Consensus Forecasts Research Articles

Modularity and composite diversity affect the collective gathering of information online

Many modern interactions happen in a digital space, where automated recommendations and homophily can shape the composition of groups interacting together and the knowledge that groups are able to tap into when operating online. Digital interactions are also characterized by different scales, from small interest groups to large online communities. Here, we manipulate the composition of groups based on a large multi-trait profiling space (including demographic, professional, psychological and relational variables) to explore the causal link between group composition and performance as a function of group size. We asked volunteers to search news online under time pressure and measured individual and group performance in forecasting real geo-political events. Our manipulation affected the correlation of forecasts made by people after online searches. Group composition interacted with group size so that composite diversity benefited individual and group performance proportionally to group size. Aggregating opinions of modular crowds composed of small independent groups achieved better forecasts than aggregating a similar number of forecasts from non-modular ones. Finally, we show differences existing among groups in terms of disagreement, speed of convergence to consensus forecasts and within-group variability in performance. The present work sheds light on the mechanisms underlying effective online information gathering in digital environments.

Stock market reactions to R&D cuts used to manage earnings

Prior research shows that stock returns are positive when firms meet or beat analysts' consensus earnings forecasts but negative when they miss. Past studies also show that managers frequently cut research and development (R&D) expenses in order to meet the consensus forecast. This study shows that the stock market penalizes this behavior and exacts a discount to the market reward if beating the forecast requires cutting R&D. However, it is only a partial discount and firms are still better off managing R&D expenditures in the short run. This study also shows that the reductions in R&D are likely temporary, as firms tend to increase R&D spending in the subsequent periods. Investors appear to recognize these short-term cuts and treat them similarly to accruals.

Does litigation change managers\u2019 beliefs about the value of voluntarily disclosing bad news?

Research suggests that earnings-disclosure-related litigation causes managers to reduce subsequent disclosure, perhaps stemming from a belief that even their good faith disclosures will cause them trouble. This paper considers unexplored dimensions of disclosure and alternative channels of disclosure to provide additional evidence that speaks to how litigation shapes managers’ disclosure strategies. Consistent with Skinner (1994)’s classic legal liability hypothesis, we find that, while managers reduce and delay forecasts of positive earnings news following litigation, they increase the frequency and timeliness of their bad news forecasts. Moreover, many managers who were nonguiders prior to facing legal scrutiny begin guiding following litigation. Managers also maintain (if not increase) the information they provide via press releases and during conference calls following litigation. Supporting the notion that managers use disclosure to walk down expectations, additional analyses document an increase in the likelihood that lawsuit firms report earnings that beat consensus forecasts in the post-lawsuit period. Collectively, our evidence suggests that following litigation managers continue to view disclosure as a valuable tool that shapes their firms’ information environments and reduces expected legal costs. In so doing, it supports an important alternative viewpoint of how firms respond to litigation as well as the effectiveness of litigation as a disciplining mechanism.

RISKS OF ACTIVITY OF UKRAINIAN ENTERPRISES IN THE CONDITIONS OF A PANDEMIC AND QUARANTINE RESTRICTIONS

The risks in the conditions of the COVID-19 pandemic for the enterprises and economy of Ukraine, which include external and internal risks, havebeen studied. It is noted that so far there is no effective mechanism for compensation of losses from quarantine restrictions to businesses andemployees and, unfortunately, there is no clear vision of the implementation of anti-crisis strategy of economic development in such conditions. Thethematic issue of the Consensus Forecast was studied and the risk assessment analysis for enterprises and the economy of Ukraine was carried out onthe basis of a survey of experts and youth. It is noted that the curtailment of Ukrainian enterprises can have devastating systemic consequences for thenational economy, namely: declining GDP and deteriorating macroeconomic imbalances, destabilization of the banking system due to insolvency ofbusiness entities, rising unemployment, increasing receivables and payables. Perspective directions of activity of the enterprises in the conditions of apandemic are defined, the basic of which is digitalization of economy and business processes which growth rates are insignificant from 1% to 5%. Atthe same time, the positive trends of digitalization will prevail in the following areas: financial, insurance, retail trade, healthcare. It is noted that theimpact of the COVID-19 pandemic on the development of processes for automation of production and implementation of digital technologies inUkraine will be uneven. The main measures to support the activities of Ukrainian enterprises by the state are identified, including the implementationof the anti-crisis Action Plan to support entrepreneurship in quarantine and economic recovery, promoting the preservation and expansion of marketsfor enterprises in conditions of restrictions, implementation of measures to prevent corruption, simplification of tax administration, stimulation ofdigital transformations.

