It took me by surprise that the R2K moved cleanly back into the 2014 range and the SP500 back above its 200MA in yesterday’s powerful up day. However, I believe odds are we still roll over here around tomorrow’s new moon only from a little higher up. That would make for a double fake: first below and then above the key technical levels. The weight of evidence still supports this being a second chance peak and that even if that is not the case then the market should first retest last week’s lows.

With history as our guide, here is the mid-2010 correction:

Compare to the current:

Both corrections began measured then collapsed in capitulative style before rallying back up from the purple circles back to the more measured downtrend. If this is repeating then the SP500 should now turn down again and tomorrow’s new moon provides the ideal timing for a reversal.

We can see how in 2010 the market based over a period with the MACD turning up. We lack such a stabilisation currently.

In last week’s analysis I argued for why we are now equivalent to late in the 2000 or 2007 topping processes, effectively January 2008 or October 2000. Both historical mirrors saw a similar collapse to a capitulative reversal point before a rally back up like now:

And in both instances price came back down to the lows, again stabilising against a rising MACD.

So all three analogs suggest price should come back down to retest the lows and stabilise around those levels over a period.

If this is not to occur then we could look to mid-2007 for the more bullish outcome:

A rally all the way to marginal new highs from a single-legged correction. However, I remind you that this corrective leg was the start of the market topping divergences in 2007, whereas now we are mature in those divergences, equivalent to January 2008. Various indicator readings just don’t match up well with mid-2007.

If we draw in the crash analogs, such as 1929 and 1987, then we are still in keeping with the second chance moves to a lower high before the true collapse, but only if we now roll over again over the next several sessions and head back down to last week’s lows. Crashes may not occur often, but I maintain that we have all the conditions in place for one to occur, and I stick with this being the most likely scenario: reversal around tomorrow’s new moon and then major crash into the beginning of November. Stabilisation to occur from much lower.

If we are not heading for a crash then the three analogs at the top of the page show that we should still make a retest of last week’s lows. I think the mid-2007 analog has the slimmest chances of reoccurrence but it will become clear over the next several sessions as if something like that were re-occurring then stocks would consolidate yesterday’s break upwards and rally higher still.

All the other analogs argue for a reversal back down without delay, and this still looks the most likely (geomagnetism is in progress, capitulative breadth is back to zero and various indicators are overbought), so I leave the analysis there for today and we see if that occurs before the week is out. If it looks like we are rolling over again then I will add again to the short positions.

I’ll cut to the summary: there were some signs of capitulation yesterday, but that was not likely the bottom. I’ve kept all my short positions, here’s the analysis:

Capitulation signs: high volume, intraday reversal hammer candles on stocks and treasury yields:

Source: Stockcharts

The dotted line on the top chart shows the YTD progress. The SP500 dipped negative yesterday and almost hit 10% down from the September peak, but closed at around 9%. By the range of my analogs, we should see a minimum 18% correction, so that’s not enough.

Rob Hannah’s capitulative breadth hit 19, an extreme level associated with bear market bottoms. The initial reaction to that is that it sounds like a golden opportunity to go long, but the considered reaction, IMO, is that given we are just 9% and 1 month down from all-time highs, this time is clearly different. He doesn’t reveal his formula so the analysis has to stop there, but we can look at other indicators for more clues.

Yesterday’s selling did not appear exhaustive as small caps held up well. As the highest risk equities class, this runs counter to previous capitulative sell-off examples, such as in 2011. That suggests there should be another sell-off ahead in which small caps participate.

Drawing together 2010 and 2011, we can see that intraday hammer candles did not necessarily mark the bottoms in equities (red circles). However, if yesterday’s hammer was indeed a bottom then we should be heading for a retest and likely lower low on positive divergences as momentum wanes (purple circles).

Source: Cam Hui / My purple annotations

If we look at the 1929 analog and assume we are not post-second-chance but at the purple circle, then a few days’ bounce before rolling over again would mirror not only 1929 but also the action after the first purple circles in the 2010 and 2011 charts above.

Source: Financial-Spread-Betting

Evidence, however, suggests that we should be post-second-chance, and so far, the steepening downtrend in price is supportive of that. If that is so, then yesterday’s candle should be quickly engulfed and a much bigger panic sell day should come to pass within days.

