Using diffusion models to forecast market size in emerging markets with applications to the Chinese car market

Marketing managers have to forecast the market size and this forecast guides strategic decisions whether to continue exporting, open new factories or expand existing production operations. Forecasting sales and the market size is a challenging task; even more so in emerging markets where data is limited and the market demand is changeable. This research proposes a novel approach that applies diffusion models using car ownership data to forecast car sales. Car ownership data may be easier to access than sales data in emerging markets but marketing managers are more interested in the sales forecast. Researchers propose using diffusion models to forecast the adoption of new products or products which are new to consumers in a market. This research demonstrates that marketing managers can use diffusion models to predict car sales in China where cars are new products to most consumers in this market. Since the majority of car buyers in China are first time buyers, car manufacturers and retailers must also forecast when the market composition will change. This effectively means predicting when first time car buying will start to slow down and repeat/replacement purchase or second hand car purchase will become more important. To forecast both sales and market composition change, marketing managers must choose a robust model. Managers want insights from models that have been tested robustly especially in less stable market conditions. In this context, this study illustrates the value of using a rolling forecast instead of a fixed horizon approach when comparing and choosing which model to use to forecast both sales and market composition change for the Chinese car market.     

Return prediction and stock selection from unidentified historical data

The experimental approach was applied to test the value of historical return series in technical prediction. Return sequences were randomly drawn cross-sectionally and over time from S&P500 records and participants were asked to predict the 13th realization from 12 preceding returns. The hypothesis that predictions (nominal or real) are randomly assigned to historical sequences is rejected in permutation tests, and the best-stock portfolios by experimental predictions significantly outperform the worst-stock portfolios in joint examination of mean return and volatility. The participants dynamically adjust their predictions to the observed series and switch from momentum riding to contrarian extrapolation when recent trends get extreme. The implicit tuning of predictions to specific series captures variabilities that could not be inferred by schematic statistical forecasting.     

Simple averaging of direct and recursive forecasts via partial pooling using machine learning

This article introduces the winning method at the M5 Accuracy competition. The presented method takes a simple manner of averaging the results of multiple base forecasting models that have been constructed via partial pooling of multi-level data. All base forecasting models of adopting direct or recursive multi-step forecasting methods are trained by the machine learning technique, LightGBM, from three different levels of data pools. At the competition, the simple averaging of the multiple direct and recursive forecasting models, called DRFAM, obtained the complementary effects between direct and recursive multi-step forecasting of the multi-level product sales to improve the accuracy and the robustness.     

基于小波神经网络的鱼类价格预测研究

鱼类市场价格是影响渔民收益的重要因素,因此如何能够准确分析和预测鱼类的价格是一个十分重要的问题.论文在考虑了BP网络原形的一些缺点和不足,尝试在传统BP算法的基础上将网络神经元中的激励函数换成小波子函数,组建成的小波神经网络.通过对鲈鱼价格的预测,验证了小波神经网络的可行性,然后基于阿里巴巴的鲈鱼价格的历史数据,验证了该方法的合理性.最后开展对新疆乌伦古湖水产综合基地三类主要鱼类价格预测的实证研究,研究结果表明河鲈和梭鲈的价格会有小幅的波动,但高白鲑的价格会有大幅的提升.因此,小波神经网络作为一种传统神经网络的改进,可用于预测鱼类价格的短期预测,为水产基地的决策提供依据.     

基于相空间重构及自适应支持向量机的短期风速预测

风速具有较强的随机性和间歇性,导致大规模风电接入电网会严重影响电力系统的安全稳定运行以及电能质量。较为准确的风速预测可以降低风能对电网的不利影响,为电网运行调度提供可靠的依据。在对风速进行混沌属性分析及相空间重构的基础上,采用自适应支持向量机进行短期风速预测,结果表明该方法的预测精度高于BP、RBF等预测模型。     

Ability of crime,demographic and business data to forecast areas of increased violence

ABSTRACT

Identifying geographic areas and time periods of increased violence is of considerable importance in prevention planning. This study compared the performance of multiple data sources to prospectively forecast areas of increased interpersonal violence. We used 2011–2014 data from a large metropolitan county on interpersonal violence (homicide, assault, rape and robbery) and forecasted violence at the level of census block-groups and over a one-month moving time window. Inputs to a Random Forest model included historical crime records from the police department, demographic data from the US Census Bureau, and administrative data on licensed businesses. Among 279 block groups, a model utilizing all data sources was found to prospectively improve the identification of the top 5% most violent block-group months (positive predictive value = 52.1%; negative predictive value = 97.5%; sensitivity = 43.4%; specificity = 98.2%). Predictive modelling with simple inputs can help communities more efficiently focus violence prevention resources geographically.     

Nonlinear deterministic forecasting of daily dollar exchange rates

We perform out-of-sample predictions on several dollar exchange rate returns by using time-delay embedding techniques and a local linear predictor. We compared our predictions with those by a mean value predictor. Some of our predictions of the exchange rate returns outperform the predictions of the same series by the mean value predictor. However, these improvements were not statistically significant. Another interesting result in this paper which was obtained by using a recently developed technique of nonlinear dynamics is that all exchange rate return series we tested have a very high embedding dimension. Additionally, evidence indicates that these series are likely generated by high dimensional systems with measurement noise or by high dimensional nonlinear stochastic systems, that is, nonlinear deterministic systems with dynamic noise.     

A NEW METHOD FOR THE ESTIMATION OF CAPACITY UTILIZATION: THEORY AND EMPIRICAL EVIDENCE FROM 14 EU COUNTRIES

In this paper, we develop a new and at the same time simple method of obtaining a measure of the rate of capacity utilization (CU) which makes use of the structural vector autoregression (SVAR) system of equation estimating technique with long‐run restrictions. The measure of CU that we derive for each of 14 EU countries replicates to a great extent the European Commission's Directorate General for Economic and Financial Affairs (ECFIN) measure. On closer examination we find that the in‐sample explanatory content with respect to the inflation rate of the SVAR measure exceeds more often than not that of the ECFIN's measure; however, the out‐of‐sample forecasting performance of the two models is approximately equivalent.     

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.     

Forecasting China's Monthly Inbound Travel Demand

Summary

China is currently expecting a growth in inbound travel demand as the result of China's “open door policy,” participation in World Trade Organization (WTO), success in hosting the Olympics in Beijing in the year 2008 and political stability. This paper focused on two issues: (1) forecasting China's monthly inbound travel demand and (2) seasonally and seasonal ARIMA model selection for monthly tourism time-series. In this paper following seasonal ARIMA models were considered: the seasonal ARIMA model with first differences and 11 seasonal dummy variables, the conventional seasonal ARIMA model with first and the fourth differences. In order to select the best forecasting model, finally both seasonal ARIMA models were compared with the AR model with fourth differences, the basic structural model (BSM) and the naive “No Change” model. In the one-step ahead forecasting comparison, the conventional seasonal ARIMA model with first and the fourth differences becomes the best forecasting model for both inbound foreign visitor demand and total visitor demand. This may be due to the nature of monthly seasonal variations in visitor arrivals, which is less marked. Our forecasts indicate that China foreign visitor arrivals and total visitor arrivals are expected to grow by 14% and 27% respectively from 2002 to 2005.