Exchange rate forecasting using a combined parametric and nonparametric self-organising modelling approach

Financial prediction has attracted a lot of interest due to the financial implications that the accurate prediction of financial markets can have. A variety of data driven modellingapproaches have been applied but their performance has produced mixed results. In this study we apply both parametric (neural networks with active neurons) and nonparametric (analog complexing) self-organisingmodelling methods for the daily prediction of the exchangerate market. We also propose acombinedapproach where the parametric and nonparametricself-organising methods are combined sequentially, exploiting the advantages of the individual methods with the aim of improving their performance. The combined method is found to produce promising results and to outperform the individual methods when tested with two exchangerates: the American Dollar and the Deutche Mark against the British Pound.

Short-horizon excess returns and exchange rate and interest rate effects

stocks. We examine the effects of foreign exchange (FX) and interest rate changes on the excess returns of U.S. stocks, for short-horizons of 1-40 days. Our new evidence shows a tendency for the volatility of both excess returns and FX rate changes to be negatively related with FX rate and interest rate effects. Both the number of firms with significant FX rate and interest rate effects and the magnitude of their exposures increase with the length of the return horizon. Our finding seems inconsistent with the view that firms hedge effectively at short-return horizons.

The sensitivity of US banks' stock returns to interest and exchange rate changes

Purpose – The purpose of this paper is to investigate the impact of foreign exchange and interest rate changes on US banks’ stock returns. Design/methodology/approach – The approach employs an EGARCH model to account for the ARCH effects in daily returns. Most prior studies have used standard OLS estimation methods with the result that the presence of ARCH effects would have affected estimation efficiency. For comparative purposes, the standard OLS estimation method is also used to measure sensitivity. Findings – The findings are as follows: under the conditional t-distributional assumption, the EGARCH model generated a much better fit to the data although the goodness-of-fit of the model is not entirely satisfactory; the market index return accounts for most of the variation in stock returns at both the individual bank and portfolio levels; and the degree of sensitivity of the stock returns to interest rate and FX rate changes is not very pronounced despite the use of high frequency data. Earlier results had indicated that daily data provided greater evidence of exposure sensitivity. Practical implications – Assuming that banks do not hedge perfectly, these findings have important financial implications as they suggest that the hedging policies of the banks are not reflected in their stock prices. Alternatively, it is possible that different GARCH-type models might be more appropriate when modelling high frequency returns. Originality/value – The paper contributes to existing knowledge in the area by showing that ARCH effects do impact on measures of sensitivity.

Forecasting the UK/US exchange rate with divisia monetary models and neural networks

This paper compares the UK/US exchange rate forecasting performance of linear and nonlinear models based on monetary fundamentals, to a random walk (RW) model. Structural breaks are identified and taken into account. The exchange rate forecasting framework is also used for assessing the relative merits of the official Simple Sum and the weighted Divisia measures of money. Overall, there are four main findings. First, the majority of the models with fundamentals are able to beat the RW model in forecasting the UK/US exchange rate. Second, the most accurate forecasts of the UK/US exchange rate are obtained with a nonlinear model. Third, taking into account structural breaks reveals that the Divisia aggregate performs better than its Simple Sum counterpart. Finally, Divisia-based models provide more accurate forecasts than Simple Sum-based models provided they are constructed within a nonlinear framework.

Forecasting exchange rates with linear and nonlinear models

In this paper, the exchange rate forecasting performance of neural network models are evaluated against the random walk, autoregressive moving average and generalised autoregressive conditional heteroskedasticity models. There are no guidelines available that can be used to choose the parameters of neural network models and therefore, the parameters are chosen according to what the researcher considers to be the best. Such an approach, however,implies that the risk of making bad decisions is extremely high, which could explain why in many studies, neural network models do not consistently perform better than their time series counterparts. In this paper, through extensive experimentation, the level of subjectivity in building neural network models is considerably reduced and therefore giving them a better chance of Forecasting exchange rates with linear and nonlinear models 415 performing well. The results show that in general, neural network models perform better than the traditionally used time series models in forecasting exchange rates.

Predicting returns in U.S. financial sector indices

This study focuses on: (i) the responsiveness of the U.S. financial sector stock indices to foreign exchange (FX) and interest rate changes; and, (ii) the extent to which good model specification can enhance the forecasts from the associated models. Three models are considered. Only the error-correction model (ECM) generated efficient and consistent coefficient estimates. Furthermore, a simple zero lag model in differences which is clearly mis-specified, generated forecasts that are better than those of the ECM, even if the ECM depicts relationships that are more consistent with economic theory. In brief, FX and interest rate changes do not impact on the return-generating process of the stock indices in any substantial way. Most of the variation in the sector stock indices is associated with past variation in the indices themselves and variation in the market-wide stock index. These results have important implications for financial and economic policies.

Forecasting exchange rates with linear and nonlinear models

In this paper the exchange rate forecasting performance of neural network models are evaluated against random walk and a range of time series models. There are no guidelines available that can be used to choose the parameters of neural network models and therefore the parameters are chosen according to what the researcher considers to be the best. Such an approach, however, implies that the risk of making bad decisions is extremely high which could explain why in many studies neural network models do not consistently perform better than their time series counterparts. In this paper through extensive experimentation the level of subjectivity in building neural network models is considerably reduced and therefore giving them a better chance of performing well. Our results show that in general neural network models perform better than traditionally used time series models in forecasting exchange rates.

