Many countries have experienced sharp currency movements over the past few years.  These fluctuations – and even just moderate exchange rate movements – can have sizable effects on output and prices. To predict how currency movements will affect inflation in the future, central banks and other forecasters frequently use rough ‘rules of thumb’. For example, in the US a 10% dollar appreciation has been estimated to result in a fall in US consumer prices of around 0.5% (equivalent to a pass-through coefficient of 5%).1

While these rules of thumb for exchange rate pass-through capture some important structural differences across countries and average effects (see Gopinath 2015), they don’t always perform well in predicting how currency movements will affect import prices and inflation within a country at a specific point in time. One reason why they do not always work well is that they do not account for the fact that exchange rate movements are not exogenous – the shock behind the exchange rate move can determine its impact. This consideration is central to answering other questions. For example, when analysing the impact of the recent fall in oil prices, Arezki and Blanchard (2015) begin by asking how much of the fall resulted from a shock to demand relative to supply. Or when analysing how exchange rate appreciations affect growth, Bussiere et al. (2015) find the impact depends on the shock causing the appreciation.

In this spirit, in a recent paper (Forbes et al. 2015), we suggest that in order to improve our estimates of how exchange rate movements affect import prices and inflation, we need to refrain from treating exchange rate movements as exogenous. Instead, we need to take a step back and model what caused the exchange rate to move in the first place.

The link between exchange rates and prices in theory
There is an extensive literature showing that firms adjust their prices and mark-ups differently after different shocks. Building on these insights, we develop a standard open-economy model to show that exporting firms’ decisions on how to adjust their prices and mark-ups in response to exchange rate movements also depend on what shock caused the exchange rate move. In the model, exporting firms set their prices in a forward-looking manner to reflect their expected marginal costs and expected demand conditions. These factors will be affected differently by different shocks – even if these shocks all lead to similar exchange rate movements. Therefore, within this relatively standard framework, the degree to which exporters pass through any move in the exchange rate to import prices – or instead adjust their mark-ups – depends on the shock which caused its move.

To illustrate this, Figure 1 depicts the change in foreign exporters’ average mark-up after a 1% appreciation of the domestic currency, caused by either a demand shock (in blue), a monetary policy shock (in green), or an exogenous shock to the uncovered interest rate equation (in red). Foreign exporters increase their mark-ups most following a positive domestic demand shock, because that shock simultaneously raises demand and inflationary pressures domestically. The foreign exporters take advantage of this situation of strong demand and low competition from domestic firms and respond to the appreciation by lowering their import prices by less (and increasing their mark-up by more) than they would have done were it not for improving domestic conditions. Conversely, foreign exporters pass more of the exchange rate movement through to import prices (i.e. change their mark-ups less) when the appreciation is caused by a monetary policy tightening, as that tightening simultaneously lowers both demand and inflationary pressures domestically.

Figure 1. Foreign exporters’ mark-up after selected shocks



Note: The figure depicts the average percentage change in foreign exporters’ mark-up (relative to domestic traded prices) following a shock which appreciates the exchange rate by 1%.

Evidence on the link between exchange rates and prices
To empirically test these theoretical predictions that exchange rate pass-through differs across shocks, we estimate the impact of different shocks to the exchange rate and prices using a structural vector autoregression (SVAR) framework for a small open economy. Specifically, we develop an identification scheme that allows us to consider how a range of shocks affect the degree of exchange rate pass through to both import and consumer prices. We estimate our empirical model on quarterly data for the UK over the period from 1993q1 through 2015q1.

What do our SVAR results imply for exchange rate pass through? Figures 2.a-d show the estimated pass-through resulting from six different shocks.2 The figures clearly indicate that different shocks imply different degrees of exchange rate pass through to import prices and consumer prices.

In line with the theoretical predictions, monetary policy shocks lead to a high degree of exchange rate pass-through to prices—with import prices falling by almost 70% of the appreciation in two quarters (and by 85% by quarter six). The domestic demand shock has the lowest degree of exchange rate pass-through to import prices – with less than 40% of the exchange rate appreciation being passed through to import prices after five quarters. Exchange rate pass-through is greatest after global shocks, although this also incorporates the effects of these global shocks on foreign export prices.

Figure 2.



Pass-through here is defined as the median ratio of cumulative impulse responses of import or consumer prices relative to the exchange rate. The x-axis displays quarters since a shock

Moving to the results on overall pass-through to consumer prices, most of the shocks generate large degrees of pass-through in the two years following the initial exchange rate movement. But once again, the magnitude of the effect varies based on the initial shock. In some cases, consumer prices fall by greater than 20% of the initial exchange rate appreciation after eight quarters, while in one case (after an appreciation caused by a positive shock to domestic demand), consumer prices tend to increase instead of decrease.3

To summarise, these impulse responses confirm that a given appreciation or depreciation could have very different effects on import prices and overall inflation depending on what caused the initial currency movement. This could explain why estimates of pass-through can change over time — even within a country — and why a fixed rule of thumb may not accurately predict the effect of an exchange rate movement on inflation.

