Technical Analysis Explained

by Martin J Pring

  A natural tendency is to use as many items as possible to calculate a diffusion indicator, but this involves maintaining a very large database. My own experience shows that the same objective can be obtained from a relatively small universe of securities. The main thing to bear in mind is that the basket of items used in the calculation reflects the diverse nature of the index’s components.

  Interpretation

  When a diffusion indicator moves to an extreme, it reflects an overbought or oversold condition. However, such readings do not in themselves constitute actual buy or sell signals. The false sell signal for the diffusion series in Chart 27.16 in 2004 is an excellent reason to await a trend confirmation signal. Obviously, the odds favor a profitable investment made at the time of a zero reading, and vice versa. However, it is usually much safer to wait for a reversal in the trend of the diffusion index, or better still, for the confirmation of a trend break in the aggregate index being monitored.

  Major Technical Principle When a diffusion indicator reverses direction from an extreme reading, the index it is monitoring usually reverses as well. If it does not and the diffusion indicator continues to correct, it is a sign that many of the securities from which the diffusion indicator is constructed are themselves correcting.

  Chart 27.17 features a diffusion indicator constructed from a basket of Dow stocks. The basis of the calculation is the percentage that are above their 50-day MA. Since that would return a fairly jagged series, the data have been smoothed with a 10-day MA. There are two extreme levels, flagged by the dashed and solid horizontal overbought/oversold levels. The arrows show that reliable buy and sell signals are often triggered when the indicator reverses direction from a position beyond the extremes marked by the solid horizontal lines. The July 2007 high was instructive for although it was slightly below the April/May high, the oscillator was distinctly weaker. That sort of combination, whether the index reached a new high or not, is typically a strong sign of a very weak market.

  CHART 27.17 DJIA, 2008–2012, and a Diffusion Indicator

  Seasonal Breadth Momentum: The Seasons Defined

  Every cycle effectively goes through four momentum stages before completion. This is shown conceptually in Figure 27.2. The first occurs after downside momentum has reached its maximum. At this point, the series turns up but is still below its equilibrium level. The second is signaled when it crosses above its zero reference line. The third phase starts when it peaks out from above zero. Finally, phase 4 is triggered when the indicator crosses below the equilibrium point.

  FIGURE 27.2 Seasonal Momentum Defined

  For simplicity’s sake, the respective stages have been labeled as spring, summer, fall, and winter.2 From both an agricultural and an investment point of view, the best results occur when planting (investing) is done in the spring and harvesting in late summer or fall.

  In effect, spring represents accumulation, summer the markup phase, fall distribution, and winter the markdown phase. In situations in which a market can be subdivided into components, it is possible to take this approach one step further by calculating a diffusion index based on the position of the seasonal momentum of its various components, e.g., industry groups for a stock market average, commodity prices for a commodity index, etc. This seasonal momentum approach has two merits. First, it helps to identify the prevailing stage in the cycle, i.e., whether the stock market is in an accumulation, markup, distribution, or mark-down phase. Second, it also helps identify major buying and selling opportunities.

  Choice of Time Span

  The choice of time span is critical for all momentum indicators, including those used in the seasonal momentum studies. For example, a series based on a smoothed 13-week ROC will have far less significance in terms of long-term investment strategy than a series based on a 48-month time span. This approach can be used for daily, weekly, and monthly data. I am sure you could expand this concept to include intraday data because the principle is the same. I have never done so, but encourage active traders to give it a try. As with most things technical, it seems that daily and weekly calculations, even when greatly smoothed, do not give as reliable a picture as calculations based on monthly data. That does not mean that shorter-term frames never work and that monthly ones always do. It’s just that longer-term frames are less determined by random events and therefore have a tendency to operate more reliably. While our explanation of seasonal momentum is focused more on U.S. equities, this approach can also be expanded to commodities, bonds, and international markets. The indicators represented in the charts included in this chapter have been constructed by finding the number of a basket of 10 S&P Industry Groups in their respective winter, spring, summer, or fall positions and then smoothing that data with a 6-month MA.

