Aug 11, 2003

http://www.sentimentrader.com

Wednesday, August 13, 2003 7:10 PM EST

One of the hallmarks of sentiment since May has been the overwhelming bullishness displayed by the four major sentiment surveys. A lesser-known survey, but one that has been just as accurate from a contrary standpoint, is the one from the website lowrisk.com. Curiously, the respondents to that survey are now showing some of the most BEARISH readings in its history.

We know that the Investor’s Intelligence survey covers investment newsletter writers; Market Vane polls Commodity Trading Advisers; Consensus covers a range of investors including brokerage house analysts and others; and AAII caters to individual investors. Lowrisk.com, on the other hand, is something of a wild card. Because it is a web-based survey, and you don’t need any particular credentials or membership status to participate, anyone can post their opinion as to whether the DJIA will be flat, or up or down by 2% or more 30 days hence.

As far as I know, responses are not tracked by IP address or any other tracking method, so one individual can post as many times as they want. So what we have is a survey where we don’t know the audience, and don’t know to what extent any one individual can skew the results. That is usually a recipe for me to stay away, but I’ve always been intrigued by the fact that the lowrisk.com survey is as (or more) accurate than the other, more established and widely-followed surveys.

For responses gathered on the site from August 4^th through August 10th, bearishness was rampant. Only 17% of the respondents thought the Dow would rise by 2% or more over the next 30 days, while 57% thought it would fall by 2% or more. This results in a “bull ratio” of 23%, which is calculated by taking the number of bulls divided by bulls plus bears. A bull ratio of 23% is 1.5 standard deviations below the mean since 1997, so it’s fair to say that it is an extreme amount of bearishness.

In the charts below, I have highlighted each prior instance in the survey’s history of a bull ratio at 23% or below. The vertical blue lines show each of these occurrences.

First, 1997 – 2000:

Now, 2001 – 2003:

We can see that out of the 10 occurrences, only 1 lead to a dramatic failure (right before 9/11/01). 6 of them were very near intermediate-term low points, such as the one in March of this year. The other 3 occurred somewhere in the midst of a rally that was already under way, and each time it lead to further upside.

These survey respondents never really bought into the Spring rally in the first place, as the bull ratio didn’t make it over 50% even once since March. While this was a good indication to stay long from a contrary perspective, it is a bit troubling in that it is a change in character in this survey, and it is not corroborated in the least by any of the other surveys that I follow. There is a possibility that because of the lack of “rules” surrounding the survey, it is perhaps open to manipulation more than the others. However, I prefer to not second-guess what an indicator is suggesting unless there is a good reason to, and right now I don’t think there is. The bearishness displayed here is one positive indication among the many negatives for the market.

There is a real lack of topics to go over tonight. The overbought condition from many of our short-term indicators on Tuesday has dissipated for the most part with today’s modest decline. The lack of interest shown by the pathetic volume of late is also being reflected in our sentiment indicators, as no real emotion is on display at the moment. As I said earlier this week, when we have conditions such as this I defer to the trend in place at the moment, which I would consider down. If we regain 1005 on the S&P, I would switch to neutral, and a close above 1015 would force me to be bullish. Unless that happens, or we see more displays of pessimism, my preference remains with the short side.

- Jason Goepfert

Disclosure: long OEX puts

This disclosure is not intended as trading advice in any form. It is meant as a note to subscribers that the author may have a position directly affected by the market outlook reflected in the commentary. Although the author takes great pains to remain objective in any commentaries, it is only fair that readers should know that the author may have taken positions in accordance with his market outlook. Positions can and do change at any time, without notice to the reader.

Monday, August 11, 2003 9:25 PM EST

In 2001, the price of gold made a major low, bottoming out at $255/oz. and rising a little over $100/oz. since then to around $362/oz. today (spot price, New York close).

During 2003, gold has been forming a nice technical triangle pattern, with a series of lower highs and higher lows. According to the textbooks, this is normally a consolidation pattern, with price expected to break in the same direction as the major trend which preceded it – in this case, higher. Perhaps because of that, or because precious metals (as an asset category) have been getting more mainstream press, I have been getting more and more requests to take a look at the sentiment picture for gold, which I will do tonight.

Sentiment is typically most useful when there are very segregated pools of traders, with ample information to tell them apart. The more liquid the market, and the more interest it holds for the most people, typically the more information there is available. The stock market holds the greatest interest for the most people, so sentiment data is readily available and getting even more so. The bond market comes a distant second, and commodity-type markets such as currencies, agricultural products and metals come third. Due to that, sentiment data becomes more difficult to obtain, and more difficult to decipher, the lower on the food chain you go. I’m not saying these other markets are not important, in fact the truth is the exact opposite, but I’m talking about the amount and the quality of sentiment data available. So, what we have to work with in gold is significantly less than what is out there for equities.

First, let’s look at a chart of how stocks, bonds and gold have moved long-term. On the left is a chart going back towards the beginning of the major bull move in equities in 1982. On the right is a chart of the most recent period, encompassing the bear market in stocks and bull markets in bonds and gold. For the purposes of these charts, bond “price” is simply the yield on 30-year Treasury Bonds subtracted from 100.

I don’t want to focus too much on the intermarket relationships here, as you can get better information on that elsewhere. Let me just say that the theory of how it “should” work is that higher gold prices are anticipatory of increasing inflation. That increasing inflation is bad for paper assets like stocks and bonds. So, when you see a falling dollar and rising gold prices, it typically means that inflation is coming down the pike, which is bad for holders of bonds and eventually for holders of stock as well. However, there are so many variables involved with these relationships, and the lead and lag times can be so great, that it is not worth pursuing in this space.