Can shock-dependence help us understand why exchange rate pass through has varied over time in the UK?
We use our framework to decompose year-on-year changes in the sterling exchange rate index into six shocks. A quick glance at Figure 3 suggests that the drivers of exchange rate movements vary across time. For example, the large depreciation during the 2007-2009 crisis was associated with larger global supply shocks (in red) and domestic supply shocks (in green), which both generate relatively higher degrees of exchange rate pass-through. In contrast, the sharp appreciation of sterling around 1996-97 was driven more by domestic demand shocks (in dark blue) and exchange rate shocks (in yellow), which exhibit substantially lower degrees of pass-through.

Figure 3. Historical decomposition of year-on-year changes in nominal sterling ERI



Note: The figure depicts the contribution of each of the six shocks to y/y changes in the ERI, in percent.

Next, we use these shock decompositions to back out the implied amount of exchange rate pass through at different points in time. We find that the shocks causing the appreciation of 1996-97 imply that a 10% exchange rate appreciation would have caused import prices to fall by around 7% and the CPI by at most 1%. In contrast, the shocks underlying the 2007-09 depreciation imply that the same 10% exchange rate movement would cause import prices to fall by 9% and the CPI by around 2%.  In other words, using the exchange rate pass through coefficients from the 1996-97 episode as a rule of thumb would have underestimated the impact of the 2007-09 depreciation on the level of UK import prices by about 20% and on the CPI by 100%!

Providing yet another contrast, sterling’s most recent appreciation from 2013-15 is associated with a relatively greater role of global shocks, making it necessary to differentiate the direct impact of  foreign export prices separately from the effects of the exchange rate (and making the estimates less precise).4 The shock decomposition suggests that a 10% exchange rate appreciation would have caused import prices to fall by 4% to 7%, and the CPI by 0.1% to 1%. Even using this broad range of estimates, implied exchange rate pass through has fallen substantively compared to that following the 2007-09 depreciation.

These results help explain two recent puzzles in the UK. Why did the 2007-09 depreciation exhibit much stronger pass-through and generate a bigger increase in inflation than expected based on pre-crisis evidence? And why has exchange rate pass through from the 2013-15 appreciation fallen? The model and estimates above suggest that these changes in pass-through result from different shocks underlying the currency movements.5

Conclusion
This series of results highlights the importance of adjusting estimates of exchange rate pass-through over time to incorporate the nature of the underlying shocks causing the exchange rate movement instead of using rules of thumb. This can help improve our estimates and help us understand why some currency movements have had surprisingly large or small effects on import prices and inflation. This framework should improve our ability to predict the impact of currency movements on inflation, and therefore improve the ability of central banks to set monetary policy appropriately in the future.

References
Arezki, R and O Blanchard (2015), “The 2014 oil price slump: Seven key questions”, VoxEU, 13 January.

Bussiere, M, C Lopez and C Tille (2015), “Exchange rate appreciations and growth: the drivers matter”, VoxEU, 7 August.

Forbes, K (2015a), “Risks around the Forecast”, speech given in London on 22 January.

Forbes, K (2015b), “Much ado about something important: How do exchange rate movements affect inflation”, Speech given at the Money, Macro and Finance Research Group Annual Conference in Cardiff on 11 September.

Forbes, K, I Hjortsoe and T Nenova (2015), “The shocks matter: Improving our estimates of exchange rate pass-through”, External MPC Unit Discussion Paper No. 43, Bank of England.

Gopinath, G (2015), “The International Price System”, Federal Reserve Bank of Kansas City Economic Symposium, Jackson Hole.

Mishkin, F (2008), “Exchange rate pass-through and monetary policy”, speech given on 7 March, 2008 at the Norges Bank Conference on Monetary Policy, Oslo, Norway.

Footnotes
1 See e.g. Mishkin (2008) for the US rule of thumb and Forbes (2015a, 2015b) for the UK rule of thumb.

2 We define pass-through as the median ratios of the impulse responses of import and consumer prices to those of the exchange rate. The figures differentiate between the effects of the four domestic and two global shocks in order to highlight that the import price movements corresponding to global shocks also incorporate the effect of the global shocks on foreign export prices.

3 Despite the appreciation following a domestic demand shock, consumer prices increase because the boost to prices from stronger demand outweighs the drag to prices from the appreciation and cheaper imports.

4 This reflects uncertainty about how much of the changes in import prices reflect the sharp movements in commodity prices over this periods or effects of sterling’s appreciation.

5 See Forbes (2015b) for more details.