Unobserved Components Model Research Articles

The 2023/24 El Niño and the Feasibility of Long-Lead ENSO Forecasting

Abstract Through its atmospheric teleconnections, El Niño–Southern Oscillation (ENSO) shifts and disrupts weather and climate patterns far beyond the equatorial Pacific where it occurs often resulting in catastrophic consequences in many countries of the world. It is also the largest source of seasonal and interannual climate predictability. Despite its huge importance, ENSO forecasting is still not performed operationally at longer leads than about 6 months ahead. At the same time, there is mounting scientific evidence that forecasts are possible even more than a year in advance. Early warning of ENSO events could substantially mitigate some of the most damaging impacts, such as floods, droughts, and harvest failure and help avoid famine, migration, and disease outbreaks. Here, we present forecasts from a statistical ENSO model of the next El Niño predicted to occur in the winter of 2023/24, at lead times between 11 and 17 months ahead of an expected peak in December 2023. We use a statistical unobserved dynamic components model (EDCM) based on subsurface ocean temperatures as well as sea surface temperatures and zonal wind stress. EDCM has been previously validated through hindcasts of the major El Niños since 1970 and through real-time forecasts of the 2015/16 and 2018/19 El Niños. Our statistical framework and results indicate that there is potential for doubling the operational predictive lead time of ENSO to at least 12 months, with additional promise for even earlier anticipation of 19 months. Such longer-lead forecasts could be of high value, because decision-making and management in a number of key socioeconomic sectors could be greatly improved.

Impact of Political Instability on Tourism in West Bengal.

This study aimed to investigate the effects of political instability on tourism-related business performance, incoming tourist flows in West Bengal, and overall tourism-based earnings. The nonlinear effects of political risk and economic growth on tourism receipts are examined in this research. Nevertheless, the secondary data that is now accessible with the unobserved component model technique detection only provides a broad picture of the issue. Experts such as managers of lodging establishments, staff members of local tourism development offices, and tourism researchers were interviewed as a result. Expert interviews exposed West Bengal's tourism industry's prospects, challenges, and potential developments in the face of the region's five years of political unrest. The findings confirm earlier research showing that, in the near run, political unrest lowers tourism. However, the long-term viewpoint has yielded various results from the expert interviews with main province cities. Political instability benefits cities with developed tourism businesses that are located outside of crisis areas. Places that relied on visitors from the countries or continents like Europe, Germany, China, USA are mostly affected. The political unrest in eastern West Bengal did not significantly affect cities with low levels of tourism development.

Trends and cycles during the COVID-19 pandemic period

We perform several trend-cycle decompositions through the lens of two unobserved components models, herein estimated for Portugal and the euro area. Our procedure copes with the COVID-19’s consequences by explicitly considering potentially larger second moments during that period. This is achieved through a set of pandemic-specific shocks affecting only the 2020–21 period and embedded into estimation through a piecewise linear Kalman filter. Our methodology generates negligible historical revisions in key smoothed variables when the sample period is expanded until 2021:4, since pandemic shocks absorb a great deal of data volatility with minimal impacts on filtered data revisions or estimated parameters. Furthermore, non-pandemic shock volatility remains largely unaffected by the pandemic period. Innovations affecting the cycle in our preferred model are the key propellers of GDP developments during the COVID-19 pandemic period.

Dynamic hysteresis effects

We study how the output gap affects potential output over time—i.e., the dynamic hysteresis effect. To do so, we introduce novel unobserved components (UC) models that consider hysteresis as a sequence of lagged effects, thus separating the long-run recession-induced adverse effects from other trend-cycle interactions. The proposed models nest several existing UC models in the literature and accommodate two key characteristics of output dynamics: non-neutrality in the long-run and time-to-build effects. Using Bayesian estimation methods, we find robust evidence supporting the presence of hysteresis effects after the 1970s, with the negative long-run effect of the Global Financial Crisis and the COVID-19 recessions robustly identified. Via Bayesian model averaging, we provide precise and intuitive output gap estimates that highlight the relationship between business cycle fluctuations and the decline in economic growth. Our findings indicate that output trend-cycle decompositions that do not consider hysteresis effects can alter stabilization policy trade-offs.

