Annual EBHS Conference, 39th Annual Economic and Business History Society Conference

Font Size: 
Did Purchasing Power Parity Hold in Medieval Europe?
Tony Moore, Adrian Bell, Chris Brooks

Last modified: 2014-03-10


In modern financial and economic theory, the ‘law of one price’ asserts that the same good cannot simultaneously trade at different prices in different locations. If so, arbitrageurs would intervene, buying where the price is lowest and selling where it is highest, thus forcing the exchange rate-adjusted prices to converge. Purchasing power parity (PPP) is an aggregate version of this and proposes that a representative basket of goods and services should cost the same (expressed in a common currency) whatever country it is purchased. However, while PPP appeals to economists’ senses of how markets should behave, empirical support for PPP in the current finance literature is mixed. Furthermore, such studies use modern data, and only a few analyses go back as far as the nineteenth century. By contrast, while economic historians have studied prices in the Middle Ages, most notably as part of the ‘Great Divergence’ debate or investigations of market integration, they do not address PPP directly. In addition, these historical studies generally use silver equivalents to compare relative prices rather than nominal prices converted using contemporary exchange rates.

This paper employs a unique, hand-collected dataset of medieval exchange rates for five major currencies (the Barcelona lira, English pound sterling, Flemish pond groot, Florentine florin and French livre tournois), extracted from business letters in the private archive of Francesco di Marco Datini of Prato (c.1335-1410). We use these rates to calculate two sets of real exchange rates: the first based on Consumer Price Indices from the historical literature on real wages and the second on wheat prices. This produces twenty sets of real exchange rates for ten currency pairs between 1383 and 1411. In the modern finance literature, PPP is found to hold where the real exchange rate fluctuates around its equilibrium value, i.e. it exhibits mean reversion and is stationary. Conversely, if the real exchange rate series is non-stationary and follows a random walk with a unit root, then PPP does not hold. We thus apply unit root and stationarity tests to each of our individual real exchange rate series. In addition, since our sample is relatively short, we also adopt a panel approach to increase the power of the analysis.

Our results find evidence of stationarity (and thus support for PPP) for half of the individual series and for almost all of the panel groups that we investigate. Moreover, we are able to distinguish between a more-integrated north-south axis running from Southern England through the Low Countries to Northern Italy and a less-integrated east-west axis between Italy, the Low Countries, France and Spain. Overall, our findings add to the growing weight of evidence that at least some medieval financial markets were well functioning and that modern economic theories can be applied even to distant historical periods.