MODELLING AUSTRALIAN BANK BILL RATES: A KALMAN FILTER APPROACH

This paper examines the applicability of the Kalman Filter technique to forecast future spot interest rates, based upon the expectation hypothesis of the term structure of interest rates, in the Australian bank bill market. In this approach, regression estimates are based on the last period's estimate together with data from the current period. In contrast to constant parameter models, this allows effective use of information underlying the process driving the evolution of the parameters. For the period tested, forecasting accuracy of such a time-varying parameter model shows marked improvement over a constant parameter model.     

An examination of causality and predictability between Australian domestic and offshore interest rates

This paper studies the causality and predictability between Australian domestic and offshore short term interest rates in both the first and second moments during the period 1987 to 1996. Causality flow is observed to be stronger from the domestic to the offshore market in the earlier sub periods but characterised by significant two-way causality flow in the latter sub-periods. Volatility tests show that the volatility in one market spills over to the other market simultaneously, which is consistent with Australian markets being well integrated with global markets. The predictability across the two markets in the first moments is examined through an error correction model, whose forecasting performance is assessed relative to a benchmark random walk model. To test the predictability of volatility, four different models are compared: A GARCH model, A GARCH model incorporating contemporaneous spillover effects, a GARCH model with lagged spillover effects, and a benchmark random walk model. Results indicate that the error correction model and the GARCH model with contemporaneous volatility spillover are the superior models for forecasting changes in interest rates and for forecasting volatility, respectively.     

Forecasting government bond yields with large Bayesian vector autoregressions

We propose a new approach to forecasting the term structure of interest rates, which allows to efficiently extract the information contained in a large panel of yields. In particular, we use a large Bayesian Vector Autoregression (BVAR) with an optimal amount of shrinkage towards univariate AR models. The optimal shrinkage is chosen by maximizing the Marginal Likelihood of the model. Focusing on the US, we provide an extensive study on the forecasting performance of the proposed model relative to most of the existing alternative specifications. While most of the existing evidence focuses on statistical measures of forecast accuracy, we also consider alternative measures based on trading schemes and portfolio allocation. We extensively check the robustness of our results, using different datasets and Monte Carlo simulations. We find that the proposed BVAR approach produces competitive forecasts, systematically more accurate than random walk forecasts, even though the gains are small.     

On forecasting interest rates: An efficient markets perspective

Efficient markets do not preclude economic agents from successfully forecasting movements in short-term interest rates. For brief forecast intervals, however, ex ante changes in long-term rates are sufficiently close to zero that economic agents are not likely to improve upon the no-change prediction of the martingale model. Economic agents, in effect, are not likely to succeed in forecasting short-term movements in long-term interest rates. This paper illustrates the closeness of the martingale approximation for two sets of Canadian interest rates data, emphasizing the importance of the forecast interval. The paper then examines three sets of recorded forecasts of Canadian interest rates and finds results consistent with the theoretical discussion.     

Accuracy of International Interest Rate Forecasts

This study exploits a unique data source with contemporaneous forecasts of three-month Euromarket interest rates for five different countries. Professional forecasts are explored in a way that avoids two limitations of previous research. First, rather than being restricted to just U.S. interest rates, data are used for five different countries: the United States, Germany, the United Kingdom, Japan, and Switzerland. Second, the study relies upon data gathered on a single date, rather than over a period of weeks. Consensus forecasts are evaluated against two naive models: a no-change model and a forward rate forecast. In general, the consensus forecasts prove superior to the no-change forecast. The consensus measures, though, are found to be inferior to the forward rate forecast. This is true even considering the dramatic success of the banks in forecasting U.S. rates in this period. However, if the spectacular, and perhaps uniquely successful U.S. results are excluded from consideration, the banks proved dramatically inferior to the forward rate of interest in forecasting interest rates. Thus, the ability of these banks to forecast three-month interest rates for these five countries exceeds that of a no-change forecast, but falls below the forecasting ability of the forward rate.     

How does Monetary Policy Influence Capital Markets? Using a Threshold Regression Model

This study indicates that the effects of interest rate changes on stock prices could be twofold and that the net effect is determined by which effect is dominant. The study employs a threshold regression model to see if, before and after the central banks cut the interest rates, there is a nonlinear relation between interest rates and the stock index. Based on traditional economic theory, stock prices should be inversely related to interest rates. However, the present study finds that as interest rates start to increase or decrease, the stock index prices are significantly and positively related to the interest rates. The changes in interest rates affect stock indexes inversely only after interest rates have crossed a certain threshold. The inverse U-shaped relationship between interest rates and stock indexes differs from the traditional wisdom. It could make interest rates more valuable in forecasting stock indexes, and it holds implications for monetary policies of central banks. To avoid the spurious regression problem, this study uses a cointegration test and an error correction model to confirm the results from the threshold regression model and finds that there is a significant cointegration relationship before and after central banks cut interest rates.     

