IAS 29 and the cost of holding money under hyperinflationary conditions

Empirical evidence is presented on the efficacy of procedures summarised in IAS 29: Financial Reporting in Hyperinflationary Economies for estimating the loss in purchasing power from holding monetary items during hyperinflationary periods. Our empirical analysis encompasses 32 hyperinflationary economies covering a wide variety of hyperinflationary conditions and spanning a period of more than 80 years. While the estimation procedures summarised in IAS 29 perform poorly under all the hyperinflationary conditions encompassed by our sample, they are especially poor when the rate of inflation accelerates towards the end of a relatively short hyperinflationary period. For these latter economies, our best estimate of the actual purchasing power loss is typically only a small fraction of the figure obtained under the IAS 29 procedures. For hyperinflations of longer duration, the IAS 29 procedures return estimated purchasing power losses that are typically around 10% larger than our best estimate of the actual losses. We also derive and empirically test a general class of ‘two point’ estimation formulae that make more efficient use of the sparse information set on which the IAS 29 estimation procedures are based. The results obtained from this procedure are encouraging and suggest it is possible to obtain reliable estimates of purchasing power losses using only sparse information sets provided realistic assumptions are made about the way monetary holdings respond to variations in the purchasing power of the currency.