The frequency of price adjustments is a key measure for monetary policy. If prices adjust often and fast, new price equilibria establish quickly after macroeconomic shocks. However, prices are typically adjusted infrequently. Rare, unsynchronised and sluggish price adjustments imply that shocks take time to feed into prices and generate inflation persistence. The resulting real distortions imply that there is room for monetary policy to enhance efficiency in well-known ways.
The recent literature on price changes in response to exchange rate shocks documents that the frequency of price adjustment is not constant but depends on the origin of the underlying shocks (e.g. Forbes et al. 2015). The Swiss National Bank’s lifting of the minimum exchange rate policy of 1.2 Swiss francs per euro on 15 January 2015 represents a rare episode that allows for a clean identification of the reaction of Swiss import prices to a sudden and large exchange rate change. In a recent paper, we analyse nominal prices adjustments of imports and exports after 15 January 2015, showing that price reactions were very fast and nominal rigidities unravelled quickly (Bonadio et al. 2016).
The large and unanticipated shock
The Swiss National Bank’s decision to discontinue its minimum exchange rate policy took financial markets by storm. It resulted in a one-day appreciation of more than 15% of the Swiss franc against the euro, before stabilising at a permanent appreciation of around 11%.
Figure 1. EURCHF evolution from 2012 to June 2015
The Swiss franc shock on 15 January 2015 was an especially large and ‘clean’ shock. It was clean in the sense that it was unanticipated by financial markets and its origin was purely nominal – that is, it was not driven by economic fundamentals that might have impacted price setting behaviour in multiple ways. This is a clear advantage over most other studies, because exchange rate movements are typically highly endogenous.
Instantaneous drop of import prices
How did nominal prices adjust in the aftermath of the Swiss franc shock? To answer this question, information on the invoicing currency is essential. Notice, for example, that the Swiss IPI, which is measured in Swiss francs, may react instantaneously in response to exchange rate shocks simply because Swiss imports are invoiced in euros and euro-denominated prices remain unchanged.1
One can, however, gauge the adjustment of nominal prices by analysing the exchange rate pass-through into unit values of import transactions that are invoiced in Swiss francs. In this case, changes in unit values can be interpreted as nominal price changes.
Figure 2. Daily reaction of import unit values for transactions invoiced in Swiss francs
Figure 2, taken from Bonadio et al. (2016), plots the estimated cumulative drop of import unit values on a daily scale since 15 January 2015 (red line) together with the confidence intervals (dotted red lines) and the cumulative change of the EURCHF exchange rate (blue line).
- The figure indicates that the unit values started to move significantly on the second working day after 15 January.
- The medium-run pass-through of roughly 50% was reached after six additional working days.2
This finding strongly suggests that the underlying price adjustments were remarkably fast and that, thus, the implied frequency of price adjustments shot up dramatically relative to normal times. Between 1999 and 2014 on average 30% of Swiss producer prices (admittedly an imprecise proxy for import prices) were adjusted every quarter, or about 4% within eight working days on average. This figure is of the same order as the 10.6 months median US import price duration found in Gopinath and Rigobon (2008). Figure 2 suggests instead that nearly all import prices that eventually changed were adjusted within eight days.
The finding of fast pass-through also differs sharply to the relatively slow pass-through results documented in Campa and Goldberg (2005) and Gopinath et al. (2010). The former study finds that most of the pass-through materialises after six months, while the latter study states that it requires at least 18 months to complete.
Two key factors may explain this difference:
- First, the EURCHF exchange rate shock was exceptionally large, so that gains from fast adjustment were prominent and did outweigh nominal frictions; and
- Second, and in addition, the shock was unanticipated – that is, firms did not start adjusting prices in anticipation of the shock.
The sharp and concentrated peak of adjustment shortly after the date of the shock was thus preserved.3
The key message of Figure 2 is that the large exchange rate shock has induced prompt changes of nominal import prices. In the face of the exchange rate shock, it appears that nominal price rigidities were very quickly overcome.
Why fast price adjustment is important
Our findings on the speed of the pass-through tend to support state-dependent pricing models by Dotsey et al. (1999) as opposed to time-dependent pricing models by Calvo (1983). They constitute strong evidence in favour of fast and universal nominal price adjustments in the face of a large shock. This fast exchange rate pass-through has two important implications.
- First, in modelling price setting behaviour, the fast adjustment of prices suggests that nominal rigidities play only a minor role in the face of large exchange rate shocks.
These considerations lead us to adopt the view that international firms demonstrate a high level of flexibility in their ability to respond to large and sudden changes in their operating environment.
- Second, we document a fast pass-through for a large shock.
In forecasting import and export prices, the new pass-through estimates highlight and confirm the view that price adjustment is heavily dependent on the nature of the exchange rate shock. Past literature has often focused on price adjustment in response to frequent and small exchange rate shocks, showing that the pass-through tends to be slow.
- Finally, we emphasise that large exchange-rate shocks happen more frequently than one might think.
Between 1994 and 2015, there have been 115 occurrences of monthly nominal effective exchange rate changes of 10% or more, concerning 19 countries, including developed economies (e.g. Japan, Australia, South Korea). While these events didn’t happen in a single day and therefore do not serve to identify the pass-through as nicely as the recent Swiss franc shock, they nevertheless do matter for firms’ optimal response to the exchange rate situation.
Authors’ note: The views expressed in this article do not necessarily reflect those of the Swiss National Bank.
Bonadio, B, A M Fischer and P Sauré (2016) “The speed of the exchange rate pass-through”, CEPR Discussion Paper, DP11195.
Calvo, G (1983) “Staggered prices in a utility-maximizing framework”, Journal of Monetary Economics, 12(3): 383-398.
Campa, J and L Goldberg (2005) “Exchange rate pass-through into import prices”, Review of Economics and Statistics, 87(4): 679-690.
Dotsey, M, R G King and A L Wolman (1999) “State-dependent pricing and the general equilibrium dynamics of money and output”, Quarterly Journal of Economics, 114(2): 665-690.
Forbes, K J, I M Hjortsoe and T Nenova (2015) “The shocks matter: Improving our estimates of exchange rate pass-through”, External MPC Unit Discussion Paper (43).
Gopinath, G, O Iskhoski and R Rigobon (2010) “Currency choice and exchange rate pass-through”, American Economic Review, 100(1): 304-336.
Gopinath, G and R Rigobon (2008) “Sticky borders”, Quarterly Journal of Economics, 123(2): 531-575.
In 2014, 66.8% of all import transactions in that period were invoiced in euro and 32.2% in Swiss francs. Furthermore, research suggests that prices invoiced in the exporter’s currency remain unaffected by exchange rate changes even in the long run, see Gopinath et al. (2010).
 The analysis is restricted to Swiss imports from the euro area and the 18 months between January 2014 and June 2015. Regressions are in levels and include time-trends for all good-postal code combinations as well as good-postal code-partner country fixed effects. See Bonadio et al. (2016) for further details.
 In addition, Bonadio et al. (2016) show that omitted variables and autocorrelation issues might bias standard estimations.
 Authors’ calculations based on the BIS broad nominal effective exchange rate.