Are central banks too reliant on SWIFT for domestic payments?

Central bank settlement systems are the the tectonic plates of the payment system: they are vitally important to our lives, but we never see them in action. All of a nations' electronic payments are ultimately completed, or settled, on these systems. If they stop working, our financial lives go on pause, or at least regress to older forms of payment.

In this post I want to introduce readers to a crucial feature of these payments tectonic plates: their reliance for domestic settlement on SWIFTNet, a financial messaging network used by banks and other financial institutions to communicate payments information. Think of SWIFTNet as a WhatsApp for banks, but exclusive and very secure. 

This reliance – or over-reliance – is best exemplified by a recent decision by the European Central Bank. The Target2 settlement system has long been the bedrock layer of the European payments universe. All domestic payment ultimately get tied-off on the system. Since it was introduced in 2007, Target2 has been solely reliant on SWIFTNet for sending and receiving messages. 

When the European Central Bank replaced Target2 with T2 earlier this year, it modified the system to have two access points: it kept SWIFTNet but added a competing messaging network, SIAnet, to the mix. As one commentator triumphantly put it, "SWIFT’s monopoly for access to the T2/T2S system is broken."

SWIFTNet is owned by the Society for Worldwide Interbank Financial Telecommunication, or SWIFT, which is structured as a cooperative society under Belgian law and is owned and governed by its 11,000 or so member financial institutions. Whenever SWIFT gets mentioned in conversations, it tends to be associated with cross-border wire payments, for which its messaging network is dominant. However, for many jurisdictions, including Europe, SWIFT is also integral to making domestic payments. It's this little-known local reliance that I'm going to explore in this post.

The dilemma faced by central banks such as the European Central Bank is that SWIFTNet is an incredibly useful messaging network. It is ubiquitous: most banks already use it for cross-border payments. And so the path of least resistance for many central banks is to outsource a nation's domestic messaging requirements to SWIFT, too. However, this reliance exposes national infrastructure to SWIFTNet-related risks like foreign control, sanctions, snooping, and system outages.

Financial messaging 101

Before going further, we need to understand why financial messaging is important. For a single electronic payment to be completed, a set of databases owned by a number of financial institutions, usually banks, must engage in an intricate dance of credits and debits. To coordinate this dance, these banks need to communicate, and that's where a messaging network is crucial.

Say, for example, that Google needs to pay Apple $10 million. Google tells its banker at Wells Fargo to make the payment. Wells Fargo first updates its own database by debiting Google's balance by $10 million. The payment now has to hop over to Google Apple, which banks at Chase. For that to happen the payment flow must progress to the core of the U.S's payments system, the database owned by the Federal Reserve, the U.S.'s central bank.

Along with most other U.S. banks, Wells Fargo has an account at the Federal Reserve. It communicates to the central bank that it wants its balance to be debited by $10 million and the account of Chase to be credited by that amount. Once Chase's account at the Federal Reserve is updated, Chase gets a notification that it can finally credit Apple for $10 million. At that point Apple can finally spend the $10 million.

This entire process takes just a second or two. For this "dance of databases" to execute properly, the Federal Reserve, Chase, and Wells Fargo need to be connected to a communications network.

The sort of messaging network to which the central bank is connected, and the stewardship of that network, is thus crucial to the entire functioning of the economy.

Proprietary messaging networks or SWIFTNet? 

The Federal Reserve is somewhat unique among central banks in that it has built its own proprietary messaging network for banks. All of the 9,000 or so financial institutions that use the Federal Reserve settlement system, Fedwire, must connect to the Fed's proprietary messaging network to make Fedwire payments. To make international payments, however, U.S. banks must still communicate via SWIFTNet.  

Let's flesh the story out by trekking north of the border. Whereas the Federal Reserve has no reliance on SWIFTNet, Canada's core piece of domestic settlement infrastructure, Lynx, relies entirely on SWIFTNet for messaging.

For example, if Toronto Dominion Bank needs to make a $10 million to Scotiabank, it enters this order into SWIFTNet, upon which SWIFT forwards the message to Lynx, which updates each banks' accounts by $10 million and sends a confirmation back to SWIFTNet, which tells Scotiabank that the payment has settled.

