Introduction

The tourism and hospitality sectors are highly vulnerable to various external factors, ranging from competition, weather conditions, natural disasters, political and legal issues, to macro economic factors. Among the existing external factors, economic variables, such as inflation and exchange rates, play a key role behind demand volatility and uncertainty in the tourism sector globally (Ongan and Gozgor 2018; Wu and Wu 2019). As these external factors are difficult to manage let alone control, it is important for industry practitioners to promptly react to those shocks, whenever they hit the market.

The role of exchange rate on tourism demand has been examined in the literature. In particular, several studies focused on the relationship between exchange rate fluctuations and international tourist demand (e.g., Aalen et al. 2019; Corgel et al. 2013; De Vita and Kyaw 2013; De Vita 2014; Karimi et al. 2019; Peng et al. 2015). Webber (2001) pointed out that tourism is one particular commodity that is likely to be affected by exchange rate fluctuations. Exchange rate fluctuations in fact may affect international tourists’ destination choice and impact their intended duration of stay and expenditure (Wang et al. 2008). Nevertheless, little research has explored how exchange rate fluctuations affect the strategic decisions (e.g., pricing) and business performance of tourism service providers (e.g., hotels). This study focuses on the special case of Switzerland and analyzes how exchange rate fluctuations are related to the pricing decisions and business performance of the Swiss hospitality sector, with a particular emphasis on hotel classes. Switzerland is a particularly interesting case because it is physically inside the European Monetary Union, but operates under its own currency, the Swiss franc. The hospitality industry is one of the important pillars of the Swiss economy. During the last decade, political and economic instability all over the world have reinforced the “safe haven” status of Switzerland, which resulted in a sequence of appreciations of the Swiss franc. In order to protect the most export-oriented sectors, the Swiss National Bank (SNB) introduced an exchange rate floor between September 2011 and January 2015, which limited currency appreciations above the limit of 1.2 Swiss franc per euro (Fig. 1). Such policy can indicate signs of reduced uncertainty to the markets and give guidance in a period of uncertainty.

Fig. 1
figure 1

Source https://www.macrotrends.net

Nominal exchange rate between the Swiss franc and the euro (number of Swiss francs for 1 euro). *A reduction in the value of the exchange rate means that the Swiss franc is appreciating.

Full size image

The goal of this paper is twofold: First, this study explores the relationship between the exchange rate and the performance of hotels categorized by class; second, it tests whether the intervention of the SNB on the exchange rate market generated a structural break in the economy, i.e., a structural change in the relationship among economic variables. In order to do so, the study compares the relationship between performance and exchange rate during the period of intervention of the SNB with the one observed when the exchange rate is completely free to fluctuate. This sheds light on whether policy interventions affect the natural relationship between market variables and create structural breaks in the economy. Furthermore, the recent COVID 19 outbreak has brought major economic disruption in the tourism industry. By analyzing the monthly performance data of hotels in Switzerland over the period from January 2000 to April 2021, this study finds the presence of two structural breaks in the economy, due-respectively- to the central bank intervention on the exchange rate market between 2011 and 2015, and to COVID 19.

The remainder of the paper is organized as follows: the next section provides a literature review on the importance of the exchange rate on international tourism demand.

The theoretical framework and the empirical model are outlined in "Theoretical framework" and "The empirical model" sections, and data and the main descriptive statistics are presented in "Data and descriptive statistics" section. "Empirical analysis" section treats the issue of instrumental variables and analyzes the results of the empirical analysis, and the last section concludes.

Related literature

One of the central themes analyzed by the macroeconomic literature is the relationship between exchange rate and production. If one country’s currency becomes stronger, this implies that the residents of the country increase their purchasing power abroad, while, at the same time, the country loses its competitivity abroad. The theory predicts that these two combined effects generate a reduction in the level of production of the country which experienced the appreciation of the currency. Nevertheless, the empirical evidence does not always confirm these theoretical predictions.

