011 here.  

As always, the question is expected return relative to expected risk. Could SPX rise well past May? Of course, but those are short odds.

To be clear, right now, there is no reason to expect 2013 to be another 2000 or 2007.  One scenario we think is likely is shown here

Notes: Since 1980, a >5% correction first took place in these months: 
  • January (15)
  • February (8
  • March (5)
  • April (1) - 2012
  • May (1) - 2006
  • July (1) - 1985
  • September (1) - 1989
  • Never (1) - 1995