Equities are bought and sold throughout the world for essentially the same reasons, so the principles of technical analysis can be applied to any stock market. Unfortunately, the degree of sophistication in statistical reporting of many countries does not permit the kind of detailed analysis that is available in the United States, although things are improving rapidly. Even so, it is possible to obtain data on price, breadth, and volume for most countries. Information on industry groups and interest rates is also widely available.
In this chapter we will concentrate on longer-term trends for the purpose of gaining perspective, but the analysis can just as easily be used to identify intermediate-term and short-term trends.
Chart 33.1a shows the Morgan Stanley Capital International (MSCI) World Stock Index, which is constructed from a selection of blue-chip stocks from many different countries weighted by capitalization.
CHART 33.1a MSCI World Stock Index, 1964–1992, Showing 4-Year Cycle Lows
This series has been adjusted to U.S. dollars and is widely published in the financial press. Other world indexes published by Dow Jones and the Financial Times can be adopted into the analysis, but the MSCI has been chosen because of its extensive history going back to the 1960s. In addition, MSCI indexes are available for direct investment in individual country and regional exchange-traded funds (ETFs), as is the World Index (symbol ACWI) itself. The World Index is a good starting point from which to analyze the cyclical trends of the various stock markets, just as the S&P Composite might be used as a starting point for the U.S. market. This is because the stock markets around the globe tend to move in the same direction, just as the majority of U.S. stocks reflect the primary trend of the S&P most of the time. Generally speaking, improvements in technology and communications have broken down geographical and trading patterns, and countries have become more interdependent, with the result that their stock markets and business cycles are now more closely related than they used to be. A giant leap in this direction appeared to take place after the 1987 crash, in which all markets participated on a synchronized basis. This was later reinforced almost 10 years later when the so-called “Asian meltdown” reverberated around the world. The introduction of international and specific country closed- and open-ended mutual funds in the 1980s and 1990s and their U.S.-based expansion in the opening decade of the current century is a striking example of this growing sense of international awareness. There are exceptions, though, because it is possible for different economies to be in a different state of expansion than others. An example might be the performance of neighbors Greece (ETF symbol GREC) and Turkey (TUR) in the 2011–2012 period in that the Greek economy was retrenching and the Turkish economy expanding. As a result of the variations in the long-term economic, financial, and political situations between countries, a good world bull market in equities may be brief or almost nonexistent for a country undergoing financial distortions, such as Hong Kong between 1986 and 1990. Country performance can also differ because of the makeup of specific markets. For example, the Swedish and Finish indexes performed superbly in the latter part of the 1990s because they were dominated by technology companies. Countries with substantial natural resources such as Canada (EWC) and Australia (EWA) tend to outperform when commodity prices are rising and so forth. An additional factor emanates from demographics. Regions and countries with a population pyramid skewed toward older people (Europe and Japan, for example) have a built-in disadvantage compared to countries such as Indonesia, India, Turkey, etc., whose population is skewed toward younger people, where growth characteristics such as family formation, consumer spending, and so forth are far more dominant.
Charts 33.1a and b show the existence of the international 4-year cycle, as indicated by the arrows. The troughs in 1962, 1966, 1970, 1974, 1978, 1982, 1986, 1990, 1994, 1998, and 2002 are all separated by approximately 4 years. I say approximately since the actual bottoms do not fall in the same month. The 1986 “bottom” was more or less nonexistent and was essentially a 6-month trading range. This demonstrates the fact that in a secular bull market, such as the 1980s, the cycle low is not so much a bottom as a buying point prior to further gains. The same sort of thing happened in 1994, where the sideways correction was more obvious, the 2006 and 2010 4-year cycle “lows” also provided good buying opportunities.
CHART 33.1b MSCI World Stock Index, 1992–2012, Showing 4-year Cycle Lows
The reason why this cycle works is because global equity prices revolve around the 4-year business cycle. Partial proof of that is provided in Chart 33.2, which compares the ACWI to a derivative of the amplitude-adjusted Organization for Economic Cooperation and Development (OECD) composite leading indicators. That derivative is a simple 1/15 price oscillator. The dashed line is a 1-period moving average (MA) of that derivative advanced forward by 3 months. The solid arrows indicate when the derivative crosses above its advanced moving average from a sub-zero position. The dashed ones indicate the same thing, but point up that equities did not respond positively to expanding economic conditions in the normal way. One of the reasons for using the price oscillator derivative calculation is that the data are reported with a 2-month time lag, so this technique allows the indicator to reverse sooner than would otherwise be the case.
