April 3, 2004 - In March 2004, the stock market changed trend for the first time since March 2003. In January and February, we saw this coming. The breakdown of the Dow Transports in January and the continued rally of the Dow Industrials began the first Dow Theory Sell Signal since the March 2003 low, and the resistance provided by Fibonacci retrace levels on several key indexes gave months of advance notice before the S&P 500 and the Industrials actually made a move lower. The NDX had also been weak from its January high, making impulsive moves to the downside and corrective rebounds. These key divergences between the indexes, along with Fibonacci and other technical indicators detailed in the last several months, sent clear signals that the trend was changing.

What remains unclear however, is the nature of what the trend has changed to: is this a correction on the same scale of December 2002 to March 2003 or the start of the next leg of the Bear? All major indexes currently remain in potential bearish wave counts for the intermediate-term, but there are also alternates. It is clear that the up-trend that began in March 2003 is over, and that we are either in a correction of that rally or in the very early stages of a much larger decline. There is a lot to cover this month in addition to the stock market, so let's get started.

Chart of the Month

Stock Market Update

The chart on page 1 of the Semiconductor Index is a good example of what you will find a lot of these days if you browse through charts of indexes and individual stocks: broken trend lines. The SOX has broken its trend line from its October 2002 low, which it had used as support for the last year from the March 2003 low. The Russell 2000 has broken its trend line from March 2003. Major stocks like Microsoft, GM, Alcoa, Boeing, Caterpillar, Citigroup, Hewlett Packard, as well as the QQQ have all recently broken major trend lines from either the 2003 or the 2002 low. Even more speculative stocks like Amazon.com have rolled over and broken its trend line from its 2003 low. These are subtle signs that the highs in first quarter of 2004 are likely more than a temporary top. And there are more hints.

In January, the Dow Transports broke down from a closing high right at its 61.8% retrace of its 1999-2003 decline [see one of the short-term Transports updates for a chart]. The Industrials continued to rally, but for the first time since a year ago the Transports did not confirm. This was the first stage of a Dow Theory Sell Signal that has since been confirmed with the breakdown of the Industrials in the last month [see the February Dow Theory summary under Recent Analysis at Sitkapacific.com].

The below chart of the Industrials is the same chart we have been looking at since late December. As we have discussed before, the Red zone in the chart is significant for two reasons: the upper limit is the Fibonacci 78.6% retrace of the 2000-2002 decline, and the lower limit is the most common internal Fibonacci relationship within these types of corrections - where wave C is 1.618 times wave A. The breakdown out of this area is very significant for the long-term wave count, as it presents us with a potential A-B-C correction from the 2002 low that is complete and internally consistent. As long as the Industrials and other major indexes remain below their 2004 highs, the long-term trend is considered down.

In the short-term, the Industrials have declined from its 2004 high in 5 waves and have since rallied to Friday's closing high at 10,470 - which is the exact 61.8% retrace of the preceding decline. As we have discussed in recent short-term updates, it is possible that the decline off the 2004 highs in several indexes looks corrective, which may hint that the recent decline to 10,007 on the Industrials is a wave A of a larger correction. However, because of the impulsive nature of the Industrial's move off the high, it remains highly likely that there will be another move below the recent low before such a correction would be complete - and we will then be able to better judge if the larger degree trend has in fact turned down. So even if this is a correction in which the Industrials will eventually make another post-2002 high, a move below 10,000 is more likely in the next several weeks.

The action of the Industrials and the Transports become even more significant when considered in the context of other indexes like the S&P 500. The recent high of the SPX was within 3 points of the 50% retrace of the 2000-2002 decline. If this rally from October 2002 has been a Bear market rally, as appears highly likely, then a high near a significant Fibonacci retrace like the one we just saw would be ideal. Since last month's update, the weekly MACD has given a clear sell signal as the SPX declined below its 10-week moving average.

If the recent high at 1163 holds, it would be a very bearish omen for the S&P 500 in the long run. The 2000-2002 decline completed at 50% off the 2000 high and the next leg down would likely be roughly equal in percentage terms to the 2000-2002 decline, which gives us a long-term target of 581. You can see a clear example of this equality on the SOX chart on page one, and on the Nikkei from the 1989 top to the 2003 low. Whether it takes a few years from now or a decade or more, the target will likely be met.

In the short-term, the rally from the low at 1087 has given the decline off the 1163 high a distinct A-B-C look. But it is also possible that this rally has been a wave C of a Wave 2 Flat correction. In fact, the SPX is responding strongly to the Fibonacci retrace levels off of the would be wave 1 low at 1105. So at this point, the short-term wave count of the decline from the 1163 high can be considered impulsive or corrective. This is also the case for the NDX.

