The Odds Favor a “Warning Shot” Scenario – but there is a “But”
As regular readers have probably noticed, we have upped the frequency of our “caution is advised” posts on the stock market in recent weeks in light of the market’s increasingly deteriorating internals. Although one never knows when exactly such warning signs may begin to matter, it is always a good bet that they eventually will. Last week the market delivered a little wake-up call to the hitherto rather complacent majority of market participants, by essentially wiping out 9 months worth of gains in more or less just four trading days.
The S&P 500, daily – obviously, this chart doesn’t look good – click to enlarge.
The sheer speed of this decline masks the fact that the S&P 500 is actually only 163.83 points or 7.67% below its all time high made in May. In other words, this decline doesn’t even amount to a routine 10% correction yet. And yet, as Zerohedge reports, cries for intervention by the Fed are amusingly already going up. We actually don’t believe that the federal purveyors of Anglo-Saxon central banking socialism will jump into the breach that quickly.
Last week’s sudden “1-800-get-me-out” moment certainly wasn’t widely expected, not least because it was an options expiration week, and expiration weeks normally tend to follow certain patterns. Either there is little net volatility, or if the market has a weak close on Wednesday, it tends to rise on Thursday and Friday. This even happened in the October 2014 sell-off, which ended on a Wednesday in an expiration week (with the SPX gaining 80 points between the Wednesday intraday low and the Friday intraday high, rendering a great many puts, VIX calls, etc. that were bought earlier in the week worthless).
As a friend of ours remarked, this time delta-hedging seems to have exacerbated the decline. Because there has been no volatility for such a long time, sellers of puts that were set to expire shortly seem to have gone into the week slightly overexposed. The only way to protect such open positions in a suddenly declining market is by either buying them back (quite possibly at a loss), or by shorting the underlying instruments in increasing quantities the closer their prices come to the strike prices concerned.
The point is that technicalities related to market structure and derivatives can influence and magnify market moves in the short term; what is remarkable about the action of last week is mainly that the “usual pattern” has failed to make an appearance.
Other than that, market participants reacted as one might expect: the VIX surged, with the VIX futures curve inverting noticeably, put-call ratios soared and so did trading in inverse ETFs. All of these symptoms are normally consistent with a short term low being very close.
The VIX futures curve has become inverted out to December – click to enlarge.
Trading volume in inverse ETFs soars – click to enlarge.
One problem with such “oversold extremes” is that they also tend to appear in the rare cases when the market is about to lose its countenance completely and crashes. An even bigger problem is that there is no way of telling in real time which type of correction one is in.
Consider for instance the CBOE equity put-call ratio shown further below. It was very high on Friday, but in the 2008 panic it went to even higher levels on two occasions (to 1.18 and 1.35 respectively), so even greater extremes are definitely possible. Also, as the chart reveals, its peaks tend to lead price lows by a few days. When prices move to a lower low, it is however not easy to tell where and when exactly they will stop and reverse.
The default assumption should be that in perhaps 14 out of 15 cases the market makes a short term low relatively quickly, likely not too far below the levels reached so far, and then reverses in the short term. However, in one out of 15 (or fewer) cases this isn’t true, and “oversold” conditions end up inviting even heavier panic selling.
It is important to be aware of this, not because it is likely, but because it can happen. One needs to have a plan as a trader or investor how one handles such situations (this is easier said than done, but having no plan at all is the worst possible approach).
If one looks at the pure price chart shown at the beginning of this article, it is clear that this is not a nice looking chart at all. On the contrary, the index has actually already broken through two support levels. Moreover, trading volume tends to be much higher in a true panic, and RSI tends to visit even lower levels. Various black-box systems are by now very likely treating this as a new downtrend and their actions may continue to influence the market regardless of various extremes in “oversold” signals.
Looking at the last three peaks in the CBOE equity p/c ratio, one can see that its peak levels tend to lead price lows by between 2 to 5 trading days. The size of these final moves lower can however not be gauged in advance (one could perhaps state that the SPX October 14 low is a viable target, but there is no way of telling how quickly it might be reached or whether it will hold if it is reached) – click to enlarge.
In short, the previous examples of put-call ratio spikes shown in the chart above may turn out not to be overly relevant, even though the odds-on bet should be that the market is merely in the process of delivering a first “warning shot” to investors. As we are writing these words, Asian and European markets are deeply in the red, and US stock index futures are once again selling off sharply – but none of this tells us how the US stock market will end the day.
