Cash may serve as an anchor of confidence during times of crisis because of its status as default-proof central bank money and the fact that it can be used in emergency and crisis situations without the need for technical infrastructure. With that in mind, this article describes international demand for banknotes during a selection of crises. The case studies are based on time series relating to banknotes in circulation in a selection of 16 countries and currency areas across all continents in the period from 1990 to 2022.

The information available on the growth in international demand for banknotes initially confirms the perception of increasing global demand for cash, despite the declining share of cash payments in many countries (paradox of banknotes). As such, the value of banknotes in circulation is rising for most of the currencies examined in this article, and is generally even increasing in relation to nominal gross domestic product (GDP). Notable exceptions to this general rule are Norway and Sweden, where the value of banknotes in circulation has fallen sharply in relation to nominal GDP. This is attributed in the literature to a declining demand for banknotes as a store of value.

In addition, case studies are used to illustrate the developments in banknote circulation during a number of different types of crises. The crises covered are the Asian financial crisis (Asian crisis) from 1997 to 1999, the millennium bug (Y2K), the terrorist attacks on 11 September 2001 (9/11), the global financial crisis of 2008, the COVID−19 pandemic and the war in Ukraine, which began with the Russian invasion on 24 February 2022. This article makes a distinction between transaction motives, store-of-value motives and precautionary motives as factors driving the demand for banknotes. The different reasons for the demand for banknotes are taken into account by analysing the circulation of small and large-denomination banknotes separately.

The analysis is descriptive in nature and is intended primarily to generate hypotheses. Consequently, responses in demand for banknotes may be temporary and localised, and in some cases, the value of banknotes in circulation may fall as a result of a crisis. Viewed overall, however, the case studies suggest that crises are likely to have contributed to the dynamic growth in demand for banknotes in the countries analysed. This applies both to small and large banknote denominations. The increased demand for banknotes as a result of a crisis is probably due mainly to precautionary motives.

Cash is central bank money and is therefore immune to any default risk. As a physical means of payment, it can also be used in emergencies and crises without the need for technical infrastructure, thus helping to ensure the continuity of payments in the event of technical disruptions or natural disasters. The role that cash plays in crises is frequently discussed with these characteristics in mind. 1 In fact, a marked increase in the value of euro banknotes in circulation was observed during the global financial crisis and the COVID−19 pandemic, for instance. 2

Against this backdrop, the Bundesbank has conducted selected case studies to investigate how demand for banknotes developed during crises. 3 The analyses are based on time series on the development of banknotes in circulation from 16 selected currencies in the period from1990 to 2022. 4 The demand for banknotes over time and in an international comparison is of fundamental interest; it is therefore presented in detail in Section 2 of the article, which focuses initially on general considerations. Meanwhile, Section 3 examines the demand for banknotes in times of crisis using a graphical analysis of a selection of countries and crises, with the focus being on six crises that occurred between 1997 and 2022. 5 The analysis differentiates between small and large-denomination banknotes to take account of their different uses. 6 For instance, small denominations are generally used to pay for goods and services (holdings for transaction purposes or transaction motives). 7 By contrast, large denominations are typically in demand for the purpose of storing value in the context of portfolio allocation (store-of-value motives). 8 The relative return, relative risk and relative liquidity of the asset class are important when selecting assets for portfolio allocation. For example, the demand for banknotes as a store of value decreases with rising opportunity costs of holding cash, i.e. the interest rate on close substitutes for cash such as sight deposits at banks. 9 To understand the demand for banknotes during crises, it is helpful to also consider precautionary motives, which can result in demand for cash both for transaction purposes and also as a store of value. In principle, precautionary cash holdings serve to provide for unplanned or unexpected expenses. Precautionary motives are also significant during times of crisis and subsequently lead to demand for banknotes as a precautionary measure in response to (major) uncertainty. This includes safeguarding liquidity in the form of small-denomination banknotes for transactions in the event of potential cash supply bottlenecks, as well as storing value by means of large-denomination banknotes in the event of fears of a banking collapse, stock market crash or political unrest. 10

The purpose of the explorative analysis is to describe the development of the demand for banknotes in a selection of crises and to derive hypotheses about the relationship between the demand for banknotes and crises. However, the descriptive approach adopted is, as such, subject to a number of significant limitations. This means that it is not suited to testing hypotheses or determining causal effects. In particular, the control for the effects of other factors, which is key in empirical work, and statements on statistically significant effects and their magnitude are reserved for inferential statistical approaches, which will be pursued at a later stage.