  Seasonal (Diffusion) Momentum for the Stock Market

  Chart 27.18 shows all four seasonal momentum curves between 1980 and 2012. A high reading in the spring series, for example, indicates that the momentum of a significant proportion of the groups is in phase 1, i.e., below zero and rising, and therefore in a position to begin a major advance.

  CHART 27.18 S&P Composite, 1980–2012, and Seasonal Momentum

  It is important to note that in most cycles there is a chronological sequence as the majority of groups move into spring from winter, subsequently landing in summer and finally fall. This is shown by the arrows. Bear market lows typically occur around the time winter momentum peaks. As with all momentum series, confirmation should come from the price, which in this case, is the S&P Composite.

  The peaking out of spring momentum is sometimes associated with the first intermediate-term peak in the bull market, but it is not a primary-trend bearish sign. It simply means that the majority of groups are moving from the spring (accumulation) to the summer (markup) phase. It is when summer peaks that we get the first sign that the trend may be topping, but because it takes longer to build than to tear down, this sign of trouble is nowhere near as reliable as the winter peaking action at major lows.

  When summer peaks, though, it does indicate that the environment has become much more selective as the smoothed momentum for more and more groups moves to the fall (distribution) phase.

  Bear Market Bottoms

  Major buying points occur when winter momentum reaches its peak and starts to turn down. Generally speaking, the higher the peak, the greater the potential for upside activity. This is because a movement out of winter momentum must flow into spring. A high and falling level in winter momentum, therefore, indicates that a significant number of groups have the potential to move into the spring position, i.e., to move to the point from which they have the greatest potential to rise. This is shown more clearly in Chart 27.18, but with a far greater history in Chart 27.19.

  CHART 27.19 S&P Composite, 1923–2012, and Monthly Winter Momentum

  CHART 27.20 S&P Composite, 2010–2012, and Daily Winter Momentum

  In this case, the universe from which the indicator is calculated is limited to 14 industry groups since that represents all of the available data. Nevertheless, reversals from above the horizontal line show consistent and reliable primary-trend buy signals. Note also that during secular bear markets, as flagged by the shaded areas, the number of groups moving into the winter position is usually far greater than in secular bull markets. Finally, you will notice that the winter position maxed out at around the same time the S&P was peaking. That was due to the fact since 1998 the market had been correcting internally as the index rallied due to the tech boom. By the time tech peaked, most groups were actually in a position to rally, which they did. The S&P sold off though, because its highly weighted tech component declined. This was a unique situation, but it does point out that while diffusion and indexes usually move in tandem, diffusion tells whether the advance or decline will be broadly based or not.

  Chart 27.20 shows another “winter” exercise, but this time using the daily Know Sure Thing (KST) formula to identify short-term buying opportunities. There is no reason
why it would not be possible to use the MACD or a smooth stochastic as a basis for constructing these indicators. In the case of the latter, the 50 level would correspond to the equilibrium zone for the KST.

  A useful exercise is to take the total of groups in a positive trend (spring + summer) as shown in Chart 27.21.

  CHART 27.21 S&P Composite, 1980–2012, and Spring and Summer Monthly Momentum

  Bull and bear markets are then signaled by reversals in this indicator. It’s not perfect, of course, but when it peaks, it is warning us to be more selective in what we buy, even though the S&P may still work its way higher, as it did in the secular bull period of the late 1990s. Generally speaking, the lower the level of summer velocity when a reversal occurs, the greater the potential for a market rise.

  Signs of a Market Peak

  Market tops are far more elusive than bottoms, but often occur at some point between the peak in summer and fall momentum. Even a topping out in the fall momentum is not always sufficient to trigger a fullfledged bear market. It is only when a large and expanding number of groups fall below their zero reference lines, i.e., move into winter, that a bear market picks up downside momentum.

