Jul 28, 2003

A few subscribers have asked me to take a look at QQQ short interest, as it has gotten a few mentions in the media recently. Short interest is the total number of shares currently being held short (borrowed by someone and sold, hoping to profit on a decline) and as of mid-July, it is the highest on record for the Nasdaq 100 tracking stock as the chart below illustrates.

We can see that short interest (the red line) has steadily grown from its inception in 1999. However, that doesn’t tell us a whole lot, as volume (the blue line) has also grown steadily, at least until the past year. As I’ve discussed with short interest on the NYSE, it’s best to view short interest figures adjusted for volume.

The chart below shows the short interest ratio for QQQ. This ratio is simply total short interest divided by average QQQ volume over the preceding month.

We can see that even after adjusting for volume, we’re still at a new all-time high in QQQ short interest. This has got to be bullish, right? If so many shares are held short, then that provides a lot of fuel for an upside explosion as these trades cover their positions. Well, in theory, that’s how it SHOULD work, and when we look at broader measures such as that for the NYSE, it DOES work much of the time, but I don’t see that type of relationship in the QQQ’s. The history here is very short, only a few years, so it’s tenuous at best to draw any conclusions from QQQ short interest data. Just looking at the chart above, I can’t see any solid relationship between extremely high or extremely low short interest and future market performance.

I’ve tried looking at this data numerous ways – using average yearly volume instead of monthly volume, looking at short interest as a percentage of total shares outstanding, etc., but the conclusions are all the same. And the conclusion is that there is no conclusion – this data appears no better than random at predicting market behavior, so today’s high short interest level in QQQ options should NOT necessarily be taken as a bullish sign. The most likely reason for this is that the QQQ’s (nicknamed “Cubes”) have become a trading and hedging vehicle for institutions. Yes, individual traders obviously still trade the Cubes, but take a look at the following table:

WHERE THE BIG BOYS PLAY QQQ Block Volume
Issue	Volume	Block Volume	Block Volume as % of Total
SUNW	60.5	16.8	28%
CSCO	43.5	8.8	20%
MSFT	41.3	10.3	25%
INTC	40.9	6.1	15%
ORCL	27.8	6.2	22%
JDSU	22.6	4.3	19%
RFMD	21.7	2.9	13%
BRCM	21.3	3.9	18%
ADCT	20.8	4.6	22%
AMAT	19.8	3.3	17%
QQQ	56.7	24.3	43%

This table shows the 10 QQQ components with the highest volume today (the volume figures are in millions of shares). Next to the “Volume” column is “Block Volume”, which is the total number of shares that were traded in single blocks of 10,000 or more shares. With the Cubes trading around $31.50 per share today, that’s a minimum trade value of $315,000. Certainly some individual traders can and do move that type of size, but not too many. If you scroll through the list, you will see that only Sun Microsystems showed block volume greater than 25% of its total. The QQQ’s, however, showed over 40% of its volume as block volume. I think this is very important to understand when looking at things like short interest or put/call values on the QQQ – it is an institutional product, and should be interpreted as such.

Speaking of put/call values on the Cubes, they continue to be quite high. Daily put/call readings over 3.0 have been common lately, but as I showed a few weeks ago, there is not a consistent relationship between high QQQ put/call ratios and future market performance. They lead to rallies as much as they lead to declines, and I don’t put any faith in them as a leading indicator. Because of the inconsistent behavior of the QQQ put/call ratios, and the fact that QQQ options have made up 20% or more of total equity option volume four times in the past month, I have added a new set of put/call ratios to the site. When you click on the link for the complete list of indicators on the site, under the put/call section you will see the daily, 10-day moving average and 21-day moving average of the equity-only put/call ratio minus QQQ options.

