Properties Of Forecasts Research Articles

Does industry co-location improve corporate information environment?

This study examines how the co-location of firms within the same industry enhances their information environments. We hypothesize that benefits arise from two effects: the industry peer effect, involving intra-industry information spillovers, and the geographic peer effect, related to shared local economic conditions with nearby non-industry peers. Our analysis evaluates how these effects, both separately and jointly, impact analysts’ forecast properties and stock price informativeness regarding future earnings. We find that the industry peer effect is negatively related to forecast dispersion and positively related to forecast accuracy, while the geographic peer effect shows no significant correlation. In addition, stock price informativeness about future earnings is positively associated with the industry peer effect but not with the geographic peer effect. We also find that greater proximity to industry or local peers enables analysts to incorporate more private (relative to public) information into their earnings forecasts, thereby enhancing the firm’s information environment. JEL CLASSIFICATION: M41, D80, G10

Book Value Forecast Quality: Financial Versus Nonfinancial Firms

ABSTRACTThis study investigates the quality and usefulness of analysts’ book value forecasts for financial firms relative to nonfinancial firms. We examine whether forecast properties (accuracy, bias, and dispersion) of book value forecasts and the market reaction to unexpected book value surprises differ between financial and nonfinancial firms. We find that book value forecasts are more accurate, less biased, and less dispersed for financial firms compared to nonfinancial firms. In addition, we show that these results are not simply a result of differences in earnings per share (EPS) forecast quality. We also provide evidence that book value forecasts are incrementally helpful to analysts in forecasting earnings for financial firms. In addition, we find evidence that the stock price drift following an earnings announcement is more pronounced when book value per share (BPS) surprises are considered in addition to EPS surprises. However, this finding only holds for financial firms. We find that this delayed reaction can form the basis for a profitable trading strategy.

Climate Disasters and Analysts’ Earnings Forecasts: Evidence from the United States

ABSTRACT We examine the relationship between climate disasters and analysts’ earnings forecasts in the United States. We find that climate disasters are associated with deteriorated analyst forecast properties proxied by forecast errors and forecast dispersion. We reason that the volatility of return on assets and of cash flows, and lower financial statement comparability, are three potential channels through which climate disasters influence analyst forecast properties. We also find that this relationship is more pronounced for firms in climate-vulnerable industries. Results from the market reaction tests further support our main findings by showing that the stock market responds less strongly to positive earnings surprises during periods of high climate disasters. Our results are robust to a battery of sensitivity tests, including a two-stage least squares approach and a difference-in-differences specification. Overall, the results shed light on the association between climate disasters and analysts’ earnings forecasts, which has significant implications for academics, investors, and standard setters.

Interpretable extreme wind speed prediction with concept bottleneck models

Concept-bottleneck models (CBMs) are a new paradigm to construct interpretable classifiers. The CBM architecture can be regarded as a neural network with a single hidden layer whose neurons are replaced by binary classifiers. Each of these binary classifiers implements a concept, that is, it attempts to answer a meaningful yes/no question: the presence or absence of a concept in the observation presented to the network. The output layer is usually implemented as a linear stage that combines the concepts into final decisions. This paper presents an application of the CBM paradigm to a problem of Extreme Wind Speed prediction, such that the forecasting properties of the system can be interpreted in terms of different concepts considered for this problem. A significant contribution of this paper is the proposal of an automated concept generation method, with controlled sparsity, in which the concepts are automatically extracted from the branches of decision trees fitted with informative features that capture the dynamics of the wind speed time series. The final forecasting system is able to predict extreme wind speed values in wind farms with good accuracy and interpretable results, as experiments over real wind speed data from a wind farm in Spain show.

