HEINO AND SUNDE ENERGY REPORTS
Description
Variable Meaning Unit Role BY Long-term sovereign bond yield Percent Dependent variable RR Repurchase rate Percent Monetary-policy control INFL Headline inflation Percent Nominal-price control DEBT Public debt-to-GDP ratio Percent of GDP Fiscal-risk control GDPG Real GDP growth Percent Real-activity control SPEI12 12-month Standardised Precipitation Evapotranspiration Index Index Physical-climate variable TA Temperature anomaly Degrees Celsius Physical-climate variable
Files
Steps to reproduce
4. Baseline specification The manuscript reports a baseline ARDL(7,7,7,0,1,2,2) model with BY as the dependent variable and RR, INFL, DEBT, GDPG, SPEI12 and TA as regressors. The selected specification is evaluated with unit-root tests, an ARDL bounds test, an error-correction representation, residual diagnostics, CUSUM stability tests and Wald tests for asymmetry. 5. Core replication targets The reported bounds-test F-statistic is 5.573. The long-run coefficients reported for RR, INFL, DEBT, GDPG, SPEI12 and TA are 0.2040, 0.0844, 0.0470, 0.0157, 0.4074 and 0.2430, respectively. The reported error-correction coefficient is −0.5271 and the baseline estimation uses 89 observations after lag losses. These values are reproduced in the Reported_Results sheet as explicit verification targets. 6. Data preparation Use the 96 quarterly observations from 2000Q1 to 2023Q4 in the Analysis_Data_Baseline sheet or the accompanying CSV. Preserve the variable names exactly. The manuscript states that monthly climate observations were aggregated to quarterly averages before estimation. Estimate the model in EViews or another package capable of ARDL bounds testing, applying the lag orders reported in the manuscript rather than reselecting a different specification if the goal is exact replication. 7. Important provenance check The uploaded source workbook contains both “BY - FRED” and “BY-SARB” columns. The “BY - FRED” column reproduces the baseline descriptive statistics reported in manuscript Table 2, while the manuscript text describes BY as the 10–15-year South African government-bond yield from the South African Reserve Bank. The Source_Series_Check sheet makes this discrepancy explicit. The authors should reconcile the final manuscript wording with the exact series used in the original EViews estimation before public archiving. 8. Medium-term robustness model The manuscript reports a 5–10-year yield robustness model, ARDL(7,3,6,0,1,0,2), with a bounds-test F-statistic of 5.494 and an error-correction coefficient of −0.4442. The uploaded source workbook does not contain the BYM series used for this specification. The Robustness_Data_Template therefore leaves BYM blank rather than inserting an unverified substitute. The exact original BYM series should be added from the authors’ EViews workfile or the confirmed South African Reserve Bank source before the replication archive is described as fully self-contained. 9. Transparency note The replication files intentionally distinguish between what is directly recoverable from the uploaded source workbook and what requires author confirmation. This prevents a plausible but undocumented replacement series from being presented as the original estimation data.
Institutions
- Namibia University of Science and TechnologyKhomas Region, Windhoek