Putting that crash scenario aside, stocks typically need to stabilise and generate positive divergences after a sell-off such as the last few days. They do not v-bounce back up to the highs. For example, below we can contrast the positive RSI divergence in 2011 and the levels of Trin washout compared to now, i.e. both are missing:

So, either yesterday’s hammer candles are quickly engulfed over the next couple of days, like the red circles in the 2010/2011 charts or post-second-chance in 1929, or we rally for a few days (perhaps into the October 23rd new moon) and then turn down again to the lows, like the purple circles in 2010/2011/1929. A v-bounce back up towards the September highs would be anomalous and only have support from the CBI, which itself has produced an anomalous reading.

Continuing, the Investors Intelligence spread has narrowed, but still needs to neutralise some way further before we could argue for capitulation. Bears are still at just 17%. Interestingly, the percentage of respondents predicting a correction rose to an extreme level last seen in late 1983:

Source: Ryan Detrick

We can see that reading was indeed followed by a correction, so was not contrarian:

Source: Stockcharts

The Russell 2000 outperformed the last couple of sessions, but has broken down the most of the indices overall. It would need to quickly recapture the 1100 level to keep any consolidation hopes alive. Otherwise, the bears are in control.

Biotech now looks to have been turned down at a double top with February.

Apple tentatively broke down from a 6 week trading range yesterday, but the pull up into the close keeps it tentative.

To sum up, the most bullish scenario is that we rally for a few days then retest yesterday’s lows again on positive divergences. The most bearish scenario is that yesterday’s range is swiftly engulfed by the bears over the next couple of days and a much bigger panic sell-off ensues.

No panic selling follow-through from Friday in the Asia and Europe sessions yesterday but weakness hit late in the US, producing the third consecutive >1% down day, taking the SP500 7% down from its peak, and strengthening the steepening down trend. It keeps us on track for tipping into the panic selling, per the analogs, unless the bulls can stop the rot and generate a relief rally, so let’s see how the indicators stand.

It has been a long time since three consecutive >1% down days, last seen in 2010, and mainly associated with bear market action:

Source: John Kicklighter

The SP500 sliced through the 200MA with little resistance. The positive divergence on the RSI has been negated with yesterday’s move.

 Source: Stockcharts

CPC put call ratio hit the 4th highest in its history. Other such instances shown below and they weren’t great markers of lows:

Source: U Karlewitz

In fact they were generally associated with negative returns on all timescales:

Source: Ryan Detrick

However, capitulative breadth hit 12, which is a washout level. Historical returns shown below:

Source: Rob Hannah

A drawdown was fairly common, sometimes large, before mainly positive results.

Trin, Nymo and Tick did not make extremes yesterday:

Year to date sector performance looks like this after yesterday:

Source: Macromon

All year the top two sectors that are associated with post-market-peak outperformance have been signalling a top, but now their relative performance is much clearer to see.

In summary, the arching-over formation and the increasing bear pressure on the stock indices looks very similar to the waterfall decline analogs (see last post) pre-collapse. With all the major indices now having broken through their key support levels, the bulls need to step in fast to prevent a crash. The CBI washout is on their side, but there may well be further selling (drawdown) before the bounce, if it works again. Other indicators are not at this point showing obvious exhaustion. Therefore, some more selling, to align indicators at contrarian bullish readings, seems the most likely, and a voluminous major distribution day, perhaps that ends as a hammer candle (intraday reversal), would further increase the odds of a bottom, if it is to come.

The question is whether some further selling here can be fairly measured and contained, or whether it would quickly spiral out of control to become a major crash day (or days), given the large imbalances in the market (leverage, bullishness, complacency, etc). Either way, I think the situation is looking fairly black and white now. Either we are cascading into panic-selling, per the historical analogs, and, having broken the key supports, that should now happen very quickly (this week if a collapse, or more sustained selling over several weeks if more like 1937). Or, the bulls can yet generate a relief rally to a second-chance peak (springboarding off that CBI) which should last a couple of weeks or so (drawing on the analogs again), before the market tips over in earnest. Right now the first option has the upper hand (closed at lows, sliced through supports, steepening down trend, a lack of alignment in indicators for exhaustion), so the onus is on the bulls to deliver something today or tomorrow.

Here’s a reminder of historical analogs (manias of similar valuations, leverage, bullishness extremes, etc), and their common topping pattern:

Source: Financial-Spread-Betting

Here is the Dax in 2014, with a similar topping pattern:

Source: Stockcharts

If we look at the Wilshire 5000, the broadest measure of US stocks, then these developments over the same timeline look like this:

The ‘second chance’ sticks out, being a higher high, so there’s the question mark.

