Browsing by Subject "forecasting"

Sort by: Order: Results:

Now showing items 1-20 of 23
  • Beniard, Henry (Helsingfors universitet, 2010)
    This thesis researches empirically, whether variables that are able to reliably predict Finnish economic activity can be found. The aim of this thesis is to find and combine several variables with predictive ability into a composite leading indicator of the Finnish economy. The target variable it attempts to predict, and thus the measure of the business cycle used, is Finnish industrial production growth. Different economic theories suggest several potential predictor variables in categories, such as consumption data, data on orders in industry, survey data, interest rates and stock price indices. Reviewing a large amount of empirical literature on economic forecasting, it is found that particularly interest rate spreads, such as the term spread on government bonds, have been useful predictors of future economic growth. However, the literature surveyed suggests that the variables found to be good predictors seem to differ depending on the economy being forecast, the model used and the forecast horizon. Based on the literature reviewed, a pool of over a hundred candidate variables is gathered. A procedure, involving both in-sample and pseudo out-of-sample forecast methods, is then developed to find the variables with the best predictive ability from this set. This procedure yields a composite leading indicator of the Finnish economy comprising of seven component series. These series are very much in line with the types of variables found useful in previous empirical research. When using the developed composite leading indicator to forecast in a sample from 2007 to 2009, a time span including the latest recession, its forecasting ability is far poorer. The same occurs when forecasting a real-time data set. It would seem, however, that individual very large forecast errors are the main reason for the poor performance of the composite leading indicator in these forecast exercises. The findings in this thesis suggest several developments to the methods adopted in order to produce more accurate forecasts. Other intriguing topics for further research are also explored.
  • Hellstrand, Julia Ingrid Sofia; Nisén, Jessica; Myrskylä, Mikko (2020)
    The ongoing period fertility decline in the Nordic countries is particularly strong in Finland, where the total fertility rate (TFR) reached an all-time low of 1.41 in 2018. We analyse the decrease in Finland's TFR in 2010–17, and assess its consequences for cohort fertility using complementary approaches. Decomposition of this fertility decline shows that first births and women aged <30 are making the largest contributions. However, women aged 30–39 are also, for the first time in decades, experiencing a sustained fertility decline. Tempo adjustments to the TFR suggest that quantum change is part of the decline. Several forecasting methods indicate that cohort fertility is likely to decline from the long-lasting level of 1.85–1.95 to 1.75 or lower among women born in the mid-1980s. Without an exceptionally strong recovery in fertility, Finnish cohort fertility is likely to decline to levels currently observed among countries with very low fertility.
  • Anttonen, Jetro (Helsingin yliopisto, 2019)
    In this thesis, a conditional BVARX forecasting model for short and medium term economic forecasting is developed. The model is especially designed for small-open economies and its performance on forecasting several Finnish economic variables is assessed. Particular attention is directed to the hyperparameter choice of the model. A novel algorithm for hyperparameter choice is proposed and it is shown to outperform the marginal likelihood based approach often encountered in the literature. Other prominent features of the model include conditioning on predictive densities and exogeneity of the global economic variables. The model is shown to outperform univariate benchmark models in terms of forecasting accuracy for forecasting horizons up to eight quarters ahead.
  • Nyberg, Henri (2007)
    In the econometric literature, there is not much research on binary time series. However, in the last couple of years, some new binary time series models have been suggested where the traditional static model is extended by different kinds of dynamic structures. In the thesis, the main goal is to forecast the economic recession periods occurred in the United States and Germany. Especially, it is interesting to consider new so called autoregressive models suggested by Kauppi and Saikkonen (2007). They also proposed a new iterative framework for multiperiod forecasts. Iterated multiperiod forecasts are compared with previously used direct forecasts. The parameter estimation of the employed dynamic models can be done by a maximum likelihood method described in the theoretical part of the thesis. Model diagnostics and forecasting procedures are as well considered. Two LM tests for the usefulness of the autoregressive part are also proposed. It is shown that the models with autoregressive part seem to outperform the static probit models in terms of in-sample and out-of-sample predictions. The best forecasting models give the distinct recession signals of the forthcoming recession which started in 2001. In the dynamic probit model with the lagged recession indicator, the iterative forecasts seem to be superior to direct forecasts. Several empirical studies have proposed that the yield curve, which is defined as a spread between long and short term interest rates, is an accurate explanatory variable in recession forecasting. As in previous studies, the domestic yield curve is an important predictor variable in both countries but also the foreign yield curve, stock market returns and, in the case of Germany, the interest rate differential between the United States and Germany are statistically significant predictors in the dynamic probit models. The most important reference is the article of Kauppi and Saikkonen (2007). Chauvet and Potter (2005) have also proposed important model variants and forecasting methods for the binary time series models. Davidson and MacKinnon's book (1993) and their article (1984) are important references for the theoretical part and especially for the proposed LM tests for the autoregressive part. The articles of Bernard and Gerlach (1998) and Estrella and Mishkin (1998) are important references for the recession forecasting. The forecasting power of the yield curves and the stock returns are also considered in numerous other articles.
