Houston, We Have A Technical Problem

As Long-Term Capital Management (LTCM) has demonstrated, history is just that — subject to change at any time. Although the case of LTCM challenges the third assumption, the true technician knows that "history repeating itself" does not mean exact repetition. After all, wars have been waged throughout history, yet no two have been identical.

Similarly, the discounting of free markets is often misinterpreted. It is frequently used to combat the set rules of "technical analysis"; if the markets discount everything and technical indicators are open information, then that information must be discounted and therefore already priced into the market. This reasoning has some merit, but the downside break of a trendline in GE will probably not show up on the front page of The Wall Street Journal, whereas GE beating earnings estimates by $0.25 a share will. In the former, only technicians will notice and perhaps take action. In the latter, the entire market will notice and take action, which would be expected to be dramatic in relation. In reality, the market is merely the output of the participants' reactions to news, earnings, economic data, and so on.

The one indisputable fact about any open market is that prices move in trends. This can be summed up as the snowball effect; as the trend — that is, the snowball — gains mass and speed, it adds to itself until the bottom of the hill — or recently, the top of the market — stops the trend and sideways action ensues.

PATTERN PROBLEMS

So what bearing does a discussion on the tenets of technical analysis have on this article? Simply put, the markets face a problem similar to the one the first US mission to the moon faced — a major technical problem. The markets are now at a long-term and significant junction with little or no guidance for what lies ahead. I say "significant" because of the confluence of major technical breakdowns and the results of past comparable bubbles.

Take a look at Figure 1, a monthly chart of the Standard & Poor's 500 index from 1987 to the present. The indicator at the bottom is the moving average convergence/divergence (MACD), an indication of the current trend's strength.

Figure 1: Long-term chart of the S&P 500.

The brown uptrend line has been the general trend beginning with the start of the recent bull run (the 1982 low). As you can see, this trend was broken in mid-2002; the June decline put the index clearly below the dominant trend. This in itself is not a concern. As the markets moved higher from 1982­2000, the trend was only tested in 1995 and post­September 2001; it spent the rest of the time well above the straight-line action.

Note the three cyan circles on the chart with the designations of "LS" (left shoulder), "H" (head), and "RS" (right shoulder). These represent a head and shoulders (H&S) pattern. Typically, this formation is considered to be indicative of a trend reversal. An inverted form of the H&S pattern is often found at market bottoms, just as this formation is synonymous with market tops.

The general theory behind the head and shoulders formation is part psychological and part technical (support/resistance-based). The psychological portion dictates that the initial high made by the market corrects itself like any other new high should. Following the brief correction, the prior uptrend resumes to make new highs. As the market heads up with these new high prices (as with the first set of new highs), along comes the expected profit-taking and correction. The problem arises when the second correction is not quickly resolved and the market does not rebound as expected. At this point, the resolve of the market is tested. A slight rally attempts to bring the market back to its prior high, but moves sideways and eventually fails.

A buy and hold investor (who make up the majority of the market) would have bought on the dip following the first high (LS), not wanting to miss out on the bull move, but not wanting to chase prices higher, either. Assuming the market continues along its bullish trend, this investor would be long at a new high (the head). Although it is possible to take the profits, it is more likely he/she will wait to buy more on the next dip to continue riding the uptrend, or simply hold on (consistent with a buy and hold strategy).

As the market moves sideways and down to consolidate the new highs, the investor may buy even more shares. Unfortunately, the dip becomes prolonged, but since there is no major decline as of yet, no harm is done. After this longer than usual dip, the market makes another attempt to take out the old highs (RS). This fails as well, which would mean the buy and hold investor faces losses and a declining market. The juncture or breaking point in an H&S formation is the "neckline" or low of the two shoulders. In this particular example, it is in the 950­960 area. Once prices broke the neckline, the winning trade became a losing one. The only options are either to sell at a loss or hang on for a much longer period.