CHART 33.2 MSCI World ETF (ACWI) Spliced, 1984–2012, and a Global Economic Indicator
Chart 33.3 shows the MSCI World ETF, spliced to the actual index prior to its listing. Also included in the chart is the long-term Know Sure Thing (KST). The light plot indicates those periods when the KST is below its 9-month MA and the ETF price is also below its 12-month MA. As you can see, this system would have sidestepped all the major bear trends. It also triggered several false negatives, as indicated by the small dashed arrows. The 12-month MA was used in the model because crossover signals are one of the most reliable over the history of the index. The model, therefore, offers an overall objective indication as to whether global equity prices are in a primary bull or bear market.
CHART 33.3 MSCI World ETF (ACWI) Spliced, 1966–2012, and a Long-Term KST
The World Index can also be used with net new high data. Chart 33.4 was constructed using the popular MetaStock program.
CHART 33.4 MSCI World ETF (ACWI) Spliced, 1996–2012, and a Net New High Oscillator
In this instance, new highs and lows are calculated for a basket of individual stock markets. Instead of the normal time span of 52 weeks, I chose a 13-week period and smoothed the data with a 5-week simple moving average. Thirteen weeks is, of course, a quarter of a year and appears to work quite well. The arrows indicate pro-trend signals, the direction of the primary trend being determined with the benefit of hindsight. So, for example, reversals to the upside are only generated when the indicator drops below zero or, in the case of 2005, very close to zero during primary bull markets. In a similar way, sell signals in bear markets are flagged with the dashed arrows and can only develop when those reversals develop from a position above zero. The shaded areas flag the two bear markets that developed during the 1997–2012 period represented on the chart. Chart 33.5, on the other hand, draws our attention to the fact that contratrend signals have a much lower chance of succeeding.
CHART 33.5 MSCI World ETF (ACWI) Spliced, 1996–2012, and a Net New High Oscillator
New lows on their own are featured in Chart 33.6. The plotted data is, in fact, a 4-week MA of a basket of individual country indexes touching a new low over a 13-week time span. The horizontal lines show two levels of tolerance. During a bull market, a reversal from the lower line triggers a buy signal. During a bear market, the requirement is a reversal from the upper horizontal line and above. Even then, the rallies are somewhat puny when compared to those signaled during bull markets.
CHART 33.6 MSCI World ETF (ACWI) Spliced, 1996–2012, and a Net New Low Indicator
Chart 33.7 shows a different representation.
In this case, it is a diffusion indicator that measures a basket of individual country indexes in local currencies that are above their 24-month MAs. The 24-month time span is used as an attempt to reflect half of the 4-year (24-month) business cycle. Buy indications develop when the indicator reverses from a below-zero reading and crosses above its 9-month MA. The 11 signals in the approximate 40-year period covered by the chart certainly reflect that 4-year business cycle idea, though, of course, they are not evenly spaced. Chart 33.8 shows the exact same diffusion series, but this time from the sell side. That is, when the indicator moves above the overbought zone and crosses through its MA on its way back to the equilibrium area. During extended bull markets, some of these signals result in false negatives, which have been flagged by the dashed arrows. Note that in neither case did the price slip decisively below its 12-month MA. On the other hand, this approach more than made up for this deficiency by calling major tops that developed in 1969, 1973, 1990, 2000, and 2007.
CHART 33.7 MSCI World ETF (ACWI) Spliced, 1969–2012, and a Diffusion Indicator Showing Buy Signals
CHART 33.8 MSCI World ETF (ACWI) Spliced, 1969–2012, and a Diffusion Indicator Showing Sell Signals
Chart 33.9 features a way of identifying trend reversals in the MSCI Emerging Market ETF (symbol EEM). It’s a diffusion indicator constructed from a basket of emerging-market ETFs.
CHART 33.9 MSCI Emerging Markets ETF (EEM), 2010–2012, and a Daily Diffusion Indicator Showing Buy Signals
The number 40 in the legend refers to the fact that we are looking at the number of individual emerging-market ETFs above their 40-day MA, and the 8 tells us that the final plot is an 8-day smoothing. The arrows flag when the indicator rallies through its oversold zone. The solid arrows reflect successful signals and the dashed ones false positives. The same exercise is undertaken in Chart 33.10, but this time from the sell side. Of course, it would be possible to construct similar measures using weekly or monthly time spans for the purpose of monitoring longer-term trends.