  • Fornaro, Paolo (Helsingfors universitet, 2011)
    In recent years, thanks to developments in information technology, large-dimensional datasets have been increasingly available. Researchers now have access to thousands of economic series and the information contained in them can be used to create accurate forecasts and to test economic theories. To exploit this large amount of information, researchers and policymakers need an appropriate econometric model.Usual time series models, vector autoregression for example, cannot incorporate more than a few variables. There are two ways to solve this problem: use variable selection procedures or gather the information contained in the series to create an index model. This thesis focuses on one of the most widespread index model, the dynamic factor model (the theory behind this model, based on previous literature, is the core of the first part of this study), and its use in forecasting Finnish macroeconomic indicators (which is the focus of the second part of the thesis). In particular, I forecast economic activity indicators (e.g. GDP) and price indicators (e.g. consumer price index), from 3 large Finnish datasets. The first dataset contains a large series of aggregated data obtained from the Statistics Finland database. The second dataset is composed by economic indicators from Bank of Finland. The last dataset is formed by disaggregated data from Statistic Finland, which I call micro dataset. The forecasts are computed following a two steps procedure: in the first step I estimate a set of common factors from the original dataset. The second step consists in formulating forecasting equations including the factors extracted previously. The predictions are evaluated using relative mean squared forecast error, where the benchmark model is a univariate autoregressive model. The results are dataset-dependent. The forecasts based on factor models are very accurate for the first dataset (the Statistics Finland one), while they are considerably worse for the Bank of Finland dataset. The forecasts derived from the micro dataset are still good, but less accurate than the ones obtained in the first case. This work leads to multiple research developments. The results here obtained can be replicated for longer datasets. The non-aggregated data can be represented in an even more disaggregated form (firm level). Finally, the use of the micro data, one of the major contributions of this thesis, can be useful in the imputation of missing values and the creation of flash estimates of macroeconomic indicator (nowcasting).
  • Päivinen, Ville (Helsingin yliopisto, 2020)
    Efficient estimation and forecasting of the cash flow is an interest of pension insurance companies. At the turn of the year 2019 Finnish national Incomes Register was introduced and the payment cycle of TyEL (Employees Pensions Act) changed substantially. TyEL payments are calculated and paid monthly by all of the employers insured under TyEL after January 1st 2019. Vector autoregressive (VAR) models are one of the most used and successful multivariate time series models. They are widely used with economic and financial data due to the good forecasting abilities and the possibility of analysing dynamic structures between the variables of the model. The aim of this thesis is to determine whether a VAR model offers a good fit for predicting the incoming TyEL cash flow of a pension insurance company. With the monthly payment cycle arises a question of seasonality of the incoming TyEL cash flow, and thus the focus is on forecasting with seasonally varying data. The essential theory of VAR models is given. The forecast abilities are tested by building a VAR model for monthly, seasonally varying time series similar than the pension insurance companies would have and could use for the particular prediction problem.
  • Widgrén, Miska (Helsingin yliopisto, 2018)
    Internet search engines produce large amounts of data. This thesis shows how the data about internet searches can be used for inflation forecasting. The internet search data is constructed from searches performed on Google. The sample covers eurozone countries over the period from January 2004 to July 2017. The performance of the internet searches is evaluated relative to traditional inflation forecasting benchmark models. The usefulness of the Google searches is evaluated by Granger causality and out-of-sample performance. Furthermore, to study the robustness of the results, the out-of-sample forecasting accuracy has been evaluated in two separate sub-samples. In this study, a simple autoregressive model augmented with internet searches is found to outperform the traditional benchmark models in predicting the month-over-month inflation of the near future. Moreover, the improvement is statistically significant in one-month ahead forecasting accuracy. The Google model also outperforms the benchmark models in year-over-year inflation forecasting. However, the improvement in year-over-year forecasting accuracy is modest. In addition, this thesis shows that the seasonally adjusted internet search data can improve the performance of the Google model slightly. This thesis is related to fast-growing research on employing Google Trends data in economic forecasting. The findings in this thesis require further research in exploiting the internet search data in macroeconomic forecasting.