At the low in July 2002, the 10-day average of the non-QQQ equity p/c ratio (the red line) rose above the equity p/c that includes QQQ options (the blue line). This shows that smaller investors were likely scrambling for downside protection to a greater degree than institutions were. Conversely, last Winter the non-QQQ ratio dropped substantially below the traditional ratio, suggesting smaller traders were significantly more optimistic than the larger traders. Now, we are seeing one of the widest divergences in the history of the two indicators – the traditional p/c ratio has been rising to a point where it is now neutral, but the non-QQQ ratio is still extremely low. The suggestion here is that once again, smaller traders have become considerably more comfortable with calls than puts. In my opinion, this is not a positive sign for the market.

As of tonight, the “score” for our short-term indicators is neutral at exactly 50%, as every one of the indicators is in some part of neutral territory. This has only happened twice in the past year (12/23/02 and 1/16/03). If we look over the past three years, then there have only been a handful of instances when 10 or more of the 14 indicators were in neutral territory, and each of them preceded a major market move within a week. With so many of our indicators in neutral, and so many important economic reports due tomorrow and Friday, the ingredients are there for a major move to occur, which should certainly move us out of the range we’ve been in for two weeks, and will probably move us out of the range we’ve been in for two months. With so many bets having been placed on both sides of the market, once we break the range, we should travel some distance as those who were wrong cover their positions and jump on the bandwagon. If a break of the range holds for the first day or two, I would not be too quick to look for a reversal. I’m sticking with my preference of short trades as long as we are under 1015 on the S&P 500, but if we rally and break and hold 1000 initially, then 1015, I am backing off the short side completely.

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.

I’ve been asked frequently lately what my thoughts are about the Rydex Nova/Ursa ratio. This ratio is very simply constructed by dividing the assets in the Nova fund (which profits when the S&P 500 rises) by the assets in the Ursa fund (which profits when the S&P 500 falls). The theory behind the indicator is that when the ratio is very high, showing a large amount of assets in Nova versus Ursa, then the Rydex mutual fund timers are very optimistic about the outlook for the market, and in true contrary manner, the market will more likely fall than rise. Conversely, when the ratio is low, then that shows a great deal of pessimism among these traders, and the market will likely rise.

Currently, the Nova/Ursa ratio is 0.28, which means that for every $1 in the Nova fund, there is $3.50 in the Ursa fund. This is very low historically, as the mean value since 2000 is 1.08, albeit with a large standard deviation of 0.66. The interpretation of this, of course, is that since these traders are so pessimistic, and we’re so near the recent highs, then there is a lot of fuel to propel the market higher and break us out of the recent trading range.

However, there is a severe problem with this type of analysis, and that problem is the Dynamic funds. In May 2000, Rydex introduced a line of funds that are leveraged 2-to-1 and can be traded once intraday. These funds have become immensely popular with Rydex timers, and it has substantially changed the way these traders move their money among the various funds.

The following chart shows the relationships among these funds. For the moment, just concentrate on the red lines (the “old” un-leveraged funds) and the blue lines (the “new” leveraged funds). Ignore the gray line for now.

It is obvious from the chart that the red lines – the un-leveraged Benchmark funds – have shown stable or declining asset bases since May 2000, while the blue lines – the sexy new leveraged Dynamic funds – have enjoyed a steady increase in assets. You will note that I haven’t mentioned anything about whether these funds are long or short the market, and it doesn’t really matter. The fact is, there has been a steady decrease in interest in the old funds, and a steady increase in interest in the new funds. There is one exception – the gray line.

That gray line is the Ursa fund, which is kind of the granddaddy for bears. Ursa is a short-oriented fund that bets against the S&P 500, and is not leveraged. I suspect the reasons why Ursa has bucked the fate of the other “old” funds is because there is a big difference between it and its “competitor” – the Tempest 500 fund, which is leveraged 2-to-1 – and also it has gained an increasing amount of interest from institutional-type money looking to hedge their portfolios. That last assertion is based on the fact that over the past few months, only Ursa has seen one-day changes in its asset base of greater than 25%. Not just once, but 5 times since May, and this is off of an asset base of nearly half a billion dollars. So we’re seeing some big-time money moving quickly in and out of the fund, and that cannot be attributed to retail traders alone.