CHART 33.10 MSCI Emerging Markets ETF (EEM), 2010–2012, and a Daily Diffusion Indicator Showing Sell Signals
The global A/D line in Chart 33.11 is a daily A/D line plotted on a weekly basis. It is constructed from 20 country and regional ETFs. The history only covers 5 years of data, but it is certainly showing some promise as a useful analytical tool. Occasionally, as at points A and B, it throws up positive and negative divergences. However, its main use comes from the fact that its jagged action allows the construction of trendlines, which when confirmed by the price itself, offer timely buy and sell indications. The indicator in the bottom panel is a price oscillator constructed by dividing an 8-week MA of the A/D line by a 30-week MA. It’s also possible to construct trendlines on this series, and we see three useful ones in the chart. The indicator also lends itself to overbought/oversold reversals more in identifying reversals in the A/D line than for the price itself.
CHART 33.11 MSCI World ETF, 2007–2012, and Two Breadth Indicators
Today there are a variety of ways in which investments or trades can be executed in specific markets. Individual stocks can be purchased through brokers with an international presence, through American depositary receipts (ADRs), etc. In recent years the exchanges of most countries have established futures markets on key indexes. Exposure can also be achieved through closed and open-ended mutual funds that specialize in individual countries or regions.
The most popular instrument, the exchange-traded fund, is now available for virtually every country in the world. Some ETF families also include sector funds for individual countries. China and Brazil are just two examples.
The key to selecting better-performing stock markets is the adoption of relative strength analysis using the principles outlined in Chapter 19. Chart 33.12 shows the relative strength (RS) line for the S&P Composite against the MSCI World Index with a long-term KST of the same series. The KST reversals were, for the most part, timely signals, though in several instances it would have been more advisable to wait for some trendline confirmation, as we can see, for instance, in the confusing 1990–1992 period. Generally speaking, the KST reversal trendline confirmation approach works very well. However, in 1987 and 2009 this approach resulted in a couple of nasty whipsaws. Both situations were involved with sharp global sell-offs, so the relative rally reflected a temporary rush to (relative) safety.
CHART 33.12 S&P Composite RS versus the World ETF, 1975–2012, and a Long-Term Monthly KST
Chart 33.13 shows the same arrangement for the Nikkei. In order to make the comparison relevant, the Nikkei has been adjusted into U.S. dollars. The most striking part of the chart lies in the two distinctive periods on either side of 1990: a secular bull prior to that date and a secular bear after. The differing characteristics for the relative KST in the bottom panel are also evident, with the secular bull rarely falling below zero and never to an oversold condition, while the opposite was true for the post-1990 period. The chart also demonstrates the power of long trendlines when they are violated, because both dashed trendline penetrations ushered in an extensive period of deterioration. As we leave the chart, the dollar-adjusted series was tantalizingly close to a 20-year upside breakout. Not only was that 1990–20?? down trendline a long one, but it had been touched or approached on numerous occasions. That means that when it is eventually violated, the Japanese stock market should, at least in nominal terms, experience a very big move. Let’s say that happened but was not accompanied by the relative strength line. In that event, it would mean that money was to be made in Japanese equities but capital could be applied more efficiently elsewhere.
CHART 33.13 Nikkei Dollar Adjusted, 1966–2012, and Two RS Indicators
Chart 33.14 shows the German DAX Index together with its 18-month rate of change (ROC). Overbought re-crossovers have a 60-year history of offering primary-trend sell signals. They are certainly not perfect, as the two dashed arrows point out, but generally speaking, downside penetrations of the +50 percent level offer reasonably consistent signals. Note also the 1960 move above the 200 percent level, cited in Chapter 23 as a sign of a secular peak. Indeed, it was because the DAX did not take out its 1960 high for 24 years. In this instance the secular bear experienced an initial sharp drop, and this was followed by a multidecade trading range, not unlike the post-1980 secular bear for the gold price.