Основные особенности крупномасштабной атмосферной циркуляции в контексте анализа консенсусного прогноза температуры воздуха и осадков на лето 2020 года по Северной Евразии

The main results of the analysis of the Northern Hemisphere large-scale atmospheric circulation features are presented for the 2020 summer. Skill scores of the consensus forecast for the 2020 Northern Eurasia summer are discussed in the context of analyzing the large-scale atmospheric circulation. The prognostic potential of the trend component in forecasting seasonal anomalies of air temperature and precipitation is noted. Keywords: air temperature, precipitation, forecast skill, trends, large-scale atmospheric circulation, sea surface temperature, NEACOF, circulation indices, Arctic ice

Evaluation of heavy rainfall warnings of India National Weather Forecasting Service for monsoon season (2002–2018)

The major objective of any national weather forecasting services is to provide weather forecast and warnings and other meteorological related information to the public and government for the safety of life and property and economic activities. The heavy rainfall causes huge loss to the public in form of flood and landslide in varying severity mainly during monsoon season (June–September). Hence its accurate prediction is essential and the accuracy of prediction needs to be verified quantitatively to evaluate its strength and weakness. The National Weather Forecasting Centre (NWFC) of India Meteorological Department (IMD) issues heavy rainfall (HR) warnings for the safety of life and property of the public. In this study, verification of operational heavy rainfall (HR) warning issued by NWFC of IMD for 36 sub-divisions of India is carried out. The verification scores presented in the study are for 24 hrs (D1), 48 hrs (D2) and 72 hrs (D3) lead period average warning skills during 2014–2018 and year-wise trend of the HR warnings for the period 2002–2018. In general, it is observed that there are significant improvements in skill scores in recent years. The improvement in D3 is at higher rate as compared to D1 scores. The improvement in the recent years is mainly due to improvement in model resolution and data assimilation in the Numerical Prediction (NWP) Models runs by Ministry of Earth Sciences (MoES), Government of India and their interpretation and utilization by the forecasters for objective consensus forecast using an objective decision support system and synoptic value addition.

Application of Multimodel Superensemble Technique on the TIGGE Suite of Operational Models

One widely recognized portal which provides numerical weather prediction forecasts is “The Observing System Research and Predictability Experiment” (THORPEX) Interactive Grand Global Ensemble (TIGGE), an initiative of WMO project. This data portal provides forecasts from 1 to 16 days (2 weeks in advance) for many variables such as rainfall, winds, geopotential height, temperature, and relative humidity. These weather forecasting centers have delivered near-real-time (with a delay of 48 hours) ensemble prediction system data to three TIGGE data archives since October 2006. This study is based on six years (2008–2013) of daily rainfall data by utilizing output from six centers, namely the European Centre for Medium-Range Weather Forecasts, the National Centre for Environmental Prediction, the Center for Weather Forecast and Climatic Studies, the China Meteorological Agency, the Canadian Meteorological Centre, and the United Kingdom Meteorological Office, and make consensus forecasts of up to 10 days lead time by utilizing the multimodal multilinear regression technique. The prediction is made over the Indian subcontinent, including the Indian Ocean. TRMM3B42 daily rainfall is used as the benchmark to construct the multimodel superensemble (SE) rainfall forecasts. Based on statistical ability ratings, the SE offers a better near-real-time forecast than any single model. On the one hand, the model from the European Centre for Medium-Range Weather Forecasting and the UK Met Office does this more reliably over the Indian domain. In a case of Indian monsoon onset, 05 June 2014, SE carries the lowest RMSE of 8.5 mm and highest correlation of 0.49 among six member models. Overall, the performance of SE remains better than any individual member model from day 1 to day 10.