In any event, the point of this whole exercise is to show that Nova has seen a long-term secular decline in its asset base, while Ursa has seen a long-term secular rise in its assets. That fact alone can account for the low Nova/Ursa ratio, regardless of what the market has done, and regardless of the trader sentiment behind the changes. Correcting the ratio for the NAV of the funds does absolutely nothing to change this phenomenon. There are an untold number of ways to correct for this bias, as the charts on the site have been. The reason this is necessary can be shown by example:

This chart shows the interaction between the S&P 500 (top), the Nova/Ursa ratio (in the middle in red) and our own Rydex Bull Flow (bottom). In August 2002, the Nova/Ursa ratio began to rise, but only reached a mediocre neutral reading at best (#1 on the chart). However, as there was a significant amount of assets flowing into the leveraged funds at the time (which went unrecognized by the Nova/Ursa ratio), our Rydex Bull Flow entered extreme territory (#2 on the chart). This tipped us off to the fact that we were seeing a huge amount of speculation by Rydex traders looking for more upside in the market, and not surprisingly we saw the market decline into October to “reward” them. Usually, the two measurements will track fairly closely. But when there is a large shift into the leveraged funds – when this Rydex information becomes the most useful – then the traditional ratios fall short.

Currently, the various Rydex ratios as posted to the site are neutral. We are NOT seeing undue pessimism on the part of retail Rydex traders, despite what the Nova/Ursa ratio suggests. We are not seeing undue optimism either. To me, that is how it should be considering the trading range we have seen, and it doesn’t give us much of a clue as to whether we will be more likely to break it to the upside or the downside.

One thing I think is interesting is that the percentage of assets in the Rydex money market is extremely low, close to the lowest seen in the past couple of years. We usually see this type of activity when traders are confident of market direction, one way or the other. Often, that has meant that traders were complacent, and the market declined soon afterward. However, it also happened in mid-May when the S&P was flirting with its breakout area around 930/940. I don’t think the low level of money market assets tells us much about direction at this point, but it does indicate that a lot of bets have been placed. When the market breaks either way, we should see some momentum carry forward, as the side that was wrong covers their positions and very likely jumps on the bandwagon.

In the weekly reports that were released this weekend, there weren’t any major changes or surprises. The sentiment surveys remained enthusiastic for the most part, as our AIM model is STILL hovering right around its too-optimistic extreme of 40%. In the Commitments of Traders report, small speculators were essentially unchanged in their market position, while commercial hedgers reduced their net short position by just over 8,000 contracts. Overall, the tone of the report remains bearish for the market, though it is not particularly troubling to me. On the NYSE, specialist shorting activity continues to be subdued, while public shorting remains high. As I’ve pointed out several times over the past couple of months, from a very long-term time frame (several years), this data is giving off one of the most bullish (for the market) configurations in its history, and it is one of the reasons I believe a break of 1015 on the S&P could carry us significantly further. I have no doubt that this data is being affected by hedge fund activity, especially in relation to the large amount of convertible bond deals that have been issued recently, but it still warrants enough respect that we should keep monitoring it.

Several of our breadth measurements are oversold or very close to it. As I said last week, how the broader market responds to this will be very telling. We’ve done nothing but see-saw so far, and the longer we go without a meaningful rally that holds, the more likely we will break support, oversold readings or not. If the bulls are to regain control, I think we must rally strongly in the next day or two. I don’t see much at all from a sentiment perspective that supports such a move, but as of tonight I don’t have much that argues against it either, at least in the short-term (under 5 days). Longer-term (out several weeks at the very least), my preference remains with shorting rallies as long as we are under the mid-June high of around 1015 on the S&P 500.

On an administrative note, beginning tonight you will notice that our charts are converting to the new format I’ve discussed before. I know many of you were used to the old ones, but this new format allows for greater clarity, and most importantly it will allow me to quickly build out an extensive chart archive. Please be patient over the next week or so as the rest of the charts are converted, and the archive gets developed. As always, constructive feedback is welcomed.