  Summary

  1. Market breadth measures the degree to which a market index is supported by a wide range of its components.

  2. It is useful from two aspects. First, it indicates whether the environment for most items in a universe (normally equities) is good or bad. Second, market breadth indicators signal major turning points through the establishment of both negative and positive divergences.

  3. Indicators constructed from breadth data include A/D lines, breadth oscillators, diffusion indicators, and net new highs.

  4. Breadth divergences are a fine concept, but should be confirmed by a trend reversal in the market averages themselves.

  5. New highs and lows can be used to indicate the underlying strength or weakness of the prevailing trend. This data can also flag divergences or serve as a measurement of trends by cumulating the plurality of the highs and lows.

  6. Seasonal momentum helps to point out major buying opportunities and explain the maturity of a primary bull and bear market.

  lArms-Equivolume Corp., 1650 University Boulevard N.E., Albuquerque, NM 87102.

  2This approach was first brought to my attention by the late Ian S. Notley, Notley Group, Yelton Fiscal, Inc., Unit 211-Executive Pavilion, 90 Grove Street, Ridgefield, CT 06877.

  Part III

  OTHER ASPECTS OF MARKET ANALYSIS

  28 INDICATORS AND RELATIONSHIPS THAT MEASURE CONFIDENCE

  A negative divergence between an A/D line and a market average is a broad measure of a subtle loss of confidence by market participants. It is also possible, though, to gain an insight into confidence levels by observing relationships that compare what we might call speculative to defensive areas, as these, too, often serve in a more direct way as an indication of growing confidence or lack thereof. When these relationships are reflecting a trend of growing optimism, it is a positive sign and is an indication of higher prices. When they are deteriorating, an omen of weakness and lower prices is signaled.

  Major Technical Principle When a confidence ratio fails to confirm a new high in a market average, it is a sign of weakness that, when confirmed by price action in the average, leads to lower prices. Conversely, when a confidence ratio fails to confirm a new low in a market average, it is a sign of strength that, when confirmed by price action in the average, is usually followed by higher prices.

  Figure 28.1 reflects the basic principles that can be applied to essentially all of the relationships described in this chapter in that the confidence ratios typically peak and trough ahead of prices. This does not happen in all instances of market turning points, nor is the lead/lag relationship constant. We should also add that not every divergence, whether negative or positive, is necessarily followed by a change in trend of the price series being monitored.

  FIGURE 28.1 Confidence versus Price

  The relationships in this chapter are focused on the U.S. market, but there is no reason why the curious reader cannot expand these principles to other equity markets, or indeed to the bond and commodity arenas.

  The Consumer Staple/Food Models

  The consumer staple sector embraces companies that produce, for want of a better term, necessities—things that consumers buy in bad times as well as good ones. Examples include manufacturers of food, household products, beverages, etc. In effect, they produce goods that people are unable or unwilling to cut out of their budgets, regardless of their financial situation. Consumer staple stocks are considered noncyclical, meaning that they are always in demand, no matter how poorly the economy is performing, because people tend to demand such products at a relatively constant level, regardless of price.

  When investors are cautious, they tend to flock to these equities for four reasons:

  1. Their lack of cyclicality makes for more accurate earnings forecasts than, say, highly cyclical and volatile mining stocks.

  2. They tend to pay better dividends.

  3. They generally have a consistent record of earnings and dividend growth.

  4. Most of these companies have a solid balance sheet.

  Because of these characteristics, consumer staples tend to do better during bear markets. Their relative action during bull markets has a tendency to be weaker as investors turn to more exciting sectors.

  This knowledge can be put to profitable use with the aid of several indicators. For example, just monitoring the relative action of staples to the S&P offers some useful buy and sell signals. Chart 28.1 compares the S&P exchange-traded fund (ETF), the SPY, with the Spider Consumer Staple (symbol XLP) relative action.

