From the dawn of the discipline, most technical analysis practitioners have applied their tools to the financial markets — usually with price and volume data as found in equities, bonds and commodities. Whatever the algorithm, whether momentum, stochastics, the relative strength index (RSI), rates of change, implied volatility, the moving average convergence-divergence (MACD), on-balance volume, parabolics or oscillators in general, the intention has been to use technical indicators to derive clues from past data to forecast future prices.
Because of the nature and historical relationship of technical analysis indicators with financial markets data and the tenuous traditional relationship of technicians with fundamental analysts — the latter who study earnings, economic data, and other "non-price-oriented" statistical information — there has always been an intellectual gulf between the two. Technical analysts have tended to assume that fundamental data have already been discounted by the time the numbers become known. Thus, fundamentals have limited value as market timing tools. On the other hand, fundamental analysts claim that fundamental factors drive all prices in the financial markets. The question remains: "Can fundamental data be used to help with market timing decisions?"
A massive store of historical fundamental data has gone largely untouched by technical analysis. Research shows that there may be more in common between the opposing schools than assumptions suggest. For example, while it has been presumed that stock market prices have a good record of discounting future events, there may also be some truth to the adage that equities predict recessions. Put another way, we know of no recession that was not presaged by stock market weakness. But on the other hand, not every serious stock market decline led to a serious economic downturn. Cases in point include 1937, 1962 and 1987. In other words, investigation of fundamental data may have provided the clues necessary for an accurate forecast.
Because of the wide availability of technical and fundamental data for the individual equity market, stocks provide a good subject for this analysis.
A wise statistician once said that "all data are relevant." It’s worth adding, "especially if the data have significant history and consistency." As a consequence, both schools of analysis, technical and fundamental, have information to offer, and not just because some believe that stocks always discount forward or that fundamentals always drive the markets higher.
Extended data sets for both stock market prices and historical economic data show that there have been points in the past where fundamental data not only coincided with equity market statistical peaks and troughs, but on occasion led the stock market. There also were instances where lingering strength in the economic indicator suggested current weakness in equities would probably not last. Both series, equity and fundamental, have information that, if analyzed properly, can provide useful clues to the future direction of both the financial markets and the economy.
The stock market can be analyzed with a top-down approach — first major indexes, then sectors, then individual stocks themselves . We can approach economic data the same way. As a general rule, four coincident indicators — non-farm payrolls, personal income, industrial production and business sales — are used to determine whether or not the economy is in a recession. Each of those components has sub-categories. For example, business sales consists of manufacturing, wholesale and retail. All that data can be analyzed separately.
While the National Bureau of Economic Research (NBER) is at the forefront of determining which economic sectors are used in the major categories, the final analysis is similar to equity research in that the eventual ingredients analyzed are not set in stone. If an analyst chooses to focus on one specific part of the economy, other ingredients than those chosen by NBER could be included in the final analytical array.
For our purposes, we select data that stretch back decades and have had high visibility in how they relate to the stock market and the economy in general. Stocks are represented by the Dow Jones Industrial Average (DJIA) monthly high-low-close data back to 1920. Then, we look at nearly 90 years of industrial production. "Housing over time" (below) examines monthly housing starts data to 1946.
The price charts include equity prices, economic data and slow stochastics. Red and green vertical lines designate points where changes in fundamental data either coincided with stock market data or led the DJIA up or down.
The indicator, stochastics in this case, levels the playing field to the extent that each series of data is treated in a similar fashion. (A 50-month moving average also is used to provide reference over time.) Slow stochastics, which was developed by George Lane in the 1950s, measures price momentum. It is a well-known indicator with a respectable history.
While the traditional use of stochastics suggests that overbought and oversold readings at the extreme higher and lower edges of the indicator are relevant, movement by the indicator above or below 0.50 usually signals a trend change. Using the TradeStation platform 8.7, we entered the following inputs into the slow stochastics algorithm:
Price inputs: For DJIA; high, low and close; and industrial production, close.
Stochastic length: 25
Stochastic length 1: 10
Stochastic length 2: 25
Smoothing type: 2
Overbought: Above 80
Oversold: Below 20
"Industrial strength stocks" (below) covers the period from January 1920 through December 1966. Notice the peak in the stock market in September 1929 when the Dow Jones Industrial Average hit 386.10. The high followed a statistical peak in stochastics in January 1929 with momentum failing over the next several months. The indicator then signaled a long-term negative with a decline below 0.50 at the end of June 1930. That signal developed after the market staged a brief countertrend rally that lasted through late April 1930 when the DJIA hit 297.25.
The DJIA declined until July 1932 and lost nearly 90% of its value in just under three years. The Great Depression was underway. Fortunes were lost and millions were out of work. At one point, 25% of the potential workforce was unemployed. The negative stochastics signal, following its oversold nadir in July 1932 at the market low, confirmed a positive reversal when the indicator moved above 0.50 at the end of December 1933 to signal that the worst decline in stock market history had ended.