Mixed-Input Residual Network for Air Temperature Forecasting by Low-Power Embedded Devices

Accuracy and model compactness are essential requirements for weather forecasting models designed for operation on low-power embedded devices. This study developed Mixed-Input Residual Network (MIRNet), a compact temperature-forecasting deep neural network model. MIRNet integrates stacked bidirectional long short-term memory layers using concatenated 1-dimensional and 2-dimensional convolutional layers to improve model accuracy. MIRNet was trained and tested on two datasets: one, IfeData, comprising historical weather data from Ile-Ife, Nigeria and the other a standard weather forecasting dataset called the Jena dataset. Training was carried out using 100 epochs of data partitioned in the standard 80:20 ratio, with an adaptable learning rate strategy. The model was tested for Nth-hour-ahead prediction for 1N24; where N are natural numbers, and performance quantified using metrics such as mean absolute percentage error (MAPE) and mean square error (MSE). The model was also implemented on a Raspberry Pi 4 device with a 1.8 GHz 64-bit quad-core ARM Cortex-A72 processor. The model achieved a MSE of 1.00 x 10-3 on the IfeData dataset, and 1.23x10-4 on the Jena dataset for 1-hour ahead forecasting. This is currently the best verifiable result achieved on the Jena dataset by any prediction model globally. For Nth hour ahead forecasting, MIRNet achieved an MSE generally below 2.0x10-3 for all values of N on the standard Jena dataset. The MSE of MIRNet for N-sequential 1-hour ahead and single Nth hour predictions using the Jena dataset reveal quadratic and linear relationships with N respectively. The model compares favourably with existing models for multi-hour predictions. The developed model is compact and has good forecasting properties.

Do AI models produce better weather forecasts than physics-based models? A quantitative evaluation case study of Storm Ciarán

There has been huge recent interest in the potential of making operational weather forecasts using machine learning techniques. As they become a part of the weather forecasting toolbox, there is a pressing need to understand how well current machine learning models can simulate high-impact weather events. We compare short to medium-range forecasts of Storm Ciarán, a European windstorm that caused sixteen deaths and extensive damage in Northern Europe, made by machine learning and numerical weather prediction models. The four machine learning models considered (FourCastNet, Pangu-Weather, GraphCast and FourCastNet-v2) produce forecasts that accurately capture the synoptic-scale structure of the cyclone including the position of the cloud head, shape of the warm sector and location of the warm conveyor belt jet, and the large-scale dynamical drivers important for the rapid storm development such as the position of the storm relative to the upper-level jet exit. However, their ability to resolve the more detailed structures important for issuing weather warnings is more mixed. All of the machine learning models underestimate the peak amplitude of winds associated with the storm, only some machine learning models resolve the warm core seclusion and none of the machine learning models capture the sharp bent-back warm frontal gradient. Our study shows there is a great deal about the performance and properties of machine learning weather forecasts that can be derived from case studies of high-impact weather events such as Storm Ciarán.

Non-Gaap Metrics and Properties of Analysts' Earnings Forecasts: The More, The Better?

Non-Gaap Metrics and Properties of Analysts' Earnings Forecasts: The More, The Better?

Product market competition, investor protection and analysts’ earnings forecasts

Purpose This study aims to examine the effects of product market competition (PMC) on analysts’ earnings forecast attributes, particularly forecast accuracy and dispersion. The authors also investigate whether investor protection moderates the relationship between PMC and forecast attributes. Design/methodology/approach The sample covers 49,578 firm-year observations from 38 countries. This study uses an ordinary least squares regression, a Heckman two-stage regression and an instrumental two-stage least squares regression. Findings This study finds that PMC is associated with higher forecast accuracy and lower dispersion. The results also show that investor protection enhances the effect of PMC on forecast accuracy and dispersion. These findings imply that countries with strong investor protection have a better information environment, as exhibited by the stronger relationship between PMC and analysts’ forecast properties. Practical implications The findings highlight the importance of strong governance mechanisms in both the country and industry environments. Policymakers, including government agencies and financial regulators, can leverage these insights to formulate regulations that promote competition, ensure investor protection and facilitate informed investment decisions. Originality/value This study advances our understanding of how PMC affects analysts’ earnings forecast attributes. In addition, it pioneers evidence of the moderating role of investor protection in the relationship between PMC and forecast attributes.

Do climate risk disclosures matter to financial analysts?

AbstractThis paper examines whether and when corporate disclosures about a firm's exposure to climate risks matter to financial analysts. More specifically, we investigate the association between climate risk disclosure (CRD) and two properties of financial analysts’ earnings forecasts (accuracy and dispersion). We predict that climate risk financial materiality at the industry level moderates this association. Using a sample of 2184 US nonfinancial firm‐year observations over the period 2010–2016, we show that CRD is associated with higher forecast precision and lower dispersion only when climate risks are perceived by investors as being financially material at the industry level. We also find that while corporate disclosures about transition risks are not associated with financial analyst forecast properties, 10‐K disclosures about climate‐related material physical risks reduce analyst forecast error and dispersion.