For payments nerds, this network setup is called a Y-copy topology. The network looks like a "Y" because the originating bank message is relayed from the sending bank via SWIFTNet, the pivot at the center of the Y, down to the settlement system, and then back up via SWIFTNet to the recipient bank. It is illustrated below in the context of the UK's payment system, with the CHAPS settlement system instead of Lynx, but the idea is the same.

A Y-copy network topology for settling central bank payments in the UK [source]

The upshot is that the Federal Reserve controls the messaging apparatus on which its domestic settlement depends, whereas Canada outsources this to a cooperative on the other side of the ocean.

Many of the world's small and middle-sized central banks have adopted the same Y-copy approach as Canada. This list includes Australia, Singapore, New Zealand, Nigeria, UK, Sweden and South Africa. However, some members of this group are starting to have second thoughts about fusing themselves so completely to SWIFT.

Removing the single point of failure

The European Central Bank is at the vanguard of this group. Prior to 2023, the European Central Bank was in the same bucket as Canada, relying entirely on SWIFTNet to settle domestic transactions. 

With its upgraded T2 system, Europe doesn't go quite as far the Fed's model, which is to build its own bespoke messaging network. Rather, European banks now have the option of either sending messages to T2 using SWIFTNet, or they can use SIAnet, a competing network owned by Nexi, a publicly-traded corporation. SIAnet stands for Societa Interbancaria per l'Automazione, a network that originally connected Italian banks but has now gone pan-European.

The reason for this design switch is that European Central Bank desires "network-agnostic connectivity." This dual access model will make things more complex for the European Central Bank. If a commercial bank originates a SIAnet message, the central bank will have to translate this over to a SWIFT message if the recipient bank uses SWIFTNet. Nevertheless, the European Central Bank believes this dual structure will offer more choice to domestic banks.

The ECB also hints at the enhanced "information security" that this new setup will provide, without providing much detail. The UK's recent efforts to update its core settlement layer sheds some extra insights into what these security improvements might be. Right now, the UK's core settlement system, CHAPS, can only be accessed by SWIFTNet, much like in Canada, so that all domestic UK payments are SWIFT-reliant.

In its roadmap for updating CHAPS, the Bank of England is proposing to allow banks to access the system via either SWIFTNet or a second network, which doesn't yet exist. The idea is to enable "resilient connectivity" to the core settlement layer, especially in periods of "operational or market disruption." Should SWIFTNet go down there would be no way for financial institutions to communicate with CHAPS, and the entire domestic economy would grind to a halt. A second network removes the "single point of failure" by allowing banks to re-route messages to CHAPS.

The Bank of England also highlights the benefits of competition, which would reduce the costs of connectivity.

This sounds great, but there are tradeoffs. Using a a single network for both domestic and international payments is valuable to the private sector because it offers standardization and efficiencies in banks' processing. Adding a second option will also complicate things for the Bank of England, since it will have to design and build a system from scratch, much like the Fed did, which could be costly. Either that or it will have to find another private option, like the ECB did with SIAnet. This second network may not be as good as SWIFTNet which, despite worries about resiliency, has been incredibly successful.

When CHAPS went down earlier this year for a few hours, for instance, it wasn't SWIFT's fault, but the Bank of England's fault. The same goes for a full day outage in 2014. 

Comparing a V-shaped network topology to Y-Copy in an Australian context [source]

The type of settlement topology that the UK is proposing is known as "V-shaped," since all messages are sent directly to the central bank settlement system for processing via any of a number of messaging networks, and then back to the recipient bank. The difference between a V-shaped topology and Y-copy is visualized in the chart above in an Australian context, but the principles apply just as well to the UK.

Sanctions and "the SWIFT affair"

The decision to make domestic payments less dependent on SWIFTNet is much more easy to make for outlier nations like Russia. SWIFT is based in Belgium and is overseen by the Belgian central bank, along with the G-10 central banks: Banca d’Italia, Bank of Canada, Bank of England, Bank of Japan, Banque de France, De Nederlandsche Bank, Deutsche Bundesbank, European Central Bank, Sveriges Riksbank, Swiss National Bank, and the Federal Reserve. That put SWIFT governance far out of Russian control.

