Usually one would say that something "falls through the cracks." But in this instance, it's rather more correct to say that the market, especially the Nasdaq (QQQ), rises through the gaps.
First, what's a gap?
I'll define a gap as the difference between the close of a session and the open of the next session. I'll be using the Nasdaq for this exercise, since due to its calculation method the Nasdaq does, indeed, reflect gaps at open due to all the stocks opening instantly. The S&P 500 (SPY), it should be known, doesn't, because it gets calculated using partial open data (some stocks already open, others using the quotes from the previous day).
Now, the extraordinary claim
Why do I say that the market rises through the gaps? Because ever since 1994, the entire Nasdaq bubble happened at night. And, amazingly, it didn't stop there. Even since March 2000, the market has risen essentially during the overnight session (between the close of one session and the open of the next one).
Let me illustrate this claim. Below, we have a Nasdaq chart starting in December 31 1994:
Now, the Nasdaq sat at 776.8 on December 31, 1994. What would happen if to this, we added every session's close minus its open? (Thus we'd build an index where gains would only come from the intraday movement - the regular session). This is what we'd get:
And what if to that start point, we added every gain made during the night? (That is, the open of today's session minus the close of the previous session.) This is what we'd get:
So basically ever since 1994 all the gains made in the Nasdaq were made in the overnight session, between the close and next session's open. And not only did this happen in the past, but it's still happening today, as we speak.
In case you were wondering if it was always like that. No, it wasn't. This is what happened during the 10 years previous to 1994:
So up until around 1994, the gains usually took place during the regular session, and the overnight session was rather chaotic.
The long-term implications
Apparently, the market structure has suffered a seismic change around 1993-1994. Ever since then, most of the market movement happens not during the regular session, but in between regular sessions - in the overnight session. The positive bias one has come to rely on and that sometimes precludes investors from assuming anything but long positions, has thus fully migrated to the overnight session.
There are a few implications to this knowledge:
- First, the overnight gain since 1994 comes to 7023 Nasdaq points, and most of them were made up until the top of the Nasdaq bubble back in March 2000. This means that the overall overnight gain is just 1.45 Nasdaq points per overnight session. This kind of movement is tradable with futures, even if it presents a rather puny gain.
- This knowledge also means that generically, someone wanting to short the market for hedging or speculative purposes ought to do it at the open and should -- if possible -- avoid being short through the overnight sessions. Doing this, one would have his short positions avoid the entire positive bias present in the market, at least until this weird behavior goes away.
- Finally, this presents an increased difficulty for long-only, intraday, traders. After all, such a trader will open and close his orders during the regular session, and as we've seen the market is lending him precisely no upside bias at all (indeed, there's a faint downside bias during the regular session). Indeed, it might be the emergence of both daytrading and after-hours trading that has led to such a phenomenon taking place.