Among the most influential studies in macroeconomics, Baxter and Stockman (1989) found that the exchange rate regime has little systematic effect on the business cycle properties of macroeconomic aggregates, including gross domestic product (GDP). Flood and Rose (1995) confirm these findings and suggest that the exchange rate “appears to have a life of its own”. Obstfeld and Rogoff (1995), observing the weak relationship between exchange rates and other macroeconomic aggregates, introduce the “exchange rate disconnect puzzle”. Unlike previous literature, they find that the degree of correlation between exchange rate and GDP varies depending on the level of development of the country. Developing countries display stronger correlations than developed ones, which might be explained by their heterogeneous economic structures and by the different types of shocks that they face. They find for all the countries a high correlation between exchange rate and trade flows (imports and exports). Finally, Di Nino et al. (2011) conduct a panel analysis including several countries for the period between 1861 and 2011. They observe asymmetric effects of (real) exchange rate movements for developing and developed countries. Additionally, in the second part of their paper, they focus their attention only on Italy, underlying how the impact of the exchange rate changes depending on the sector.

The importance of the exchange rate is widely recognized in the literature on international tourism demand. International tourists tend to choose destinations in which the exchange rate is more favorable (Wang et al. 2008) and the variations of exchange rates may affect tourists’ expenditure and duration of stay (Gao et al. 2018). A substantial number of studies have included some form of exchange rate variable in their models of tourism demand (e.g., Bond et al. 1977; Gao et al. 2018; Karimi et al. 2019; Webber 2001), but the research has reached mixed conclusions. According to Crouch (1994a)’s analysis of the literature on the topic, while many studies find that the exchange rate has significant effects on tourism demand (e.g., Cheng et al. 2013; Eilat and Einav 2004; Saayman and Saayman 2013; Vogt 2008), other studies find no relationship between exchange rate and tourism demand (e.g., Gao, et al. 2018; Vanegas and Croes 2000; Quadri and Zheng 2010). Furthermore, some scholars argue that exchange rate effects are different, depending on the countries considered (e.g., Croes and Vanegas 2005).

De Vita (2014) and Webber (2001) look at the exchange rate regime and at the volatility of the exchange rate and emphasize the importance of maintaining a relatively stable exchange rate in order to attract international tourist arrivals. Quadri and Zheng (2010) study how the bilateral exchange rate affects international tourist arrivals to Italy and find no effect in 11 out of the 19 countries being studied. Falk (2014) specifically analyzes the case of Switzerland and studies how exchange rate appreciations affected Swiss alpine resorts over the period 2007–2011. His findings suggest that Swiss franc appreciations have strongly affected Alpine destinations, while cities and lake destinations have registered a much weaker effect. In his analysis, he looks at the nominal exchange rate and at a modified measure of real exchange rate, including the price index of competitors instead of the price index of the origin country. Aalen et al. (2019), using a panel of hotel overnight data from Norway, find that the nominal exchange rate affects hotel occupancy with a lag of two/three quarters. Vogt (2008) concludes that exchange rate volatilities along with real income greatly influence inbound tourist arrivals in the U.S. Stettler (2017), using highly disaggregated data collected in 141 Swiss communities, finds that real exchange rate appreciations do not affect touristic demand in cities nearly as much as they do in touristic communities. Furthermore, he underlines that, among European tourists, German, Dutch and Belgian visitors seem to react to the Swiss franc’s appreciations much more strongly than French and Italian ones. Finally, Corgel et al. (2013) study how demand for U.S. hotel rooms responded to (real) dollar fluctuations. Their findings suggest that exchange rates have a minor effect on hotel demand and that the exchange rate affects luxury, upper upscale and upscale segments more than lower-price hotels.

Empirical studies in the hospitality and tourism literature have extensively discussed the role of the exchange rate, but, to the best of the authors’ knowledge, they have all agreed on one main point: the dependent variable. The dependent variable considered has always been a quantity, mostly represented by tourist arrivals and hotel occupancy. Unlike the previous literature, this study explores how exchange rate fluctuations affect hoteliers’ response and their business performance.

