Wednesday, Aug 6: 8:30 AM - 10:20 AM
4143
Contributed Papers
Music City Center
Room: CC-212
Main Sponsor
Business and Economic Statistics Section
Presentations
Inflation has been a significant issue since the onset of the Global Pandemic in 2020, but there are renewed fears that inflation could be reignited as new tariffs are imposed. This paper aims to analyze the potential impacts of tariffs on US inflation using machine learning and traditional regression methods. The paper will use machine learning methodologies such as XGBoost, Random Forest, and Facebook Prophet. After using the latter methods, we will compare the efficiencies using traditional methods such as autoregressive integrated moving averages (ARIMA) and vector autoregression from January 2000 to April 2025. Then, we compare the efficiencies of each forecasting methodology. Using the best forecasting method based on sound ethical and professional analysis, we will try to understand where the direction of inflation could be heading for the United States given the imposed tariffs on several countries throughout the world.
Keywords
Machine Learning
Tariffs
Forecasting
XGBoost
Facebook Prophet
The economic recovery after the response to the Covid-19 pandemic varied considerably across US counties. Using data from Opportunity Insights, we employ functional data techniques to model the recovery of employment among counties as a function of time. First, we regress the functional employment data on pre-2020 American Community Survey data to estimate how county demographics and industry distribution affected the course of the recovery. We compare this functional regression model with more standard panel regression techniques, like the random effects model, and demonstrate the benefits of a functional data approach. We also use cluster analysis on the functional employment data to find different recovery patterns across counties. The results from these analyses can be used to help municipalities plan and anticipate the economic impact of future similar disasters.
Keywords
Functional Data Analysis
COVID-19
County-Level Employment
Cluster Analysis
Crop yields and harvest prices are often considered to be negatively correlated, thus acting as a natural risk management hedge through stabilizing revenues. Storage theory gives reason to believe that the correlation is an increasing function of stocks carried over from previous years. In this paper, we use semi-parametric quantile regression (SQR) with penalized B-splines to estimate a stock-conditioned joint distribution of yield and price. The method enables sampling from the empirical joint distribution using SQR. Then it is applied to approximate the stock-conditioned correlation for both corn and soybeans in the United States. For both crops, Cornbelt core regions have more negative correlations than do peripheral regions. We find strong evidence that correlation becomes less negative as stocks increase and also upon moving north. We suggest three channels through which stocks can predict revenue. The first two channels are currently addressed in premium rate-setting procedures. The third is not and we provide yield auto-correlation evidence to suggest that this could be a concern. We conduct a rating game to evaluate our methodology for assessing premium rates.
Keywords
Crop Insurance
Quantile Regression
Agricultural Economics
Price-yield correlation
Insurance premium rate setting
After the stock market crash in 2020, the US stock market was almost out of control, with the S&P 500 index soaring 173.5% in less than five years. Based on the log-periodic power law singularity (LPPLS) method, we systematically investigate the bubble status of sectors with different total market capitalization levels of U.S. stocks through four major U.S. stock market indexes, including the Wilshire 5000 Total Market index, the S&P 500 index, the S&P MidCap 400 index, and the Russell 2000 index, which represent the overall stock market, large-cap stocks, mid-cap stocks and small-cap stocks, respectively. We find that the peak confidence indicator of these four indexes all exceed 19% after November 2024, which indicates that the price trajectories of these four stock market indexes have clearly featured the obvious LPPLS bubble pattern of the faster-than-exponential growth corrected by the accelerating logarithm-periodic oscillations and are indeed in a positive bubble regime. The accelerating growth trends of these four indexes are likely unsustainable, and the positive bubble regime may change in the form of a sharp crash or volatile sideway plateaus.
Keywords
Stock market bubble
Log-periodic power law singularity (LPPLS)
LPPLS confidence indicator
Financial bubble and crash
As one of only eight countries in the world that lie partially within the Arctic Circle, Finland exhibits very interesting air passenger traffic flows. The three airports with the greatest passenger volumes are Helsinki-Vantaa, Oulu, and Rovaniemi. At the Helsinki-Vantaa Airport, domestic passenger flows peak during the cold weather months, while international traffic peaks during the summer months. Both domestic and international travel reach their respective apexes during the Spring and Fall months in Oulu. In Rovaniemi, the coldest destination of the three areas, air passenger flows reach their maxima every year in December. To date, there have been very few formal studies of air travel demand in Finland. Because of unique geographic features, location, and tourism patterns, the three principal destination markets in this country merit additional research. Air traffic has grown substantially in recent years and econometric modeling analysis may yield interesting results that may differ from those of other regions. Santa Claus and Aurora Borealis effects are examples of seasonal and cyclical variables that potentially affect air travel to, and within, this Nordic economy.
Keywords
Air Passenger Traffic
Business Cycles
Polar Tourism