A generalised arbitrage-free Nelson–Siegel model: The impact of unspanned stochastic volatility

Although statistical term structure models provide exceptional in-sample fitting and out-of-sample forecasting of interest rates, the lack of theoretical background is criticized by academics and practitioners, such as the absent of arbitrage free. In this paper we develop a general Arbitrage-Free Nelson–Siegel model under the HJM framework. It features unspanned stochastic volatility factors while maintaining a Nelson–Siegel factor loading structure. This paper also exploits the potential to jointly model the interest rates and their derivatives.     

Extrapolations of yields over the short run: Forecast or folly?

This paper is concerned with the usefulness of extrapolations in short-term forecasts of long-term interest rates. Four extrapolative models were tested for accuracy in forecasting average yields of new issues of A-rated public utility bonds for forward periods of less than one year. The period 1953–1971 was chosen for study. The best model was found to be the one which assumed that each predicted yield was the same as the current yield at the time the prediction was made. Thus, the best forecasts were forecasts of no change in yields. This finding is supportive of the efficient capital-market thesis.     

预测中国宏观经济变量:专家与模型的组合预测

本文使用组合预测方法,探究以“朗润预测”为代表的专家预测以及计量模型对于中国宏观经济变量的预测效果,并研究对不同预测进行组合预测是否有助于改进预测效果。本文发现,对我国CPI和GDP的增长率,专家预测效果总体上优于模型预测。从原因看,一方面,专家在预测时已经考虑了计量模型的预测信息;另一方面,在经济出现“拐点”的时期,专家通过对实际经济环境和政策的把握,得出更准确的经济预测。组合预测有助于提升预测精度,对专家预测进行组合得到的预测效果优于大多数的专家预测,“模型—专家”组合预测的效果也优于所有的模型和大部分专家预测。     

THE RESPONSE OF FORWARD EXCHANGE RATES TO INTEREST RATE FORECASTING ERRORS

This paper examines the effects of interest rate news on changes in forward foreign exchange rates. Virtually none of the errors in forecasting forward exchange rates are explained by interest rate forecasting errors. The results are consistent with a conjecture that the forward exchange rate is not an estimate of the expected spot exchange rate.     

REGIME-SWITCHING IN AUSTRALIAN SHORT-TERM INTEREST RATES

This paper applies a generalized regime-switching (GRS) model of the short-term interest rate to Australian data. The model allows the short rate to exhibit both mean reversion and conditional heteroscedasticity and nests the popular generalized autoregressive conditional heteroscedasticity (GARCH) and regime-switching specifications. It is shown that empirical estimates of many popular interest rate models provide curious results which imply that innovations to the short rate process are extremely persistent, and that the short rate is potentially non-stationary. The source of these curious results, which are also present in US and European interest rates, is identified in the context of the GRS model, which is shown, via specification and forecasting tests, to capture the features of Australian short-term interest rate data better than existing models. The stochastic process of short-term interest rates in Australia is compared with evidence from the US and Europe, highlighting a number of important differences.     

The October 1979 Change in the U.S. Monetary Regime: Its Impact on the Forecastability of Canadian Interest Rates

Subsequent to the October 1979 shift in monetary policy in the United States, interest rates in North America not only reached unprecedented levels but also exhibited unprecedented volatility. Using Canadian data, the authors show that anticipated quarterly changes in long-term rates associated with the rational-expectations model have remained small during this post-shift period. The authors examine three sets of recorded forecasts of long-term interest rates in Canada and note their failure to improve upon the no-change prediction. The “perverse” relationship between the slope of the yield curve and the subsequent movement in long-term rates exists in the Canadian data but is of only modest value in a forecasting context. The excess returns on long-term bonds implicit in the recorded forecasts of the level of interest rates vary sharply, yet there is little evidence that forecasters have identified a predictable component of time-varying term premia.     

Forecasting Japanese Stock Returns with Financial Ratios and Other Variables

This paper extends the previous analyses of the forecastability of Japanese stock market returns in two directions. First, we carefully construct smoothed market price–earnings ratios and examine their predictive ability. We find that the empirical performance of the price–earnings ratio in forecasting stock returns in Japan is generally weaker than both the price–earnings ratio in comparable US studies and the price dividend ratio. Second, we also examine the performance of several other forecasting variables, including lagged stock returns and interest rates. We find that both variables are useful in predicting aggregate stock returns when using Japanese data. However, while we find that the interest rate variable is useful in early subsamples in this regard, it loses its predictive ability in more recent subsamples. This is because of the extremely limited variability in interest rates associated with operation of the Bank of Japan’s zero interest policy since the late 1990s. In contrast, the importance of lagged returns increases in subsamples starting from the 2000s. Overall, a combination of logged price dividend ratios, lagged stock returns, and interest rates yield the most stable performance when forecasting Japanese stock market returns.     