Macro-financial imbalances and cyclical systemic risk dynamics: understanding the factors driving the financial cycle in the presence of non-linearities

Abstract This paper develops a multivariate filter based on an unobserved component model to estimate the financial cycle. Our model features: (1) a dynamic relationship between the financial cycle and key variables; (2) time-varying shock volatility for trend and cycle components. We demonstrate that our approach not only exhibits superior early warning properties for banking crises but also outperforms commonly used indicators in terms of data fit for decomposition exercises, as evidenced by the higher marginal likelihood. We document three important properties of the financial cycle. First, the sensitivity of the financial cycle to changes in real estate valuations increased during the post-90s period. Second, the sensitivity of the cycle to changes in financial conditions displays volatility and country specificities. Finally, our reduced form estimates suggest that the banking crisis of 1988 was preceded by positive contributions from the risk appetite shock, while the primary source of vulnerabilities emanated from the housing market in the run-up to the Global Financial Crisis.

Estimation of the TFP Gap for the Largest Five EMU Countries

AbstractIn this paper we extend the Bayesian unobserved components model of the EU Commission to estimate the cyclical component of total factor productivity (TFP gap) with a factor structure that includes a wide array of business cycle indicators. We demonstrate that this model extension considerably stabilizes the estimate of the TFP gap for the largest five EMU countries. Specifically, consider the usual autumn forecast of the EU Commission in October of a year: Using the model extension, we reduce the year-to-year revisions in the TFP gap estimates of the current year and the two previous years by up to 30 percent. Revision reductions for the two years ahead also considered by the EU Commission are quantitatively smaller (up to 10 percent) but still relevant. The results do vary across countries but are qualitatively robust with respect to different indicator sets, model specifications, and vintages considered.

Multivariate Trend‐Cycle‐Seasonal Decompositions with Correlated Innovations*

AbstractMultivariate analysis can help to focus on important phenomena, including trend and cyclical movements, but any economic information in seasonality is typically ignored. The present paper aims to more fully exploit time series information through a multivariate unobserved component model for quarterly data that exhibits seasonality together with cross‐variable component correlations. We show that economic restrictions, including common trends, common cycles and common seasonals can aid identification. The approach is illustrated using Italian GDP and consumption data.

Are Long Waves 50 Years? Reexamining Economic and Financial Long Wave Periodicities in Kondratieff and Schumpeter

ABSTRACT In 1925, the Russian economist, Nikolai D. Kondratieff first presented his identification and analysis of a ‘long wave’ cycle of approximately 50 years in twenty-five economic and financial time series across the major capitalist economies. The statistical evidence for their existence was based on nine-year centered moving averages of residuals from econometric time-trend models for eight English and five French time series that spanned the 18th, 19th and early 20th centuries. Schumpeter supported and promoted Kondratieff's estimate of a long wave periodicity which was consistent with his trigonometric models published in Business Cycles (1939). While Schumpeter never attempted to measure the statistical association between his theoretical values and historical observations, he identified long waves in the numerous graphs and charts in Business Cycles. Kondratieff's original data is used to estimate long wave periodicities by replicating his published models and smoothed residuals to verify specific years of turning points. ‘Unobserved component models’ are also used to extract long wave periodicities from Kondratieff’s data as well as from new long-term time series recently published by the Bank of England. The new estimates confirm an endogenously-propagated long cycle of about fifty ears.

Productivity Shock and Labour Input: Evidence from Correlated Unobserved Component Model

This note assesses the relationship between labour productivity and employment in the framework of the unobserved component model as well as the vector autoregressive model. In the case of the organized/formal manufacturing sector in India, the transitory increase in productivity is seen to reduce the man-days in the short run, though it is not statistically significant. Permanent shock to productivity decreases labour inputs permanently, and vice versa and their association is also statistically significant. Incentivizing the firms to expand their activities, diversify production and process secondary products may help restore employment in the face of a permanent productivity shock. JEL Classification: C11, C32, E32

The race between education and technology in Chile and its impact on the skill premium

While studies using pre-2000 data for Chile implied that technological change drives the skill premium evolution, post-2000 data suggests that the supply of skilled labour expansion has reduced this premium. In contrast, literature shows a consistent link between technology and the growing demand for skilled labour, despite educational improvements, leading to skill premium increases. We examine these interactions under the race between education and technology model (RBET) for Chile between 1980 and 2018 considering the inconsistent findings (e.g., negative elasticities) reported by past studies. We also find inconsistencies using cointegration techniques. Alternatively, we apply an Unobserved Component Model using a Bayesian estimation that enforces the required economic and theoretical conditions. We find that, before the 2000s, the technological change dominated. However, in the 2000s, the demand was surpassed by educational supply increases, reducing the premium. The estimated elasticity of substitution between skilled and unskilled labour is 6.5, implying that both are more substitutable than commonly thought.