The Relationship Between Changes in Analyst Consensus Earnings Forecasts and Subsequent Share Returns on the Johannesburg Stock Exchange

Sell-side company analysts are most likely to be closest to a company’s financial health, will be privy to new information, and run models that process new information into meaningful expected changes to future earnings per share. Importantly, data vendors collate analyst’s individual Eps forecasts and publish consensus median Eps forecasts for the next three financial year-ends. A change in a median consensus forecast should be followed directly by a commensurate change in share price. We run a style analysis over 16 years, ranking quintile portfolios by the year-on-year change to Eps forecasts three months after portfolio formation. We find strong and consistent evidence that subsequent return is highly positively correlated with Eps upgrade. We vary both look-back and look-forward periods from one month to 12 months, confirming that three months forward, and 12 months backward, are optimal settings to maximise alpha. The effect is concentrated in the outer quintiles, with significant alpha generated when we move to decile portfolios.

Aggregating Artificially Intelligent Earnings Forecasts

This paper proposes a new artificial intelligence-based framework for aggregating the opinions of sell-side equity analysts. Our key conjecture is that analyst-level individual errors are more predictable and can be corrected more effectively than the aggregate consensus forecast. Artificial neural network models are able to learn the individual-level biases of sell-side equity analysts and can compensate for their biases more effectively. Using the bias-corrected consensus earnings forecasts and a neural network-based aggregation method, we find that the accuracy of the consensus forecast improves by 24% relative to a benchmark linear model. The neural network-based model can predict the correct sign for the earnings surprise 68% of the time. A portfolio strategy that uses the improved consensus forecasts earns a monthly abnormal return of 0.50% on a risk-adjusted basis.

Forecasting macroeconomic risks

We construct risks around consensus forecasts of real GDP growth, unemployment, and inflation. We find that risks are time-varying, asymmetric, and partly predictable. Tight financial conditions forecast downside growth risk, upside unemployment risk, and increased uncertainty around the inflation forecast. Growth vulnerability arises as the conditional mean and conditional variance of GDP growth are negatively correlated: downside risks are driven by lower mean and higher variance when financial conditions tighten. Similarly, employment vulnerability arises as the conditional mean and conditional variance of unemployment are positively correlated, with tighter financial conditions corresponding to higher forecasted unemployment and higher variance around the consensus forecast.

Methodical approaches to forecasting and early detection of economic cyclicity

The analysis and generalization of theoretical preconditions of formation of the advancing economic indicators which at different times in different countries of the world are used as indicators of change of a phase of an economic cycle is carried out. Based on the analysis of literature sources, it is established that the methodology of economic forecasting of cyclicality in the scientific literature includes two main areas, namely: a) direct forecasting of the dynamics of key macroeconomic indicators based on extrapolation, correlation and regression analysis, expert surveys (consensus forecasts) and surveys of economic entities; b) the use of leading economic indicators, which by their nature can be pro-cyclical, counter-cyclical and acyclical, and in relation to the phase of the economic cycle - ahead, late and synchronous.
 The most common leading indicators are considered, among which the Index of leading economic indicators of the USA, Composite leading index of the Organization for Economic Cooperation and Development (OECD), Indicator of business confidence by types of economic activity, consumer confidence, business climate and economic sentiment in accordance with the Special Data Dissemination Standard of the International Monetary Fund. Since 2006, the National Bank of Ukraine has been conducting a survey on the status and prospects of business activity, and similar quarterly indicators have been developed by the State Statistics Service of Ukraine and implemented since 2013. At the same time, a separate indicator is proposed for forecasting business activity in construction - the indicator of business confidence in construction (IDVB). However, the feasibility of using these indicators has yet to be proven in practice. To increase the accuracy of forecasting changes in the phases of economic cycles, it is proposed to develop separate indicators at the sectoral and regional levels. Theoretical generalization and substantiation of causes, consequences, methods of counteracting economic cyclicality should become the basis for further applied research. This approach is a theoretical basis for the development of tools for countercyclical management of enterprises, sectors of the economy and national economies in conditions of fluctuations in business activity.

