检索结果-内蒙古大学图书馆

Deep learning on mixed frequency data

JOURNAL OF FORECASTING 2023年第8期42卷 2099-2120页

作者： Xu, Qifa Wang, Zezhou Jiang, Cuixia Liu, Yezheng Hefei Univ Technol Sch Management Hefei Peoples R China Minist Educ Key Lab Proc Optimizat & Intelligent Decis Making Hefei Peoples R China Hefei Univ Technol Sch Management Tunxi Rd 193 Hefei 230009 Anhui Peoples R China

In deep learning, it is common to encounter data observed at different frequencies. mixed data sampling (MIDAS) is an efficient technique for handling mixed frequency data, where a high frequency predictor is converted into a set of low frequency variables using the frequency alignment approach and parametric function constraints. This efficiently prevents the proliferation of parameters, ensuring the consistency of data frequency. We introduce the MIDAS technique into the deep learning architecture and develop a novel deep learning-MIDAS (DL-MIDAS) model, which enables to conduct deep learning on raw mixed frequency data directly. Its efficacy is then illustrated through extensive Monte Carlo simulations and a real-world application. The simulation experiments show that the DL-MIDAS model is able to explore nonlinear patterns in mixed frequency data and achieves more stable and accurate prediction results than several competing models, such as the artificial neural network for mixed frequency data (ANN-MIDAS), long short-term memory (LSTM), and MIDAS regressions. Additionally, the real-world application of predicting the inflation rate of China also confirms the strength of DL-MIDAS. The model can exploit high frequency information contained in financial market to produce timely and accurate prediction results on the inflation rate.

关键词： deep learning DL-MIDAS inflation rate forecasting mixed data sampling nonlinear pattern timely prediction

来源：评论

学校读者我要写书评

暂无评论

Asymmetric responses in the timing, and magnitude, of changes in Australian monthly petrol prices to daily oil price changes

引用

ENERGY ECONOMICS 2018年 69卷 89-100页

作者： Valadkhani, Abbas Smyth, Russell Swinburne Univ Technol Swinburne Business Sch Hawthorn Vic 3122 Australia Monash Univ Dept Econ Clayton Vic 3800 Australia

This article investigates whether uncompetitive pricing tactics are being employed in the retail petrol market in Australia through examining the effect of a change in daily oil prices on monthly petrol prices. To do so, we incorporate asymmetry into the coefficients of a normalised beta weighting function within an Asymmetric mixed data sampling (AMIDAS) framework. This enables us to examine both the timing, and the lagged marginal effects, of a change in retail petrol prices in response to a change in the oil price. We find evidence of asymmetries in both the timing, and magnitude, of retail petrol prices to a change in the oil price. Specifically, we find that while price falls are slowly and symmetrically passed onto consumers, price increases are more delayed, but higher in intensity over time. Depending on the capital city, when retailers eventually do increase petrol prices, in response to an oil price rise with delay, the price rise is between 2.1 and 3.4 times more than when retailers reduce petrol prices in response to a fall in the price of oil. This finding is consistent with retailers delaying substantial petrol price rises in order to mask the existence of uncompetitive practices. (C) 2017 Elsevier B.V. All rights reserved.

关键词： Petrol price Retail mixed data sampling Australia

来源：评论

学校读者我要写书评

暂无评论

Systemic risk-shifting in US commercial banking

引用

REVIEW OF QUANTITATIVE FINANCE AND ACCOUNTING 2020年第2期54卷 517-539页

作者： Kanas, Angelos Zervopoulos, Panagiotis D. Univ Piraeus Piraeus Greece Univ Sharjah Sharjah U Arab Emirates Hellen Parliament Sci Comm Parliamentary Budget Off Athens Greece

This paper puts forward the proposition that U.S. commercial banks use dividends as a mechanism to shift systemic risk to debt-holders and the deposit insurer. Using a mixed data sampling modeling approach, it is shown that monthly systemic risk factors are associated with a positive effect on future quarterly bank dividends indicating systemic risk-shifting. These factors include absorption (Kritzman et al. in MIT working paper, 2010), catfin (Allen et al. in Rev Financ Stud 25:3000-3036, 2012), covar (Adrian and Brunnermeier in CoVaR. NBER Working Paper 17454. National Bureau Economic Research, Cambridge, MA, 2011), delta_covar (Adrian and Brunnermeier 2011, mes (Acharya et al. in Rev Financ Stud 24:2166-2205, 2011b), real_vol (Giglio et al. in J Financ Econ 119:457-471, 2016), and size_con (Giglio et al. 2016). In addition, they can now-cast the upward trend in systemic risk-shifting in the 1990s and the downward trend from the early 2000s to 2007. The findings suggest that the rules governing bank dividends need be revised, support the imposition of a dividend tax to mitigate the negative externalities of dividends as a risk-shifting mechanism, and document a reduced effectiveness of Prompt Corrective Action in controlling risk-shifting.

关键词： Systemic risk Dividend payout mixed data sampling Risk-shifting

来源：评论

学校读者我要写书评

暂无评论

Do measures of systemic risk predict US corporate bond default rates?

