It seems you have JavaScript disabled.

Ummm.. Yeah... I'm going to have to ask you to turn Javascript back on... Yeah... Thanks.

Sentiment Addendum: Gold, Big Money Poll & More at Trader’s Narrative

While the weekly sentiment overview on Friday covered a lot of ground, a few data points came up afterward that warrant mention before another weekly sentiment walkthrough.

Gold Sentiment
First, a reader inquired about the sentiment landscape for gold right now. I wrote about gold about a week ago: Gold Reacts To Bullish Extreme Sentiment (Finally). At that time we looked at the gold DSI, the Rydex market timers views on gold and the premium/discount on a gold closed-end fund. I concluded:

Based on these various sentiment gauges it is possible that we have seen a major top for gold but that in response gold will not decline heavily.

So far, that has been the case. Gold has refused to fall below $1325 and seems to have put in a bounce. Total funds in the Rydex Precious Metals fund have also rebounded from their recent low of $243 million to $268 million (after a peak of $282 million in mid-October).

The weekly Bloomberg gold sentiment survey is a new indicator that we haven’t discussed before. I’ve smoothed out the data with a 4 week moving average since it can be a bit jittery. As you would assume, it can provide an edge as a contrarian indicator. In the past, when the bull ratio of this survey has been low, gold has tended to find a bottom. And when the bull ratio is high, a top.

bloomberg gold sentiment Oct 2010

I was surprised to see that the short and shallow correction in gold was enough to provoke a 24.5% point drop in the bull ratio - it fell from its high of 82.5% in late September to just 58% last week. Even though the nominal level of the bull ratio is not as low as previous bottoms, the precipitous fall does suggest that a lot of bullish optimism has been swiftly wrung out of the gold market. And it also shows that investors are continuing to be skittish even in the presence of a secular bull market.

Barron’s Big Money Poll
The semi-annual Big Money Poll from Barron’s got the usual amount of attention this weekend. This isn’t surprising when you consider that the participants are some of the brightest and most successful money managers in the US. But are we to perceive the sentiment data it provides directly or through a contrarian lens?

From the latest Barron’s Big Money Poll, 4% are very bullish, 56% are bullish, 3% are very bearish and 12% are bearish (with the rest being neutral). This provides a bull ratio 80% (that is, taking the total bulls as relative to both bulls and bears). This is a slight increase from the last Big Money poll in April 2010 (75%).

barrons big money poll Nov 2010

Unfortunately, data from this sentiment poll isn’t really useful as it does not provide an edge one way or another. We can use a survey that historically has provided us with a direct signal, such as the S&P 100 put/call ratio. Or we can use the AAII as a contrarian measure. But the most exasperating indicator is one that is neither and appears more random than anything else. Sadly, this happens from time to time as we saw with TickerSense’s blogger sentiment and Citi’s Panic Euphoria model. And that’s the case with this indicator as well.

Before we put it to sleep, I’ll give you a few examples to show you what I mean. In April 2002, in reaction to the bear market, the Big Money got much less bullish suddenly. Their bull ratio fell to approximately 60%. That may give you the idea it is a good contrary indicator. But then consider that a year later, the bull ratio fell hard once again. This time however the S&P 500 index kept climbing.

Or consider that bullishness for bonds is nonexistent in the current Big Money poll: US Treasuries at 72% bears and only 5% bulls. That is not surprising as the “Big Money” has been almost constantly and extremely bearish on US treasury bonds since 2001. Of course, yields have been falling and bond prices rising throughout that time.

Mutual Fund Flows
According to Lipper FMI, equity funds received $2 billion of inflows this past week. That brings the total for October to $6.2 billion. This is the first month of positive inflows since April 2010. Keep in mind that this is for all equity funds (domestic and foreign) and does not include ETF flows. As expected, taxable bond funds continue to be the favorites garnering an additional $3.3 billion of inflows last week.

Hedge Fund Exposure
According to a recent report from JP Morgan (mentioned earlier today in the Morning Notes), hedge funds have increased their equity exposure to levels approaching the most bullish for the past two years:

hedge fund exposure Nov 2010

This chart approximates equity exposure by measuring the rolling betas of daily hedge fund returns compared to global equities (MSCI World index).

TSP Sentiment
Finally, a retail sentiment survey which I’ve mentioned a few times before is increasingly bullish. This week the TSP Sentiment is 55% bullish and 35% bearish. That gives us a bull/bear ratio of 1.57 - close to the August 1.61 bull/bear ratio which coincided with that short-term top.

tsp sentiment survey Oct 2010

There have been a few wild spike highs but usually it takes little more to provide an excess bullish sentiment in this indicator.

Ned Davis “Crowd Sentiment”
This is proprietary mix of several sentiment surveys and indicators as concocted by the firm of Ned Davis Research. Unfortunately I don’t have the actual figure but according to a brief mention in Barron’s over the weekend, it is in “the excessive bullish zone”. Based on its historical precedent, I would assume that it is near the 70% level since that has usually corresponded with excess bullishness and market tops. For a chart, see the July 9th Sentiment Overview.

Enjoyed this? Don't miss the next one, grab the feed  or 

                               subscribe through email:  

6 Responses to “Sentiment Addendum: Gold, Big Money Poll & More”  

  1. 1 Farhan

    Nice post Babak.

    I also consider myself as a contrarion, but sentiment only works best when in conjunction with wave count or the context. However, the problem with wave count is that it is continuously evolving and not even the Elliott Wave people know the correct wave count.

    This is why I admire the market so much. Because of its intelligence and evolutions. I want to discuss this intelligence with some really active and avid traders like you. I hope that you will visit my blog, which is designed specifically for this discussion.

  2. 2 Jim

    Your historical perspective on the Big Money Poll was most helpful.

  3. 3 Stan

    Why smooth the gold data? Us poor souls out here want to see the raw data and be very aware of what might happen in any given week. Please don’t hide it in a “smoothing” attempt. We can take it!!!! (grin).

  4. 4 Babak

    Stan, the data isn’t really hidden as much as clearer this way (smoothed). The raw weekly data is so volatile that it looks like chicken scratchings on its own. The same technique is applied to the AAII or Investors Intelligence weekly surveys for similar reasons.

  5. 5 Stan

    Would rather consider plurality of data sets than smoothing data. One could easily miss key timing indications by smoothing, whereas, strong simultaneous agreement on multiple data sets (indicators) can have key meaning. My data sets for turning point timing are dependent on daily data. Of the nine I follow, if 5 give a daily signal within a one-two day period, it’s usually a key reversal indication. Timing is everything!

  6. 6 Babak

    Stan, you make a compelling case so here’s the raw data:
    bloomberg gold sentiment raw Oct 2010

Leave a Reply