You can see why this could be a problem for Russia. Imagine that only way to settle domestic Russian payments was by communicating through SWIFTNet. If Russia was subsequently cut off from that network for violating international law, that would mean that all Russian domestic payments would suddenly cease to work. It would be a disaster.

Needless to say, the Central Bank of Russia has ensured that it doesn't depend on SWIFTNet for communications. It has its own domestic messaging network known as Sistema peredachi finansovykh soobscheniy, or System for Transfer of Financial Messages (SPFS), which was built in 2014 after the invasion of Crimea. Prior to then, it appears that "almost all" domestic Russian transactions passed through SWIFTNet – a dangerous proposition for a country about to face sanctions.

Mind you, while Russia has protected its domestic payments from SWIFTNet-related risk, it can't do the same for its international payments. SWIFTNet remains the dominant network for making a cross border wire. There is no network the Russians can create that will get around this.

I'm pretty sure that most larger developing states and/or rogue nations have long-since built independent domestic financial messaging systems to avoid SWIFTNet risk. I believe China has done so. Brazil has the National Financial System Network, or Rede do sistema financeiro nacional (RSFN). India also has its own system, the Structured Financial Messaging System (SFMS), built in 2001. India is even trying to export SFMS as a SWIFT competitor.

The Japanese were typically way ahead on this. The Bank of Japan built its messaging network, the Zengin Data Telecommunication System, back in 1973, several years before SWIFT was founded.

The last SWIFTNet risk is snooping risk. This gets us into the so-called SWIFT affair. After 9/11, the U.S. intelligence agencies were able to pry open SWIFT through secret broad administrative subpoenas. They had the jurisdiction to do so because one of SWIFT's two main data centres was located in the U.S.

To ensure data integrity, SWIFT had been mirroring European data held in its data centre in Belgium at its U.S. site. That effectively gave U.S. intelligence access to not only SWIFT's U.S. payments information, but  also information on foreign payments sourced from Europe or directed to Europe. Worse, it also provided spooks with data on domestic European payments. Recall that the European Central Bank's Target2 settlement system, which settles all digital domestic payments in Europe, was entirely reliant on SWIFTNet for communications.

When the U.S.'s snooping arrangement was made public by the New York Times in 2006, it caused a huge controversy in Europe. SWIFT tried to placate Europe by building a third data warehouse in Switzerland to house Europe's back-up data. But the precedent was set: SWIFT is not 100% trustworthy. And that may be part of the reason why the European Central Bank chose to downgrade its reliance on SWIFTNet when it introduced its new system, and is surely why other nations want to entirely hive their domestic systems off from it.

---

In sum, central banks face a host of complicated decisions in how to bolt on messaging capabilities to their key settlement systems. SWIFTNet is a top notch network. However, too much SWIFT-related risk may be perceived as having negative implications for national security. For large nations with extensive banking industries, building a proprietary domestic messaging alternative seems to be the preferred option. It also seems to be the default choice for rogue states like Russia.

Another alternative is to fallback on using multiple independent networks for access, of which one is SWIFTNet, and thus mitigating exposure to SWIFT-related problems. This is the approach taken by Europe and the UK.

For smaller nations that comply with the global consensus, like Canada, the calculus is different. Building an alternative communications network is likely to be costly. The risk of sanctions and censorship are negligible while the benefits of using a high-quality ubiquitous network for both domestic and foreign payments messaging are significant. Given these factors, it may be worthwhile to bear all SWIFT-related risks and adopt the Y-copy model.

Another Indian payments idea: a lightweight portable payments system for catastrophes

India is so far ahead of most of the world when it comes to payments technology that when a new Indian payments idea emerges, we should probably pay attention.

In its recent annual report, the Reserve Bank of India, the country's central bank, put forward the idea of a super-resilient back-up payments rail, which it calls the Lightweight Payment and Settlement System (LPSS). If a war or natural disaster cripples core payments infrastructure such as India's real-time gross settlement system, then the LPSS would kick in to process essential payments. The LPSS would be "independent of conventional technologies," portable, and capable of being operated "anywhere by a bare minimum staff."