Analysts’ Earnings per Share Forecasts: The Effects of Forecast Uncertainty and Forecast Precision on Investor Judgements

This study uses controlled experiments to investigate the joint effects of forecast uncertainty and forecast precision on investor judgements. It finds that forecast precision moderates the effects of forecast uncertainty on investors’ forecast reliability judgements such that the effects of forecast uncertainty on investors’ judgements of forecast reliability are more negative when an analyst's point earnings per share (EPS) forecast is rounded than when it is precise. In addition, the relationship between forecast precision and investors’ judgements of forecast reliability is mediated by investors’ perceptions of forecast attributes. The evidence also suggests that while forecast uncertainty exerts a negative effect on investment judgements, forecast precision does not play a role in mitigating these negative effects. Using a supplementary within‐participants experiment, the study further finds that investors may not be consciously aware of how forecast precision influences their judgements of forecast reliability.

The NOAA Weather Prediction Center’s Use and Evaluation of Experimental Warn-on-Forecast System Guidance

This study examines use of experimental Warn-on-Forecast System (WoFS) guidance for short-term flash flood prediction at the NOAA Weather Prediction Center’s Meteorological Watch (Metwatch) desk. The WoFS guidance provides storm-scale ensemble forecasts for individual thunderstorms out to six hours and has previously shown great promise in its predictive skill for heavy rainfall events. Its operational utility was examined during 2019 and 2020 in a formal collaboration between Warn-on-Forecast scientists and Metwatch meteorologists. During that time, Metwatch meteorologists integrated real-time WoFS guidance into their Mesoscale Precipitation Discussion forecast processes and provided evaluations via a post-event survey. The survey queried impacts of WoFS guidance on their situational awareness, workload, and confidence, and Metwatch meteorologists also reported subjective assessments of model performance. Survey results highlighted the importance of viewing consistency in WoFS guidance across runs and agreement between WoFS guidance with conceptual models, other numerical weather prediction guidance, and observations. The use of WoFS tended to either maintain or slightly increase Metwatch meteorologists’ workload, while also increasing their confidence (notably for events perceived as better predicted). Of the different forecast attributes evaluated, Metwatch meteorologists reported convective mode as the attribute best predicted by WoFS. Use of WoFS guidance supported Mesoscale Precipitation Discussion decision making, including the placement and spatial extent of the product and the level of specificity provided about the related flash flood threat(s).

The Consequences of Abandoning the Quarterly Reporting Mandate in the Prime Market Segment

We examine the determinants and consequences of firms’ interim reporting choices in response to the gradual deregulation of quarterly reporting in the prime market segment of the Vienna Stock Exchange (VSE). Following the deregulation, only a few firms terminate quarterly reporting entirely, while most firms reduce the content of quarterly reports predominantly by omitting the notes disclosures. Our interview and empirical evidence suggest that firms primarily consider the information needs of their key stakeholders, such as institutional (foreign) investors and analysts, as well as their transparency preferences in their interim reporting decisions. Reporting and proprietary costs are less important to prime market firms. Firms that terminate quarterly reporting or substantially reduce the content of quarterly reports experience an economically significant reduction in liquidity, but not a deterioration in analyst forecast properties. Collectively, our results suggest that stakeholders demand quarterly reporting, but not necessarily full quarterly reports.

Probabilistic vs deterministic forecasts – interpreting skill statistics for the benefit of users

Owing to probabilistic uncertainties associated with seasonal forecasts, especially over areas such as southern Africa where forecast skill is limited, non-climatologists and users of such forecasts frequently prefer them to be presented or distributed in terms of the likelihood (expressed as a probability) of certain categories occurring or thresholds being exceeded. Probabilistic forecast verification is needed to verify such forecasts. Whilst the resulting verification statistics can provide clear insights into forecast attributes, they are often difficult to understand, which might hinder forecast uptake and use. This problem can be addressed by issuing forecasts with some understandable evidence of skill, with the purpose of reflecting how similar forecasts may have performed in the past. In this paper, we present a range of different probabilistic forecast verification scores, and determine if these statistics can be readily compared to more commonly known and understood ‘ordinary’ correlations between forecasts and their associated observations – assuming that ordinary correlations are more intuitively understood and informative to seasonal forecast users. Of the range of scores considered, the relative operating characteristics (ROC) was found to be the most intrinsically similar to correlation.

