Exploring the representativeness of the M5 competition data

The main objective of the M5 competition, which focused on forecasting the hierarchical unit sales of Walmart, was to evaluate the accuracy and uncertainty of forecasting methods in the field to identify best practices and highlight their practical implications. However, can the findings of the M5 competition be generalized and exploited by retail firms to better support their decisions and operation? This depends on the extent to which M5 data is sufficiently similar to unit sales data of retailers operating in different regions selling different product types and considering different marketing strategies. To answer this question, we analyze the characteristics of the M5 time series and compare them with those of two grocery retailers, namely Corporación Favorita and a major Greek supermarket chain, using feature spaces. Our results suggest only minor discrepancies between the examined data sets, supporting the representativeness of the M5 data.     

Online adaptive lasso estimation in vector autoregressive models for high dimensional wind power forecasting

Wind power forecasts with lead times of up to a few hours are essential to the optimal and economical operation of power systems and markets. Vector autoregression (VAR) is a framework that has been shown to be well suited to predicting for several wind farms simultaneously by considering the spatio-temporal dependencies in their time series. Lasso penalisation yields sparse models and can avoid overfitting the large numbers of coefficients in higher dimensional settings. However, estimation in VAR models usually does not account for changes in the spatio-temporal wind power dynamics that are related to factors such as seasons or wind farm setup changes, for example. This paper tackles this problem by proposing a time-adaptive lasso estimator and an efficient coordinate descent algorithm for updating the VAR model parameters recursively online. The approach shows good abilities to track changes in the multivariate time series dynamics on simulated data. Furthermore, in two case studies it shows clearly better predictive performances than the non-adaptive lasso VAR and univariate autoregression.     

Are China’s regional agricultural productivities converging: How and why?

This article tests the hypotheses of convergence to a single level of total factor productivity (TFP), and a steady state of TFP growth rate in China’s agricultural sector. Based on multilateral TFP estimates we found that China’s agricultural sector has rebounded in recent years from a slower TFP growth in the 2005–2007 period. While convergence test results confirm a “catch-up” effect that provinces with lower TFP levels tend to grow faster than others, estimated rates of β convergence are conditional on how we capture the heterogeneity effect across regions. The rates of β convergence range from 0.016 to 0.039 under different model specifications. Estimates show that higher growth rates of educational attainment, R&D, and intermediate goods density (per unit of labor) can enhance TFP growth. Unfortunately, there is no evidence of an overall σ convergence, indicating that TFP levels are not converging except in the South region. It implies that to catch up with leading provinces, it would require extra efforts for those lagging behind by increasing their region-specific research investment, promoting rural educational attainment, and enhancing embodied technical change.     

The labour income share and the relative price of investment in the US: an empirical investigation

Several studies argue that the recent decline in the secular trend of the labour income share is mostly driven by capital-embodied technological progress which is typically identified with trend reductions in the relative price of investment. In this paper, I use data from the United States to assess the nature of the relationship between trends in the labour share and the relative price of investment. Results from co-integration tests reveal that the share and the relative price of investment are most likely not co-integrated. However, co-variation tests indicate that both time series share a common stochastic component, and additional tests of structural breaks point at the presence of a common change in the mean or trend of both series. These results suggest that capital-embodied or investment-specific technological progress may have played an important role in the decline of the secular trend of the labour share.     

Hedging and pricing early-exercise options with complex fourier series expansion

We introduce a new numerical method called the complex Fourier series (CFS) method proposed by Chan (2017) to price options with an early-exercise feature—American, Bermudan and discretely monitored barrier options—under exponential Lévy asset dynamics. This new method allows us to quickly and accurately compute the values of early-exercise options and their Greeks. We also provide an error analysis to demonstrate that, in many cases, we can achieve an exponential convergence rate in the pricing method as long as we choose the correct truncated computational interval. Our numerical analysis indicates that the CFS method is computationally more comparable or favourable than the methods currently available. Finally, the superiority of the CFS method is illustrated with real financial data by considering Standard & Poor’s depositary receipts (SPDR) exchange-traded fund (ETF) on the S&P 500® index options, which are American options traded from November 2017 to February 2018 and from 30 January 2019 to 21 June 2019.     