The 16 countries selected are spread across all continents. Besides the most widely used currencies (US dollar, euro, Swiss franc, pound sterling and yen), major South American and Asian emerging market economies and their currencies are also included. 11 The charts show the ratio of banknotes in circulation to nominal GDP (banknote ratio) for these countries, grouped by continent, as a measure of banknote utilisation. The trajectories indicate both the similarities and specificities of individual countries.

The banknote ratio in Morocco more than doubled during the period under investigation. After increasing temporarily to 29 % in 2020 as a result of the COVID−19 pandemic, the ratio stood at just under 27 % in 2022. With the exception of Japan, this is therefore significantly higher than the corresponding ratio for developed advanced economies. It is likely that the rapid rise in the banknote ratio in Morocco was driven by the demand for large denominations as a store of value. Saidi et al. (2023) estimate that the hoarded share of the two largest banknote denominations increased from around 20 % at the start of the millennium to between 60 % and 80 % in 2021. 12

Of the four Asian countries analysed, the ratio of banknotes in circulation to GDP as a measure of cash use is relatively stable in Indonesia. In Türkiye, the banknote ratio rose between 2005 and 2016 before dropping back to its starting level by 2022. 13 At the same time, in terms of sales, the share of cash transactions at the point of sale remained essentially unchanged in the period between 2012 and 2019, i.e. cash spending increased in proportion to consumer spending. 14 The banknote ratio in Japan shows a marked upward trend and stood at an exceptionally high level of 22.5 % in 2022. Main drivers of this development are likely to have been low interest rates, falling prices and concerns about bank stability after the real estate market bubble burst in the early 1990s. 15 Fujiki and Nakashima (2019) explain the increase in the ratio of currency in circulation to GDP in Japan as being due to strong hoarding demand, despite advances in the area of cashless payments. They add that this offset the negative effect of substituting cash with cashless payment instruments for day-to-day payments. 16 In India, at just under 9 %, the ratio of banknotes in circulation to GDP in 1990 was on a par with Japan. It has mostly been on the rise since then, reaching just under 13 % in 2022. The temporary dip in 2016 came on the back of demonetisation on 8 November that year, when the Indian government took the surprise step of withdrawing the two largest rupee denominations (500 and 1,000 Indian rupee banknotes) in the Mahatma Gandhi series from circulation, which came into effect at midnight on the same day. 17 By taking this measure, the government sought to seize undeclared income and increase the tax base. 18 Demonetisation triggered a decline in cash transactions at the point of sale. In terms of sales, this share fell by around 30 percentage points between 2012 and 2019, with the three years following demonetisation accounting for 17 percentage points alone. 19

Overall, the banknote ratio in Australia moved sideways with a number of fluctuations. Of particular note are the temporary crisis-related increases (peaks) in 1999 (Y2K), 2001 (9/11), 2008 (global financial crisis) and 2020 (COVID−19 pandemic). 20 Having reached an all-time high of just under 5 % in 2020, the ratio of banknotes in circulation to GDP remained at a moderately elevated level by historical standards, standing at 4.2 % in 2022. 21 The demand for banknotes for transaction purposes is falling. The share of cash payments in the total volume of payments (by value) in the retail sector dropped from 38 % in 2007 to 11 % in 2019. 22 Apart from the holdings maintained as a store of value, currency holdings outside Australia are likely to have played a role in the increase in total circulation. For instance, there is evidence that the banknote denomination with the highest face value (A$100) is in demand outside Australia. 23