In the hospitality industry, the room rate (average daily rate, ADR) has a crucial role in the performance of hotels. High occupancy (i.e., total number of overnight stays) does not always indicate a positive impact on business performance (i.e., revenue and profit). Therefore, estimating the impact of exchange rate on hotel demand may show only limited implications. Instead, Revenue per Available Room (RevPAR), which is calculated by multiplying a hotel's ADR by its occupancy rate, is commonly used as a performance metric in the hospitality industry. This study focuses on how exchange rate affects the hotelier’s pricing decision (ADR) and, indirectly, the business performance (RevPAR, which is highly correlated with ADR, as shown in Table 3) of hotels categorized by class in Switzerland. We find it interesting to study the pricing strategy, because occupancy and ADR are influenced by different sets of agents. Occupancy can be seen as the result of consumers’ decisions, while ADR is a strategic variable chosen by hotel managers. As the literature has mostly focused on occupancy rates, our analysis will allow us to better understand whether different sets of agents react differently to the same shock. Last but not least, the Swiss experience is used to infer how hoteliers’ incentives are affected by an exchange rate that is under the control of the monetary authority. Switzerland represents an interesting case also because it is a small open economy highly focused on exports and imports. To the best of the authors’ knowledge, the role of exchange rate related to monetary policy in the hospitality sector has not received a lot of attention yet. Blengini and Heo (2020) explored how hotels adopting different business models react to macroeconomic shocks, including exchange rate fluctuations. According to Blengini and Heo (2020) mostly chains and independent hotels react to exchange rate fluctuations, while franchise hotels are not affected in any significant way. Finally, Inchausti-Sintes and Pérez-Granja (2022), where, adopting a dynamic stochastic general equilibrium modeling (DSGE) approach, they compare the effects of different monetary policies on the macroeconomic variables of three touristic islands. Rather than considering a macroeconomic setting, this study follows a more microeconomic approach, specifically focusing on the hospitality industry and looking at the effects of the monetary policy for exchange rate on hoteliers’ pricing decision and on equilibrium occupancy rates.

Theoretical framework

In order to justify the empirical model, the study moves from the structural form of a simple theoretical model, where the individual hotel operates in monopolistic competition. The demand function is the following:

$${\mathrm{Demand}{\text{:}} Q}_{t}^{\mathrm{D}}= a-b{p}_{t}+{cd}_{t},$$
(1)

where \({Q}_{t}^{\mathrm{D}}\) represents the quantity of rooms demanded, \(a\) is a constant, \(p\) is the price of a room, and \(b\) the slope of the demand function. \({d}_{t}\) represents time-varying variables (for example macroeconomic variables like GDP and exchange rate that can affect demand and supply) and \(c\) are the coefficients associated with these time-varying variables.

In equilibrium, ADR (the equilibrium price) and Occupancy (the equilibrium quantity normalized by the number of rooms available) can be expressed as follows:

$${\mathrm{ADR}}_{t}={\alpha }_{1}+{{f}_{1}\delta }_{t},$$
(2)
$${\mathrm{Occupancy}}_{t}={\alpha }_{2}+{{f}_{2}\delta }_{t},$$
(3)

where \({\alpha }_{1}\) and \({\alpha }_{2}\) are the two intercepts, while \({{f}_{1}\delta }_{t}\) and \({{f}_{2}\delta }_{t}\) are the time-varying macroeconomic variables and their coefficients.

Note that all the coefficients and variables appearing in the reduced form model are a combination of the ones initially used in the structural form model.

More precisely:

$${\alpha }_{1}=\frac{a}{2b} +\frac{m}{2} {{f}_{1}\delta }_{t}=\frac{c}{2b}{d}_{t},$$
$${\alpha }_{2}=\frac{a}{2}-\frac{bm}{2} {{f}_{2}\delta }_{t}=\frac{c}{2}{d}_{t},$$

where m is a constant marginal cost.

The empirical model

The analysis focuses on the equilibrium room price, ADR. As already shown, ADR can be expressed as a function of a set of constants, time-varying terms, including macroeconomic variables that shift hotel demand and supply, plus an error term.

$${\mathrm{ADR}}_{t}={\alpha }_{1}+{f}_{1}{\delta }_{t}+{\varepsilon }_{t}.$$
(4)

In the empirical model, we underline heterogeneity among classes and focus on the relationship between the dependent variable, ADR, and a specific macroeconomic time-varying variable: the exchange rate.Footnote 1 We emphasize the role of class heterogeneity because they define different sub-markets characterized by heterogeneous levels of price elasticities. As a result, the economic intuition suggests that the same exchange rate shock should have significantly different effects on the prices fixed in the different classes, mainly depending on firms’ market power and on consumers’ willingness to pay.