  • Velling, Pirkko; Tigerstedt, P. M. A. (Suomen metsätieteellinen seura, 1984)
  • Kokkonen, Paavo (Helsingin yliopisto, 2019)
    House prices have a very important role in the economy. House prices have strong influence to the economy especially in Finland, where around one-half of the value of households' total assets is coming from households' own dwellings. The real estate investment market is large in proportion in Finland when compared internationally to the size of the economy. Surprisingly, there are not many papers discussing the relationship between house prices and output in Finland. This paper intends to enrich the recent literature about this topic. Primary research question in this paper was do house prices affect output in Finland. Secondary interests were transmission mechanisms. The methods used in this thesis are typical in vector autoregression (VAR) analysis in recent literature. First, the time series are analysed visually and with unit root tests. Then, the optimal VAR model was chosen by using different information criterion tests and correlation tests. After selecting the optimal VAR model, Granger causality was tested with Toda-Yamamoto causality test. Other methods utilized in this paper were cointegration tests, forecasting, impulse responses and forecast error variance decomposition. These empirical methods were computed in intention to answer the research question. The most important empirical results of the paper were following. The results of Toda-Yamamoto causality test suggested that there are unidirectional Granger causality going from real house prices to real GDP per capita. This indicates that house prices could have significant explanatory power for GDP. Cointegration tests implied that the series are not cointegrated. This suggests that the series do not share a common stochastic trend for the long-run. The results of forecasting supported the results of Toda-Yamamoto causality test and it seemed that house prices might be a useful predictor when forecasting output. This result implied that the house prices have an effect on output. The analysis of impulse responses suggested that a house price shock have a positive and persistent effect on output. Forecast error variance decomposition intimated that after 15 quarters 63 percent of the output variation can be explained by the house price shock which was suspiciously strong result. The conclusion were made based on the results of the empirical analysis. Answer to the primary research question were house prices seem to have effect on output in Finland. The results of this paper supported the theory behind the wealth effect. If policy makers have a desire to stabilize output in Finland, they might need to consider stabilizing the house prices to further the stabilization of the output. It is necessary to understand the effects of housing prices to the business cycle for an efficient housing policy strategy.
  • Scolini, C.; Chané, E.; Pomoell, J.; Rodriguez, L.; Poedts, S. (2020)
    Predictions of the impact of coronal mass ejections (CMEs) in the heliosphere mostly rely on cone CME models, whose performances are optimized for locations in the ecliptic plane and at 1 AU (e.g., at Earth). Progresses in the exploration of the inner heliosphere, however, advocate the need to assess their performances at both higher latitudes and smaller heliocentric distances. In this work, we perform 3-D magnetohydrodynamics simulations of artificial cone CMEs using the EUropean Heliospheric FORecasting Information Asset (EUHFORIA), investigating the performances of cone models in the case of CMEs launched at high latitudes. We compare results obtained initializing CMEs using a commonly applied approximated (Euclidean) distance relation and using a proper (great circle) distance relation. Results show that initializing high-latitude CMEs using the Euclidean approximation results in a teardrop-shaped CME cross section at the model inner boundary that fails in reproducing the initial shape of high-latitude cone CMEs as a circular cross section. Modeling errors arising from the use of an inappropriate distance relation at the inner boundary eventually propagate to the heliospheric domain. Errors are most prominent in simulations of high-latitude CMEs and at the location of spacecraft at high latitudes and/or small distances from the Sun, with locations impacted by the CME flanks being the most error sensitive. This work shows that the low-latitude approximations commonly employed in cone models, if not corrected, may significantly affect CME predictions at various locations compatible with the orbit of space missions such as Parker Solar Probe, Ulysses, and Solar Orbiter.