The countries selected in Europe can be divided into three groups. The first group comprises the two Scandinavian countries Norway and Sweden. Banknotes in circulation have fallen significantly in Sweden since 2010 and in Norway since 2016. The banknote ratio in both countries has dropped significantly since the early 1990s, and most recently stood at a mere 1 %. Despite the growing significance of digital means of payment, Claussen et al. (2023) attribute this deviation from the paradox of banknotes 24 primarily to declining demand for cash as a store of value and not to a drop in demand for cash for transaction purposes. 25 They put the decline in the circulation of large denominations (in demand as a store of value) in Sweden down to the general public’s positive experience of the government’s crisis management in the country. According to Claussen et al., this has reduced incentives to hold large banknotes as a means of hedging against uncertainty. The second group comprises the United Kingdom. The share of cash payments in total payment transactions there fell from just over 80 % in 1990 to just under 20 % in 2020. 26 Nonetheless, the ratio of banknotes in circulation to GDP has been trending upwards in the period under investigation and is now at a historically high level. This points to higher demand for banknotes as a store of value, which can probably be explained by historically low interest rates and the associated low opportunity costs of holding money, for instance. 27 Demand for the largest denomination (50 pound banknote) from other countries is also likely to have played a role in the growing banknote ratio. 28 The third group comprises the euro area, Poland and Switzerland. This group has seen the banknote ratio rise sharply, but this trend has only been observed in Switzerland since the global financial crisis in 2008. Standing at 11.8 % in the euro area and Poland, and 10.6 % in Switzerland in 2022, this ratio was high by international standards. In the euro area, strong foreign demand for euro banknotes was a contributing factor. It is estimated that between 30 % and 50 % of the (net) value of euro banknotes issued by the Eurosystem are in circulation abroad. 29 Foreign demand is also likely to play a role in the case of Swiss franc banknotes. 30

In the three North American countries analysed, the banknote ratios have been rising significantly both in Mexico since 1997, and in the United States since the global financial crisis in 2008, reaching just under 9 % in both countries in 2022. The reason for this comparatively elevated figure in Mexico may be the fact that cash plays an important role in the country, both as a store of value and for payments. 31 During the global financial crisis, the banknote ratio rose steadily by just under 1 percentage point before flattening out in the period up to 2011. Antón et al. (2021) explain that the additional increase observed from 2013 onwards was due to a series of measures introduced that year to improve tax enforcement in Mexico, such as making electronic invoices mandatory and increasing the number of taxpayers. They argue that intensifying tax enforcement in this way may lead to attempted tax avoidance, resulting in an increase in the cash in circulation. The ratio of banknotes in circulation to GDP in Mexico was boosted further during the COVID−19 pandemic, when it rose by 2 percentage points. The increasing banknote ratio in the United States reflects the ongoing growth in overall demand for cash in the country. The reason for this is presumably that cash is being used increasingly as a store of value while being used less as a means of payment. 32 The trend in the banknote ratio in the United States is moving in line with the circulation of US dollar banknotes abroad. For example, the different estimation methods in Judson (2017) point to strong international demand in the 1990s, a decline in the early 2000s and a sustained resurgence since the collapse of the bank Lehman Brothers. According to their estimates, the share of US dollar banknotes in circulation abroad in the total circulation of US dollar banknotes rose in the period leading up to the global financial crisis and has been at around 50 % since then, albeit showing a slight downward trend. In Canada, the banknote ratio ranged between 3 % and 4 % until the start of the COVID−19 pandemic. It has been on a slightly upward trend since the financial crisis. Thus, in contrast to a comparable country such as Sweden, the diminishing role played by cash reserved for transaction purposes in Canada has not yet led to a sustained reduction in demand for banknotes. 33 According to Claussen et al. (2023), this can be explained by the ongoing rise in large-denomination banknotes in circulation in Canada, which may have been driven by foreign demand in recent years. As in the United States, crises also caused the banknote ratio in Canada to rise temporarily in 1999 (Y2K) and 2020 (COVID−19 pandemic) (peaks).

In Brazil, the banknote ratio rose from just under 1.9 % in 1995 to 3.9 % in 2009. A temporary increase in 2002 was likely due to political uncertainty in connection with the presidential elections in October 2002, which led to a steep rise in inflation and the devaluation of the Brazilian real. Excluding the temporary surge during the COVID−19 pandemic, the ratio of banknotes in circulation to GDP fell slightly following the global financial crisis. The sharp decline in the use of cash at the point of sale is likely to have been a contributing factor. In terms of sales, it fell by around 40 percentage points between 2013 and 2019. 34 Nonetheless, cash continues to play an important role in payment transactions. For example, in the latest survey conducted by Banco Central do Brasil (as at 2018), around 60 % of respondents cited cash as their most frequently used means of payment. 35

In spite of the falling share of cash payments at the point of sale, banknotes in circulation in terms of value are on an upward trend in the majority of the countries presented here and are even increasing in relation to gross domestic product (paradox of banknotes). This means that demand for the physical form of central bank money, i.e. cash, remains high. In the discussion surrounding the underlying factors, it has already become clear that crises can stimulate demand for banknotes. For instance, since the global financial crisis, the number of banknotes in circulation has been growing faster than before in many advanced economies. The next section takes a detailed look at demand for banknotes during times of crisis.