The reduced form equation can be rewritten as follows:

$${\mathrm{lnADR}}_{it}=\alpha +\beta {\mathrm{lnER}}_{t}+{\gamma }_{i}{\mathrm{lnER}}_{t}+\theta {\mu }_{t}+{\varepsilon }_{it}.$$

With i = Luxury, Upper Upscale, Upscale, Upper Midscale, Midscale and Economy.

The regression includes a constant, \(\alpha ,\) the exchange rate, ER, which is a time-varying variable that does not change with classes, the parameter \(\beta\) that describes the relationship between the dependent variable and the exchange rate, which is common to all the classes; \({\gamma }_{i}\) describes the specific relationship between the dependent variable and the exchange rate for each class i. In practice, we test if the slope of each regression is significantly different among classes or not. Consider for example the luxury class. The relationship (the slope) between the dependent variable and the exchange rate is defined by (\(\beta\) \(+ {\gamma }_{\mathrm{luxury}}\)). And that logic applies to all the other classes. In this specification, each class has a specific slope. Finally \({\mu }_{t}\)—which is a time fixed effect, i.e., a dummy variable common to all classes—is included. Being aware that other macroeconomic variables besides the exchange rate could affect our dependent variable, our modelling choice to address this issue was to use time (monthly) dummies instead of specific macroeconomic variables, which is a highly flexible and less arbitrary approach, as the literature suggests.

The most interesting feature of the model is represented by the interaction term \({\gamma }_{i}{\mathrm{lnER}}_{t}\). This is the model that allows us to answer the first question: do exchange rates appreciations affect hotel classes differently?

Data and descriptive statistics

The analysis included herein used monthly data for Switzerland over the period running from January 2000 to April 2021. The period considered is composed of 256 months (1353 observations). The data concerning hotel performance indicators have been supplied by Smith Travel Research (STR). ADR is denominated in local currency, the Swiss franc, and categorized by class, according to STR classification: Luxury, Upper Upscale, Upscale, Upper Midscale, Midscale and Economy. We have observations for the whole period for luxury and upper scale hotels, while the number of observations is smaller for midscale and economy hotels.

As shown in Table 1, the data suggest that during the period considered the average occupancy rate for the hotels in the sample was 62%, while average ADR was around CHF 220 CHF and average RevPAR, which combines occupancy and average price, was approximately CHF 133. Looking more in detail at the summary statistics by class, the highest average ADR and RevPAR are observed in luxury hotels immediately followed by Upper Upscale hotels. In terms of rooms sold, Midscale hotels register the highest occupancy ratio (around 66%) followed by Upper Upscale, Upscale and Economy hotels.Footnote 2

Table 1 Summary statistics
Full size table

This study’s main exogenous variable is the real exchange rate between Switzerland and its main trading partners. The real effective exchange rate index is calculated by the Swiss National Bank (SNB) as a weighted average of bilateral exchange rates, using a variable group of countries that is updated annually. The weighting method applied uses the International Monetary Fund (IMF) approach and takes into account export and import flows as well as so-called third-market effects. The index formula used is a chained Törnqvist index.Footnote 3 The real exchange rate (RER) index measures the real external value of the Swiss franc and is calculated as the nominal exchange rate (NER) index adjusted for price developments in Switzerland and abroad. RER is frequently used as an indicator for assessing the price competitiveness of an economy in the literature. A rise in the index value indicates an appreciation in the Swiss franc.

As Table 2 shows, the data suggest that the highest average values for the three performance indicators are observed at the very beginning of the sample in 2000Footnote 4 and 2001. Not surprisingly, 2020 looks as the worst year in terms of occupancy rate (at around 30%), ADR and RevPAR (respectively at around CHF 197 and CHF 60). It is also worth noting that in 2020, there is an increase in the real exchange rate (RER) index, indicating a strengthening of the Swiss franc.