CHANGING STRATEGIES

Most of you probably found yourselves on one side of this formation. The "buy on dips" strategy that worked for almost 15 years would have had you buying after the 1998 Asian crisis, especially after seeing the market only go down for one month. The head, or top, is never obvious. There was still rampant bullishness throughout late 2000 into 2001. Following September 2001, the lows in the market appeared to be solid, and occurred at a time when excess needed to be removed from the market. The rally that followed through to March 2002 was a continuation of the general Wall Street bullishness and the "buy on dips" mentality. If you have been in the market over the last five years, you have found yourself at one, if not all three, of these junctures.

Look at Figure 1 and think back to the mindset at the left shoulder and head. Think about the post-September bullishness that prompted many to invest because they thought the worst was over. Now think about today: How many stocks in your portfolio have been positive over a two- or even five-year time frame?

RETRACEMENT PROBLEMS

You may have noticed the mention of Fibonacci retracements on Figure 1. The following is from John Murphy's Technical Analysis Of The Financial Markets concerning the work of Leonardo de Pisa, also known as "Fibonacci":

. . . That number sequence has become identified with its discoverer and is commonly referred to as Fibonacci numbers. The number sequence is 1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89, 144 . . . and so on to infinity. The sequence has a number of interesting properties, not the least of which is an almost constant relationship between the numbers.

• The sum of any two consecutive numbers equals the next higher number. For example, 3 and 5 equal 8, 5 and 8 equal 13, and so on.
• The ratio of any number to its next higher number approaches .618 after the first four numbers. For example, 1/1 is 1.00, 1/2 is .50, 2/3 is .67, 3/5 is .6, 5/8 is .625, 8/13 is .615, 13/21 is .619, and so on. Notice how these early ratio values fluctuate above and below .618 in narrowing amplitude. Also, notice the values of 1.00, .50, and .67.
• The ratio of any number to its next lower number is approximately 1.618, or the inverse of .618. For example, 13/8 is 1.625, 21/13 is 1.615, 34/21 is 1.619. The higher the numbers become, the closer they come to the values of .618 and 1.618.
• The ratios of alternate numbers approach 2.618 or its inverse, .382. For example, 13/34 is .382, 34/13 is 2.615.

These numbers and ratios are apparent in everything from the pyramids of Egypt to flora in nature to the human body (as demonstrated by another Leonardo, this time DaVinci, in his drawings of the perfect man). Technicians have adopted Fibonacci ratios as a tool because of the numbers' proven usefulness and forecasting ability.

To the right of Figure 1, you will notice three red lines drawn. These are the 38.2%, 50%, and 61.8% retracement levels of prices from the 1962 low to the 2000 high. Notice how 40 years after the fact, measuring from the end of the last major depression period to the end of the most recent expansion, the 38.2% retracement mark supported the S&P 500 following the September 2001 terrorist attacks (first blue arrow, left to right). Move ahead to the second blue arrow (June 2002). Again, the markets declined below the 38.2% retracement mark (955), only to rally before the end of the month. This is a technical positive for the market: Fibonacci retracements are used as bounce points to gauge the extent of a decline from a top. If the 38.2% level is broken, the next point to watch would be the 50% mark, then the 61.8% level.

As you begin to assemble the puzzle, however, so many confluent factors occurring nearly simultaneously begin to look less and less like a coincidence. The June 2002 decline also marked another important point: the downside break of the long-term uptrend. That by itself is of minor significance, and the late-month strength may have signaled the market was ready to resume its prior uptrend. Unfortunately, along came July and the hopes of holding the 38.2% mark were limited. The market bounced off that level, which you can see is in the same area as the neckline of the H&S formation.

This break of the trendline becomes a much larger problem. Combine the downside break of a long-term trendline, the completion of a five-year head and shoulders reversal, and the downside break of the 38.2% retracement, and the coincidence factor begins to wane. As Sherlock Holmes said, "When you have eliminated all possibilities, whatever remains, no matter how improbable, must be the truth." Unfortunately, the market has slowly been eliminating the bullish signs, leaving only bearish truths.

Moving on to the third blue arrow, not only do the lows in September 2001 and June 2002 happen to occur right near the 950 mark (with Fibonacci importance), but also, once that mark was broken, the index moved in a nearly straight line to what would be expected, the 50% retracement level (785). We have since seen a sharp bounce off this mark.