Industrial production reached an actual and stochastics-based high a month before equity prices in July 1929. It then confirmed a negative reversal below 0.50 at the end of April 1930, a signal that remained negative until the end of August 1933 when both the market and industrial production had moved to the upside. Then things got interesting. Industrial production rallied to a new high in December 1936, stalled a bit into the 1938 lows, and began rallying without a serious look back for the next 70-plus years. On the other hand, equity market prices rallied off of the 1932 lows but did not eclipse their 1929 highs until November 1954 and 18 years after the 1929 peak.
There are other examples of industrial production leading the stock market in the post-1929 era: It rose from late 1939 until the beginning of World War II when the stock market remained in a modest downtrend. Again, during the 1961-62 bear market, industrial production remained dramatically positive while equity market prices were sinking. In late 1974, when the stock market was making new lows below the worst levels it hit in 1970 and many were calling for a new Depression, the fundamental measure failed to confirm that action and held well above its 1970 low.
One of the most notable positive industrial production divergences occurred in 1987. Heading into the fall of that year, the stock market had just completed all-time highs when the DJIA reached an unheard of price of 2752.07 in August. Within two months, the bloom came off the rose and the DJIA had plummeted more than 40%. As in 1929, there was investor panic. As in 1974, there were calls for the start of a new Depression with suggestions by some that the index was headed back to the 1974 lows (616.08) or worse.
But unlike 1929, industrial production was rising as it had done in 1937 and 1962 when the stock market was doing its best to eviscerate portfolios. When the stock market was undergoing a massive consolidation from 1966 until the upside breakout in 1982, industrial production gained nearly 50% during the same 16-year period. Later, the measure rallied to a new all-time high in August 1988 and nearly a year before the stock market hit new highs in August 1989. Another instance of a major diversion from the stock market occurred when industrial production hit a statistical bottom at the end of November 2001 and more than a year before the equity market reached its lows in October 2002 (see "Chasing production," below).
This brings us up to the most recent stock market and economic cycle, the period from the fall of 2007 to date. Industrial production data peaked in September 2007, a month before equity market data stalled out in early October. Stock market prices hit their lows in March 2009 while industrial production data were late with a low hit three months later at the end of June 2009. Since then, both series of data have risen, with an exception.
While stock market prices as measured by the DJIA have threatened to penetrate the trailing 50-month moving average, industrial production data have, as yet, been unable to move upward through the same average. While, admittedly, industrial production has only a short distance to go before it penetrates its average to the upside, not since the 1929 crash has industrial production taken so long to do so. While that hesitation could be preceding more strength in both the market and industrial production, as neither series is overbought as yet, the disparity nevertheless bears watching in the fourth quarter of 2010.
Exceptions make a difference
When looking for linkage between data sets, be prepared for exceptions to the rule. When the linkage works, and the data makes a statistical bottom at the same time as stock market prices and then stochastics rise, so much the better. If equities are weak when industrial production data are strong, for example, different conclusions can be inferred. Sometimes, though, the fundamental data signals will be out of synch with equity market prices. The analyst then tries to determine the importance of the disparity by looking for the next point of synchronicity, which could offer new opportunities.
With analysis underway, it is important to know that when combined with a second fundamental data stream, such as housing starts, two separate markets are being considered. Improvement in housing starts, for example, could have a positive effect on the earnings of home builders, on employment in general and on retail sales as consumers ramp up home-related purchases leading to a commensurate rise in the prices of related equities. Then there are those points when the relationship dissipates, which happened in early 2006. Starts peaked and housing construction and real estate prices sank. But it wasn’t until nearly a year and a half later that the broad stock market began a major decline.
It is possible that economic data can lead equities and be measured by slow stochastics or other classic technical analysis indicators, despite historical presumptions. When both economic data and stocks are oversold, the odds are good both will rise thereafter on the longer term. Notice also that when industrial production remains strong and the stock market is declining, equity weakness may be suspect. Such a bias has been the case many times with industrial production since 1920. While this information has little practical short-term value, knowing the status of major trends can help you avoid some pitfalls of long-term investing.
While technical analysis has been traditionally applied to data in the financial markets, primarily equities, there is a wide array of economic and fundamental data that can hold a wealth of information when technical indicators are similarly applied to that data. Knowing the status of an economic series as compared to a traditional stock market index, such as the DJIA or the S&P 500, can provide a useful perspective on not only the stock market, but also the economy. Simply put, while the argument will no doubt continue as to which market approach, technical or fundamental, offers the best approach for traders, statistical knowledge based on the application of the same indicators to apparently unrelated streams of data on a similar cycle could provide answers that no true believer in either camp might imagine — or deny.
Robert McCurtain is a technical analyst, market timer and private investor based in New York City. He can be reached at email@example.com.