引用

INTERNATIONAL REVIEW OF FINANCIAL ANALYSIS 2020年 71卷

作者： Kanas, Angelos Molyneux, Philip Univ Piraeus Piraeus Greece Univ Sharjah Sharjah U Arab Emirates Hellen Parliament Parliamentary Budget Off Athens Greece

Using univariate and multivariate mixed data sampling (MIDAS) and LASSO estimation methodologies, we explore whether the U.S. annual average corporate bond default rate can be predicted by 12 monthly systemic risk measures proposed in the literature. We find that nearly all of the systemic risk indicators have predictive power for the default rate. Granger causality tests based on multivariate mixed frequency VAR models further support this conclusion. On the basis of MIDAS models, we illustrate that five of these indicators are able to forecast out-of-sample the 2009 corporate default crisis. Using a LASSO multivariate model, it is further shown that the systemic risk indicators can forecast out-of-sample both the 2009 default rate and the default rates during the buildup before the crisis and in the aftermath of the crisis. Institution-specific and volatility systemic risk measures are the most relevant for modeling U.S. corporate bond default rates, with the Conditional VaR measure of Adrian and Brunnermeier (2016) exhibiting the best performance.

关键词： Systemic risk Corporate bond defaults mixed data sampling LASSO Predictive ability

来源：评论

学校读者我要写书评

暂无评论

Nowcasting Unemployment Rates with Smartphone GPS data 1

引用

1st International Workshop on Multiple-Aspect Analysis of Semantic Trajectories (MASTER)

作者： Moriwaki, Daisuke CyberAgent Inc Shibuya Ku 1-12-1 Dogenzaka Tokyo Japan

ISBN: (数字)9783030380816

ISBN: (纸本)9783030380816;9783030380809

Unemployment rate is one of the most important macroeconomic indicators. Central governments and market participants heavily rely on the index to assess the economies. However, official statistics of unemployment rate are released infrequently with substantial delay. Prediction of official statistics of labor market will be helpful for these authorities as well as private companies and even workers. In this paper, we combine massive location data coming from smartphones and mixed data sampling (MIDAS) techniques to predict current unemployment rate in Japan. We found GPS data is very useful to predict the status of labor markets.

关键词： GPS data MIDAS mixed data sampling Location data Unemployment rate Time series analysis Macroeconomic policy Nowcasting Forecasting

来源：评论

学校读者我要写书评

暂无评论

The effectiveness of fiscal policy in Brazil through the MIDAS Lens

引用

JOURNAL OF POLICY MODELING 2024年第1期46卷 113-128页

作者： Alves, Renan Santos Palma, Andreza A. Rabobank Macro Strategist Sao Paulo Brazil Fed Univ Sao Carlos UFSCar Dept Econ Sao Carlos Brazil

This paper aims to examine the effects of fiscal policy on economic growth in Brazil between 1999 and 2017. For this purpose, a novel methodology is applied, using a Vector Autoregressive with mixed Frequency (MIDAS-VAR) model, proposed by Ghysels (2016), which allows for the estimation of the spending multiplier by measuring the impact of high-frequency data at low-frequency and vice versa. The impact of various types of spending on the gross domestic product (GDP) is analyzed, including primary expenditure, personnel, social benefits, subsidies, investment, and costing, while the central government's primary revenue is used as the revenue variable. The expenditure and tax revenue are the high-frequency variables (monthly observations), whereas GDP is a low-frequency series (quarterly). The estimated fiscal multipliers for primary spending are less than one, suggesting no significant Keynesian effect on output, with particular attention given to the investment multiplier, which is estimated to be close to zero. Our results suggest that the frequency of data matters, and government expenditure has no significant impact on real GDP growth in Brazil. Therefore, the ability of Brazilian fiscal policy to influence economic growth may be limited. (c) 2023 The Society for Policy Modeling. Published by Elsevier Inc. All rights reserved.

关键词： Fiscal multiplier mixed frequency vector autoregression mixed data sampling Fiscal Policy

来源：评论

学校读者我要写书评

暂无评论

Asymmetric causality between Bitcoin and tech stocks in the US market using mixed frequency data

引用

JOURNAL OF ECONOMIC STUDIES 2024年第3期51卷 569-586页

作者： Valadkhani, Abbas Swinburne Univ Technol Dept Accounting Econ & Finance Hawthorn Australia

