عملکرد بازار بورس و قابلیت ساخت صنایع دوره پنجم طولانی

Nearly 80 years ago Russian economist, Kondratieff, introduced the theory of economic long-cycles. Since from the start, this theory has faced controversial acceptance; for example, in the future studies researchers have used it to develop further specific applications, but in economics some leading scientists reject the entire idea still. Although, this theory is well developed, there does not exist research from the examination of relation between stock market performance, and leading innovation cycle industries manufacturing capacity addition and utilization. Based on the system dynamics model, called world dynamics, capacity addition and utilization have earlier been identified as the leading indicators of long-cycles. Our research results in this paper indicate that capacity utilization of computer manufacturing in US, and in some cases of US semiconductors, has influence on the stock market indexes of Nasdaq, S&P500 and Dow. However, it should be noted that capacity investment changes of these three examined industries (semiconductors, computers and telecommunications) are involved in the proposed regression models too. Further analysis reveals, that we are able to build regression models for all three stock indexes, containing only two variables. Notably, these two variables are capacity addition change in semiconductors and computers. This observation further increases discussion, whether we should be interested only about capacity addition changes of innovation wave industries, and possibly give secondary importance for the utilization.

We are currently living from the fruits of the fifth innovation wave, which gave the needed technological base for the fifth economic long-cycle; as logic was originally proposed by Russian economist Kondratieff. Despite of the significant amount of criticism towards Kondratieff's work in his home country at the time as being reported by Garvy [1], and especially some technical flaws (e.g. wage index deflated by wholesale prices makes the entire long-cycle to disappear), this theory has received considerable attention in the scientific research. For example, Futures journal devoted whole issue in the early 1980s for this issue alone. Research has faced difficulties to identify the fifth innovation wave “leading industries”, and especially its starting year (or the end of fourth cycle). Explanation for this lack of direction could be explained by the foundation of Kondratieff's [2] long cycle theory. In his seminal article, the most important data series to verify this new theoretical construction was price development of France, UK and US, starting from the time of industrial revolution. Kondratieff further argued that long-cycles are only a by-product of industrial capitalistic system. However, Goldstein [3, pp. 195–210] has shown that long wave type of behavior has been around since the early days of price index recording; his data starting from 16th and 17th century strongly supports K-wave theory. However, if the situation is compared to the development occurred after the birth of the fourth long wave, we could identify that in all of the developed countries price indexes have increased considerably. For example, in year 1945 consumer price index of US was approx. 50, and at the moment the very same index has reached level of 580. Therefore, cycles seem to have diminished; Ayres [4] shares the same opinion that in recent decades the effects of long-cycles have severely weakened. However, economists have found that the use of price changes is now repeating the well-founded long-cycle pattern [5] and [6]. So, in one sense long-cycles are history (as compared to ordinary life), but in other way around, they are pretty useful in economic research and model building purposes. In the first appeared article Kondratieff [7] argued that economics is more like a dynamic model, changing with respect of time, and cycle theories follow pattern prosperity—crisis—depression. One year after Kondratieff [2] introduced long wave theory in more detailed manner, and identified prosperity part takes approx. 24–25 years, while decline showed more variation, and it was impossible for him to give exact number for the wave length, since the first long wave lasted for 60 years, and the second one for 47 years. Ayres [5] and [6] presented comprehensive analysis concerning this length issue, but still the same uncertainty exist. However, it could be generalized that long-cycles have 25 years up-swing, and similar length of downward movement. Devezas and Corredine [8] represent impressive analysis of biological systems (e.g. development of human from newborn, ancient civilizations, economics, solar system), and argue that long-cycle length to be around 54 years. Earlier Marchetti [9] has shown that even infrastructures follow the same cycle length. Other contemporary research works in the field has not found any opposite directions to cycle duration ( [10] and [11]. As uncertainties exist in both of these cycle phases, total length shows higher variation. Thus, on possible solution and explanation for long wave length was provided in Dator's [12] research; “Historical Societal Generation” has been estimated to have duration of 39.2 years (±4.2 years). Researchers in social science have strong belief that this is the driving force of evolution, including also economics (Table 1). Table 1. Five economic long-cycles and leading industries. Sources: Ayres [5] and [6], Dator [12] Long-cycle Years Leading industries First 1792–1847 Steam power, iron and cotton textiles Second 1847–1893 Railroadization Third 1893–1939 Electricity Fourth 1939–1986a Nuclear power, basic electronics (radio, TV and simple computer) Fifth 1986–2033a Information technology a Authors own estimation. Table options Kondratieff [2] argued in his seminal work, that in the forthcoming economic wave driving innovations occur in the decline phase of the previous wave. However, verifying statistical analysis was gained roughly half a century later from Goldstein's [3, pp. 219–226] research, where he found out evidence that innovations cluster five years (or in some cases 10) year before the next wave starts upward movement. As innovations cluster around innovation industry, Sahal [13] was able to provide an statistical example that older innovation cycles lack innovations after its boom cycle. For example, in locomotives inventions per year halved between periods of 1800–1899, and 1900–1957. Generally research has agreed that world has faced five economic long-cycles, and number of papers have given specific years for all three first long-cycles, and starting point of the fourth cycle. However, research giving end-point for the fourth wave, and correspondingly starting point of fifth wave is still an open issue. There simply does not exist research, which could justify, where this point locates in the time-scale. Thus, researchers agree that fifth wave concerns about microelectronics, computers and telecommunication, and these all together could be put under the broader term of “Information Technology” [12], [14], [15] and [16]. As research agrees that this is the common nominator for fifth innovation wave, we could proceed further with deductive logic. According to Forrester [17] and Sterman [18], consumer markets are vital part of the innovation wave industries. Also all the previous leading industries in the former waves have been related to consumer markets, (except some applications of nuclear power). So, we just need to date back into early consumer applications of IT to estimate the starting point. One good example is IBM's introduction of PC, which happened during year 1981. However, business-wise organizations followed couple of years later. One of the most successful companies, Compaq, made its revenue break-through during years 1984–1986. Company was included in the Fortune 500 list during year 1986, and this could be considered as an acid test of real business prosperity. Another fast growing technology start-up in PCs was Apple, and correspondingly their real break-through products, containing graphical user interface, were introduced during the same years as Compaq started to be successful. This case dependent argumentation is also supported by the analysis of larger statistical sample by Modis and Debecker [19]; they argue that years 1982–1983 represent peak in the new PC models brought to the markets. As two former long-cycles have lasted 46 years, we could rightly argue by the given company examples, as well as with previous realized data that starting year of fifth innovation cycle is somewhere in the mid 1980s, most probable is the year 1986. One of the ‘longwave’ website opinions concerning fifth innovation cycle, as reported by Dator [12], have suggested this same year as a momentum bottom. However, Wymbs [16] have argued that fifth wave started four years later, but he does not provide any detailed evidence to support this argument. Linstone [11] and Goldstein [3] both argue that the starting moment of the fifth innovation cycle is even 10 years later! However, this author has strong belief that fifth economic long wave started in the mid 1980s rather more than during 1990s. Therefore, we could expect to have a turning point in this fifth cycle in somewhere around 2010, and this is in line with stock market/economy analysis of Dent [20, p. 36], who claims, based on population development, that after 2008 world economy will start approx. 10 years economic decline (Interestingly Dent's research was already incorporated on Devezas and Corredine [8] research work concerning biological determinants of long-cycles.). This argument is further supported by the adaptation rate of fifth economic wave innovations; logistic curves show that around 2010 world will face saturation points in several Internet based IT products [21]. This paper is structured as follows: in the following section we will highlight the complex nature of manufacturing utilization in macro as well as micro level. As our analysis shows, manufacturing layout evolution from garage to continuous production type hinder interesting dynamics inside, and capacity addition is more irregular than could be expected from demand data alone. In the third section we will correspondingly review the system dynamics research concerning long-cycle theory, and present major affecting factors of wave like economic behavior of three identified wave types. Notably, system dynamics research argues that capacity investment as well as labor addition delays cause cyclic nature. Following section introduces research environment of the fifth leading long-cycle industries in US as well as stock market indexes of this economy. In the fifth section we will present results from the regression model building, and we are able to provide high explanation power models with either many or few leading innovation cycle manufacturing indicators concerning resource utilization and capacity addition. Based on this analysis we speculate that most important industry of our analysis, semiconductor manufacturing, is now boldly investing for new capacity, although utilization levels are low. In the last section of our paper, we will conclude our research work, and provide some avenues for further research.

