How unusual are this market's intraday bounces?

The W survives to live another day.   While Tuesday's equity market weakness looked set to continue for reasons that weren't entirely obvious (yeah, the Markit services PMI was weak, but so were stocks before the release), just before mid-day Spooz turned tail and chugged higher for reasons that also weren't entirely obvious.   To be sure, the failure of oil to sell off after inventory data was a relief, but it probably doesn't explain the magnitude of the recovery in stocks.

Cue the usual allegations of manipulation and other chicanery from the tin-foil hat crowd, to whom Occam's Razor is as much an alien concept as the mundane variety was to W.G. Grace.  Having had a go at investigating official central bank buying of equities last month, Macro Man decided to take a crack at determining whether the degree of intraday rallies recently observed in stock markets really is that unusual.

At first blush, it seems as if the conspiracy theorists have a point.   Macro Man looked at the daily difference between the SPX cash close and the intraday low; the data is understandably noisy, so some smoothing is required.  Lo and behold, when we look at a 30 day moving average of this "rip factor", we see a sharp uptick over the last six months or so.

OK, but what if we put this measure into context against a larger sample, while also accounting for the general level of volatility?  It turns out that the tendency to bounce off of the lows this hard isn't unusual, and is largely a function of the general volatility environment.  It's remarkable:  higher close-to-close vol begets higher intraday volatility!

If there were truly something unusual going on to push equities higher whenever they fall, we should expect to see the blue line above rise relative to the red line, i.e. the magnitude of bounces off the lows should be disproportionate relative to the level of close-to-close volatility.  Macro Man calculated this ratio going back to 1982; the chart is pretty noisy so he doubled the observation window to 60 days.

As you can see, the ratio exhibits a strong tendency to mean-revert, with a long-term average ratio of 0.71.  The latest datapoint in the chart above is 0.67, which actually represents a below average magnitude of intraday bounces relative to the general level of vol.  In fairness, the shorter 30 day window does register an above average level of 0.77, but that is well within the bounds of normality as you can see above.

An argument can be made, however, that while intraday volatility is a function of the general vol environment, using the latter in the denominator can distort the signal lower in high vol environments.  This is most easily observed in the sharp blip lower in 1987, which of course represents the volatility engendered by Black Monday.  Macro Man solved this problem by calculating the intraday rip (i.e., difference between the close and the low) as a proportion of the intraday trading range (the high minus the low).  Macro Man can honestly say that he learned something after performing this study.

It turns out that the SPX generally closes closer to the high of the day than the low of the day, a tendency that has persisted for at least the last 30 years.  Without giving it much thought, Macro Man would have supposed that the average close would be right in the middle of the daily range rather than skewed towards the highs.   Upon further reflection, this could reflect the tendency of asset managers to trade disproportionately high volumes in the last hour of the day.   Regardless, it's an interesting phenomenon: since 1982 the SPX has on average closed in the 55th percentile of the day's trading range.

What about recently?   Again, it's difficult to discern much in the way of sinister behaviour.   The 60 day moving average of the indicator is below the historical average, registering 0.52.   The 30 day moving average is above the historical average by roughly the same magnitude, posting 0.58.  Either way, it's difficult to suggest that recent intraday price patterns are in any way unusual.

So there you have it.   Over the very recent past intraday bounces have been very slightly larger than normal, but nothing to excite interest.   Indeed, over a slightly longer window of time the bounces have been smaller than normal.  The answer to the question posed in this post's title, therefore, is pretty clearly "not very".