Emergency department visits and boarding for pediatric patients with suicidality before and during the COVID-19 pandemic.

To quantify the increase in pediatric patients presenting to the emergency department with suicidality before and during the COVID-19 pandemic, and the subsequent impact on emergency department length of stay and boarding. This retrospective cohort study from June 1, 2016, to October 31, 2022, identified patients ages 6 to 21 presenting to the emergency department at a pediatric academic medical center with suicidality using ICD-10 codes. Number of emergency department encounters for suicidality, demographic characteristics of patients with suicidality, and emergency department length of stay were compared before and during the COVID-19 pandemic. Unobserved components models were used to describe monthly counts of emergency department encounters for suicidality. There were 179,736 patient encounters to the emergency department during the study period, 6,215 (3.5%) for suicidality. There were, on average, more encounters for suicidality each month during the COVID-19 pandemic than before the COVID-19 pandemic. A time series unobserved components model demonstrated a temporary drop of 32.7 encounters for suicidality in April and May of 2020 (p<0.001), followed by a sustained increase of 31.2 encounters starting in July 2020 (p = 0.003). The average length of stay for patients that boarded in the emergency department with a diagnosis of suicidality was 37.4 hours longer during the COVID-19 pandemic compared to before the COVID-19 pandemic (p<0.001). The number of encounters for suicidality among pediatric patients and the emergency department length of stay for psychiatry boarders has increased during the COVID-19 pandemic. There is a need for acute care mental health services and solutions to emergency department capacity issues.

Fitting Time Series Models to Fisheries Data to Ascertain Age

The ability of government agencies to assign accurate ages of fish is important to fisheries management. Accurate ageing allows for most reliable age-based models to be used to support sustainability and maximize economic benefit. Assigning age relies on validating putative annual marks by evaluating accretional material laid down in patterns in fish ear bones, typically by marginal increment analysis. These patterns often take the shape of a sawtooth wave with an abrupt drop in accretion yearly to form an annual band and are typically validated qualitatively. Researchers have shown key interest in modeling marginal increments to verify the marks do, in fact, occur yearly. However, it has been challenging in finding the best model to predict this sawtooth wave pattern. We propose three new applications of time series models to validate the existence of the yearly sawtooth wave patterned data: autoregressive integrated moving average (ARIMA), unobserved component, and copula. These methods are expected to enable the identification of yearly patterns in accretion. ARIMA and unobserved components account for the dependence of observations and error, while copula incorporates a variety of marginal distributions and dependence structures. The unobserved component model produced the best results (AIC: −123.7, MSE 0.00626), followed by the time series model (AIC: −117.292, MSE: 0.0081), and then the copula model (AIC: −96.62, Kendall’s tau: −0.5503). The unobserved component model performed best due to the completeness of the dataset. In conclusion, all three models are effective tools to validate yearly accretional patterns in fish ear bones despite their differences in constraints and assumptions.

Should I open to forecast? Implications from a multi-country unobserved components model with sparse factor stochastic volatility

In this paper, we assess whether and when multi-country studies pay off for forecasting inflation and output growth. Factor stochastic volatility is adopted to allow for cross-country linkages and model economies jointly. We estimate factors and rely on post-processing, rather than expert judgement, to obtain an estimate for the number of factors. This is different from most existing two-step approaches in the factor literature. Our approach is then used to extend the existing unobserved components model, which assumes that 34 economies are independent. The results suggest that allowing for cross-country linkages yields inflation and output growth forecasts that are highly competitive with those obtained from estimating economies independently. Zooming into the forecast performance over time reveals that allowing for cross-country linkages is of particular importance when interest centres on forecasting periods of uncertainty. Another key finding is that the estimated global factors are correlated with the domestic business cycle. We interpret this to mean that part of the variation captured in global factors reflects a global business cycle.

Recent Advances in the Use of the Mortality Syndromic Surveillance System-New York City, 2015-2020.

