Understanding the difference between banknotes used for transactions and those used for non-transactional purposes is a perennial issue for central banks. A new paper issued by the Sveriges Riksbank adds to our understanding by comparing five similar countries using the seasonal method to understand their differences 1.
The five countries are all open, high income economies. They have mature payment markets with banked populations who are financially literate.
There are well documented seasonal variations in payments. Seasonal demand is different from other motives for holding cash and so if seasonal variations are removed, it is possible to identify the size of non-seasonal demand. ‘Easy’ assumptions about the use of low and high value notes can be avoided and there is no reliance on accessing time series data.
The study found that transactional balances fell in all countries. In Sweden and Norway this was driven by a reduction in non-transactional cash balances, a result that had not been anticipated. In Canada and Iceland there was a clear increase in the demand for non-transactional cash. Denmark has a more stable cash profile, but it is, overall, closer to that of Canada and Iceland.
The study paid particular attention to Sweden, where cash demand peaked in 2007 and has fallen by nearly 50% today. This is unusual given that most OECD countries have rising nominal values overall but falling transactional usage. For Sweden empirical cash demand models don’t work and the study believes that Sweden has a unique combination of events and policies.
The study looked for reasons and evidence about what might reduce the demand for non-transactional cash holdings.
1. Different legal tender status and cash validity
If central banks do not allow co-circulation of different series of banknote design, the public have a reason to return old notes to banks quickly.
Canada and Denmark both allow older series of banknotes to co-circulate with the latest design. Iceland, Norway and Sweden limit the period of time olds series are valid, usually within one year of issue. Norway and Sweden enforced this strictly in their respective 2017 and 2013 change-overs and this was clearly identifiable in their cash data for those years, with a big dip a year after the launch of the new series.
2. Acceptance of cash at the POS
If cash is not accepted at the point of sale (POS), which makes it harder to use, then there is less motivation to hoard cash.
Sweden has no legal requirement for cash to be accepted at the POS. In 2022 12% of Swedish shops did not accept cash. Cash is seldom accepted in hotels, restaurants and on public transport. Norway is similar to Sweden.
In Canada only 3% of shops don’t accept cash at the POS. In Denmark in most circumstances cash must be accepted at the POS. In Iceland cash is usually accepted.
3. Differences in foreign demand for cash
Foreign demand for a country’s cash is shown as ‘hoarded’ cash in the data. Because it is not circulating it does not show in the seasonal amplitudes and is included with the non-transactional cash balances.
There is no evidence that there is demand for Scandinavian banknotes.
There was evidence of demand for Canadian notes. For example, in 2016 5% of all $100 banknotes went to Hong Kong. The study compared Norway and Canada’s seasonal demand, which allowed an estimate of Canadian foreign demand.
From 2007 to 2015 foreign demand for Canadian notes was important. In 2015 25% of Canadian notes were abroad. Since 2016 this demand has reduced, reaching 6% in 2020. These changes were reflected in the data for domestic hoarding. This reduced from 50% in 2007 to 40% in 2015, but by 2020 it was 65%.
4. Different experience of crisis
Iceland was severely affected by the 2008 global financial crisis. This was reflected in hoarding increasing by 60% in 2008 and reaching 200% of the pre-crisis figure by the end of 2011. The other countries did not have the same experience in that crisis.
In the pandemic, Denmark, Sweden and Canada did see an increase in hoarding, but Norway and Iceland did not.
5. Different access to cash
It is plausible that hoarding is less likely if it is hard to access high value banknotes.
In Sweden and Norway, it is hard to get high value notes. The number of ATMs per person has been rising in Canada but falling in Norway. Denmark and Iceland have more than twice the number of ATMs per person compared with Sweden and Norway.
Norway and Sweden do not put high value notes into ATMs. Their highest denomination used in ATMs, the Kr500, is worth about €50, whereas some Canadian ATMs have $100 notes and in Denmark some have Kr1,000. Data for Iceland was not available.
Norway and Sweden also have both a low number of bank branches and over the counter (OTC) services have been reduced. In Sweden in 2012 only 40% of branches offered OTC services, today none do. Norway is similar.
6. Shadow economy
Crime is not very seasonal. Again, the result is that cash used for illegal activities does not show in the seasonal data and, therefore, is included in non-transactional data.
For these five countries the data on their shadow economies shows that it is falling. In 1991 the shadow economy was about 20% of GDP, plus or minus about 5%. In 2021 it was 12%, plus or minus about 3%.
Cash as a proportion of GDP has risen in Canada, Iceland and Denmark while transactional cash has fallen. In line with the well-known cash paradox, non-transactional cash has risen in Canada and Iceland, and slightly in Denmark.
Cash as a proportion of GDP has fallen in Sweden and Norway. Although transactional cash has fallen, so too has non-transactional cash. The study found that rather than Sweden being special and its cash decline being ahead of other countries and an indicator of what others will now go on to experience, the situation is less clear. The negative supply side effects are special to Sweden and Norway. Hoarding increased where there were specific shocks in all of the countries.
Understanding non-transactional demand is important for the design, production and provision of high value denomination notes. They don’t wear out as fast and appropriate infrastructure is needed for people to be able to withdraw, exchange and deposit them. The paper concludes by suggesting that perhaps CBDCs need to be designed to function as a store of value.
1 - Sveriges Riksbank Working Paper 427. Cash for Transactions or Store-of-Value? A comparative study on Sweden ad peer countries. Carl Andreas Claussen, Björn Segendorf, Franz Seitz.