Predicting housing prices in China based on modified Holt's exponential smoothing incorporating whale optimization algorithm

The forecast of the real estate market is an important part of studying the Chinese economic market. Most existing methods have strict requirements on input variables and are complex in parameter estimation. To obtain better prediction results, a modified Holt's exponential smoothing (MHES) method was proposed to predict the housing price by using historical data. Unlike the traditional exponential smoothing models, MHES sets different weights on historical data and the smoothing parameters depend on the sample size. Meanwhile, the proposed MHES incorporates the whale optimization algorithm (WOA) to obtain the optimal parameters. Housing price data from Kunming, Changchun, Xuzhou and Handan were used to test the performance of the model. The housing prices results of four cities indicate that the proposed method has a smaller prediction error and shorter computation time than that of other traditional models. Therefore, WOA-MHES can be applied efficiently to housing price forecasting and can be a reliable tool for market investors and policy makers.     

Semiparametric time series models driven by latent factor

We introduce a class of semiparametric time series models (SemiParTS) driven by a latent factor process. The proposed SemiParTS class is flexible because, given the latent process, only the conditional mean and variance of the time series are specified. These are the primary features of SemiParTS: (i) no parametric form is assumed for the conditional distribution of the time series given the latent process; (ii) it is suitable for a wide range of data: non-negative, count, bounded, binary, and real-valued time series; (iii) it does not constrain the dispersion parameter to be known. The quasi-likelihood inference is employed in order to estimate the parameters in the mean function. Here, we derive explicit expressions for the marginal moments and for the autocorrelation function of the time series process so that a method of moments can be employed to estimate the dispersion parameter and also the parameters related to the latent process. Simulated results that aim to check the proposed estimation procedure are presented. Forecasting procedures are proposed and evaluated in simulated and real data. Analyses of the number of admissions in a hospital due to asthma and a total insolation time series illustrate the potential for practical situations that involve the proposed models.     

Destabilizing role of futures markets on North American hard red spring wheat spot prices

The objective of the paper is to determine if the futures prices of hard red spring wheat (HRSW) have stabilizing or destabilizing impact on spot HRSW price in North America. Several important results emerge from thorough empirical analysis. First, both Granger causality tests and directed acyclic graph algorithms (DAGs) point to two-way causality between futures and spot HRSW prices and thus endogeneity in both prices formation. To the contrary, both procedures suggest that ending stocks are exogenous to spot and futures HRSW prices. Both vector error correction model and impulse response functions point to a large and long-lasting impact of a shock to futures price on spot price level. Finally, variance decomposition analysis indicates that futures prices are responsible for the bulk of spot price volatility in both short and long run. Our result is consistent with those of theoretical models suggesting that when production (supply side) is the dominant disturbance, spot price is destabilized in both the short and the long run by futures prices. An important implication of this research is the need for alternative market mechanisms or alternative farm policy measures that would mitigate price risk and ensure sustainable farming of American HRSW farmers.     

Correlated daily time series and forecasting in the M4 competition

We participated in the M4 competition for time series forecasting and here describe our methods for forecasting daily time series. We used an ensemble of five statistical forecasting methods and a method that we refer to as the correlator. Our retrospective analysis using the ground truth values published by the M4 organisers after the competition demonstrates that the correlator was responsible for most of our gains over the naïve constant forecasting method. We identify data leakage as one reason for its success, due partly to test data selected from different time intervals, and partly to quality issues with the original time series. We suggest that future forecasting competitions should provide actual dates for the time series so that some of these leakages could be avoided by participants.     

An empirical investigation of water consumption forecasting methods

Many regions on earth face daily limitations in the quantity and quality of the water resources available. As a result, it is necessary to implement reliable methodologies for water consumption forecasting that will enable the better management and planning of water resources. This research analyses, for the first time, a large database containing data from 2 million water meters in 274 unique postal codes, in one of the most densely populated areas of Europe, which faces issues of droughts and overconsumption in the hot summer months. Using the R programming language, we built and tested three alternative forecasting methodologies, employing univariate forecasting techniques including a machine-learning algorithm, with very promising results.