Economic studies point to a statistically significant link between crises and banknotes in circulation. 36 For the graphical analyses in this article, crises are first identified in which there was a noticeable change in banknote circulation in the selected countries. In line with Rösl and Seitz (2024), the crises can be assigned to the following categories: uncertainty with respect to digital infrastructure, crises of confidence in the financial system, natural disasters, political uncertainty, inflation crises and exchange rate crises. The results for some selected crises are presented below. 37 In general, only countries where there was a significant change in the value of banknotes in circulation during a crisis are included. A uniform, objective criterion for defining what is meant by a “significant” change in this context will be established in this article. In doing so, it is important to avoid setting a threshold that is too low or too high. Setting the value too low increases the risk of including cases in which the fluctuation in the circulation of banknotes during a crisis was merely random. By contrast, setting the threshold too high could result in the exclusion of countries where there were actual changes in banknote circulation due to a crisis. The definition chosen here is based on unusually large changes in the annual growth rate of banknotes in circulation. “Large change” in this context means “amplitude of growth” of at least 5 percentage points. Amplitude of growth refers to the maximum increase (or decrease) in the annual growth rate 38 during a crisis, i.e. the difference between the highest (or lowest) value during the crisis and the value recorded immediately before the crisis began. 39 The “response period” refers to the period in which there is an unusually sharp rise or fall in banknote circulation in a country during a crisis. The response period begins with the month in which the annual growth rate of banknotes in circulation in value terms deviates by at least one percentage point from its level of the previous month, and ends in the month in which the annual growth rate has stabilised at a new level (the pre-crisis level, for instance). The response period varies from country to country. The “aggregated” response periods shown in the charts comprise the individual response periods of the countries depicted and therefore indicate the period in which banknote circulation was unusual in at least one of these countries. In order to take account of the different motives behind the demand for banknotes, the annual growth rates of both small and large-denomination banknotes are examined.

This approach is used to describe the development of banknote demand in several selected crises. Hypotheses about the relationship between banknote demand and crises can be drawn from the cases presented here, but it is not the aim of this article to make any statements about systematic relationships. In particular, it is important to note that although the focus on case studies with clear developments in the demand for banknotes increases the clarity of the presentation, this should not be taken to mean that crises always have a strong impact on banknote demand. Instead, the results show that the global crises outlined in this article do not necessarily have visible significant effects in all the scenarios analysed.

The demand for banknotes during crises due to uncertainty with respect to digital infrastructure is described on the basis of the year 2000 problem (Y2K, millennium bug). 40 This involved making changes in the way years were represented in computer programs in order to display years in the new millennium correctly. There were fears that because the year had been shortened to two digits in program codes, computer programs would not be able to operate correctly after the turn of the millennium, with various possible negative impacts, such as public utilities being switched off or the smooth operation of payment systems and cash withdrawals from ATMs being impaired. 41 Ultimately, however, there are no known major failures associated with Y2K.

The annual growth rates of large-denomination banknotes in circulation rose significantly during YK2 in 10 of the 16 countries analysed. This increase was likely driven by precaution, as investors may have expanded their banknote holdings to hedge against growing uncertainty about still being able to access their digitally stored savings at short notice. 42 Mexico was the first country to be affected, in September 1999, and the crisis lasted the longest in Sweden (until April 2000). The greatest impact of Y2K was felt in Indonesia, Mexico and Türkiye, which are also the countries with the highest annual growth rates. For example, the annual growth rate's amplitude of growth in Mexico (in the period from September 1999 to January 2001) was just under 45 percentage points, and in December 1999, the annual growth rate in Indonesia rose by around 33 percentage points. In all countries, the biggest increase was seen in December 1999. After the turn of the millennium, the one-off period of uncertainty associated with the millennium bug dissipated. 43

In ten countries, Y2K brought about a sharp increase in the annual growth rate of small-denomination banknotes as well. This was probably also due to precautionary measures, which led to an increase in transaction balances in view of growing uncertainty regarding the availability of the infrastructure for digital payments. 44 As with large denominations, the strongest effects for small denominations were observed in all countries in December 1999. The highest amplitude of growth was seen in Indonesia (around 41 percentage points) and Canada (around 24 percentage points). 45 Overall, Y2K had about the same impact on small banknote denominations as on large denominations. Y2K is a prime example of the importance that users attach to cash in the face of impending technical crises. Cash is the only means of payment that can be used independently of digital infrastructure.