New York City's automated mortality syndromic surveillance system monitors temporal and spatial patterns in mortality. To describe the use of the syndromic surveillance system, we used the system to find mortality patterns for the 15 leading causes of death and for deaths from rare and reportable diseases in New York City from February 2015 through June 2020. We used results to find aberrations that indicate threats to public health. We used unobserved components models to analyze time series of mortality counts for leading causes of death, historical limits methods for rare and reportable diseases, and SaTScan for temporal-spatial cluster analysis. We obtained data on the number of deaths from the electronic death registry system maintained by the city's Bureau of Vital Statistics. The mortality syndromic surveillance system detected an increase in the number of deaths from heart disease by April 1, 2020, when 75.0 deaths occurred on March 24, 2020, instead of an expected 45.8 deaths (95% upper prediction limit of 61.0) and an increase in the number of deaths from all causes on March 20, 2020, when 194.0 deaths were observed while 150.1 deaths were expected (95% upper prediction limit of 178.0). The number of deaths from all causes returned to normal the week beginning June 14, 2020, when 990.0 deaths were observed and 998.8 deaths were expected. When compared with efforts from New York City to provide yearly vital statistics, the automated mortality syndromic surveillance system can provide timely mortality data with fewer resources and raise the capacity to detect anomalous increases in mortality.

Increase in hospitalisation-associated methadone intoxication in France following first COVID-19 lockdown

ObjectivesAt the beginning of the COVID-19 pandemic, the French Addictovigilance Network drew attention to the need to facilitate access to methadone while ensuring its safe use, in order to avoid the occurrence of overdoses and deaths. The objectives of the study were to assess the impact of the lockdowns on the incidence of methadone-use-related hospitalisations (MUHs) and describe the characteristics of patients and hospitalisations. Study designAn interrupted time series using the unobserved components model was performed to predict the monthly incidence of MUHs in 2020 on the basis of previous years' data and compared with MUHs observed. Data were presented with prediction intervals (PI95%). MethodsThis retrospective study was conducted on patients hospitalised in France for methadone between 2014 and 2020, using the French national database hospital discharge database. Patients’ characteristics and hospitalisations were described over four periods: before lockdown, first lockdown, after first lockdown, and second lockdown. ResultsCompared to the predictions, a higher incidence of MUH was found during the first lockdown, especially in March 2020 (66 cases vs. 51.3; PI95%: 34–65), and there was a large increase during the month following the end of the first lockdown (79 cases vs. 61; PI95%: 46–75). Coconsumptions (alcohol, cannabis, cocaine) were more frequent during the first lockdown, whereas patients aged over 30 years were more concerned thereafter. The second lockdown did not present any particularity. ConclusionsThe first lockdown had a significant impact on the incidence of MUHs. These results confirm the data from the reinforced national monitoring during first lockdown published in 2020, where methadone was the primary substance involved in overdoses and deaths.

Forecasting Rural and Urban Otolaryngologists, Radiation Oncologists, and Oropharyngeal Carcinoma.

To forecast oropharyngeal carcinoma (OPC) incidence with otolaryngologist and radiation oncologist numbers per population by rural and urban counties through 2030. Incident OPC cases were abstracted from the Surveillance, Epidemiology, and End Results 19 database, and otolaryngologists and radiation oncologists from the Area Health Resources File by county from 2000 to 2018. Variables were analyzed by metropolitan counties with over 1,000,000 people (large metros), rural counties adjacent to a metro (rural adjacent), and rural counties not adjacent to a metro (rural not adjacent). Data were forecasted via an unobserved components model with regression slope comparisons. Per 100,000 population, forecasted OPC incidence increased from 2000 to 2030 (large metro: 3.6 to 10.6 cases; rural adjacent: 4.2 to 11.9; rural not adjacent: 4.3 to 10.1). Otolaryngologists remained stable for large metros (2.9 to 2.9) but declined in rural adjacent (0.7 to 0.2) and rural not adjacent (0.8 to 0.7). Radiation oncologists increased from 1.0 to 1.3 in large metros, while rural adjacent remained similar (0.2 to 0.2) and rural not adjacent increased (0.2 to 0.6). Compared to large metros, regression slope comparisons indicated similar forecasted OPC incidence for rural not adjacent (p = 0.58), but greater for rural adjacent (p < 0.001, r = 0.96). Otolaryngologists declined for rural regions (p < 0.001 and p < 0.001, r = -0.56, and r = -0.58 for rural adjacent and not adjacent, respectively). Radiation oncologists declined in rural adjacent (p < 0.001, r = -0.61), while increasing at a lesser rate for rural not adjacent (p = 0.002, r = 0.96). Rural OPC incidence disparities will grow while the relevant, rural health care workforce declines. NA Laryngoscope, 134:136-142, 2024.