Testing bias in professional forecasts

AbstractProfessional forecasters can rely on econometric models, on their personal expertise or on both. To accommodate for adjustments to model forecasts, this paper proposes to use two stage least squares (TSLS) (and not ordinary least squares [OLS]) for the familiar Mincer–Zarnowitz regression when examining bias in professional forecasts, where the instrumental variable is the consensus forecast. An illustration for 15 professional forecasters with their quotes for real gross domestic product (GDP) growth, inflation and unemployment for the United States documents the usefulness of this new estimation method. It also shows that TSLS suggests less bias than OLS does.

Financial Analysts’ Forecasts Have Improved Significantly in the Post-Reg FD Period

Reducing the amount of private information in corporate disclosures does not necessarily reduce the accuracy of analysts’ forecasts. This paper applies model-based earnings forecasts as a benchmark that is immune from disclosure of private information and evaluates the relative performance of analysts’ forecasts of earnings against the benchmark. It finds that the I/B/E/S consensus forecasts in general outperform the benchmark forecasts in the post-Reg FD period, while they underperform the benchmark in the pre-Reg FD period. It seems that Reg FD is a watershed. The difference-in-difference analysis confirms that the accuracy of analysts’ consensus forecasts of earnings has improved significantly following the passage of Reg FD.

Orange is the New Black: Changing Landscapes of Earnings Surprises and the Market Reaction

In this paper, we examine the changing time-series properties of the earnings-return relation and explore the implications of its changing landscape for the literature. We document strikingly opposite time-series patterns of earnings surprises and associated market reaction. Earnings surprises have been increasing over time, with the mean analyst forecast error rising from negative 8 cents in 1990 to positive 1.4 cents in 2019. However, average earnings announcement returns have declined from 0.30% in 1990 to -0.35% in 2019, turning negative in the past 15 years. As firms strive to meet or beat earnings expectations, strong past performance of a firm and strong performance of other firms raise the market’s earnings expectation above analysts’ consensus forecast, leading to investors’ disappointment upon earnings announcements. Our evidence has broad implications for appropriate earnings benchmarking, for empirical design when examining the earnings-return relation, for a disappearing earnings announcement premium, and for disappearing discontinuity in the earnings surprise distribution around zero.

Revenue recognition in achieving consensus on analysts’ forecasts for revenue, operating income and net earnings: the role of implementing IFRS 15. Evidence from Poland

In the study, we confirm that reporting entities increase discretionary revenue when their pre-managed operating income and net earnings are slightly lower than the consensus forecast by analysts for the fourth quarter. The novelty of this article is that introducing IFRS 15 significantly mitigates the increase of discretionary revenue when pre-managed operating income is slightly lower than the fourth quarter consensus on operating income. We do not find the implementation of IFRS 15 to have an impact on discretionary revenue if reporting entities are unable to meet the fourth quarter revenue and net earnings consensus by a small amount.