Performance commitments and the properties of analyst earnings forecasts: Evidence from Chinese reverse merger firms

We investigate whether the performance commitments in Chinese reverse merger (RM) transactions affect the properties of analyst earnings forecasts. All RM firms in China are required to make performance commitments for a limited number of years after being publicly listed. As performance commitment is an important piece of public information, it can influence analysts' understanding of firms and their efforts to forecast earnings. Using manually assembled information on RM transactions, we find that, in comparison to the control firms, RM firms exhibit an increase in analyst forecast error and dispersion after the end of performance commitment. This effect is more pronounced in firms with lower levels of information transparency. We also document that the public information contents of analyst forecasts decrease and forecast revisions increase in the post-commitment period, while the private information content of analyst forecasts and the number of their firm visits remain unchanged. Overall, our findings suggest that analysts rely greatly on public information; they have important implications for academics and policymakers in understanding how performance commitments in RM transactions affect the market information environment.

Classification shifting using income-decreasing special items: measurement and valuation issues

Research suggests that the standard model used to detect opportunistic shifting of core expenses to special items is potentially biased. Such bias has been attributed to the use of accruals, including special item related accruals, as a control for the impact of performance on core earnings in this model. This paper provides an improved classification shifting model which both tests for such accruals-related bias and controls for other sources of error in the measurement of shifting. The paper also modifies conventional market rationality tests in accounting research to examine new dimensions of rationality in relation to measurement and valuation of shifting. The main empirical findings are as follows. First, the improved classification shifting model provides strong evidence of shifting and rejects the hypothesis that inclusion of accruals in the model causes bias. Second, estimates of shifted core expenses generated by the improved model exhibit forecasting properties of shifted earnings. Third, rationality test results are broadly consistent with rationality in relation to shifted core expenses but indicate possible partial (ir)rationality in relation to adjusted special items (i.e., special items excluding shifted core expenses). Further analysis of the latter findings, however, suggests they are more likely related to risk than irrationality. Overall, the paper contributes to improved measurement of shifting and highlights the importance of considering rational expectations when examining stock returns associated with shifting.

Have housing value indicators changed during COVID? Housing value prediction based on unemployment, construction spending, and housing consumer price index

PurposeHousing market research involves observing the relationships between housing value and its indicators. However, recent literature indicates that the disruption of the COVID-19 pandemic could have an impact on the forecasting properties of some of the housing indicators. This paper aims to observe the relationships between the home value index and three potential indicators to verify their forecasting properties pre- and post-COVID-19 and provide general recommendations for time series research post-pandemic.Design/methodology/approachThis study features three vector autoregression (VAR) models constructed using the home value index of the USA, together with three indicators that are of interest according to recent literature: the national unemployment rate, private residential construction spending (PRCS) and the housing consumer price index (HCPI).FindingsUnemployment, one of the prevalent indicators for housing values, was compromised as a result of the COVID-19 pandemic, and a new indicator for housing value in the USA, PRCS, whose relationship with housing value is robust even during the COVID-19 pandemic and HCPI is a more significant indicator for housing value than the prevalently cited All-Item consumer price index (CPI).Originality/valueThe study adds residential construction spending into the pool of housing indicators, proves that the finding of region-specific study indicating the unbounding of housing prices from unemployment is applicable to the aggregate housing market in the USA, and improves upon such widely accepted belief that overall inflation is a key indicator for housing prices and proves that the CPI for housing is a vastly more significant indicator.

