Several years ago, there were large billboards alongside main roads outside Accra in Ghana. They showed a well-endowed woman shoving bank-notes inside her bra, with the message "Keep your currency clean." It was not a warning against organised crime but against the practice of hiding bank-notes on one's person. One woman, a manager in a large enterprise, said "I keep notes in my shoes so that whatever happens, I have enough money for a taxi the office, home or even a hospital." Accra was, then, a city of two nations: the safe and the not safe. The real problems that the government was concerned with were two-fold. The first was the fact that bank-notes were filthy: literally dirty money. The risk of disease was ever present. One visitor said "I've got this 2,000 cedi note: it's disgusting, I'll give it to a beggar at the next junction." To the visitor, the note had no useable value but carried risk. The driver said "I'll have it. I can buy packs of sugar with it." But that was not the only problem Ghana's government was struggling with: at the time, it was estimated that some 70% of currency was held outside the banks. That meant that there was no data as to the circulation of money and no information as to who had it and how they spent it. Outside tax-related issues, that meant that economic forecasting and planning were impossible. It also meant that corruption could not be, effectively, contained. Ironically, relative to many of the countries nearby, at that time Ghana had a relatively small corruption problem.

Several years later, in 2007. The Ghanain Cedi was re-denominated (some mistakenly say "revalued"): today, one USD buys about 4.5 cedi. The re-denomination was brutal: according to local media at the time, a litre of petrol at the fixed price of 7,913.25 cedis would in future cost 0.79 cedi. The change was so severe that decimal division of the cedi was re-introduced, the Pesewa (Gp) and the reappearance of coins. One of the reasons for the change was the high cost of processing payments for even small transactions. The hidden message was that those costs were a disincentive for traders to put their cash takings into their bank accounts. And it was right that the quantity of notes used would collapse: one thousand old cedis would be worth just one tenth of a new cedi, and there was a coin for that.

This asks questions of those other countries where the nominal value of currency is so large that transactions are counted in vast numbers. In Indonesia, one USD buys more than 13,500; in Vietnam, one USD is worth about 22,700.

In India, the USD/Rupee exchange is about 1:65 which explains why the change in bank-notes did not, of necessity, mean a change in denomination.

The use of high-denomination notes in cash-based economies creates the opportunity for exploitation and fraud. Micro-economies of shifting populations are created by tourists, a term which means anyone who is not known in the immediate locality.

A recent trip to Indonesia provided additional insight. One of the lessons of buying anything in Indonesia is the ease with which unmarked prices are customer, not product, related. It is common all over the world for retailers and f&b operators to apply different prices to locals and foreigners. In fact, in many countries, access to publicly funded venues, e.g. art galleries, a supplement is applied to those who do not present a residence card. Often the increase is for an amount that, for the customer is insignificant but it still wrankles to know that the table next door pays less.

In pre-euro days, a (supposed) Shanghaiese restaurant in the back streets of a random town in Italy maintained one price list for locals and another for non-locals. Stupidly, they gave some diners the local menu, then charged according to the foreign menu. The difference on the bill was a huge number. It was only after an argument and threats to call the police that someone worked out that the uplift was a paltry GBP1.50 - and a change of attitude to laugh at the ridiculousness of ripping off patrons for such a tiny amount. But the euro helped Italy by making the numbers manageable.

In Indonesia, for example, today however, the situation is no different: applying the high nominal value of the currency means that IDR1.5 million is less than GBP 90, at current rates.

Adding IDR10,000 to the price of a product seems insignificant because so many products are charged in 1000 units. A bottle of Bintang beer in a food stall is 40,000 when having a drink with a local. Two days later, a foreigner alone, was charged 50,000 just two stalls away. There is no apology, just a shrug and "O.K." and an acceptance of the lower price. Examples abound. Each time, the uplift was in multiples of 10,000. A lens cleaning cloth: 50,000. No, 30,000 (already far too expensive). Settled on 40. At 15,000 to the pound, the cloth was outrageously expensive but the chances were no other customers would be in that shop that hour and, sometimes, one just has to spread a little goodwill around. He no doubt felt he'd negotiated well: the purchaser felt he'd helped the small trader in a poorly maintained shop with little stock keep the lights on just a little bit longer.

The reason the uplift trick works is simple: the visitor's mind is overwhelmed by the size of the numbers involved. It simply discounts numbers it thinks are small. They are not that small: that 10,000 is almost 80p, more than USD1. In the real world, customers change their shopping habits for that kind of difference on a single item.

Yet, unless the visitor has a clear point of reference, he probably won't notice, even when he finds himself visiting the ATM more often than expected.

We are not, here, talking about hyper-inflation reducing the value of currency within months, days or even hours although ultimately the results are similar: eventually, the currency ends up like that of Zimbabwe, to all intents and purposes abolished by popular demand while trade takes place in (officially illegal) US dollars. That causes its own problems, as in Russia in the early 1990s, where the ruble became so useless that the de facto daily currency was US dollars and the then finance minister admitted that the Russian economy would collapse if the fake USD100 notes were suddenly withdrawn from circulation (see How not to be a money launderer by our boss, Nigel Morris-Cotterill, for the story.

There are enormous hazards in non-denominational, small value currencies from a systemic perspective. The locals lose respect for the currency: if that's the case why bother with it at all - why not move to a crypto-currency? Or resort to barter? While it is true that economic development tends to push people towards bank accounts and, for some shoppers, towards some form of electronic payment, it is also true that there remains a large proportion of retail and F&B that runs on cash. While governments try to find ways to reduce the amount of cash in circulation and to bring cash within reporting systems, the simplest solution is also one of the most difficult to introduce. The wholesale swapping out of one series of bank notes for another over a short period of time places extraordinary burdens on banks. But, as the Irish found out when the Irish Pound was swapped for the Euro, the amount of "mattress money" that flowed into banks was astonishing - but so, it is said, was the amount that flowed over the border to be converted into British Pounds, which were then converted by the banks, bureaux de change, etc. that received it. There are always unintended consequences but that does not mean they are unanticipated.

A change in currency can also force the repatriation of funds held, in cash, offshore. Indonesia has long been looking to find ways of forcing its citizens to bring back moneys held around the world. There are many other countries in a similar position, for a wide range of reasons.