I read somewhere on Monday (sorry, I can't remember the source) which stated that most "healthy" corrections in the Nasdaq Composite are limited to 8%.  If it corrects more than that, then it is highly likely to continue to correct another 5% - 15% before putting in any kind of meaningful low.

That just begs to be tested, because today it violated that threshold.

We're using the Nasdaq Composite here instead of the Nasdaq 100, because of its broader base and longer history, dating back to 1971.

What we're going to look for is any time the Composite hit a 52-week high sometime in the past 2 months.  Then it corrected more than 8% from that high.

We'll see how long it took - and how much more in losses - before the Composite put in an intermediate-term low.  "Intermediate-term low" in this case is a day whose close was not violated for at least the next three months.

It turns out that the "5% - 15%" range was pretty accurate.  The median amount that the Composite lost, in addition to the 8% it had already shed, was -7.2% before it put in a low.

Of the 28 precedents, 9 led to additional losses of -10% or more and 16 led to losses greater than -5%.  Of those, it took a median of 42 trading days (about two months) before the index formed a meaningful low.

Of course, that means that 12 of them, or 43% of the total, had losses of less than -5% before putting in a low.  6 of those bottomed the very day the correction exceeded 8%, but on average they took 13 trading days before bottoming.

So is it a bad sign that the Composite has exceeded a correction of 8%?  On average, yes, it probably is.  But the variation among the precedents is very high, and as noted nearly half of the instances suffered additional losses of less than -5% before bottoming.

If the Composite closed below its March low of 2616 for multiple sessions, that would be a much worse sign.