  • Kuusela, Kullervo (Suomen metsätieteellinen seura, 1964)
  • Morley, S. K.; Brito, T. V.; Welling, D. T. (2018)
    Quantitative assessment of modeling and forecasting of continuous quantities uses a variety of approaches. We review existing literature describing metrics for forecast accuracy and bias, concentrating on those based on relative errors and percentage errors. Of these accuracy metrics, the mean absolute percentage error (MAPE) is one of the most common across many fields and has been widely applied in recent space science literature and we highlight the benefits and drawbacks of MAPE and proposed alternatives. We then introduce the log accuracy ratio and derive from it two metrics: the median symmetric accuracy and the symmetric signed percentage bias. Robust methods for estimating the spread of a multiplicative linear model using the log accuracy ratio are also presented. The developed metrics are shown to be easy to interpret, robust, and to mitigate the key drawbacks of their more widely used counterparts based on relative errors and percentage errors. Their use is illustrated with radiation belt electron flux modeling examples.
  • Karakozova, Olga (Svenska handelshögskolan, 2005)
    Economics and Society
    Recently, focus of real estate investment has expanded from the building-specific level to the aggregate portfolio level. The portfolio perspective requires investment analysis for real estate which is comparable with that of other asset classes, such as stocks and bonds. Thus, despite its distinctive features, such as heterogeneity, high unit value, illiquidity and the use of valuations to measure performance, real estate should not be considered in isolation. This means that techniques which are widely used for other assets classes can also be applied to real estate. An important part of investment strategies which support decisions on multi-asset portfolios is identifying the fundamentals of movements in property rents and returns, and predicting them on the basis of these fundamentals. The main objective of this thesis is to find the key drivers and the best methods for modelling and forecasting property rents and returns in markets which have experienced structural changes. The Finnish property market, which is a small European market with structural changes and limited property data, is used as a case study. The findings in the thesis show that is it possible to use modern econometric tools for modelling and forecasting property markets. The thesis consists of an introduction part and four essays. Essays 1 and 3 model Helsinki office rents and returns, and assess the suitability of alternative techniques for forecasting these series. Simple time series techniques are able to account for structural changes in the way markets operate, and thus provide the best forecasting tool. Theory-based econometric models, in particular error correction models, which are constrained by long-run information, are better for explaining past movements in rents and returns than for predicting their future movements. Essay 2 proceeds by examining the key drivers of rent movements for several property types in a number of Finnish property markets. The essay shows that commercial rents in local markets can be modelled using national macroeconomic variables and a panel approach. Finally, Essay 4 investigates whether forecasting models can be improved by accounting for asymmetric responses of office returns to the business cycle. The essay finds that the forecast performance of time series models can be improved by introducing asymmetries, and the improvement is sufficient to justify the extra computational time and effort associated with the application of these techniques.
  • Larsson, Aron (Helsingin yliopisto, 2021)
    The science of fish stock assessment is one that is very resource and labor intensive, with stock assessment models historically being based on data that causes a model to overestimate the strength of a population, sometimes with drastic consequences. The need of cost-effective assessment models and approaches increases, which is why I looked into using Bayesian modeling and networks as an approach not often used in fisheries science. I wanted to determine if it could be used to predict both recruitment and spawning stock biomass of four fish species in the north Atlantic, cod, haddock, pollock and capelin, based on no other evidence other than the recruitment or biomass data of the other species and if these results could be used to lower the uncertanties of fish stock models. I used data available on the RAM legacy database to produce four different models with the statistical software R, based on four different Bayes algorithms found in the R-package bnlearn, two based on continuous data and two based on discrete data. What I found was that there is much potential in the Bayesian approach to stock prediction and forecasting, as our prediction error percentage ranged between 1 and 40 percent. The best predictions were made when the species used as evidence had a high correlation coefficient with the target species, which was the case with cod and haddock biomass, which had a unusually high correlation of 0.96. As such, this approach could be used to make preliminary models of interactions between a high amount of species in a specific area, where there is data abundantly available and these models could be used to lower the uncertanties of the stock assessments. However, more research into the applicability for this approach to other species and areas needs to be conducted.