Table 2 Annual averages of occupancy, ADR, RevPAR, NER index and RER index, between January 2000 and April 2021
Full size table

In the depths of the financial crisis, which started in 2007 in the US and then moved to Europe until 2012/2013, there is a contraction in occupancy rate, ADR and RevPAR, especially between 2009 and 2014. Additionally, between September 2011 and January 2015, the SNB intervened on the financial markets introducing an exchange rate floor to limit Swiss franc appreciations versus the euro. During that period, the real exchange rate stabilizes and slightly depreciates. Furthermore, during the SNB intervention, there is a strengthening of the occupancy rate, which goes from 63 to 66% and a slight weakening of the ADR, which moves from CHF 217 in 2011 to CHF 210/211 towards the end of the intervention. During this period, RevPAR slightly fluctuates and reaches CHF 135 in 2014. After the Swiss National Bank abandoned its exchange rate floor in January 2015, and the economic recovery in Europe and in the US, the Swiss franc appreciates markedly. In this period the occupancy rate, as well as ADR and RevPAR slightly weaken, to recover in the following years.

Table 3 displays pairwise correlations between variables. Intuitively there is a negative correlation between occupancy and average price (ADR). RevPAR displays a small and positive correlation with occupancy, while it is strongly correlated with ADR, which is a result that we also constantly observe in the analysis. RER is negatively correlated with the three performance indicators. More precisely, they show a weak correlation with ADR and RevPAR and a stronger one with Occupancy. The table also shows the nominal exchange rate (NER), which displays a very high correlation with the RER. Additionally, the correlation between performance indicators and RER is very similar to the correlation between performance indicators and NER, which explains our choice to include only the RER in our regressions.

Table 3 Pairwise correlations
Full size table

As already mentioned, in the analysis time dummy variables have been included. This allows to control for seasonality and, at the same time, avoids introducing other macroeconomic variables, still controlling for changes in the environment that took place over time.

Empirical analysis

After controlling for autocorrelation, i.e., the correlation between one variable and its previous values, we found that the dependent variable as well as the exchange rate (the independent variable) have a problem of autocorrelation. To find the optimal number of lags to be included in the regressions, Schwarz's Bayesian information criterion (SBIC) was used, the Akaike's information criterion (AIC), and the Hannan and Quinn information criterion (HQIC), which suggested to use one lag for the dependent as well as for the independent variable. Hence, the regressions have been modified as follows:

$${\mathrm{ln} \mathrm{ADR}}_{it }=\alpha +\rho {\mathrm{lnADR}}_{it-1}+\beta {\mathrm{lnER}}_{t-1}+{\gamma }_{i}{\mathrm{lnER}}_{t-1}+\theta {\mu }_{t}+{\varepsilon }_{it}.$$

With i = Luxury, Upper Upscale, Upscale, Upper Midscale, Midscale and Economy.

Given that in principle exchange rates are daily variables, it seems reasonable to imagine that they affect the hotel performance indicators with a one-month delay. Even though the goal of the paper is not to test for the causality between exchange rate and hotel performance, it is worth noting that the model nests a Granger test. If the parameters \(\beta\) and \(\gamma\) prove significant, de facto we are saying that the exchange rate Granger-causes the endogenous variable under study. We do not test for reverse causation, i.e., causality going from hotel performance indicators to exchange rate, because economically it is unreasonable to imagine that one sector in the economy is able to affect a macroeconomic variable like the exchange rate.

Room demand and supply are included in the specifications as control variables. The combination of demand and supply, as defined by STR, determine the occupancy rate, which affects ADR, as shown in our simple theoretical model. We preferred to include demand and supply separately instead of working directly with the occupancy rate to better evaluate the effects of the two components on the dependent variable.

Given that study sample includes the period of the pandemic, we ran the same regression over the whole period as well as over the two sub-periods before January 2020 and from January 2020 and April 2021, which we named “pre-COVID” and “COVID” period.

Table 4 shows that results depend on the period considered. If we consider the whole sample, including pre-COVID and COVID sub-periods, a Swiss franc appreciation significantly increases ADR in the luxury segment and reduces it in the economy segment. Nevertheless, it is worth underlying that during the COVID period, many of the traditional economic relationships broke up. In fact, the results suggest that during this period exchange rate appreciations, room demand and supply do not produce any effects on ADR. When only the pre-COVID period is considered, a Swiss franc appreciation negatively affects ADR for midscale and economy hotels.