The demand for banknotes during crises of confidence in the financial system is analysed on the basis of the 2008 global financial crisis. The crisis began in 2007 in the US subprime market and reached its peak with the collapse of the major bank Lehman Brothers on 15 September 2008. This brought the interbank market to a standstill worldwide. Investors feared that their bank deposits were not secure. This was followed by a global recession and the launch of bank rescue packages, which triggered sovereign debt crises due to rising risk premia on government bonds.

The 2008 global financial crisis led primarily to an increase in demand for large-denomination banknotes, which are particularly suitable for store-of-value purposes. Given the extensive crises of confidence, the increase in precautionary cash holdings is likely to have played an important role in this regard. 46 Annual growth rates of large-denomination banknotes in circulation rose sharply in half of the 16 countries analysed here. The first effects were seen in Indonesia, Switzerland and Türkiye in September 2008 and in the other countries in October 2008. The effects lasted longest in the United States (from October 2008 to December 2009), the country where the crisis had originated. The two countries with the highest annual growth rates, Indonesia and Türkiye, stand out from the other six countries as the effects of the crisis lasted for the shortest period of time there: in Indonesia they were seen only in September 2008 and in Türkiye from September 2008 to March 2009. In the other countries, the annual growth rate responded to the crisis until at least September 2009. In addition, the annual growth rate was strongest in Türkiye, with an amplitude of growth of roughly 35 percentage points, and in Indonesia, with an amplitude of growth of around 14 percentage points. Ultimately, the increase in the circulation of large-denomination banknotes (in national currency) was only temporary in Indonesia and Türkiye, while it remained permanently above its pre-crisis level in Australia, the euro area, Mexico, Switzerland, the United States and the United Kingdom. The annual growth rate profile in the euro area corresponds to the level of a one-time, permanent increase in the circulation of large-denomination banknotes (in euro) in October 2008. This also roughly applies to Switzerland. 47 In principle, it is not unusual for the effects of crises to persist. 48

Only Australia, Indonesia, Mexico and Türkiye saw a significant increase in the annual growth rate of small-denomination banknotes in circulation during the global financial crisis. In the light of the crisis of confidence in the banking system, the main reason for this was probably to secure liquidity for transactions as a precautionary measure. The increase in Mexico – an amplitude of growth of around 8 percentage points – was weaker than in the other three countries, which had an amplitude of growth of around 18 percentage points. In Indonesia and Türkiye, the annual growth rate only responded in September 2009, with the circulation of small-denomination banknotes in these two countries (in Indonesian rupiah and Turkish lira, respectively) increasing only temporarily in that month. In October, it returned to its pre-crisis level. The global financial crisis had a much longer impact on the annual growth rate of small-denomination banknotes in circulation in Australia (from October 2008 to September 2009) and Mexico (from October 2008 to June 2009). Australia was the only country to see a sustained increase in small-denomination banknotes in circulation (in Australian dollars) as a result of the global financial crisis. 49

The demand for banknotes during natural disasters is illustrated using the example of the COVID−19 pandemic, which broke out at the start of 2020. 50 The onset of the pandemic was rapidly followed by comprehensive quarantine measures, lockdowns and a worldwide recession. The repercussions for banknote circulation are described in Deutsche Bundesbank (2022). This shows that banknote circulation in value terms grew strongly during the pandemic, even though the online trade gained market share and consumers increasingly paid using cashless instruments in shops due to their – ultimately unfounded 51  – fear that the virus could be transmitted via banknotes. 52 The accumulation of banknotes during the first few months of the pandemic is therefore likely to be attributable to precautionary behaviour with regard to potential logistical restrictions on the supply of cash. The increases declined in the ensuing period, though banknote demand remained at an elevated level internationally. The sharp rise in the value of large-denomination banknotes in circulation was not just a case of a general increase in cash holdings as a store of value, a phenomenon typical of crises (precautionary motive). In this case, the storing of cash is also likely to have been partly caused by a lack of spending opportunities.

