Here's a little exercise from my Excel lab, where I investigate historical patterns:
Recall the post on relative gap, where we look at the opening gap (the distance between today's open and yesterday's close) as a function of recent market volatility. If we look historically, what constitutes a large opening gap vs. a normal one?
I went back to the start of 1995 in the S&P 500 Index (SPY) and calculated each day's gap as a function of the previous 20 days' median daily high-low price range. It turns out that, over that period, the median gap is close to zero (just as many days gap up as gap down), with a standard deviation of about 40%.
What that means is that about 2/3 of all opening gaps will be within 40% of the median volatility of the prior 20 days. Given that current 20-day median range is running about 2.1%, that means that about 2/3 of all days will fall within a gap of +.84% and -.84% (about 7-3/4 ES points or .75 SPY point).
So why is that important? If we get an opening gap running much above or below that point, we have overnight movement that is relatively unusual from a statistical vantage point. This tells us that events overseas and/or preopening economic reports are moving the market to an unusual degree--a worthwhile dynamic to note for the open.
Interestingly if we go back to 1995, we find that the size of the overnight gap correlates .30 with the coming day's relative high-low price range (the day's range sized as a function of prior 20-days' volatility). That is not a huge correlation, but it is significant and worth noting. One of the first heads up we get as to whether or not today's session will be busy and volatile comes from overnight trade. For instance, when the opening gap is larger than 50% of the median daily trading range over the past 20 days, the coming day session averages a range of 138% of that range. When the opening gap is less than 10% of that median range, the coming day session averages 102% of the range.
If we combine overnight gap with early relative volume in the ES futures, we can arrive at a surprisingly good estimate for the day's volatility--key to helping short term traders understand whether a market is likely to hold much or little opportunity.