Table 4 Basic Model: relationship between ADR and RER, by class over three different sub-periods
Full size table

Overall, midscale and economy hotels react to exchange rate appreciations in a quite sophisticated way. Knowing that their guests are relatively highly sensitive to price increases, midscale and economy hotels may reduce their ADR when it is objectively necessary, which is to say when they are losing competitive edge abroad.

Instrumental variable analysis

To control for any endogeneity problems,Footnote 5 we use instrumental variables (IV) to approximate the RER. Following the literature, we use Swiss international official reserves (IR) to predict exchange rate movements. IR pass all the weak instruments tests, indicating that they are good instruments for RER. The correlation between RER and IR is quite high (0.83). Additionally, we conduct a weak instrument test to demonstrate that IR is a good and strong instrument to explain our endogenous variable, RER (Table 5).

Table 5 Weak-instruments tests
Full size table

In Table 6 we show the results of our two-stage-least-squares regressions (2SLS) done over the three periods, as already done in Table 4. We controlled for heteroskedasticity-robust standard errors. The results suggest that a Swiss franc appreciation generates significant effects only if we consider the pre-COVID period. A 1% appreciation reduces ADR in all the classes, with the only exception of the luxury class. Data suggest that the ADR reduction is stronger in lower classes. Intuitively, classes characterized by a more price-elastic demand reduce their prices in a relatively stronger way to compensate for the currency appreciation.

Table 6 IV regressions over three different periods
Full size table

Exchange rate fluctuations: monetary policy

As mentioned in the introduction, one of the goals of this study is to study the effects of the exchange rate fluctuations on the hospitality industry. In Switzerland, the minimum exchange rate of CHF 1.20 per euro between 6 September 2011 and 15 January 2015 was the major monetary policy instrument that the SNB used to limit the appreciation of the Swiss franc. In this last section of the paper, we conduct a Chow test in order to verify whether the SNB’s intervention between September 2011 and January 2015 generated a structural break in the parameters of the model. A Chow test is a test of whether the true coefficients in two linear regressions on different data sets are equal (Chow 1960). This approach is typically used in the field of econometrics with time series data to determine if there is a structural break in the data at some point. By structural break we mean a change in the fundamental relationship between variables. Did the awareness that the SNB was protecting the Swiss franc from extreme appreciations, change the relationship between exchange rate movements and hotel performance? In other terms, during the SNB intervention, did hotels change their pricing strategies, knowing that the risk of franc appreciation was relatively low? Our results confirm the existence of a structural break in the relationship between the exchange rate and ADR in the pre-COVID period.

Table 7 shows the results of OLS, IV, IV with SNB intervention and IV without SNB intervention regressions, only considering the period before COVID. When the exchange rate is free to fluctuate (non-intervention period), exchange rate appreciations lower the ADR in all the classes, with the only exception of the luxury class. During the period of the SNB intervention, instead, paradoxically we observe an increase in ADR in all the classes but the Economy class. This increase is stronger the higher the hotel class. A possible interpretation of this result is that the SNB intervention created an expectation of stability in the hotel industry that reduced the responsiveness of hoteliers to Swiss franc appreciations. Surprisingly enough, during the central bank intervention, hoteliers not only did not react in the expected way to Swiss franc appreciations, but actually took advantage of those to increase their own prices.

Table 7 Table containing OLS, IV (2SLS), IV with SNB intervention (2SLS_SNB) and IV without SNB intervention (2SLS_noSNB) in the pre-COVID period
Full size table

Conclusion

Several scholars have studied the role of the exchange rate on international tourism demand, while the topic has been less explored in the hospitality literature. In a break with extant literature, this study focuses on ADR, which represent a more strategic variable controlled by hotel managers. Furthermore, we categorize hotels by class to verify whether the same shock produces differentiated effects in the industry, as the economic intuition would suggest. We study the special case of Switzerland, which experienced a sequence of important currency appreciations as well as an explicit and strong intervention of the Central Bank to limit them.