In the majority of countries analysed, the COVID−19 pandemic had a clear impact on the annual growth rates of banknotes in circulation. Small-denomination banknotes responded to this crisis in a similar number of countries and – when measured in terms of the amplitude of growth – roughly as strongly as large-denomination banknotes did. The growth of the former was presumably driven by precautionary reasons, while in the case of the latter, store-of-value motives probably loomed large as well. 53 In most countries, the initial impact on the annual growth rate of large-denomination banknotes in circulation became apparent in March 2020. The crisis proved most prolonged in Canada (lasting until August 2022). The annual growth rate of large-denomination banknotes in circulation was greatest in Türkiye (amplitude of growth of around 66 percentage points), followed by Brazil (amplitude of growth of approximately 42 percentage points). Brazil was also confronted with a strong rise in inflation and therefore also in the opportunity costs of holding cash. 54 Unlike in the other crises, the annual growth rate of banknotes in circulation also fell (strongly) in a few countries, notably Morocco and Norway. That said, the value of large-denomination banknotes in circulation in Morocco (in Moroccan dirham) continued to rise, albeit less strongly than before the pandemic. By contrast, in Norway the pandemic had the effect of accelerating a downward trend in the circulation of large-denomination banknotes (in Norwegian krone) that dated back to 2007. In Australia, Brazil, the euro area, Indonesia, Japan, Canada, Mexico, Türkiye and the United States, the value of large-denomination banknotes in circulation (in national currency) increased in a lasting way. 55

The COVID−19 pandemic had the most prolonged impact on the annual growth rate of small-denomination banknotes in circulation in the United Kingdom (May 2020 to April 2022). This annual growth rate exhibited by far the greatest amplitude of growth in Türkiye (around 57 percentage points) and Brazil (around 40 percentage points) on account of the COVID−19 pandemic. Unlike in other crises, there were also countries – namely Japan and Morocco – where this annual growth rate declined. 56 By contrast, in Australia, Brazil, the euro area, Canada, Mexico, Norway, Türkiye, the United States and the United Kingdom, the additional demand resulting from precautionary motives more than offset the decline in cash payments at point of sale (transaction motive) and, with the sole exception of Brazil, the value of small-denomination banknotes in circulation increased in these countries permanently. 57

The impact of crises on the demand for banknotes due to political uncertainty is illustrated using the examples of the terrorist attacks of 11 September 2001 and – more recently – the war in Ukraine.

The terrorist attacks of 11 September 2001 (9/11) had far-reaching political and economic consequences. On this day, Islamist terrorists brought catastrophe to New York by flying two hijacked aeroplanes into the Twin Towers of the World Trade Center. A third aircraft crashed into the Pentagon, the headquarters of the US Department of Defense. A total of around 3,000 people were killed by these terrorist attacks.

Only in Australia did the annual growth rate of large-denomination banknotes in circulation respond significantly to the events of 9/11. Here, it exhibited an amplitude of growth of around 6 percentage points, rising between October 2001 and March 2002, before then falling back to its pre-crisis level by September 2002. The rise in the circulation of these banknotes (in Australian dollars) was a lasting one. It seems likely that the precautionary holdings of these banknotes were stocked up on account of the political uncertainty that followed the terrorist attacks of 11 September 2001. In contrast to the other charts, it seems appropriate here to show countries such as the United States and Sweden, where the crisis did affect the value of banknotes in circulation, but – as per the definition applied here – had no significant impact. Following the terrorist attacks of 11 September 2001, the annual growth rate of large-denomination banknotes in circulation in Sweden (September 2001 to November 2001) and the United States (October 2001 to October 2002) initially increased with an amplitude of growth of just under 4 percentage points, before subsequently falling back to its pre-crisis level. The corresponding rise in the value of large-denomination banknotes in circulation (in national currency) was probably not a lasting one in Sweden, whereas in the United States – the target country of the terrorist attacks of 9/11 – it presumably was. 58

The annual growth rate of small-denomination banknotes in circulation rose significantly in Australia and Sweden. As was the case for large-denomination banknotes, this rise was probably attributable to precautionary motives. The development in both countries was similar to that of the annual growth rate of large-denomination banknotes in circulation, but the amplitude of growth was significantly greater, namely 31 percentage points in Australia and just under 10 percentage points in Sweden. As with large-denomination banknotes, only in Australia – and not in Sweden – did the circulation of small-denomination banknotes (in national currency) remain permanently above its pre-crisis level. 59

In summary, it may be asserted that the additional demand for banknotes as a result of the terrorist attacks of 11 September 2001 mainly affected small denominations. Given the background of political uncertainty, this was presumably attributable to precautionary motives. Relevant effects were only seen in Australia and Sweden, with the rise in the circulation of small denominations proving to be permanent in Australia only. In the United States, the target country of the terrorist attacks, banknote circulation did not change significantly.