That WLSH chart shows the importance of last week’s action. After 18 months of higher lows and levitation above the 200MA,  we now have a lower low and a break of the moving average.

The SP500 ended Friday pretty much on the 200MA and at a double bottom with the August low, so hasn’t broken yet, but the leading action in RUT, DAX, FTSE and others, plus the selling into the close on Friday suggests the SP500 should follow suit and print the break and lower low.

Historically we have seen a phenomenon of strong selling into the close on a Friday bringing about major Monday sell-offs due to weekend stewing. The Nikkei chart above shows some examples. Meanwhile, the waterfall declines post-second-chance on the historical analogs occurred when a technical breakdown of a key level in price took place. Combined, they present the potential for this Monday, 13th Oct, to be a panic selling day, if buyers do not step in early and defend the SP500 $1900 support. So back to the question mark of whether we are post-second-chance.

I’ve shown before that various behind-the-scenes indicators suggest we are indeed post-second-chance, e.g. HYG:TLT, NAAIM, Investor Intelligence, Rydex and margin debt. So, despite US large caps having made higher nominal highs through to September, these indicators plus charts such as RUT and DAX suggest the September highs were rather a second chance peak. If so, then a quick look back up the page at the historical analogs show things rapidly fell apart after that point. Bears engulfed bulls in a steepening trend which snapped once key supports were broken and became several days of panic selling. It happened very fast. The steepening engulfing trend is what we saw in the markets last week, which adds weight to us being now on the cusp of the panic selling.

Therefore, if we are indeed post-second-chance and closed Friday under strong selling at the key twin support on the SP500 of $1900, then there really is a compelling set-up for Monday to be a major sell-off – a crash – should that support fail to hold from the outset, as weekend stewers press sell. So, all eyes on how the markets open overnight Sunday-Monday. If we do see the panic selling on Monday then the analogs suggest a rapid collapse could ensue over several days, taking us very swiftly to my targets. As a reminder I am looking for a minimum 18% off the peaks and large caps closed Friday about 6% down.

If, conversely, we are not post-second-chance on US large caps, and price needs to make a true lower nominal high, then we should instead see no panic selling and either the SP500 defended at $1900 or at a lower level, before being bid up again to perhaps make a ‘right shoulder’  beneath $2000. Whilst this may be considered the more likely option by many, I just reiterate that indicators and other indices instead position us post-second-chance, with the September higher nominal high in large caps being the ‘red-herring’.

Below is the longer term picture for the historical analogs. After the swift and devastating waterfall declines came a multi-month partial retrace of the falls, before the long bear market kicked in properly.

My aim is to try to exit my short trades after the waterfall declines and before that recovery rally erupts. To this end I will be looking to indicators for signs of exhaustion. At the end of Friday, capitulative breadth was still below 10 so the short term does not look washed out, whilst indicators such as stocks above 50MA show we would need to drop some way yet for a true correction:

Source: Charlie Bilello

Because we have levitated in a mania for so long, postponing any meaningful correction, I believe the correction has to be nasty when it hits. For example, with this intensity and duration of investors intelligence skewing can we really avoid a repair like 1987?:

Source: Fat-Pitch

If we draw in the 1937 analog we see a less panicky decline post-second-chance over a more measured 8 weeks:

However, I believe the odds favour a more sudden collapse like the other historical analogs, because we can see in asset allocations, leverage, sentiment, volume and complacency that a limited group of buyers are all-in and leveraged with no bears left to convert, in an extremely lop-sided positioning historically. Therefore, if selling erupts it will be like an exit through a key hole and margin calls will quickly generate further selling in a series of forced redemptions.

Nonetheless, we can draw a range from the various analogs displayed of anything up to 8 weeks for the waterfall declines to complete, whilst the average falls were 30% with a minimum 18%. With this time and price guide in mind I am then looking to cross-reference with indicators that can signal exhaustion (e.g. CBI, II, stocks above MA, positive divergences, a volume surge etc) and see if we can pick the bottom before the recovery rally emerges. Important to get these parameters in perspective in case things occur very fast this coming week. If they don’t, then maybe the bulls can come back and generate a lower high in large caps. Anything more bullish than that now looks improbable, given the breakdowns in RUT, JNK and others.

So a big week ahead. Opex is usually positive, but this week’s proper kick-off of US earnings ought to be negative. If the panic selling erupts on Monday then neither will matter.