Thursday, July 19, 2007

Opening Gaps to the Upside in the Stock Market

Here is a replication of an earlier study relevant to the current market:

As I write, we're headed toward the opposite of yesterday, with the prospect of a large opening gap to the upside. Going back to the start of 2004 (N = 890 trading days), we've had 74 upside opening gaps that have exceeded .40% in the S&P 500 Index (SPY). From the open to that same day's close, we don't see any edge for the strong opening gap day. When we look from the open to the *following day's close*, however, the average gain in SPY has been .14% (43 up, 31 down). That is stronger than the average open to next day close for the remainder of the sample of .04% (437 up, 379 down).

I then broke down the strong upside gap days on the basis of whether SPY had been down or up the prior day. When we've had a strong upside gap following an up day (N = 27), the average gain from the open to the following day's close has been an impressive .33% (19 up, 8 down) for SPY. When we've had a strong upside gap following a down day (N = 47), the average gain in SPY from the open to the next day's close has been only .03% (23 up, 24 down).

In sum, as in the prior study, I don't see any open-to-close edge following a strong market open. Indeed, returns even to the next day show no bullish edge when, as currently, the prior trading day had been down. There is some indication, however, that strong upside gaps following an up market may lead to superior returns in the near term--an apparent momentum effect. This is a pattern of interest to many stockpickers, but may require more than a day trade to fully exploit.


Do Opening Gaps Tend to Fill?


Humble1 said...

a lot depends on where the opening gap occurs on the chart. Some open at resistance others open above it, both give different results.

Jeff said...

Very illuminating post, Dr. Brett!
I wonder if you may find it interesting to take into consideration the NYSE A/D at the open as well.

Dr Bruce Hong said...

Hi Dr Brett

I have a slightly different take. Yesterday, we were down substantially, only to see a late day rally. While it didn't take us to positive territory, it did display that the market was willing to assume risk at the close.

As you pointed out in a previous post, that had bullish implications for the market as a whole. I couldn't find the exact post, but I believe that this had an effect that was over more than a day.

I also think that we've had an island reversal (if we don't close the gap.) I wonder if you could do an analysis of island reversals. I think that that would also be very instructive.

Thanks for all your good work on our behalf!

Brett Steenbarger, Ph.D. said...

Hi Jeff,

Good point; the intraday NYSE A/D--and shift in A/D--do provide useful information. It's well correlated with NYSE TICK. If the odds are for continuation of strength and I detect firmness in TICK and A/D as the market trades early in the day, that adds confidence to my analysis.


Brett Steenbarger, Ph.D. said...

Hi Bruce,

I also can't come up with the specific post you're referring to. Because Wednesday's market had weaker momentum and strength numbers, I normally would not have expected bullish follow through the next day.

I've never used chart patterns in my trading and so have never looked into island reversals. Perhaps Bulkowski has studied them; he's done interesting work with chart patterns.

Thanks for the note--