Volatility transmission between stock and bond markets

A two-factor no-arbitrage model is used to provide a theoretical link between stock and bond market volatility. While this model suggests that short-term interest rate volatility may, at least in part, drive both stock and bond market volatility, the empirical evidence suggests that past bond market volatility affects both markets and feeds back into short-term yield volatility. The empirical modelling goes on to examine the (time-varying) correlation structure between volatility in the stock and bond markets and finds that the sign of this correlation has reversed over the last 20 years. This has important implications far portfolio selection in financial markets. © 2005 Elsevier B.V. All rights reserved.

A two-factor model of the U.K. yield curve

I model the forward premium in the U.K. gilt-edged market over the period 1982–96 using a two-factor general equilibrium model of the term structure of interest rates. The model permits the decomposition of the forward premium into separate components representing interest rate expectations, the risk premia associated with each of the underlying factors, and terms capturing the direct impact of the variances of the factors on the shape of the forward curve.

Econometrics of yield spreads in the money market:a note

The literature on bond markets and interest rates has focused largely on the term structure of interest rates, specifically, on the so-called expectations hypothesis. At the same time, little is known about the nature of the spread of the interest rates in the money market beyond the fact that such spreads are generally unstable. However, with the evolution of complex financial instruments, it has become imperative to identify the time series process that can help one accurately forecast such spreads into the future. This article explores the nature of the time series process underlying the spread between three-month and one-year US rates, and concludes that the movements in this spread over time is best captured by a GARCH(1,1) process. It also suggests the use of a relatively long term measure of interest rate volatility as an explanatory variable. This exercise has gained added importance in view of the revelation that GARCH based estimates of option prices consistently outperform the corresponding estimates based on the stylized Black-Scholes algorithm.

Accommodating energy price volatility in life cycle cost analysis of asphalt pavements

Energy price is related to more than half of the total life cycle cost of asphalt pavements. Furthermore, the fluctuation related to price of energy has been much higher than the general inflation and interest rate. This makes the energy price inflation an important variable that should be addressed when performing life cycle cost (LCC) studies re- garding asphalt pavements. The present value of future costs is highly sensitive to the selected discount rate. Therefore, the choice of the discount rate is the most critical element in LCC analysis during the life time of a project. The objective of the paper is to present a discount rate for asphalt pavement projects as a function of interest rate, general inflation and energy price inflation. The discount rate is defined based on the portion of the energy related costs during the life time of the pavement. Consequently, it can reflect the financial risks related to the energy price in asphalt pavement projects. It is suggested that a discount rate sensitivity analysis for asphalt pavements in Sweden should range between –20 and 30%.

Leading edge forecasting techniques for exchange rate prediction

This paper describes how modern machine learning techniques can be used in conjunction with statistical methods to forecast short term movements in exchange rates, producing models suitable for use in trading. It compares the results achieved by two different techniques, and shows how they can be used in a complementary fashion. The paper draws on experience of both inter- and intra-day forecasting taken from earlier studies conducted by Logica and Chemical Bank Quantitative Research and Trading (QRT) group's experience in developing trading models.

Steric stabilization and dissolution characteristics of aqueous suspensions of pharmaceutical interest

The adsorption of two qroups of nonionic surface active agents and a series of hiqh molecular weiqht hydrophilic polymer fractions onto a polystyrene latex and a drug substance diloxanide furoate B.P. has been investigated. The presence of pores within the drug surface has been demonstrated and this is shown to increase the adsorption of low molecular weight polymer species. Differences in the maximum amount of polymer adsorbed at both solid-solution interfaces have been ascribed to the different hydrophobicities of the surface as determined by contact angle measurements. Adsorbed layer thicknesses of polymer on polystyrene latex have been determined by three techniques: microelectrophoresis, intensity fluctuation spectroscopy and by viscometric means. These results, in combination with adsorption data, were used to interpret the configuration of the adsorbed polymer molecules at the interface. The type of druq suspension produced on adsorbing the different polymers in the absence of electrostatic stabilization was correlated with theoretical prediuctions of suspension characteristics deduced from potential energy diagrams, The agreement was good for the adsorption of short chain length surfactants, but for the polyvinylalcohols, discrepancies were found between experiment and theory. This was attributed to the inappropriate use of a mean segment density approximation within the adsorbed layer to calculate attractive potentials between particles. A maximum in the redispersibility values for suspensions coated with adsorbed nonylphenylethoxylates was attributed to "partial static stabilization" of the particles in conjunction with the attractive forces operating in the sediment between bare surface patches on neighbouring particles. No significant change in the dissolution of the drug was observed when nonylphenylethoxylates were adsorbed due to desorption upon contact with the dissolution medium. Pluronic F68 and all the polyvinylalcohol fractions caused a reduction in the dissolution rate which is explained by the decreased diffusion of drug' through the adsorbed polymer layer.

Trade balance and real exchange rate:new evidence from Mauritius-UK trade

Two main questions are addressed here: is there a long-run relationship between trade balance and real exchange rate for the bilateral trade between Mauritius and UK? Does a J-curve exist for this bilateral trade? Our findings suggest that the real exchange rate is cointegrated with the trade balance and we find evidence of a J-curve effect. We also find bidirectional causality between the trade balance and the real exchange rate in the long-run. The real exchange rate also causes the trade balance in the short-run. In an out-of-sample forecasting experiment, we also find that real exchange rate contains useful information that can explain future movements in the trade balance.