The war in Ukraine started with the Russian invasion of Ukraine on 24 February 2022 and continues to this day. This is the largest armed conflict in Europe since the Second World War. The West imposed severe sanctions on Russia. Among the serious economic consequences of this war and the associated sanctions for the global economy, price increases for energy and food – in some cases significant ones – were the most evident.

The annual growth rate of large-denomination banknotes in circulation rose significantly in Indonesia, Morocco, Norway and Türkiye as the war in Ukraine unfolded. This is likely to have been driven by precautionary motives in the face of economic and political uncertainty triggered by this conflict. In the case of Norway and Türkiye, geographical proximity will also have played a role. In Morocco, as in other countries of the Maghreb region, this war led to sharp price rises and a shortage of essential foods. 60 The crisis-related increase in demand for large-denomination banknotes in Morocco could therefore be attributable to transaction motives as well as precautionary motives. The first effects of the war in Ukraine on the annual growth rate of banknotes in circulation manifested themselves in Indonesia (from February 2022 to April 2022). The amplitudes of growth of large-denomination banknotes in circulation ranged from 5 percentage points in Norway to just over 19 percentage points in Türkiye. In Indonesia and Türkiye, the circulation of large-denomination banknotes (in national currency) persisted at an elevated level even after the above-mentioned periods. 61

The war in Ukraine also caused a rise in the annual growth rate of small-domination banknotes in circulation in Indonesia, Morocco and Türkiye. As was the case for large denominations, this additional demand was probably attributable to people stocking up their cash reserves for precautionary reasons. 62 Norway was not (significantly) affected, but its neighbour Sweden was. In Indonesia, Morocco (from April 2022 to December 2022), and Türkiye (in April 2022), the crisis had an impact over the same period as it did for large-denomination banknotes, while for Sweden this impact lasted from March to December 2022. Amplitudes of growth ranged from just under 10 percentage points in Morocco and Türkiye to around 15 percentage points in Indonesia. From June 2022 onwards, the circulation of small-denomination banknotes in Indonesia (in Indonesian rupiah) reverted back to its pre-crisis level. In Türkiye, this circulation (in Turkish lira) remained above the pre-crisis level. 63

The visible repercussions appear to have been the same for small and large-denomination banknotes. The additional demand for large-denomination banknotes in Norway and small-denomination banknotes in Sweden is striking given that both these countries have made considerable progress along the path to a cashless society. The exposed geographical situation of these two countries no doubt proved a contributory factor. 64

The impact of exchange rate crises on banknotes in circulation is illustrated using the example of the Asian crisis, 65 which started off as an exchange rate crisis and then impacted on other areas of the economy. 66 First of all this crisis is summarised using the example of Indonesia, as it was here that the repercussions were the most severe. 67 The crisis first manifested itself in Thailand in March 1997 before spreading to other countries such as Indonesia. In July 1997, the Indonesian rupiah came under pressure. In August 1997, the peg to the US dollar was removed, prompting the Indonesian rupiah to slump in value. Following multiple market interventions by Bank Indonesia, which also tightened monetary policy, the crisis then spread to the banking sector. In October 1997, 16 banks were closed. After a certain time lag, the real economy was also affected as a result of banks restricting lending and lending rates rising dramatically. The economic recession in turn provoked social unrest and a loss of confidence in the government. In November 1997, the IMF approved a Standby Credit Facility arrangement amounting to US$10.1 billion. In January 1998, the Indonesian Bank Restructuring Agency (IBRA) was founded in order to relieve the financial sector of its high debt burden. In July 1999, the Indonesian government announced a plan to liquidate the IBRA banks.