With gold likely to break out of its triangle at some point soon, does sentiment support the technical argument that the most likely direction is up? Not surprisingly for an asset that is already up so much, the answer is “not really”.

First, let’s look at the Commitments of Traders data for gold futures. This information is interpreted in exactly the same manner as it is for equity futures – commercial hedgers are the big boys, the smart money. We typically want to follow their lead, although they can often take a long time to build up their positions. Small speculators, on the other hand, are the “dumb” money, and should be used as a contrary indicator when they become extreme in their positions.

First, this chart shows the entire history of the C.O.T. data for gold futures:

We can see that there are four periods in history where small specs were significantly longer than commercial hedgers for a considerable amount of time – 1987, 1993, 1996 and now. In the past, such wide gulfs between the two trader sets have resulted in long-term stagnation or declining prices for gold. The current period looks most like 1993, when it looked like gold was going to break out of a long-term triangle pattern to the upside, only to go nowhere for years on end before ultimately breaking down once again.

The next chart looks more closely at the most recent time period (please note the different scales for commercial traders and small specs, which was done to more clearly show the interaction between the two groups as opposed to the absolute levels of their positions):

Here, we can see that this information is often useful on a smaller time frame as well. Large, sudden spikes in opposite directions between the traders often coincide with intermediate-term turning points in gold, most clearly seen in the first couple of months of 2003.

At that time, small specs showed a large jump in bullishness, building their net long position up to a new record high. At the same time, not surprisingly, commercial hedgers were doing the exact opposite, becoming the most net short in their history. The result was a nearly $60/oz. drop in gold over the next couple of months. Once again in June of this year, commercial traders built up a large net short position, and once again gold dropped into mid-July. Now, with the rally in gold over the past few weeks, commercials have again been building up their shorts, and small specs have been building longs, though not to a very large degree. From the history of these trader groups, it appears as though any further rally in gold will be met by selling from large, informed traders and buying from “dumb” money. If so, that would put a cap on outsized gains in the metal. Currently, I view this data as moderately bearish for gold, with the possibility of it becoming extremely bearish in a very short amount of time.

Speaking of dumb money, the hands-down all-time favorite award has to go to the mutual fund timers who frequent the Rydex funds. I’m familiar with the argument that these funds are used by professional hedge fund managers for speculation or hedging, and I know that to be true to some extent, but it cannot be argued that these traders, considered as a group, get in and out of the markets at anything but the least opportune times.

The chart below shows the assets in the Rydex Precious Metals fund, corrected for the NAV of the fund itself. Correcting for NAV is done so that we can get a truer picture of the sentiment behind the asset moves. If assets in the fund hit a new all-time high, that doesn’t necessarily tell us much if the fund itself is also hitting new all-time highs.

We can see from the chart that these assets moves tend to track the movements in gold itself. When pessimism becomes high (the blue line is on the lower end of its trading range, showing a lack of enthusiasm among Rydex traders toward gold), then the yellow metal has often formed at least a short-term low. On the other hand, when enthusiasm is greatest, and assets are pouring into the fund more than what could be justified by the performance of the fund, then we often see a setback in the price of gold.

This NAV-adjusted ratio is very close to a new three-year high, despite the fact that gold itself is well off the highs seen in February. While there has been a trend higher in this ratio over the past year (following gold itself), the fact that assets in the Precious Metals fund are hitting new highs while gold is not has to be looked at as a bout of speculation by those very often wrong on market direction. I would consider this a minor negative for gold.

Lastly, let’s look at a proprietary put/call ratio that I’ve constructed for gold. This is ratio is really an index of put/call ratios on several gold stocks and indices, weighted by their (negative) correlation to movements in gold. The chart below is a 10-day average.

Like most put/call ratios on securities or commodities other than US stock indices, I don’t find this consistently effective and I don’t give its readings too much weight. In any event, the put/call ratio is quite low, telling us call volume has been heavy in relation to puts. This is usually a sign of investor confidence in rising prices (or at least not dramatically declining ones), and from a contrary point of view should be bearish. I would view this as a very minor negative for gold.

Overall, I don’t see anything from a sentiment perspective that would have me excited about being long gold. I’m well aware of all the anecdotal evidence suggesting “nobody respects gold” and because of that, it should rocket higher. However, when we move away from anecdotes and put some numbers to the guessing, it appears as though not only do many traders respect gold, they expect it to perform quite well in the future. Those in a position to know well whether that is likely – commercial hedgers – do not share that enthusiasm and in fact are betting quite heavily against it. I don’t pretend to be an expert in metals, and admittedly have little idea about the underlying supply/demand forces at work behind gold, but from my limited perspective, an investment in gold seems like it would be better from a risk/reward point of view when it is a little less obvious.

I’ve been asked to include the Indicator Bias and Model Glance tables from the site in these emails, so you’ll now find them at the beginning as you’ve no doubt noticed. As you can see, we’re in neutral across a broad cross-section of indicators and all the models. At times like this, I defer to whatever trend is in place on whatever time frame I’m trading. Currently, I would have to say that the trend is down except on a weekly time frame (which I would consider neutral), so my preference remains with the short side at this point.

- Jason Goepfert

Disclosure: long OEX puts