Bottom Line: If this decline has been the start of a large correction and not the start of a large degree decline, it is highly likely that recent low is just the bottom of the first leg down. Such a correction would probably last for several more months, be highly volatile, and leave the December 2002 highs on the SPX and the Industrials unbroken. But if this is the start of a larger move down we may move down quickly in the following weeks and months. We are entering a seasonally weak period that lasts through October, and some long-term cycles are turning down this summer as well (the 4-Year cycle is discussed below). Unless the 2004 highs are broken, short positions on the indexes should be held.

The S&P 500/VIX Ratio

In last month's Market Focus, the S&P 500/VIX ratio was highlighted as the market approached the March high. Now that we have had a decent decline from the high, it seemed a good time to update those charts to see what kind of progress had been made.

For those of you who have not seen this ratio before, it can best be thought of as an indicator of when the market is too pessimistic (which is a good time to buy) or when the market is too optimistic (which is a good time to sell). You can see in the charts below the past history of buy (green) and sell (red) signals, in both bull and bear markets. Using only this indicator, you would have been long the market for almost the entire rally from 1995 to 1999, minus a few months during 1995-1996 and during the 1998 decline. Having exited in 1999, the market continued higher for another year generating increasingly stronger sell signals. Although this was an early exit, in hindsight it was a very good time to step aside. The first major buy signal came at the lows after 9/11, but was quickly followed by another sell signal in March 2002, just before the waterfall decline to the 2002 lows. The 2002 lows and the March 2003 low generated buy signals, which remained in effect until December 2003.

If you heeded the last sell signal, you were well ahead of the recent decline. The SPX/VIX is at levels not seen since the 2000 top, which is even more ominous this time since the SPX (the numerator in the ratio) is 26% below its peak - this implies the market is even more complacent than it was in 2000. If you look closely at the charts below, you'll see that the March decline had only a negligible effect on the SPX/VIX ratio, which remains high in "Sell" territory. 401k and other long-only equity accounts should be in cash until the next buy signal is generated.

An Advance/Decline Sell Signal

The cumulative Advance/Decline line put in a negative divergence on its MACD last month as the March decline got underway, thereby giving another sell signal to compliment the SPX/VIX ratio. You can see that these sell signals were very good at signaling the start of a significant declines during the 2000-2002 period, but have been short-lived during the post-2002 period.

As of Friday, the A/D line has retraced its March decline to close within a few hundred points of the recent high (as a side note, Friday's jobs-related rally had only 167 more advancers than decliners, a weak showing which you can see on the chart below). It has been wise over the past year to remove sell signals from this indicator when the A/D line moves on to a new high, and a continued rally next would almost certainly produce a new high. A decline from Monday onwards, as short term charts of the indexes suggests is likely, would preserve the sell signal.

Approaching a 4-Year Cycle High

If the wave count presented on the page 2 Dow chart is correct, the 2002 low was Wave A of a larger A-B-C decline. And if the rally from the 2002 low is an A-B-C correction of that decline, it would correlate very well with the 4-year cycle that is scheduled to top in 2004. You can see a very long history of the 4-Year cycle on the Dow Industrials at Sitkapacific.com under Recent Analysis, and it is also shown on the S&P 500 from 1990 in the above chart.

In most non-bubble years from 1900-2002, the 4-Year cycle has been a good price cycle, in that the price of the market usually trended up and down with the rising and declining phases of the cycle. In strong bull markets like from 1995-2000, the declining phase of the cycle had little effect on the trend, but it had a major effect on volatility. If you compare 1995-1996 to 1997-1998 and 1999-2000 to 2001-2002, you can see how the volatility increased markedly after the cycle highs (shown by the dashed lines). The last cycle low was in 2002, 4 years from the volatile summer of 1998. Since the 2002 low we've had a rally that has been characterized by low volatility, if nothing else. Technical indicators like Average True Range are at their lowest levels since 1996, which indicates inter-day volatility has come way down from its peak in 2000. However, if the 4-Year cycle stays true to form, that is about to change.

If the rally from October 2002 has been a correction that is now complete or near completion and the larger degree trend from the 2000 high remains pointed down, the following 2 years may be as volatile as the 2001-2002 period was. The fact that 2004 is the top of this cycle also gives added weight to the current extreme SPX/VIX ratio levels. In 2000, the high of the S&P 500 was seen in March even though the market attempted to chop its way higher through the summer as the 4-Year cycle was topping. Of course the risk at that time was very high, and those who ignored the signals like the 4-Year cycle, the SPX/VIX ratio, the MACD and others did not have to wait too long before the opportunity to get out was gone. There are many signs that the market is in a similar situation now. If you are a trader and can profit from another short-term high, then there may be more profit on the long side if the market decides to make a new high this Spring. But if you are an investor, this is a time to take a step back from the short-term and see the larger picture for what it is.

Bond Market Update

The economic numbers on Friday gave the stock market a bounce, but the far more important action was in the Bond market. The yield on the 10-Year Treasury gained over 29 basis points on Friday, which is a significant move by any measure. More importantly, it raises the possibility that the corrective decline in yields from last summer's high is over. Let's take a look at the chart on the next page.