This study finds the presence of a structural break in the economy, due to the central bank intervention on the exchange rate market. During the period when the exchange rate floor was in place, hotels’ performance was not affected by exchange rate appreciations. On the contrary, when the exchange rate was free to fluctuate, hoteliers promptly reacted to currency appreciations. The awareness that the SNB was protecting the Swiss franc from strong appreciations, induced hoteliers not to reduce ADR, which stands in contrast to their behavior when the exchange rate has been free to fluctuate. Interestingly enough, this finding suggests that hoteliers pay close attention to economic variables' fluctuations as well as to economic policies. This implies that the policymaker, using her policy instruments, can affect hoteliers’ incentives and induce them to adopt optimal behaviors. Further, given that the last period of observations includes the first wave of COVID 19, this study verifies the stability of the study’s results over the COVID (i.e., between January 2020 and April 2021) and non-COVID period (i.e., before January 2020) suggesting the existence of a structural break in the economy between these two periods. In other words, during the COVID period the traditional relationships between economic variables broke.

Discussion

This study’s main results suggest that the sample of hotels act in a quite sophisticated way in terms of pricing decisions. Hoteliers belonging to different classes on average react differently to real exchange rate appreciations. In the pre pandemic period we observe that currency appreciations induce all classes, but the luxury one, to reduce their prices. Nevertheless, during the pandemic, strong price reductions have been observed in the luxury sector. These reductions were most likely not related to currency appreciations, but rather to other strategic considerations.

We find that hotel managers’ expectations concerning exchange rate fluctuations have a huge impact on the performance of hotels. We find a structural break in the economy, indicating that prices are negatively affected by exchange rate appreciations during periods of high uncertainty. On the other hand, during the intervention of the SNB to limit the Swiss franc’s appreciations, hoteliers felt less exposed to currency risk and reduced their tendency to lower prices despite the Swiss franc’s appreciations.

The COVID 19 pandemic has triggered a massive shock in the tourism industry and the tourism and hospitality sector continue to be one of the sectors hardest hit by the COVID 19 pandemic. While many countries started to develop recovery measures to support the tourism and hospitality sector, policy makers already should consider the longer-term implications of the COVID 19 pandemic and prepare for actions to support the sustainable recovery of tourism. Therefore, the findings of our study are especially interesting in the current period, which has been characterized by strong policy interventions aimed at containing the diffusion of COVID 19. The findings of this research suggest that government’s policies could be used to reduce economic uncertainty. This study is valuable in the sense that it allows us to gain awareness on how hoteliers’ pricing decisions are related to exchange rate fluctuations as well as to policy interventions to manage economic uncertainty. On the one hand this analysis can help the individual hotelier gain awareness and understand how on average the market is behaving during policy interventions; on the other hand, it can shed light on the potential for policymaking.

Nevertheless, the paper presents some limits that should be explored in future research. First, the size of the sample is relatively small, and hotels are not fairly represented across classes. STR data are collected on a voluntary basis which that STR can provide data only for a group of hotels that decide to participate. For questions of privacy, STR does not provide disaggregated data, but only an average by group, under the condition to have at least five observations for each group of hotels. Overall, the data tend to be biased towards upper scale classes. Our database represents around 35% of the total number of upper scale hotels in Switzerland (luxury, upper upscale and upscale), while only 7% of the total number of lower scale hotels in Switzerland (upper midscale, midscale and economy). In the analysis, we focus our attention on the specific case of Switzerland. From a certain point of view, Switzerland is a very interesting case, because it is a small open economy, extremely focused on the international trade and on the exchange rate policy. At the same time, we recognize that it could be interesting to extend the analysis to other countries, which could allow at the same time to verify the solidity of our findings and reduce the aforementioned sample issue.

We encourage future study to extend the analysis in several directions. For example, one interesting extension of the study would consist in including additional regressors besides the exchange rate, like Airbnb’s presence on the market and oil prices, and some features that specifically characterize the hotels considered (size, location, business model), which is commonplace in the hedonic pricing literature. One other possibility, especially in the light of the most recent events, would be to include other policy interventions and evaluate if and how they have affected the hospitality industry performance.