If this is a “warning shot” scenario in an emerging downtrend, then one should expect a low to be put in over the coming days, followed by a rally that fails at a lower high. Thereafter the selling should resume.
If the panic continues and deteriorates into a mini-crash right away, then any upcoming low will likely be at a much lower level, but will also likely be more durable. This is especially so as commodities are also severely oversold and appear overdue to at least produce a counter-trend move at some point soon. Such a “panic without pause” is usually the by far less probable outcome, but we wouldn’t rule it out (DJIA futures are down 700 points as we write this, so the odds are clearly moving in the direction of “panic”).
That the potential for such a panic exists is mainly given by the unusual nature of this decline – its commencement during expiration week, and its suddenness and ferocity to date.
A Word on Insider Buying
Below you can see a table we have taken from the Wall Street Journal’s weekly insider spotlight pages. Although insider selling of course still outweighs insider buying by a huge margin, there has been more insider buying last week than in any previous week this year.
However, just as we suspected when discussing the market last week, much of the buying is concentrated in relatively unimportant (to the market as a whole) sectors, such as raw materials. For instance, there has been noteworthy insider buying in Freeport McMoran’s shares, which has greatly pushed up the “basic industries” total. There was also increased buying in the energy sector, finance and health care but in the latter two industries, both actual and planned sales dwarf the buying that was recorded last week.
Friday, August 21, 2015
Ranked by market value. Sector classification by Thomson Reuters. Planned sales from Form 144 filings.
| PURCHASES | SALES | PLANNED SALES | ||||||||
| Rank | Sector | Market value |
Rank | Sector | Market value |
Rank | Sector | Market value |
||
| 1 | Basic Industries | $12,370,268 | 1 | Health Care | $152,403,488 | 1 | Health Care | $119,376,889 | ||
| 2 | Finance | 10,976,788 | 2 | Finance | 152,316,678 | 2 | Finance | 116,324,867 | ||
| 3 | Health Care | 10,207,871 | 3 | Technology | 123,277,009 | 3 | Consumer Services | 89,162,650 | ||
| 4 | Consumer Services | 8,090,039 | 4 | Industrial | 100,208,126 | 4 | Consumer Non Durables | 87,388,082 | ||
| 5 | Energy | 6,380,830 | 5 | Consumer Non Durables | 100,157,552 | 5 | Technology | 80,081,540 | ||
| 6 | Technology | 4,177,717 | 6 | Basic Industries | 56,465,249 | 6 | Business Services | 38,601,108 | ||
| 7 | Consumer Non Durables | 2,065,356 | 7 | Transportation | 49,187,252 | 7 | Industrial | 37,095,147 | ||
| 8 | Industrial | 1,675,405 | 8 | Consumer Services | 46,602,217 | 8 | Basic Industries | 31,531,439 | ||
| 9 | Undesignated | 588,758 | 9 | Business Services | 26,921,727 | 9 | Utilities | 31,116,275 | ||
| 10 | Consumer Durables | 370,648 | 10 | Utilities | 20,885,797 | 10 | Media | 18,107,777 | ||
n.a. = Not Available
Insider buying surges, but actual and planned sales still dwarf it.
In short, we wouldn’t make too much of this, although the buying in basic industries is noteworthy – after all, “everybody knows” commodities can only go much lower from here …
Conclusion
The next few days will undoubtedly make clear whether this was just a major shakeout, a warning shot, or whether the market’s “panic potential” is realized right away. From a statistical perspective, the extremes recorded in short term positioning and sentiment indicators last week argue for a short term low to be put in soon. Depending on the level at which such a low is put in, it may merely postpone the panic potential, but not remove it.
Note also that if especially strong rebounds occur (in the form of very large one-day rallies), this would actually be another negative signal, as the biggest one day rallies tend to happen shortly before truly gut-wrenching sell-offs begin (recall in this context the more than 1,000 point advance in the DJIA in 2008 after short selling of financial stocks was banned – it proved to be the prelude to the crash. Similar moves could be observed prior to other crashes as well).
Although it is notoriously impossible to forecast a market crash, one must always remain aware of the fact that crashes do occur from oversold conditions. The fact that markets around the world are breaking important support levels and that the price structure of the SPX looks decidedly negative at the moment is certainly a good reason not to rule any particular outcome out.
Charts and tables by: StockCharts, SentimenTrader, vixcentral.com
