PPP and the return to valuation

macroeconomic purchasing power parity

30 Jan 2014

One of the cornerstones in international economics is the theory of purchasing power parity (PPP), which simplified states that price levels in two countries should be identical after converting prices into a common currency. PPP is an appealing approach for many long-term investors in currency markets as it serves as an anchor when exchange rates are moving rapidly, as at present. But with various PPP measures, which one should we look at and why? Which currencies are farthest away from equilibrium levels and may therefore be subject to weakening or strengthening going forward? Has the past year’s sharp movements erased the misalignments in the G10 currency universe and are we now close to equilibrium?

Basically, two versions of PPP exist – an absolute version and a relative one. Absolute PPP theory states that a basket of goods costs the same at home and abroad if goods prices are converted into a common currency. To put it differently, PPP assumes that purchasing power is equal between countries. By contrast, relative PPP theory does not compare levels of purchasing power at home and abroad, but rather focuses on changes in purchasing power. Accordingly, relative PPP states that inverse changes in the nominal exchange rate that offset inflation differentials between countries are such that the PPP ratio remains constant. The validity of absolute PPP theory therefore implies the validity of relative PPP theory, but not vice versa.

The PPP theory is based on the law of one price and there are a number of caveats to the theory. First, transportation costs, barriers to trade and other transaction costs can be significant. Second, there must be competitive markets for the goods and services in both countries. Third, the law of one price only applies to tradable goods – immobile goods such as houses, and many services that are local, are of course not traded between countries. This might elucidates why PPP explains movements in exchange rates over short or medium time periods quite poorly. But over very long periods of time PPP does seem to have better luck in explaining FX movements. However, Rogoff (1996) found that the speed of convergence to PPP is extremely slow; deviations appear to dampen out at a rate of roughly 15%.

macroeconomic purchasing power parity deviations against bigmac

The simplest way to calculate purchasing power parity between two countries is to compare the price of ‘standard’ goods that are identical across countries. The Economist publishes a light-hearted version of PPP: ‘The Big Mac index’ which compares the price of a McDonald’s hamburger around the world. This index suggests that a Norwegian Big Mac is too costly compared with an Australian one.

For practical FX analysis, there are a number of characteristics we would ideally like our PPP-estimates to possess: First, a good PPP estimate encompasses a large number of goods and services – not just the price of a burger. Accordingly, CPI- or PPI-based measures are preferred. Second, we need PPP estimates to be comparable – the best way to ensure the use of price data from a single source, measure exchange rates at the same date and time, and evaluate estimates over the same base period. Third, availability is a must. If estimates are difficult to get hold of and access is cumbersome, most investors look elsewhere for alternative valuation measures. Finally, we like it if markets agree on a common measure. This equates to less confusion and less potential misunderstanding.

An easily accessible way to get a quick overview on PPP estimates is via Bloomberg’s PPP <GO>. If one accepts the pre-chosen base period (January 1982-June 2000) it is convenient to dive into a small PPP study and many insights are obtained quickly and straightforwardly.

macroeconomic purchasing power parity deviations from ppp

By looking at long-term valuation against the US dollar in the G10 universe derived from CPIs we observe that Danish krone, euros and Swiss francs are currently the most overvalued currencies while sterling, the New Zealand dollar and especially the Swedish krona are the most undervalued. This picture is, however, very different compared with one year ago, when all G10 currencies were overvalued against the US dollar, except the yen. We have, in other words, seen a huge adjustment toward long-term estimates in the past year. The picture is very similar when using PPIs instead of CPIs. This makes us conclude that we gain little additional insight when applying a different price metric within this framework.

macroeconomic purchasing power parity total deviation from ppp

On the one hand, we observe that both the average total actual and absolute under- and over-valuations within the G10 universe have fallen. Specifically, the average total actual valuation misalignment is currently close to zero, implying the G10 universe is close to equilibrium. This is not the same as all spot rates being at their equilibrium levels, but merely an indication that the sum of deviations from equilibrium is small at present. This is also confirmed by the fact that the average total absolute deviation is below its historical average.

macroeconomic purchasing power parity deviation from ppp 2

On the other hand, we observe that differences among actual and absolute under- and over-valuations, measured by standard deviations, have risen and both are above historical averages. Additional to that, differences between the largest and smallest valuation misalignments have similarly risen and are likewise above historical averages.

macroeconomic purchasing power parity fiference from ppp

To sum up, a CPI-based PPP estimate, which uses a long-run averaging methodology, makes it easy to derive long-term valuations in FX markets. According to PPP analysis, EUR/SEK could very well see a downward swing in excess of 40% if the spot rate adjusts to the current PPP estimate. Such a large move might seem outrageous, but it would in fact not be far greater than the spike previously seen in EUR/GBP. Triggers for a downward move in EUR/SEK could be an improvement in risk aversion, a pick-up in equity performance and a bettering of global economic growth prospects. We are not able to draw any strong conclusions as to whether the G10 currency universe is in equilibrium or not, as statistical measures point in various directions. We are, however, able to state that previous misalignments have become considerably smaller while deviations among under- and over-valuations have risen lately and now remain above historical averages.

John Hydeskov