OTHER INDICATIONS

As you can see, the MACD indicator became wildly bullish and then turned bearish, giving yet another nod to the bearishness of the trend. Further, note the two blue downtrend lines to the right of the chart. The steep one is representative of the decline since March 2002; the one with a lesser slope marks the intermediate-term downtrend beginning with the all-time highs. You already know that prices, as with everything else, tend to move in a trend. So which trend is controlling the market in this chart?

This takes us to the "Houston" problem. First, the technical facts: a full head and shoulders reversal pattern (950) has been completed, multiple new downtrends have begun, an uptrend that had endured for 13 years has been broken, and the first Fibonacci retracement of an advance (950 again — coincidence?) has been broken. Then a new problem arises. Although prices have bounced off the 50% retracement mark, it occurred recently and quickly. The concern stems from the implications of the possibility of breaking the 50% retracement mark.

Theory suggests the market would move down to test the final 61.8% level. If that level is broken, a 100% retracement of the move would be in order. Now the overriding downtrend lines and the recent five-year reversal pattern indicate a down market, so the problem lies in two points, and a few downtrend lines. Does the 50% level hold? If not, does the 61.8%? At this point, the market appears to be in the same situation as the first lunar mission was early in its journey. It can survive, but one or two mistakes could end the mission: the capital markets' version of a technical problem.

CYCLE PROBLEMS

A historical reference may come in handy at this point. The most recent, comparable event to the market bubble would have to be the Nikkei's rise to over 30,000. The period of "popping" the bubble has taken more than 10 years and may not be over yet, as the index has not resumed what could be characterized as a bull trend. Another interesting fact is the Kondratieff cycle, which suggests that the time from trough to trough of major economic cycles is an observed cycle of 54 years. Although this cycle may not have held up exactly, it does match, more or less, the performance of the economic cycles, following the ups and downs through the Industrial Revolution, the Great Depression, and the recent bull run (1960 and on).

Assume the US exhibits a similar pattern to Japan's and the market at best trades sideways, at worst goes down for 10 years from the peak of 2000. That would forecast this bear or volatile market lasting until 2010 at the earliest. Applying the 54-year Kondratieff cycle to the 1960 low would put the "trough to trough" measurement of 54 years at 2014. Just applying the blank template of the 10-year Nikkei bear market to the US markets already puts it at 50 years through the cycle (2000 high + 10 years = 2010).

NARROWING IN: DRIFTING DOWN

Technicians attempt to gather all the clues and decide what the market's next course of action will be. Sadly, the only conclusion from this evidence is that you could expect a prolonged bear market resembling the 1970s, or the depression that holds back Japan to this day. The coincidence factor at some point begins to be too much to ignore.

I do not like the conclusions I have come to; I hope I am mistaken. But occasionally a dose of the truth is necessary. Remember the great bear markets in history have all been marked with early bullishness, just as all major tops have been marked with complacency. Does this mean the market will return to its 1960 levels? That's not likely. Does the combination of these technical factors make for an extended bear market? At this point, that is very likely.

There are two dominant downtrend lines that show the market's direction. The chances of catching the low tick, day, week, or month are slim and will prove futile — the short sellers found this out on the way up; buyers will discover it on the way down. One of the big secrets of the market resides in the statement that "prices move in trends." In looking at that chart, how many bottoms were there to buy? The trend was up; there were no places where buying the index would have produced losses (pre-1997).

There is no need or justification in trying to find the bottom. To quote the movie Wall Street, "If I'm right, quarters and eighths won't matter." Concern yourself less with trying to find the bottom, and wait for the market to go up again (ideally, a new uptrend with two points to form a trendline or a monthly close above 980 would be a bullish signal). Quarters and eighths exist no longer, but the point is the same: Making money in the markets requires catching a major surge, not a small push.

Nick Porcella is a technical analyst with E&J Securities.

RELATED READING

Charts courtesy of TradeStation (TradeStation Technologies)