During the Asian crisis, the annual growth rates of large-denomination banknotes in circulation increased noticeably in Indonesia, Japan and Türkiye. As was the case in the global financial crisis of 2008, this additional demand was probably attributable to both precautionary and store-of-value motives. The crisis first had an impact on Japan (November 1997 to October 1998) and the effects were felt for the longest duration in Indonesia (May 1998 to May 1999). Türkiye was affected from January 1998 to March 1999. The rise in the annual growth rate was particularly pronounced in Indonesia (amplitude of growth of around 102 percentage points) and in Türkiye (amplitude of growth of around 61 percentage points). In Indonesia, this increase went hand in hand with a sharp rise in the rate of inflation. 68 Owing to high inflation, even large-denomination banknotes could have been used increasingly for transaction purposes. 69 Compared with Indonesia and Türkiye, the effects in Japan were small (amplitude of growth of around 6 percentage points). Japan was not affected by the Asian crisis directly, but rather indirectly via its trading partners. The increase in the circulation of large-denomination banknotes (in national currency) during the Asian financial crisis is likely to have been a lasting one in all three countries. 70

According to the criteria outlined at the beginning of this article, only in Indonesia did the annual growth rate of small-domination banknotes in circulation respond to the Asian crisis, doing so over the entire duration of the crisis (March 1997 to August 1999). This rate increased very sharply with an amplitude of growth of around 69 percentage points. After the crisis, the circulation of small denominations (in Indonesian rupiah) remained clearly above its pre-crisis level.

Overall, the Asian financial crisis (Asian crisis) from 1997 to 1999 had an impact on only a few of the countries analysed here, resulting primarily in an increase in demand for large-denomination banknotes. Both precautionary and store-of-value motives were the likely driving forces in this case. As a result of this crisis, the circulation of these banknotes increased in a lasting way in Indonesia, Japan and Türkiye.

This article explores global developments in demand for banknotes with the aim, in particular, of describing banknote demand during crises using selected examples. The general analysis of banknote demand initially confirms the paradox of banknotes already described in the literature, namely that demand for cash has risen strongly since the start of the 1990s, even though consumers are increasingly using cashless means of payment. 71

The analysis of a selection of countries and crises in conjunction with striking developments in banknote circulation makes it possible to derive hypotheses on the relationship between crises and demand for banknotes. It emerges that crises – even if they deliver a general shock to all countries, such as was the case with the global financial crisis, for example – can have different repercussions for banknote circulation in different countries. This is true of the timeframe, magnitude, and sign of a possible response, but it also applies to the question of whether they have any effect on the value of banknotes in circulation at all. Indeed, there are also examples of crises not having had any noticeable effect on banknote circulation. Second, crises can have an impact on both small and large-denomination banknotes in circulation. This is likely to be attributable to precautionary motives in the face of uncertainty, and in the case of large denominations, also in some instances to store-of-value motives due to risk considerations. Third, alongside temporary effects, it appears that crises also have lasting effects on banknotes in circulation in terms of value.

The descriptive approach chosen for this article does not permit these hypotheses to be verified or systematic relationships to be identified and quantified. Nonetheless, it can be concluded overall that demand for banknotes changed in an unusual way in the wake of a number of crises. For example, the circulation of both small and large-denomination banknotes rose sharply in 10 of the 16 countries considered here during Y2K. During the global financial crisis of 2008, the value of large-denomination banknotes in circulation increased in eight countries. Over the course of the COVID−19 pandemic, the annual growth rate of large and small-denomination banknotes in circulation changed significantly in 11 of the countries considered, while this rate fell in some cases. This therefore suggests that crises can play a role in explaining developments in banknote demand.

Empirical analyses aside, it can be maintained that cash is the only means of payment that is largely independent of technical infrastructure, and can therefore be used for payment transactions in the event of power outages or technical disruptions, for example. As a form of central bank money, cash is also a safe haven during turbulent times given its capacity as a default-proof store of value. Accordingly, cash can be expected to make a key contribution to the resilience of a society in the face of natural disasters or disruptions to technical infrastructure, for instance. However, in order to be able to fulfil this function even in times of crises, the acceptance and availability of cash must always be guaranteed – including during crisis-free periods. 72 As part of the Eurosystem’s cash strategy, the Bundesbank is committed to ensuring that euro banknotes and coins remain widely available and generally accepted as a means of payment and store of value in the future, too.