Dynamic Relationship between Housing Values and Interest Rates in the Korean Housing Market

The goal of this study is to identify the long-term relationship between housing values and interest rates in the Korean housing market, using the cointegration test and spectral analysis. The result shows a long-term negative (–) equilibrium relationship between housing values and interest rates. Moreover, the Granger causality test for confirming the short-term dynamic relationship between these variables shows one-way causality from interest rate to the growth rate of housing values, while the transfer function model demonstrates concretely the causal structure of this relationship. These findings suggest that the interest rate adjustment policy in the Korean housing market can work very effectively and will contribute to forecasting the growth rate of future housing values. This study was supported from the 2003 Daegu University Research Fund     

The information in the Mexican term structure of interest rates: capital market implications

This paper employs the term structure approach to examine Mexican security markets during the recent period of political and economic turmoil. We investigate the characteristics of these markets and the forecast applicability of the pure expectations hypothesis to interest rates in Mexico. We find that both forward rates and spot rate spreads are found to have significant forecasting ability for future spot rates for Mexico. Both forecasting approaches suggest greater predictive ability during the period of higher interest rates and general economic volatility (1995–1996) than the more stable economic environment of the early 1990s (1991–1994).     

Yield curve modeling and forecasting using semiparametric factor dynamics

Using a dynamic semiparametric factor model (DSFM) we investigate the term structure of interest rates. The proposed methodology is applied to monthly interest rates for four southern European countries: Greece, Italy, Portugal and Spain from the introduction of the Euro to the recent European sovereign-debt crisis. Analyzing this extraordinary period, we compare our approach with the standard market method – dynamic Nelson–Siegel model. Our findings show that two nonparametric factors capture the spatial structure of the yield curve for each of the bond markets separately. We attributed both factors to the slope of the yield curve. For panel term structure data, three nonparametric factors are necessary to explain 95% variation. The estimated factor loadings are unit root processes and reveal high persistency. In comparison with the benchmark model, the DSFM technique shows superior short-term forecasting in times of financial distress.     

The generation of stock market cycles

This paper demonstrates that the relation between stock market and business cycle dynamics can be conceptualized using a dividend discount model. The interaction of changes in earnings and interest rates throughout the economic cycle are shown to cause changes in the level of stock prices. This implies that monitoring and forecasting these factors can help explain and possibly predict stock price behavior over time.     

Real term structure forecasts of consumption growth

This paper employs an empirically tractable affine term structure model of real interest rates to examine the predictive ability of the real short-term interest rate and its term spread with a longer-term interest rate to predict future real consumption growth. The estimates of the model provide support of the consumption smoothing hypothesis. The paper shows that the real term structure is spanned by two mean-reverting state variables. The mean-reverting property of these variables can consistently explain the forecasting ability of the short-term real rate and term spread to forecast future consumption growth rate, over different horizons ahead. Although the risks associated with changes in these variables are both priced in the market, they are not volatile enough to obscure the information of the real term structure about future real consumption growth.     

An adaptive hierarchical fuzzy logic system for modelling of financial systems

In this paper an intelligent hierarchical fuzzy logic system using genetic algorithms for the prediction and modelling of interest rates in Australia is developed. The proposed system uses a hierarchical fuzzy logic system in which a genetic algorithm is used as a training method for learning the fuzzy rules knowledge bases that are used for prediction of interest rates in Australia. A hierarchical fuzzy logic system is developed to model and predict three‐month (quarterly) interest rate ?uctuations. The system is further trained to model and predict interest rates for six‐month and one‐year periods. The proposed system is developed with ?rst two, three, then four and ?nally ?ve hierarchical knowledge bases to model and predict interest rates. A novel architecture called a feed forward fuzzy logic system using fuzzy logic and genetic algorithms is also developed to predict interest rates. A back‐propagation hierarchical neural network system is also developed to predict interest rates for three‐month, six‐month and one‐year periods. The results obtained from these two systems are then compared with the hierarchical fuzzy logic system results and conclusions are drown on the accuracy of all systems for prediction of interest rates in Australia. Copyright © 2004 John Wiley & Sons, Ltd.     

A melting pot — Gold price forecasts under model and parameter uncertainty

Gold is special as it is influenced by a wide range of factors such as commodity prices, interest rates, inflation expectations, exchange rate changes and stock market volatility. Hence, forecasting the price of gold is a difficult task and the main problem a researcher faces is to select the relevant regressors at each point in time. This model uncertainty in combination with parameter uncertainty is explicitly accounted for by Dynamic Model Averaging (DMA) which allows both the forecasting model and the coefficients to change over time. Based on this framework, we systematically evaluate a large set of possible gold price determinants and find that DMA (1) improves forecasts compared to other frameworks, (2) yields strong time-variation of gold price predictors and (3) favors parsimonious models. The results also show that typical in-sample features of gold such as its hedge property are weaker in an out-of-sample context.