Currency

The well-known document ISO 4217:2015 defines a standard for currencies. It formalises a listing of official currencies, and is designed to inter-operate with other ISO standards. In it, we find the notion of currency designators, also known as alphabetic codes, which are unique identifiers composed of three letters¹. These can be classified into roughly three kinds of buckets. Firstly, for "regular" fiat currencies, the two initial letters make up an ISO 3166 Alpha-2 country code, and the last designates the currency itself. However, there are exceptions to this rule (e.g. USN for US next day, etc).

Secondly, for "other currencies"², the currency code must always begin with the letter X, followed by a two digit code. Then, the rules become even fuzzier:

• there are commodities such as gold (XAU), silver (XAG) and the like, which use the periodic table element code;
• there are currencies of countries or other territories (XPF for "franc Pacifique" and so forth);
• there are "operational" entries such XTS, which is a test currency, and XXX which stands for "no currency" — these are intended to be used for testing and setup purposes;
• and, of course, there are the special cases such as XDR, which are IMF's Special Drawing Rights.

Thirdly, and to make matters even more interesting, there are codes that do not fit any particular pattern such as EUR for the Euro. After all, the EU is not actually a country, though of course it is a legal and political entity; nor is R representative of the currency itself, called the "Euro".

In summary, though it is a cornerstone of the modern financial system, it is important to avoid ascribing too much method to currency codes. There is some, clearly, but the diversity of structure and purpose is perhaps best understood if one thinks of this document as a standard which has evolved over time, mapping to a very complex and constantly shifting geo-political status quo, and where, in the end, pragmatism had to win over rigour. A creature far more of politics than science.

A closing word is perhaps warranted on the remaining properties of each currency. As per Table iso_4217, in addition to currency codes there is also a corresponding Currency Name, for the official name of the currency, and a three-digit numeric code which can also be used to uniquely identify a currency. Finally, as depicted in the "Digits" column, the standard also captures the minor units for a given currency — i.e. whether it divides into 100 or 1000, etc.

It is important not to confuse the notion of an "ISO standardised currency" with the notion of a "currency". Pretty much anything can be a currency — just like anything can be used as money — provided the previously stated requirements are met. The fact that a currency is part of the ISO standard gives it a "seal of approval", together with respectability and availability to a set of economic actors, but it does not make non-standardised currencies any less of a currency from a conceptual perspective. And, as it happens, there are many in this position.

Traditionally, these have been fiat currencies for smaller nations or territories, or currency codes for specific purposes such as restrictions in circulation³. Now, however, there is a new kid on the block (if the pun may be pardoned). Cryptocurrencies, also known as cryptos, offer in our opinion the most prominent example of non-standard currencies due to their apparent popularity within certain segments⁴. But can they be rightfully considered currencies?

Wikipedia defines cryptos as follows (emphasis ours):

Definition 2.9: A cryptocurrency (or crypto currency) is a digital asset that is used as a medium of exchange wherein individual digital token coin ownership records are stored in a digital ledger or computerized database using strong cryptography to secure financial transaction record entries, to control the creation of additional digital token coin records, and to verify the transfer of token coin ownership. It typically does not exist in physical form (like paper money) and is typically not issued by a central authority.

Firstly, this definition is teeming with terms that require further elaboration but, sadly, we need to defer their discussion, as it would get us too bogged down on the machinery of crypto implementation. Secondly, notice how carefully the definition is worded in order to avoid stating whether cryptos are or aren't currencies; this is no coincidence. That cryptos are a very different beast from their regular fiat cousins is patently clear, since they belong to no country or territory — it's probably more apt to state that the Internet is their area of circulation — and, thus far, there is no government support for their continued existence. It is also obvious they possess certain properties of money but not all — e.g., at the time of this writing, you cannot pay your taxes with cryptos, though that may change soon. In addition, unlike regular currencies, they are not associated directly with state actors, to our knowledge, nor are they considered to be fiat — at least under certain definitions of the term.

A compelling argument can be made to demonstrate that cryptos satisfy most of the definitions presented thus far and, therefore, should be considered currencies. Trouble is, it is also entirely possible to argue the converse, and with an equal measure of success. The subject is very complex, straddling economic theory, law and even philosophy, and so an authoritative discussion of this topic lies well beyond the remit of the present work. Instead, for our purposes, we shall take the easy path and "declare" cryptos to be currencies. This is clearly an unsatisfactory stance for certain use cases, and that is the crux of the argument: the answer to this question is deeply rooted on the context in which cryptos are used. From a general Computational Finance perspective, there little that is special about cryptos to prevent their use; and, since they bear a close resemblance to currencies, they can be treated in very much the same way for great many cases. This rationale may not apply to other specific use cases, but "thereof we must remain silent" (White 2006).

With that dealt with, we can now look at the schemes for the identification of cryptos. Due to their suis generis existence, cryptocurrencies do not have a seat at the ISO table, which means they do not get to have an ISO currency code ⁵. However, since inception there have been informal "currency codes", at times called symbols, which do not necessarily follow the ISO 4217:2015 conventions described above. Though there is no formal standardisation process for these currencies, the codes in use are de facto standards within the community, and have stood the (admittedly limited) test of time. In this material, we shall be making heavy use of these de facto standards, with the list defined in Wikipedia forming a base for our universe of cryptos.

Table list_cryptocurrencies provides a small sample of the Wikipedia page's dataset. Its worth noticing that, unlike ISO 4217, some of the entries have more than one symbol and at least one of these symbols has more than 3 characters (e.g., DOGE). Note also that the Wikipedia page is but a small subset of all known cryptocurrencies; for a more representative range see the CoinMarketCap website⁶. Last but not least, please note that some cryptocurrency symbols clash with currency codes defined within ISO 4217.

With both standard and non-standard currencies been introduced, we can now shift our attention to the different stages in the life of a currency.

As Definition 2.7 already hinted, the most obvious physical manifestation of a currency are the notes and coins associated with it. These have a lifecycle: they enter circulation, circulate amongst the economic agents for a period of time, fulfilling their functions, and, ultimately, are retired and stop circulating — perhaps because of damage or due to a decision by the issuer. Interestingly, the exact same pattern is fractal-like, repeating itself at three separate levels:

1. the currency itself;
2. the types of notes and coins, called denominations; and
3. the instances of bank notes and coins.

For the purposes of Computational Finance, we're particularly interested in the first item of this list. Just as with a bank note, currencies themselves have a lifecycle: they are "born" when they are created and instances of it enter into circulation — that is, when they are made available to a set of economic agents for general use; they "live", that is, they are in circulation for a period of time; and eventually, they "die" when taken out of circulation. However, even after "death" records must still be kept regarding transactions that were made during their life time.

Whilst it may appear that the lifecycle of a currency is far too long to be of relevance to those developing IT systems, in practice they occur much more often than one may think, and the circumstances under which these transitions take place don't always allow for careful operational planning. Due to this, events on a currency lifecycle can be extremely disruptive to systems which are not designed to take them into account.