Martti Luoma

THE PARAMETERS OF THE GRAVITY MODEL ARE CHANGING -- HOW AND WHY?.

Abstract Friction of distance in spatial interaction grows when the distance increases. This is a well-known fact and almost self-evident on the basis of the gravity model analogy and numerous empirical studies. Because this friction should apparently decrease over time, the change of the distance–decay parameter should be monotonic along time. We present empirical and theoretical evidence that this is only partially true. We also investigate the role of another parameter which is the scale parameter and the factors influencing on it. An empirical and theoretical explanation will be given as to why the changes of that parameter are rather contrary to the changes of the distance parameter.

The linear and non-linear dependence of stock returns and trading volume in the Finnish stock market

This paper investigates the dynamic linkages between stock returns and trading volume in a small stock market, i.e. the Helsinki Stock Exchange in Finland during the period 1977–88. Both linear and...

THE THRESHOLD GRAVITY MODEL AND TRANSPORT GEOGRAPHY: HOW TRANSPORT DEVELOPMENT INFLUENCES THE DISTANCE-DECAY PARAMETER OF THE GRAVITY MODEL.

A threshold gravity model and its possibilities as a planning and forecasting tool for transport problems are presented. The models are based on the earlier papers of the authors. The principal ideas concerning the development of the threshold value gravity model are presented. The new findings are the behaviour of models in representing the real mobility patterns of the population seeking central services. There is an attempt to predict some of the future travel intensities at different distances from the central places.

An analysis of lead-lag structures using a frequency domain approach: Empirical evidence from the Finnish and Swedish stock markets

Abstract The aim of the paper is to analyse the lead and lag structures of two closely related small stock markets: the Finnish and the Swedish. The approach taken is univariate spectral analysis and cross-spectral analysis. Hence the purpose is to study the differences in the spectral characteristics between the two markets and to capture the lead and lag structure between the markets as well as the changes in the spectral characteristics of the market return series over time. The empirical results clearly indicate differences between the return spectra of the two markets. The more volatile Swedish market exhibits a two-day periodicity and autoregressive dependence of about two weeks. The cross-spectrum of the two return series shows a Swedish lead of about 10 days, which decreases to 5 days for the latter part of the observation series. The nonlinearity of the phase, however, indicates a compound effect of several leading terms.

On the usefulness of standard industrial classifications in comparative financial statement analysis

Abstract It is shown that the comparison of the financial ratios between firms should be done with extreme caution even when the firms belong to the same official International Standard Industrial Classification (ISIC) category. A new measure of financial ratio closeness is developed in this paper. The homogeneity of each International Standard Industrial Classification category is established. This is done by comparing the homogeneity of the actual classification to a homogeneous clustering of the firms based on the new closeness measure. The empirical results based on all the Finnish publicly traded firms indicate non-homogeneity of the industry classification categories in terms of financial ratios.

Testing for common autocorrelation features of two scandinavian stock markets

Abstract This paper empirically investigates, in the spirit of Engle and Kozicki (1993) and Engle and Susmel (1993), the long-run persistence of a common serial correlation feature in the index return series of two closely related Scandinavian equity markets; the Finnish and the Swedish stock markets. The paper covers the period January 1920 through December 1993. Monthly index quotations for the period are analyzed as a complete series and for four structurally different subperiods. The return series for both Finland and Sweden seem to have an autocorrelation component present both before and after World War II, but this feature is more pronounced in the Helsinki return series. The strongest common autocorrelation feature is found in the period after the 1974 oil crisis. Nevertheless, the feature does not seem to be common over all the subperiods considered. The common codependence in the last subperiod can be interpreted as a sign of increasing integration between the markets.

An Alternative Estimation Method of the Equity Risk Premium using Financial Statements and Market Data

Abstract This paper develops a method to estimate the equity risk premium. The method exploits the Earn Back Period (EBP) formula presented by Luoma and Ruuhela (2001), which is a generalization of the P/E ratio. The EBP has a clear theoretical interpretation and can be used to compare stocks with different earnings growth rates, while the P/E ratio is not useful if stocks have substantially different growth rates. Since growth is taken into account, differences in EBPs are due to risk. Using this property, a stocks risk premium is derived from the stocks current P/E ratio and from its growth rate of earnings. For investors, this offers a practical method for evaluating stocks.

A pseudo criterion for security betas in the Finnish stock market

Because of nonsynchronous trading, the traditional estimates for systematic risk are biased especially in small security markets. A pseudo criterion is applied to evaluate ten different systematic risk estimates (beta coefficients) in the thin Finnish stock market.

Different beta estimation techniques in infrequently traded and inefficient stock markets

This paper investigates the characteristics of different beta estimation techniques in infrequently traded and inefficient stock markets. These markets are artificially created from actual stock market data by removing return observations and by delaying the information transfer from market returns to individual stocks. Alternative beta estimation techniques are reported to behave differently in different types of markets.

Adjusting for the interval effect bias in beta coefficients on a thin security market: application of a lag distribution model

The problems of measuring the systematic risk of a security are discussed. Prior research indicates that estimates for systematic risk, i.e. beta coefficients, are greatly affected by infrequent trading and the selected return interval. This is the case especially in thin stock markets. A lag distribution model to estimate betas is introduced. In addition, the empirical properties of these betas are compared with several alternative beta estimates using data from a thin security market. The empirical evidence suggests that the estimated betas based on this model are less biased by infrequent trading than the betas based on several other estimation procedures proposed in the literature.