We could argue that Kondratieff's theory of long-cycles as being applied in stock markets brings favorable results: Fifth innovation cycle industries eventually seem to explain the behavior of stock market indices in US. Interestingly, our research work indicated that capacity addition is the major factor to explain stock market development, and manufacturing utilization plays only minor importance. This finding is giving great importance, since capacity utilization alone is a complex issue, as was shown in section two. Secondly, we could argue that semiconductors and computers are leading manufacturing industries in the fifth innovation wave. Although telecommunications has gained plenty of attention in recent years, these two identified industries seem to have taken the leading place. We explained this issue by the integrating nature of telecommunication products; knowledge and semi-finished items of different industries are integrated in their products to represent complete package. Semiconductors are one of the most important component types in telecom products, thus fostering the importance of this sector. There exists number of avenues for further research in the topic area of this research work. For example, leading innovation industries in US are dominated by small number of companies. It would be extremely interesting to analyze their press releases concerning capacity addition, and possibly build regression models, which are able to predict stock index behavior in the future. Our models proposed in this paper concern only current period's capacity addition as well as utilization, which are used to explain stock index development. However, building leading indicators from company based investment plans would give more visibility concerning the future state of the market. Second, interesting avenue for further research would be the previous fourth long wave industries, and stock index development at that corresponding time. This research work would most probably give more evidence for the theory developed for this seminal research work. Third avenue for future research would be the leading industries of the sixth innovation and economic long wave, namely nano- and biotechnology (e.g. Traore [31]), and their product development as well as market enlargement and investment plans. The rise of the sixth wave depends greatly in their ready-made products and planned investments. We expect this sixth long wave to appear somewhere around 2030, but it would be great benefit for the whole world to have it available sooner, (further dampening the depression effect of the fifth cycle).