  CHART 28.1 S&P ETF, 1998–2012, and Consumer Staples RS Showing Divergences

  In view of the fact that the relative strength (RS) line moves in the opposite direction to the S&P, it has been plotted inversely in the chart. This relationship often gives us advance warning of a change in trend as the inversely plotted RS line fails to confirm new highs and lows in the S&P. These divergences represent a subtle way in which confidence changes ahead of the overall market. Chart 28.2 codifies this in the sense that the divergences at A, B, C, and D were all confirmed with joint trendline breaks. Note that each instance was followed by a worthwhile move. The only exception was the potential divergence in the 2011–2012 period, but that had not been confirmed as the manuscript for this book was being submitted in mid-2013.

  CHART 28.2 S&P ETF, 1998–2012, and Consumer Staples RS Showing Trendline Violations

  Another technique using the XLP/SPY relationship is to compare the momentum of the XLP relative action to that of the SPY. I use the daily Know Sure Thing (KST) formula presented in Chapter 15, but there is no reason why two moving-average convergence divergence (MACD) or stochastic indicators could not be overlaid to achieve similar results. An example using KSTs is shown in Chart 28.3, where the dotted line represents the XLP relative momentum and the solid one the KST for the SPY.

  CHART 28.3 S&P ETF, 2006–2012, Comparing Two Momentum Indicators

  The light highlights indicate when the relative XLP momentum is above that for the SPY and when the SPY itself is responding by being below its 50-day MA advanced by 10 days. For a description on the technique of advancing moving averages, please refer to Chapter 11. During the bullish 2009–2012 period, this technique was not particularly helpful in that the declines were fairly truncated and therefore the sell signals developed fairly close to the final intermediate lows. On the other hand, for the most part, the bullish periods indicated by the dark highlights caught most of the upside action. Alternatively, the light shadings that occurred in the 2007–2009 bear market tell us that the model would have offered great protection against downside action, except for a relatively small decline at the start of 2009. The bottom line is that this approach works best when the signals are triggered in the direction of the prevailing primary trend.

  An a
lternative method of presenting the same information is to subtract the S&P momentum from that of the relative XLP (Chart 28.4). In that way positive KST crossovers are represented on the chart by a move above the zero equilibrium line. Since this arrangement tells us when the relationship is overstretched, warnings of trend reversals, ahead of the actual zero crossovers, can be generated when the ratio reverses direction. Some examples have been flagged with the arrows.

  CHART 28.4 S&P ETF Daily, 2008–2012, and a Differential of Two Momentum Indicators

  However, you can also see that other extreme reading reversals fail completely in their trend signaling abilities. This will typically happen at the start of a new bull market, where momentum is particularly strong and countercyclical signals benign. The two “sell” signals in May and August 2009 are prime examples.

  Finally, it’s possible to extend the time frame to an intermediate one, as shown in Charts 28.5 and 28.6.

  CHART 28.5 S&P ETF Weekly, 1977–1994, and a Differential of Two Momentum Indicators

  CHART 28.6 S&P ETF Weekly, 1995–2012, and a Differential of Two Momentum Indicators

  In these instances, the S&P Food Index, which is a member of the consumer staple sector, has been substituted for the consumer staples themselves because more history is available. The differential indicator in the lower panel is constructed by subtracting the intermediate KST of the S&P Composite from that constructed from the relative action of the food group. As you can see, reversals in the indicator from an extreme level, i.e., at or beyond +170 and –170, usually provide reliable signals of intermediate-trend reversals unless they develop at the start of a new primary trend. The light highlights reflect bearish trends when two conditions are in force. First, the relative food KST is above that for the S&P and the S&P itself is below its 40-week moving average (MA), which itself has been advanced by seven periods. Again, the main criticism would come from bearish signals that are triggered in bull markets, where limited declines mean that the signals are triggered very near to the final low of the move, the 1987 decline being a principal example. Some additional instances have been flagged in the chart by the small solid arrows. The results using this system between 1976 and August 2012 returned annualized gains of 9.65 percent when this technique was in a positive mode and –0.60 percent when negative.