The Role of Stickiness, Extrapolation and Past Consensus Forecasts in Macroeconomic Expectations

We propose a simple model of expectation formation with three distinct deviations from fully rational expectations. In particular, forecasters’ expectations are sticky, extrapolate the most recent news about the current period, and depend on the lagged consensus forecast about the period being forecast. We find that all three biases are present in the Survey of Professional Forecasters as well as in the Survey of Professional Forecasters as well as in the Livingston Survey, and that their magnitudes depend on the forecasting horizon.Moreover, in an over-identified econometric specification, we find that the restriction on coefficients implied by our model is always close to being satisfied and in most cases not rejected. We also stress the point that using the past consensus forecast to form expectations is a reasonable thing to do if a forecaster is not able to come up with fully rational expectations all by herself.

Earnings Expectations and the Dispersion Anomaly

We show that investors do not fully unravel sell-side analyst forecast pessimism, as measures of prior consensus and individual analyst forecast pessimism predict both the sign of earnings surprises and the stock returns around earnings announcements. Firms with a high probability of forecast pessimism experience significantly higher announcement returns than those with a low probability. Importantly, our findings are driven by pessimistic short-term forecasts as opposed to optimistic longer-term forecasts. We further find that this mispricing is related to the difficulty for investors of identifying differences in expected forecast pessimism. We demonstrate how our findings provide a new explanation for the puzzling phenomenon that the dispersion in analysts’ earnings forecasts, which is also correlated with the probability of forecast pessimism, is negatively associated with future returns. Overall, we conclude that market prices do not fully reflect the conditional probability that a firm meets or beats earnings expectations as a result of analysts’ pessimistically biased short-term forecasts.

Agree to Disagree? Predictions of U.S. Nonfarm Payroll Changes between 2008 and 2020 and the Impact of the COVID19 Labor Shock

We analyze an unbalanced panel of monthly predictions of nonfarm payroll (NFP) changes between January 2008 and December 2020 sourced from Bloomberg. Unsurprisingly, we find that prediction quality varies across economists and we reject the hypothesis of equal predictive ability. In an error decomposition, we find evidence of significantly biased forecasts. Participation rate in the survey is affecting this bias. We find that survey participants under-predict job losses in times of market turmoil while also under-predicting the recovery thereafter, especially during the COVID19 labor shock. For prediction of NFP changes, autoregressive models are outperformed by a deep learning long short-term memory network. However, the consensus forecast yields better forecasts than model-based approaches and are further improved by combining the forecasts of the best performing economists. The COVID19 labor shock is shown to have adverse effects on the prediction performance of economists. However, not all economists are affected equally.

О результатах Девятнадцатой сессии Северо-Евразийского климатического форума (СЕАКОФ-19), 17–18 ноября 2020 г

The main results of the 19th session of the North Eurasian Climate Forum (NEACOF- 19) held online at the Hydrometeorological Center of Russia (Moscow) in November 2020 are presented. A brief overview of presentations of the leading Russian and international experts on the implementation of the WMO Climate Services Information System, the development of methods and technologies for climate prediction, vulnerability of transport infrastructure in the Arctic zone exposed to climatic hazards is given. Keywords: North Eurasian Climate Forum, North Eurasian Climate Center, Climate Services Information System, climate prediction technologies, climate model, climate warming, atmospheric circulation, consensus forecast

Consensus Forecast of Rainfall Using Hybrid Climate Learning Model

Rainfall event forecasting is prominently done using climate models (CMs) to produce multiple forecasts for the same rainfall event. The best forecast is complicated to find and hence has not yet been explored in the CMs. Recent advances in deep learning methods have provided an exceptional ability to investigate intricate weather patterns from big climate data. In this article, a hybrid climate learning model (HCLM) is proposed that utilizes both the CM and the deep learning models for improving the rainfall forecast. More specifically, a probabilistic multilayer perceptron (PMLP) network evaluates multiple forecasts from the CM-generated forecasts and selects the best one. The selected forecast is next passed onto a hybrid deep long short-term memory (HD-LSTM) network, which looks back and learns the relationship of the selected forecast with corresponding rainfall and temperature observations to produce the next-day rainfall forecast. The experimental results from various climate zones in Australia show that the HCLM outperforms existing state-of-the-art climate and deep learning models.