CHART 33.14 German DAX, 1950–2012, and an 18-month ROC
Breadth data are now published for many countries, but in many instances A/D lines appear to have a downward bias, probably because many stocks included in the breadth numbers are highly illiquid. For this reason, I prefer to produce my own data selected from a basket of stocks reflecting broad industry participation in each country. Unfortunately, space is limited in this book, so we are only able to scratch the surface. I certainly advise readers in emerging countries to experiment with some of the ideas expressed here and in Chapter 27.
Chart 33.15 features a Brazilian A/D line constructed from 30 leading Brazilian stocks.
CHART 33.15 Bovespa, 1999–2009, and a Brazilian A/D Line
You can see how it was possible to construct trendlines for both this indicator and the Brazilian index, the Bovespa, and observe divergences. At point A1 two things are happening. First, the A/D line makes a new high but the Bovespa does not. This is usually a bullish phenomenon, and we would normally expect the Bovespa to follow through on the upside. However, in technical analysis, a discrepancy is a discrepancy, so when both dashed up trendlines are violated, we need to respect that, and as you can see, both series declined. As time moved on, the A/D strength extended as the line formed a reverse head-and-shoulders pattern but the price trended down. Note the two trendline breaks at A2. Finally, at the top in 2008, it is the A/D line’s turn to be weaker, as you can see from the two small arrows. Then at B both series violate major up trendlines and a bear market was signaled at B. Note how the line lagged the Bovespa at the bear market low. This did not tell us much, as it is normal for breadth indicators to lag the market average.
Chart 33.16 shows a 45-day A/D ratio.
CHART 33.16 Bovespa, 2008–2009, and a Brazilian Breadth Oscillator
This relatively long-term parameter means that the indicator reflects intermediate trends and is quite deliberate in its trajectory. It is, therefore, more prone to allowing trendline construction. Point A is especially interesting because the oscillator was clearly in a constructive mode as it was experiencing many positive divergences with the price during August and most of September. Then the ratio broke below its up trendline and the Bovespa broke to a new low. This is what I call a destructive breakdown, as the potentially very bullish action by the momentum indicator is ruptured. Such situations are typically followed by a sharp decline, which often turns out to be the final one for the move. At points B, C, and D we see joint trendline breaks, which are then followed by a rally of some kind. That at B was a weak one because the basing action was not really complete at that point.
Chart 33.17 features an A/D line I developed for the Indian market because the National Stock Exchange’s published figures resulted in a downward bias and, therefore, not helpful for analytical purposes. Chart 33.17 also shows a 65-day oscillator. When all three indicators break trend, there is a greater probability that the signals will be valid. That was certainly true for those triggered at A, C, and D. The signal at B was a bit more problematic because it developed under the context of the strong 2003–2008 primary bull market. As a result, the decline was quite scary but relatively brief.
CHART 33.17 The Nifty, 2005–2006, and Two Breadth Indicators
Diffusion indicators, of course, are another possibility. In this respect, Chart 33.18 features an indicator that monitors a basket of Saudi stocks that are above their 40-day MAs. The arrows flag the overbought and oversold crossovers, the signals for which are expected to have an effect for 4 to 6 weeks. In most of the instances, this is the case. Sometimes these buy signals prove to be premature, as we can see from the two ellipses. However, in most cases, there is the possibility of constructing a trendline and waiting for its violation as a confirmation.
CHART 33.18 The Saudi General, 2006–2008, and a Saudi Diffusion Indicator
Finally, a useful technique for any market is to calculate the number of stocks in a selected basket whose intermediate KST is below zero (winter and spring positions). When that series reverses from a high reading, it indicates that many stocks are moving to the mark-up phase, i.e., starting a new intermediate uptrend. Such activity, therefore, offers an indication of a strong technical position. Of course, it’s also possible to undertake the same operation using the moving-average convergence divergence (MACD) or a stochastic.
Chart 33.19 shows such an exercise based on a basket of Indian stocks compared to the principal Indian market average, the Nifty. In this case, the number 8 in the legend tells us that the raw data have been smoothed with an 8-week MA.
CHART 33.19 The Nifty, 2004–2009, and a Momentum Indicator
1. There is a definite 4-year global equity cycle.
2. Recent technological innovation and other factors have led to a much closer relationship between equity markets around the world.
3. Diffusion indexes, net new highs, and other breadth-based indicators incorporating individual country indexes can be used for identifying trend reversals in the World Index.
4. Relative strength is the best tool for identifying markets that are likely to outperform or underperform the world stock indexes.