Can Government Gold Be Put to Better Use?  Qualitative and Quantitative Effects of Alternative Policies Dale W. Henderson, John S. Irons, Stephen W. Salant, and Sebastian Thomas International Finance Discussion Papers (1997-582) See below for instructions and definition of terms. (Note, program may take several minutes to download.)

 
 

Instructions. The simulation is run in three stages. First you must calibrate the model to match the current year's data. Second, you run the simulation of a gold sale. The final stage is to display the results. Calibrate. At the right of the screen above, choose which parameters you would like to calibrate, and press the Calibrate button. (If the calibration does not converge, try other parameter values.)   The first option allows you to choose values independently for the elasticity parameters epsilon and rho then computes a value of hbar to match the other assumptions. The second option allows you to specify a value for hbar and computes the value of epsilon and rho, with the constraint that epsilon = rho. The final option allows you to specify hbar and rho, and computes epsilon.   Calculate. If the calibration is successful, you are ready to simulate a gold sale by pressing the Calculate button. You may alter the year of the sale and/or the amount of the sale to run alternate senarios.  If you wish to run the simultion with no gold sale, enter -1 into the Sale Year box.  If you change parameters other than sale year or gBar, you must again press the Calibrate button. If the simulation does not converge, try other parameter values. Results. The results box displays the price of gold in the year of  the sale as well as total consumer and/or producer surplus for each of the groups. More results can be graphed and displayed by selecting the desired time-series in the box (default is set to Price(p)) and entering the number of years to display. (See warnings below). Notes/warnings: The reset button will reset all parameters to their default values. The simulation may not converge for all combinations of paramters. After calibration, you can change the sale year or gBar without having to recalibrate the model. If you change any other parameter you must first press the Calibrate button to determine a new value for the calibrated parameter(s). Surplus results should not be compared across calibrations, only across alternative sale dates and amounts. In the Years to Display box, a maximum of 297 years may be graphed at a single time - all years will still be displayed properly for values > 297. The results from the simulation above may differ slightly from those reported in previous versions of the paper. Problems with the program? Contact jsirons@amherst.edu. See the full paper for more details.

 

Definitions / Details.   Sale year: Enter year of the gold sale. Should be between 1 and <horizon>.  For no sale enter a value of -1. gBar: The amount of the government gold sale to simulate. (Troy ounces). Interest rate: The real rate of interest as a decimal (e.g. 2.5% is 0.025) Price t=0: The initial price of gold. (US$ per troy ounce). Horizon: Final year of simulation. q: Depletion demand in the initial year (troy ounces) cost: Cost of mining (US$ per troy ounce). Population growth: Initial growth rate of the world population as a decimal (e.g. 4% is 0.04, see See the full paper for more details.) Abar: Initial quantity of gold held by service users: above-ground stock. (Troy ounces) Epsilon: Depletion demand parameter: q = a * P ^ (-epsilon) where P is the price, q the depletion demand, and a is a  (time varying) constant. Rho: Rental service demand parameter: A = b * R ^ (-rho) where R is the rental rate, A the rental demand, and b is a (time varying) constant. hBar: Total stock of gold below ground (not yet mined). Years to display: Number of years to graph, display for the selected series (initially set to Price(p), but you may choose other series). See the full paper for more details.