1. Investors Intelligence
II Bears (advisor sentiment) is sub 15 again this week, previous clusters shown:
Here are those sub 15 readings on the 80s Dow:
The 1983 extreme II reading occurred 12 months after the new bull market kicked off, and is comparable to the spike down in early 2010 in the first chart that didn’t quite reach <15 bears. The 1986 and 1987 double (circa 5 years post bull launch) is comparable to the Dec 2013 and current extreme pair, but note how stocks rallied for 5 months after the 1987 extreme before collapsing.
Rydex asset ratios (Rydex family of funds) current extremes are most similar to the 2000 period, namely a time band of extremely skewed holdings lasting for around 7 months in 2000, as we are seeing in 2014 (9 months mature).
But note that within that 2000 time band we saw two periods of major sell-offs: March to May and September to December. We have not experienced either so far in 2014.
NAAIM exposure (managers US equities exposure) made a pattern of extremes-plus-divergence at both the 2007 and 2011 market peaks:
Source: Acting-Man / My annotations
We see similar extreme high readings plus the same gradual divergence since November 2013. The pattern is most similar to 2007, where it lasted 10 months before the market finally entered a bear market. Here in 2014 the pattern has also been in play for 10 months.
The high yield to treasury bond divergence (a risk off measure) lasted 4 months, 5 months and 5 months at the 2007, 2010 and 2011 market peaks. The 2014 divergence has been running 8 months and is therefore excessively mature.
Source: Jesse Felder
The persistent extremes in Skew (risk of outsized move) beats any historic parallel, but we could point to 1990, and the most recent cluster in 2011, both which produced 20% sell offs in the market. The 1990 cluster lasted 7 months which makes the 2014 cluster, at 10 months, again excessively mature.
Source: Dana Lyons
6. Q ratio
Q ratio valuation (replacement cost of stock index companies) shows that flirtations with an extreme value of 1 were historically swiftly repelled (by bear markets) with the exception of 1996-2000.
Source: D Short
With the stock market now 2 years above the 1-level, the closest parallel is that period into 2000. Note there was a 20% correction half way in 1998. However, the run up to 2000 was a demographic peak, this is not, and valuations need demographic context. I believe 1937 is the most applicable mirror which puts us ripe to fall.
7. Household Equity Allocations
Equities as a percentage of US household financial assets have historically signalled market peaks once flirting with the 30% region, with the exception of 1997-2000. Again, demographics do not support this indicator rising to higher levels, so I would mark this indicator as similar in outlook to the one above.
8. Margin Debt
Margin debt (investor leverage) surged for 15 months into 2000’s peak and 10 months into 2007’s peak whilst the surge into the Feb 2014 peak lasted 18 months, so relatively mature. Whilst the Feb 2014 margin debt peak is only tentative currently, it occurred 6 months ago, versus 5 months and 4 months pre-peak in 2000 and 2007, so also relatively mature.
Source: D Short / U Karlewitz
9. Sornette Bubble
Sornette’s bubble end flag calculation looks like this currently on the SP500 and the US tech sector:
Compared to historic examples, the SP500 has not flagged at as higher intensity, whereas the tech sector now has. On the other hand, the tech sector flag looks fairly ‘new’, whereas the SP500 flagging has built up gradually. Neither have spiked multiple times in an extended period like in 1929 or 2000 (which both lasted around 2 years). At this point, the picture is most similar to 1987 of the three.
Source: Financial Crisis Observatory
10. Solar maximum
By most solar models, the smoothed solar maximum is behind us, circa March 2014. If this is so, then the last four solar maxima delivered the asset mania peak within 5 months, which makes the current peak ripe.
There are question marks over how close the timing ought to be (looking further back in history), and we cannot yet be sure the smoothed solar max is behind us. However, this is where cross-referencing comes in useful.
Investors Intelligence argues the top should be before 2014 is out, whilst NAAIM calls for one straight away. Rydex, Skew and HYG:TLT all argue a sharp correction circa 20% or a bear market is overdue, and margin debt tentatively does too. Meanwhile, Q ratio and household allocations argue for a sharp correction circa 20% imminently or a bear market too.
All these indicators cast great doubt on the speculation peak being some way ahead in 2015. Collectively they argue for us to be in the last gasps of a topping process that began at the turn of the year. If stocks were to continue to rally into 2015 then we would print major anomalies in all these indicators: they all worked historically, but this time is different. I don’t buy that.
The Sornette bubble either places the market at a 1987 style peak now, or will come again at a higher intensity any time up to the end of 2015. The solar cycle places us either ripe to fall, 5-6 months post smoothed solar max, or also may allow for a peak up to the end of 2015.
I therefore believe that the logical case is for the eight market indicators to work once again and deliver a peak now, which then fits with the 1987-style Sornette reading and the most likely smoothed solar max / speculation peak lag combination. It all fits together, it all fits with history.