  • Uusitalo, Jori (The Society of Forestry in Finland - The Finnish Forest Research Institute, 1997)
    To enhance the utilization of the wood, the sawmills are forced to place more emphasis on planning to master the whole production chain from the forest to the end product. One significant obstacle to integrating the forest-sawmill-market production chain is the lack of appropriate information about forest stands. Since the wood procurement point of view in forest planning systems has been almost totally disregarded there has been a great need to develop an easy and efficient pre-harvest measurement method, allowing separate measurement of stands prior to harvesting. The main purpose of this study was to develop a measurement method for pine stands which forest managers could use in describing the properties of the standing trees for sawing production planning. Study materials were collected from ten Scots pine stands (Pinus sylvestris) located in North Häme and South Pohjanmaa, in southern Finland. The data comprise test sawing data on 314 pine stems, dbh and height measures of all trees and measures of the quality parameters of pine sawlog stems in all ten study stands as well as the locations of all trees in six stands. The study was divided into four sub-studies which deal with pine quality prediction, construction of diameter and dead branch height distributions, sampling designs and applying height and crown height models. The final proposal for the pre-harvest measurement method is a synthesis of the individual sub-studies. Quality analysis resulted in choosing dbh, distance from stump height to the first dead branch (dead branch height), crown height and tree height as the most appropriate quality characteristics of Scots pine. Dbh and dead branch height are measured from each pine sample tree while height and crown height are derived from dbh measures by aid of mixed height and crown height models. Pine and spruce diameter distribution as well as dead branch height distribution are most effectively predicted by the kernel function. Roughly 25 sample trees seems to be appropriate in pure pine stands. In mixed stands the number of sample trees needs to be increased in proportion to the intensity of pines in order to attain the same level of accuracy.
  • Riihinen, Päiviö (Suomen metsätieteellinen seura, 1962)
  • Lerviks, Alf-Erik (Svenska handelshögskolan, 2004)
    Research Reports
    A diffusion/replacement model for new consumer durables designed to be used as a long-term forecasting tool is developed. The model simulates new demand as well as replacement demand over time. The model is called DEMSIM and is built upon a counteractive adoption model specifying the basic forces affecting the adoption behaviour of individual consumers. These forces are the promoting forces and the resisting forces. The promoting forces are further divided into internal and external influences. These influences are operationalized within a multi-segmental diffusion model generating the adoption behaviour of the consumers in each segment as an expected value. This diffusion model is combined with a replacement model built upon the same segmental structure as the diffusion model. This model generates, in turn, the expected replacement behaviour in each segment. To be able to use DEMSIM as a forecasting tool in early stages of a diffusion process estimates of the model parameters are needed as soon as possible after product launch. However, traditional statistical techniques are not very helpful in estimating such parameters in early stages of a diffusion process. To enable early parameter calibration an optimization algorithm is developed by which the main parameters of the diffusion model can be estimated on the basis of very few sales observations. The optimization is carried out in iterative simulation runs. Empirical validations using the optimization algorithm reveal that the diffusion model performs well in early long-term sales forecasts, especially as it comes to the timing of future sales peaks.
  • Kilpua, E. K. J.; Fontaine, D.; Moissard, C.; Ala-Lahti, M.; Palmerio, E.; Yordanova, E.; Good, S. W.; Lumme, E.; Osmane, A.; Palmroth, M.; Turc, L.; Kalliokoski, Milla (2019)
    We present a statistical study of interplanetary conditions and geospace response to 89 coronal mass ejection-driven sheaths observed during Solar Cycles 23 and 24. We investigate in particular the dependencies on the driver properties and variations across the sheath. We find that the ejecta speed principally controls the sheath geoeffectiveness and shows the highest correlations with sheath parameters, in particular in the region closest to the shock. Sheaths of fast ejecta have on average high solar wind speeds, magnetic (B) field magnitudes, and fluctuations, and they generate efficiently strong out-of-ecliptic fields. Slow-ejecta sheaths are considerably slower and have weaker fields and field fluctuations, and therefore they cause primarily moderate geospace activity. Sheaths of weak and strong B field ejecta have distinct properties, but differences in their geoeffectiveness are less drastic. Sheaths of fast and strong ejecta push the subsolar magnetopause significantly earthward, often even beyond geostationary orbit. Slow-ejecta sheaths also compress the magnetopause significantly due to their large densities that are likely a result of their relatively long propagation times and source near the streamer belt. We find the regions near the shock and ejecta leading edge to be the most geoeffective parts of the sheath. These regions are also associated with the largest B field magnitudes, out-of-ecliptic fields, and field fluctuations as well as largest speeds and densities. The variations, however, depend on driver properties. Forecasting sheath properties is challenging due to their variable nature, but the dependence on ejecta properties determined in this work could help to estimate sheath geoeffectiveness through remote-sensing coronal mass ejection observations.
  • Ervasti, Seppo; Heikinheimo, Lauri; Holopainen, Viljo; Kuusela, Kullervo; Sirén, Gustaf (Suomen metsätieteellinen seura, 1965)