Here’s this week’s comprehensive sentiment overview:
Let’s start by looking at the AAII sentiment survey: this week, 42.9% of respondents believe that the market will be higher 6 months later and 30.4% are bearish (with the rest undecided).
Since it is easy to get lost in the minutia of the weekly changes here’s a look at the bull ratio going back to 2007 (the bull ratio is calculated by dividing the weekly bullish respondents by the sum of bullish and bearish respondents):
Within this time frame, whenever the bullish ratio reaches 68% or higher, the stock market either faces a short or major decline not long afterward. Right now the ratio is at 59% which is slightly higher than the 10 year average. While we are seeing a considerable amount of bullishness, the AAII bull ratio is still not at levels which correspond to market tops.
Keep in mind that this is a short term look back through the AAII history which goes as far as 1987. A longer term chart would show the bull ratio reach much higher levels. For example, during the rapid market recovery in 2003 the bull ratio reached +80% as it did back in 2000 during the bubble mania. The 3 year average of the bull ratio is 48% but the 10 year average is 56%.
This tells us that we are actually seeing much less excitement about the stock market from retail investors than we did in previous cyclical bull markets. Which is not surprising as it confirms the retail investors preference for bonds instead of equities. So while I’m watching the 68%-70% level for this metric, I’m also keeping in mind that the definition of “extreme” can easily be reset.
This week’s II is also showing an increased level of bullishness: 48.9% expecting higher prices, only 18.9% expecting lower stock prices and 32.2% expecting a correction. Similar to the AAII sentiment indicator, the newsletter editors’ sentiment has also continued to inch ever more bullish - but it has yet to reach extremes that correspond to market tops.
Currently there are 2.59 bulls for every 1 bear. In the past few years, market tops like the October 2007 high and the January 2010 high have seen slightly higher than 3 bulls for every 1 bear in the Investors Intelligence survey. So this sentiment indicator suggests that we still have a little bit of work on the upside.
Hulbert Newsletter Sentiment
According to Mark Hulbert, keeper of the eponymous Hulbert Stock Newsletter Sentiment Index, we are seeing a worrisome level of bullishness from newsletter writers which puts in jeopardy a further rise in the market. The current HSNSI is 56% which is the highest level since May 2007 - with the exception of a short term spike up to 65% in Januray 2010 which precipitated the correction into February 2010.
Just like retail investors and newsletter editors, the average active money manager also got more bullish. The NAAIM survey this week pushed higher to 86% bullish (average) and 95% bullish (median) for the second week in a row.
As I mentioned before when we looked at the history of this sentiment indicator, the NAAIM survey can stay at very elevated levels for multiple consecutive weeks before the market finally crumbles from the weight of so many bulls. The important thing is how the managers react if we see the market give back some of the gains it has made. Will they remain stubbornly bullish or quickly abandon the long side?
Daily Sentiment Index
The DSI continues to hover around 85% bullish for the S&P 500 index which is a few points lower than it was at the peak in October 2007 (96%). Also, according to the DSI 93% are bullish on the US Dollar while only 5% feel the same way about the Euro. This is a huge swing from late last year when the bullish percentage for the US Dollar was in the single digits (7%).
Corporate insiders may be using the money in their companies’ coffers to buy back their corporations’ stocks but they are not putting their own money to work with the same conviction. According to TrimTabs, insider activity in March showed them selling $6.9 billion while only purchasing $0.83 billion. This may mean that insiders are more skeptical of the rally and prefer to participate on the upside through stock performance based compensation (by pushing up share prices through buy-backs) rather than risking their own dime.
There are the first few signs of a nascent consumer recovery. According to SpendingPulse, a division of MasterCard (MA), retail sales in March 2010 were robust with the online and luxury enjoying especially strong growth.
Mutual Fund Flows
According to the ICI, March 2010 had inflows of $519 million into US equity funds and inflows of $41.3 billion into bond funds. So the trend that we have observed for some time continues unabated.
According to TrimTabs, March had inflows of $11.7 billion into US equity funds and February had $110 million (compared to outflows of -$6.3 billion according to ICI’s February data). I’m not sure how to explain the huge discrepancy between the two. One possibility is that TrimTabs is estimating the monthly data based on a sample.
Another way to measure what the retail trader is doing through the asset flows in the Rydex mutual fund family. Total money market assets reached a peak of almost $4 billion at the March 2009 bottom and about half of that was held by Rydex traders:
But since then they have slowly moved out of cash and suddenly accelerate their exit in the last few months. Right now they are holding the lowest level since February 2001 (not shown in the above chart). So what have they done with the cash? Much of it has been destined to the leveraged Nova fund which provides 1.5 times the upside of the S&P 500 index:
The current assets of the Nova fund indicate a new found confidence as it is the highest since September 2008 - before the worst cascading declines in the S&P 500. The current Nova/Ursa ratio (bull/bear funds) also indicates a very high level of confidence.
The CBOE volatility index has been slowly melting for more than a few months but with Friday’s close it is as low as it has been since October 2007. That is making many traders sit up and take notice. Wayne passed on these thoughts:
- Average VIX over the last decade was 22.97
- The lowest VIX was 9.89 on January 24th 2007 with the S&P at 1440.13
- The highest VIX was 50.17 on March 5th 2009 with the S&P at 682.55
I looked at S&P performance as a function of the VIX level and couldn’t come up with anything statistically significant. I then looked at S&P performance as a function of the 3 month percent change in the VIX and there was some modest correlation between the direction of VIX and the S&P, but not anything worth writing home about.
The one data point that I did find of particular interest was that on March 12th 2010, the VIX had declined 30% over the previous 3 months to a level of 17.58. There was only one other period in this last decade of data where VIX was low (less than 20) and had dropped by 30% in a three month period; that was for four days in the middle of October 2007.
Even at the start of this week it was obvious that the option traders were on a call buying binge. The rest of the week only served to confirm and detail that fact even more.
The 10 day equity only CBOE put call ratio fell to 0.50 implying that traders were buying twice as many calls in the past 2 trading weeks as puts:
Going back to 2004, there were only two other times that we saw option traders positioned in a similar extreme bullish fashion: January 2004 and July 2005. While they are only 2 time periods, none of them were a smart time to be long the market.
The S&P 500 struggled for much of 2004, basically churning sideways for the majority of the year. It wasn’t until November 2004 that the S&P 500 index was able to reach new highs for the year. In the summer of 2005 the S&P 500 was able to climb another 10-20 points before falling from 1245 to 1177 (-5.5%).
The ISE sentiment index continued to push higher this week with the 10 day equity only ratio hitting 230 (slightly more than 230 calls purchased by retail option traders for every 100 put options purchased to open a trade):
That is even higher than the reading we saw in October 2007. To find a higher level, we’d have to go back to mid-July 2007 when the ISE index peaked at 245. But remember that the July 2007 top in the S&P 500 was only about 5 points lower than the October 2007 high.
Enjoyed this? Don't miss the next one, grab the feed or