What can be learned from the historical trend of crude oil prices? An ensemble approach for crude oil price forecasting

Crude oil price series are nonlinear and highly volatile, making it difficult to obtain satisfactory performance for traditional statistical-based forecasting methods. To improve forecasting accuracy, this study proposes a novel learning paradigm by integrating the trajectory similarity method with machine learning models based on the decomposition-ensemble framework. In the proposed learning paradigm, raw data of international crude oil prices are first decomposed using variational mode decomposition (VMD), after which, using sample entropy (SE), the resulting essential modal functions are divided into high and low frequencies. The process aims to reorganize the data by using the forecasting properties of different models. Finally, to obtain the final forecasting results, two models, i.e., the trajectory similarity method (TS) and long short term memory neural network (LSTM), are applied to predict and sum up the low and high-frequency subseries, respectively. As sample data for validation, this study selected the international crude oil price series of West Texas Intermediate (WTI) and Brent. Experimental results showed that the proposed VMD-SE-TS/LSTM learning paradigm significantly outperforms all other benchmark models, including the single models without decomposition and the hybrid models with decomposition. The proposed approach performs best in different evaluation metrics and statistical tests under different horizons, indicating that the proposed VMD-SE-TS/LSTM learning paradigm is effective and robust in crude oil price forecasting.

Public Information Priorities across Weather Hazards and Time Scales

Abstract As the abundance of weather forecast guidance continues to grow, communicators will have to prioritize what types of information to pass on to decision-makers. This work aims to evaluate how members of the public prioritize weather forecast attributes (including information about location, timing, chance, severity, impacts, and protective actions) on average and across event timelines in the severe, tropical, and winter weather domains. Data from three demographically representative surveys of U.S. adults indicate that members of the public generally prioritize information about event location, timing, and severity when evaluating the importance of forecast attributes. This pattern is largely consistent across hazard domains but varies across event timelines. In early stages of a forecast (such as the outlook time scale), people generally prioritize information about chance and location. In middle stages (watch time scale), event timing and severity become more important. In late stages (warning time scale), information about protective actions is a higher priority, especially for people with less exposure to a hazard.

Mandatory CSR disclosure and analyst forecast properties: Evidence from a quasi-natural experiment in China

Based on a quasi-natural experiment that mandates a subset of listed firms to issue standalone corporate social responsibility (CSR) reports, we examine whether mandatory CSR disclosure improves analysts’ information environment. We focus on two properties of analysts’ earnings forecasts: forecast error and forecast dispersion. We find that the mandatory issuance of standalone CSR reports is related to less forecast error and less dispersed forecasts, and the effect varies with the firm-level information environment and province-level marketization. Additional tests show that the improvement in forecast properties is mainly driven by CSR reports that i)are of high quality and ii)contain more long-term-oriented information than other CSR reports. Our findings provide evidence that mandatory CSR disclosure plays an important informational role for financial analysts.

ИСПОЛЬЗОВАНИЕ МОДЕЛИ МНОЖЕСТВЕННОЙ ЛИНЕЙНОЙ РЕГРЕССИИ ДЛЯ КРАТКОСРОЧНОГО ПРОГНОЗИРОВАНИЯ ЦЕН НА КАЛИЙНУЮ ПРОДУКЦИЮ

The modern economy instability due to global and regional conflicts and contradictions leads to significant fluctuations in commodity markets. The demand and prices unpredictability leads to the raw materials in-dustries risks increase and limits their successful functioning and development. The purpose of the study. In this paper, we made an attempt of a multiple linear regression model application to obtain a forecast price of potash products acceptable quality. Despite the wide application and simplicity of construction and interpretation, the forecasting properties of such models are usually unsatisfactory. Nevertheless, an adequate selection of factors and the sample size used to estimate unknown parameters of the model, al-lows to achieve an acceptable forecasting quality. Materials and methods. In the issue the evaluation of the multiple linear regression model unknown parameters is performed on the basis of United States of America Geological Survey data. The sample contains information about the American potash market in the period from 1900 to 2020. The quality of forecasting is tested using the post-forecast method for years 2019 and 2020. The model built using the entire dataset gives an unsatisfactory relative prediction error. To reduce its value, a search of the sample volumes of data used for model parameters estimation was per-formed, and the one with the minimal forecast error was selected as optimal. The calculations were per-formed using MS Excel and Jupyter Notebook 6.1.4 environment for Python 3.8.5. Results. The conducted research allowed us to determine that in order to minimize the US domestic market potash products price relative error, it is necessary to use from 9 to 13 years time intervals for modeling. By doing that it is possi-ble to improve the 118-year time interval forecast for 2019 year by 23.9%, and a similar forecast for 2020 by 83.70%. Conclusion. Based on the results of the work done, it can be stated that the multiple linear re-gression model can be successfully used for short-term forecasting of the potash products price, and by an adequate modeling interval length choice it is possible to achieve acceptable forecasting quality.