Search Results

You are looking at 1 - 10 of 34 items for :

  • "bootstrapped sample" x
Clear All
Klaus-Peter Hellwig

using bootstrapped samples. Selection frequencies for same-year forecasts and five-year ahead forecasts are plotted in Figure 10 . Points close to a 45-degree line would suggest that WEO forecasters’ choice of predictors is consistent with their actual relevance. The point in the top right corner indicates that, regarding the most robust predictors, the behavior of WEO forecasts is consistent with that of actual growth outcomes. However, especially at the longer horizon, several variables are above a 45-degree line, indicating that they receive too much attention

Klaus-Peter Hellwig
I regress real GDP growth rates on the IMF’s growth forecasts and find that IMF forecasts behave similarly to those generated by overfitted models, placing too much weight on observable predictors and underestimating the forces of mean reversion. I identify several such variables that explain forecasts well but are not predictors of actual growth. I show that, at long horizons, IMF forecasts are little better than a forecasting rule that uses no information other than the historical global sample average growth rate (i.e., a constant). Given the large noise component in forecasts, particularly at longer horizons, the paper calls into question the usefulness of judgment-based medium and long-run forecasts for policy analysis, including for debt sustainability assessments, and points to statistical methods to improve forecast accuracy by taking into account the risk of overfitting.
Francesco Manaresi and Mr. Nicola Pierri

: Distribution of γ from equation ∆ω i,t =ψ i +ψ p,s,t +γ·φ i,t +η i,t . See section 5 for details. Distribution is computed from 50 (firm-level) bootstrapped sample. Industry level production function and firm level productivity growth is re-estimated for each bootstrapped sample. Estimates are all above zero (red vertical line) for all samples. Figure A.8: Distribution of γ from equation ∆ω i,t =ψ p,s,t +γ·ITBK i,2006 +η i,t . See section 6 for details. Distribution is computed from 50 (firm-level) bootstrapped sample. Industry-specific production function and firm

Mr. Raphael A Espinoza, Mr. Ananthakrishnan Prasad, and Mr. Gene L. Leon
Motivated by the global inflation episode of 2007-08 and concern that high levels of inflation could undermine growth, this paper uses a panel of 165 countries and data for 1960-2007 to revisit the nexus between inflation and growth. We use a smooth transition model to investigate the speed at which inflation beyond a threshold becomes harmful to growth, an important consideration in the policy response to rising inflation as the world economy recovers. We estimate that for all country groups (except for advanced countries) inflation above a threshold of about 10 percent quickly becomes harmful to growth, suggesting the need for a prompt policy response to inflation at or above the relevant threshold. For the advanced economies, the threshold is much lower. For oil exporting countries, the estimates are less robust, possibly reflecting heterogeneity among oil producers, but the effect of higher inflation for oil producers is found to be stronger.
Mr. Raphael A Espinoza, Mr. Ananthakrishnan Prasad, and Mr. Gene L. Leon

checks in the inflation-growth literature—that explain the range of estimates found in the literature as inherent to the data process as much as the estimation technique. Indeed, although the bootstrapping exercises suggest that the value of the threshold is robust to outliers, the range of bootstrap estimates turns out to be almost as large as the range of estimates in the literature. For instance, looking at developing countries, for more than 90 percent of the bootstrapped sample, the estimated inflation threshold lies between 7 and 13 percent. Furthermore, the

Mr. Olumuyiwa S Adedeji, Mr. Calixte Ahokpossi, Claudio Battiati, and Mrs. Mai Farid

true standard deviation. Non-parametric approach Equation  ( 4 )  can be written as:  p r o b [ d p > d D C ( 1 + λ α − μ λ ) ] = α In this equation, d DC is known for each country (as explained above). After approximating μλ by the bootstrapped sample mean, we conduct a stochastic simulation by drawing with replacement λ 10,000 times from the country-specific sample and calculate the equivalent prudent

Majid Bazarbash

trees 17 . The first principle, bagging , is a general-purpose procedure in which the base model (a decision tree) is estimated on many randomly drawn subsamples with replacement from the train dataset where the subsamples have the same size as the train dataset (also called “bootstrappedsampling). The predicted values by estimated trees are then averaged to yield the final prediction. In classification, a majority vote is taken to determine the final prediction 18 . Based on the second principle, decorrelation , only a subset of features is considered at each

Majid Bazarbash
Recent advances in digital technology and big data have allowed FinTech (financial technology) lending to emerge as a potentially promising solution to reduce the cost of credit and increase financial inclusion. However, machine learning (ML) methods that lie at the heart of FinTech credit have remained largely a black box for the nontechnical audience. This paper contributes to the literature by discussing potential strengths and weaknesses of ML-based credit assessment through (1) presenting core ideas and the most common techniques in ML for the nontechnical audience; and (2) discussing the fundamental challenges in credit risk analysis. FinTech credit has the potential to enhance financial inclusion and outperform traditional credit scoring by (1) leveraging nontraditional data sources to improve the assessment of the borrower’s track record; (2) appraising collateral value; (3) forecasting income prospects; and (4) predicting changes in general conditions. However, because of the central role of data in ML-based analysis, data relevance should be ensured, especially in situations when a deep structural change occurs, when borrowers could counterfeit certain indicators, and when agency problems arising from information asymmetry could not be resolved. To avoid digital financial exclusion and redlining, variables that trigger discrimination should not be used to assess credit rating.