PurposeThis study is the first to investigate the causal relationship between Bitcoin and equity price returns by sectors. Previous studies have focused on aggregated indices such as S & P500, Nasdaq and Dow Jones, but this study uses mixed frequency and disaggregated data at the sectoral level. This allows the authors to examine the nature, direction and strength of causality between Bitcoin and equity prices in different sectors in more ***/methodology/approachThis paper utilizes an Unrestricted Asymmetric mixed data sampling (U-AMIDAS) model to investigate the effect of high-frequency Bitcoin returns on a low-frequency series equity returns. This study also examines causality running from equity to Bitcoin returns by sector. The sample period covers United States (US) data from 3 Jan 2011 to 14 April 2023 across nine sectors: materials, energy, financial, industrial, technology, consumer staples, utilities, health and consumer *** study found that there is no causality running from Bitcoin to equity returns in any sector except for the technology sector. In the tech sector, lagged Bitcoin returns Granger cause changes in future equity prices asymmetrically. This means that falling Bitcoin prices significantly influence the tech sector during market pullbacks, but the opposite cannot be said during market rallies. The findings are consistent with those of other studies that have established that during market pullbacks, individual asset prices have a tendency to decline together, whereas during market rallies, they have a tendency to rise independently. In contrast, this study finds evidence of causality running from all sectors of the equity market to *** implicationsThe findings have significant implications for investors and fund managers, emphasizing the need to consider the asymmetric causality between Bitcoin and the tech sector. Investors should avoid excessive exposure to both Bitcoin and tech stocks in their p

关键词： Exchange-traded funds Bitcoin Stock market Causality mixed data sampling E44 G11 G23 G32

来源：评论

学校读者我要写书评

暂无评论

Market states and the risk-return tradeoff

引用

QUARTERLY REVIEW OF ECONOMICS AND FINANCE 2017年 65卷 314-327页

作者： Wang, Zijun Khan, M. Moosa Univ Texas San Antonio Coll Business Dept Finance San Antonio TX 78249 USA Prairie View A&M Univ Prairie View TX USA

We re-examine the risk-return tradeoff in the U.S. equity market by allowing for time variation in the tradeoff and estimating conditional variance by the new mixed data sampling method. The main finding is that the risk-return tradeoff is strongly time-varying with the state of the market and the average of the time-varying tradeoff estimates is 1.43. The lagged market return is found to be the best indicator of market states. The empirical finding holds true for a battery of robustness checks during the post-Compustat sample period. The evidence from the international markets is similar to the U.S. one. (C) 2016 Board of Trustees of the University of Illinois. Published by Elsevier Inc. All rights reserved.

关键词： Risk-return tradeoff Time variation Market states mixed data sampling

来源：评论

学校读者我要写书评

暂无评论

mixed frequency composite indicators for measuring public sentiment in the EU

引用

Quality and Quantity 2023年第3期57卷 2357-2382页

作者： Mattera, Raffaele Misuraca, Michelangelo Spano, Maria Scepi, Germana Department of Social and Economic Sciences Sapienza University of Rome Rome Italy Department of Economics and Statistics University of Naples “Federico II” Naples Italy Department of Business Administration and Law University of Calabria Rende Italy

Monitoring the state of the economy in a short time is a crucial aspect for designing appropriate and timely policy responses in the presence of shocks and crises. Short-term confidence indicators can help policymakers in evaluating both the effect of policies and the economic activity condition. The indicator commonly used in the EU to evaluate the public opinion orientation is the Economic Sentiment Indicator (ESI). Nevertheless, the ESI shows some drawbacks, particularly in the adopted weighting scheme that is static and not country-specific. This paper proposes an approach to construct novel composite confidence indicators, focusing on both the weights and the information set to use. We evaluate these indicators by studying their response to the policies introduced to contain the COVID-19 pandemic in some selected EU countries. Furthermore, we carry out an experimental study where the proposed indicators are used to forecast economic activity. © 2022, The Author(s).

关键词： Composite indicator Confidence indicator Media sentiment mixed data sampling Public opinion

来源：评论

学校读者我要写书评

暂无评论

Forecasting VIX with time-varying risk aversion

引用

INTERNATIONAL REVIEW OF ECONOMICS & FINANCE 2023年 88卷 458-475页

作者： Wu, Xinyu He, Qizhi Xie, Haibin Anhui Univ Finance & Econ Sch Finance Bengbu 233030 Peoples R China Zhejiang Gongshang Univ Sch Stat & Math Hangzhou 310018 Peoples R China Univ Int Business & Econ Sch Banking & Finance Beijing 100029 Peoples R China

In this paper, we investigate the predictive value of time-varying risk aversion (RA) for VIX via the realized EGARCH-mixed-data sampling model incorporating RA (henceforth REGARCHMIDAS-RA). The REGARCH-MIDAS-RA model builds on the REGARCH model, which takes into account the high-frequency information by including the realized measure of volatility. Moreover, the model provides a convenient framework to model the long-run variance, which responds to changes in RA. We obtain the risk-neutralization of the REGARCH-MIDAS-RA model and derive the model-implied VIX formula. Our empirical results show that realized measure and RA possess predictive value for VIX. The REGARCH-MIDAS-RA model yields more accurate VIX forecasts compared to a range of competing models, including the GARCH, GJR-GARCH, nonlinear GARCH, EGARCH, REGARCH and REGARCH-MIDAS. In summary, our findings highlight the importance of incorporating the realized measure as well as RA in forecasting VIX.

关键词： VIX forecasting Time-varying risk aversion Realized EGARCH mixed data sampling Realized measure

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：