The larger picture of the bond market from May 2003 in bullish for yields. Since the 4Q of 2003, it has been clear that the bond market decline in yields from the 3Q 2003 high has been a corrective one. This corrective decline followed a impulsive rise in yields from the 2003 low of 3.074% to 4.668% on the 10-Year, which was the first strong hint that the trend had turned up. Since last summer, the corrective decline in yields has responded very well to Fibonacci retrace levels and has been strongly overlapping, further suggesting a break down below the 4.668% low in 2003 was likely - it has just been a matter of waiting for the correction to end.

Back in January this newsletter identified the bond market as the clearest potential short setup of 2004, and it is likely that the time to trade is approaching (this will be detailed in a Trading Focus update). From 2000, yields have been declining in an expanding wedge shown by the green trend lines below. A break of the upper trend line will likely lead to a rise in yields to the upper blue trend line, highlighted by the red area. Given last year's throw-over below the lower blue trend line amid frenzied buying, it's also possible that the larger-degree trend has turned up and yields will eventually continue up out of the channel. We'll cross that bridge when we get to it, but for now a rise in the 10-Year yield to the 5.8% area seems likely after a break above the upper green trend line.

The economic implications of a rise in bond yields may be significant. As is well known, the rise in housing prices during the bursting of the stock market bubble helped hold off a much more severe recession in the economy. However, even as housing prices have risen to new highs in many places, homeowner's equity on a national scale is at an all time low as the money that was extracted during refinancing supported consumer spending. A rise in interest rates in 2004 may be enough to reduce the momentum of the housing prices, which may eventually reduce an important support of this post-bubble economy.

It has been suggested by many that housing prices have a much bigger influence on consumer behavior than the stock market. If true, rising yields in 2004 may not be as inflationary as some think. On the chart above it is clear that bond yields remain in a long-term deflationary down trend, and a rise to 6% or so on the 10-Year will not change break that trend. In fact, it may be just the thing to tip the US into outright deflation as consumer demand wanes. How the bond market would react to deflation in the US is anyone's guess. Comparing a potential US deflationary environment to that of Japan's ultra-low interest rate economy is dangerous, as they have one of the highest domestic savings rates in the world accompanied by huge trade surpluses - the opposite of the situation here. We will just follow the charts, and for now the charts suggest yields will more likely rise beyond the 2003 high over the next year than fall to the 2003 low.

US Dollar Update

A quick look at the US Dollar Index suggests that the decline in the dollar from its 2001 high is coming to an end, if it has not ended already. The February-March rally in the dollar completed a positive divergence on the Weekly chart below, and there are wave counts that show a complete 5 wave decline at the February low. The Daily chart at the bottom of the page shows a potential Head and Shoulders bottom, which would be quite bullish over the next month or two.

If the dollar breaks the upper blue down trend line, it will be likely that the decline from the 121 high is complete and a large correction has begun. Such a correction is not expected to approach the old highs, but it may last for some years and retrace a significant percentage of the preceding decline. If the February low was not the end of the decline from 2001, the next low will probably put in a stronger positive divergence on the Weekly MACD. Those who have ridden the commodities and foreign currency bull over the last three years should take note of charts like these. The Canadian dollar chart on the following page is one of many currencies that are looking bearish against the dollar.

Gold and $XAU Update

The last update for this month's Market Focus is the XAU. At the turn of the year, it became clear from the chart below that Gold bullion would out-perform gold stocks for the time being. That has definitely been the case over the last tree months, as the HUI and XAU have remained range-bound while the metal has gone on to new highs. As always, the charts reveal the trends and right now the trend remains for Gold to out-perform gold stocks.

The past 18 months have been favorable to mining stocks, as the stock market advance and the decline in the US dollar provided all the needed momentum for the HUI and the XAU to advance strongly. However, with the probability of a rally in the Dollar within the next few years accompanied by a likely decline in the stock market, those favorable conditions may be fading. Of course this is not mere conjecture; the charts are already dropping hints that the trend may be changing. Mining stocks can under perform in an advance and a decline, and the XAU charts on the following page suggest a decline in on its way.

In the short term, the XAU declined in 5 waves from the early January 2004 high, and has since traced out a corrective-looking rise to the 61.8% retrace. Friday's decline dropped out of a rising wedge, complete with a negative divergence on the MACD. A continued decline below the wave B (blue) low would be bearish.

The long-term chart below shows that the double top high completed in January was not at a random level: it topped just one point beyond the 61.8% retrace of the 1996-2000 decline. Needless to say, the below chart is potentially very bearish for the XAU. In the short-term, the bears may have the upper hand. Long term bulls will need the XAU to eventually rise above the January high and make rally from the 2000 low look less like an A-B-C correction. For now, a solid break and retest of 113 on the XAU would be a very good buy signal. Without such a beak, the trend appears to be down.