Age and gender differentials in unemployment and hysteresis

Abstract We use a time-varying panel unobserved components model to estimate unemployment gaps disaggregated by age and gender. Recessions before COVID affected men’s labor market outcomes more than women’s; however, the reverse was true for the COVID recession, with effects amplified for younger workers. We introduce time-variation in both the hysteresis dynamics and the Phillips-curve coefficients on labor market slack. The aggregate Phillips curve flattens over time and hysteresis is countercyclical for all groups. We find heterogeneity in both the Phillips curve and hysteresis coefficients, with wages responding more to workers with an outside option (high school- and retirement-age) and larger effects of hysteresis for younger workers.

Modelling output gaps in the Euro Area with structural breaks: The COVID-19 recession

This paper proposes a novel unobserved component model with a COVID-19 structural break in the trend growth rate to model output gaps. Using historical real GDP data for the Euro Area between 1995Q1 and 2022Q1, we test our framework against a battery of competing models, including a standard unobserved components model, a correlated model with a second-order Markov process, a Hodrick–Prescott filter and an augmented version of it. To examine the impact on the fitting performance, we test the inclusion and exclusion of pandemic quarters and we also extend the estimation to a country-level detail. We find that: (i) our suggested model outperforms the competing ones; (ii) when excluding pandemic quarters, the standard unobserved component model outperforms their counterparts; (iii) our model yields the best fitting performance for most of the Euro Area countries and (iv) the Hodrick–Prescott filter model has the poorest fitting performance.

Potenciální produkt a mezera výstupu v období ekonomických krizí

The article investigates potential output and output gap modelling and estimation in the Czech Republic in the period 1996-2021, including the global recession from 2008 and the recent crisis caused by government measures against the COVID-19 pandemic. The unobserved components (UC) methodology is applied, coefficients are estimated by the maximum likelihood method, unobserved variables are estimated using the Kalman filter. The standard UC model is modified in an original way to nonlinearly describe the hysteresis effect by allowing the output gap to have an asymmetrical influence on potential output. The econometric model verification proved significance of the hysteresis effect and showed a substantial inertia of negative consequences of both crises. Predictions of an impact of the War in Ukraine on the gap were also calculated and the uncertainty associated with these predictions was quantified.

Estimating the Business Cycle of Ukraine Under the Conditions of Large External Compound Shocks

Abstract. Introduction. Conventional business cycle estimation methods typically rely on the assumption that the shocks to an economy are normally distributed. However, global social and political instability can result in external shocks that are more severe than purely economic shocks, thus hampering the ability of these estimation methods to separate cyclical behavior from long-run dynamics. Since effective economic policy is dependent upon the ability to make accurate forecasts, an understanding of the properties of business cycle estimation methods in the presence of large shocks in the data is of first order importance. Purpose. The purpose of this article was to compare the ability of various business cycle estimation methods — and in particular detrending filters — to account for large external compound shocks (i.e. those containing both economic and noneconomic components) when extracting the cyclical component of the real GDP series of Ukraine. In view of Ukraine’s policy of European integration, a secondary goal was to investigate the performance of various detrending filters in deriving a measure of business cycle co-movement of Ukraine vis-à-vis the EU that is robust to external compound shocks. Results. Using an unobserved components model with external geopolitical shocks as a benchmark, it was found that the application of the boosted Hodrick-Prescott filter, the Christiano-Fitzgerald filter, and the Hamilton regression filter to the real GDP data of Ukraine produced cyclical components that exhibited spurious dynamics, particularly from 2020 and onward. Conclusions. It was shown that no single business cycle estimation method performed the best in application to the real GDP series of Ukraine. While the unobserved components model relied on extensive researcher-specified assumptions and an ad hoc approach to identifying external compound shocks, the use of data filtering resulted in series that did not accurately reflect cyclical and trend dynamics in Ukraine toward the end of the sample. Thus, there is a practical need to develop new business cycle estimation models that would be able to account for the distorting influence of large external compound shocks.