Figure 1 - uploaded by Foteini Kyriazi
Content may be subject to copyright.
Source publication
There is ample literature on foreign exchange rate predictability and the results point out towards entirely different directions, some of it claiming that foreign exchanges are unpredictable random walks, some other that they are mean-reverting and coin-tegrating and yet some other still that they can be best modelled and forecasted by non-linear...
Contexts in source publication
Context 1
... in all, we have collected results on the following explanatory variables: the monthly return of the GBP/USD, the realized volatilities of the EUR/USD and the GBP/USD, the realized correlation between the EUR/USD and the GBP/USD, and then the rolling-window volatilities and the rolling window correlation, for a sum of 7 explanatory variables. We present the levels of the two series of interest and their estimated realized correlation in Figure 1; that they two series move together is apparent from the top panel and that they maintain a consis- highly negative territory). All these salient features of the data are useful in explaining our forecasting results later on. ...
Context 2
... took the best model for each panel of Table 11 and run the MZ-regressions as before, summarizing our results in It is interesting to see, for this last set of top performing models, and in continuation of the predictability analysis of the full-sample results, the rolling estimates of R 2 's. We present them in three figures, Figures 9, 10 and 11, with the models appearing in the same order as in Table 12 that holds the MZ-regression results. There are two issues to consider here, the magnitude of the R 2 and its direction. ...
Citations
... We then "jump" to papers on or after 2020's just to illustrate the continued strong research interest on this topic. For example, Dai et al. (2020) [17] who consider technical indicators for forecasting stock returns, Dichtl (2020) [18] on forecasting excess returns of the gold market, Kyriazi and Thomakos (2020) [42] on the existence and interpretation of predictability in currencies, Liang et al. (2022) [48] on the use of dimensionality reduction methods for predicting market returns, Li et al. (2022) [49] on forecasting US stock market returns and Lv and Qi (2022) [52] that examine return predictability from a combination forecasting perspective. A stream of very recent papers includes Brennan and Taylor (2023) [8] on the interplay of expected returns and risk in the stock market, Casta (2023) [10] on the relationship between inflation, interest rates and the predictability of asset returns, Chen et al. (2023) [11] on use of economic policy uncertainty indices as predictors of market returns -with additional work on the topic by Huang et al. (2023) [33], Cotter et al. (2023) [16] on commodity futures return, Guerard et al. (2023) [29] on the long-term historical predictability of the S&P500 and Dow Jones indices, Haase and Neuenkirch (2023) [30] on the predictability of market returns during different regimes, Li and Sun (2023) [50] on predicting market returns using aggregate credit risk, Qiu et al. ...
... We then "jump" to papers on or after 2020's just to illustrate the continued strong research interest on this topic. For example, Dai et al. (2020) [17] who consider technical indicators for forecasting stock returns, Dichtl (2020) [18] on forecasting excess returns of the gold market, Kyriazi and Thomakos (2020) [42] on the existence and interpretation of predictability in currencies, Liang et al. (2022) [48] on the use of dimensionality reduction methods for predicting market returns, Li et al. (2022) [49] on forecasting US stock market returns and Lv and Qi (2022) [52] that examine return predictability from a combination forecasting perspective. A stream of very recent papers includes Brennan and Taylor (2023) [8] on the interplay of expected returns and risk in the stock market, Casta (2023) [10] on the relationship between inflation, interest rates and the predictability of asset returns, Chen et al. (2023) [11] on use of economic policy uncertainty indices as predictors of market returns -with additional work on the topic by Huang et al. (2023) [33], Cotter et al. (2023) [16] on commodity futures return, Guerard et al. (2023) [29] on the long-term historical predictability of the S&P500 and Dow Jones indices, Haase and Neuenkirch (2023) [30] on the predictability of market returns during different regimes, Li and Sun (2023) [50] on predicting market returns using aggregate credit risk, Qiu et al. ...
We propose a new method for selecting the local estimation window for forecasting and trading financial returns. It is built around one particular definition of predictive complexity and we apply it in the simplest of predictors, the sample mean. We derive the exact conditions for the process of optimally selecting the local estimation window among a, theoretically found, grid of potential values of it. We use different loss functions, statistical and financial, which are considered individually and then pooled under two selection concepts, stochastic dominance and minimum description length, and find exact expressions as to how the associated complexities and their combinations can be derived and applied. Our results are based on a set of probabilistic assumptions for the time series under study and, based on those, we offer an inferential procedure for testing the presence of excess trading returns. Our empirical illustration on a set of diverse exchange trade funds (ETFs) across different asset classes suggests that the method works very well in practice and that it can generate both statistical and financial performance enhancements. A number of extensions to different predictors and different underlying assumptions is discussed. JEL Codes: C53, C58, G17. MSC2020 : 60G25, 68Q30, 90.
The disruption of supply chain due to Covid-19 and the war in Ukraine, render the prediction of agricultural output a determinant factor of economic life. We consider the predictability of agricultural output based on a set of explanatory variables, that include agricultural input, prices and consumer demand among others, for Greece and explore the usefulness of these variables compared to standard, univariate, forecasting methods. We evaluate the impact of using combined information in the form of principal components, and the use of averaging for producing accurate forecasts. Our results indicate that agricultural output is predictable and, moreover, we identify the factors that, for the case of Greece, lead to such predictability. Our outcomes can be used in a variety of ways, the least of which can be scenario analysis that might be very useful in real-world policy making.
Accepted by: Konstantinos Nikolopoulos
We explore the ability of market timing, model-based forecasts that include incremental information over and above momentum, to create long-short portfolios that consistently provide superior returns to traditional benchmarks, and frame our analysis in the idea of “prediction-led prescription” for forward-looking decision making. Our methodology consists of linear and non-linear forecasts with model selection, that predict the direction of future returns and apply then into the construction of portfolios. Our results strongly recommend that predictive-based portfolio allocation can produce excess returns and higher Sharpe ratios, above and beyond the individual return series that are utilized to create the forecasts and traditional portfolio benchmarks. The use of “predictive methods in portfolio construction reinforces the idea of prediction-led prescription” as portfolios are prescriptions for investment management and have an exclusively forward-looking nature.
With the adoption of flexible exchange rates in 1973, international capital markets have become more completely integrated. This chapter discusses portfolio selection of international equities and how international diversification lowers total risk of portfolios. Particular attention is paid to the diversification implications of Asian stocks, other emerging markets, and Latin American securities. The US equity selection model developed and estimated in chapter “Risk and Return of Equity and the Capital Asset Pricing Model” is used to rank global (ex-US) securities and produces statistically significant information coefficients and excess returns. An investor owns foreign stocks because their inclusion into portfolios produces higher Sharpe ratios than using only domestic securities. Global and (domestic) US securities may produce portfolios of higher returns for a given level of risk.