Here is the full description:

Source: RBI

In my admittedly not-very-technical head, what I'm imagining is some sort of truck filled with computers and other apparatus for processing payments, sort of like a mobile police command vehicle. I could imagine a fleet of them. And there would be multiple way for these mobile payments bunkers to communicate with commercial banks, some of which wouldn't rely on the internet, like... ham radio? Transactions recorded on DVD and delivered by hand? (Yes, apparently that's a real thing). Dunno. I'm sure others will have a better idea how the LPSS works.

Central banks already plan for disasters by creating layers of payments redundancy, specifically back-up data centres in fixed locations.

In the U.S., for instance, the key Fedwire system is hosted at the Fed's East Rutherford Operations Center in New Jersey. Should disaster strike at the East Rutherford hub, a secondary back up data center at the Federal Reserve Bank of Richmond is designed to resume Fedwire operationality 60-90 minutes later. A third backup center exists at the Federal Reserve Bank of Dallas. (The source for this is here.)

As for Canada, our new Lynx high-value payments system is hosted at a primary site (I don't know where) with a secondary site ready to turn on as a back-up, as explained in this document. If both sites go down, then the system moves to good ol' fashioned manual processing.

I assume that India currently uses the same fixed primary/secondary site strategy as Canada and the U.S.

Ultimately, every single digital payment gets settled via transfers of central bank deposits over core payments infrastructure. So ensuring that these systems are available is crucial. No central bank wants to fall victim to something like the infamous 2014 breakdown of CHAPS, the UK's wholesale payments system, which prevented people from making crucial payments for ten hours.

Ages ago, some of us suggested that blockchain-based systems might provide a suitable layer of redundancy. If key centralized infrastructure like India's RTGS, U.S.'s Fedwire, or Canada's Lynx were to stop working thanks to a disaster, a central bank could temporarily fail-over to a decentralized option until the primary system is restored. This model has never been tried out. Perhaps in a world with ever-increasing natural disasters and renewed geo-political tensions, an LPSS, some sort of truck-full-of-payments-processing-machinery, is the best solution for central banks.

Indians' "ill-informed notions" concerning the legitimacy of ₹10 coins

The BBC has an interesting story about India's coinage. Apparently more and more Indians  believe that the ₹10 coin is not real, or that it has been banned by the authorities, and as a result they are unwilling to accept them in trade. Doubts about the ₹10 coin have been emerging for several years now: Amol Agarwal has covered the story here, here, and here.

This is an excerpt from the BBC article:

"Nobody accepts the coins - grocery shops, tea stalls, nobody accepts it", an auto rickshaw driver in the southern state of Tamil Nadu told BBC Tamil.

In the southern city of Hyderabad, a young girl told BBC Telugu she had been saving up to buy her brother a gift but several shop owners wouldn't take her 10 rupee coins.

A man on his way to a job interview was forced to get off the bus because the conductor wouldn't accept 10 rupee coins, the only currency he had.

"They say it's because the other passengers don't accept the coins in return", explains a shop owner who also said bus conductors wouldn't take the coins.

The Reserve Bank of India (RBI) has twice addressed the public's worries about the ₹10 coin. In a 2016 announcement it begged Indians to ignore "ill-informed notions" concerning the legitimacy of ₹10 coins and to continue to "accept these coins as legal tender in all their transactions without any hesitation." More recently, in a January notice, we learn that the RBI has issued fourteen different designs for the ₹10, all of which are "legal tender and can be accepted for transactions."

What are the underlying reasons for Indians' fears? One interesting fact about the ₹10 coin is that it is relatively new, having been introduced back in 2009. People are always skeptical about new monetary instruments, which generally take a long time to acquire trust.

Another interesting fact is that in addition to minting a ₹10 coin, the RBI also prints a ₹10 banknote. The ₹10 banknote has a long history, having debuted before independence in 1947. Below is a chart showing how many of each instrument is in circulation. The year-over-year net increase in banknotes continues to outpace the increase in coins by a large amount, indicating that  Indian's have a preference for the paper version of the ₹10.

I think there is an easy explanation for the ₹10 coin's loss of currency. Because the ₹10 coin and ₹10 note are perfect substitutes, and converting between them incurs no conversion costs, there is no disciplining mechanism to prevent irrational worries about the newer of these two instruments from crippling its usage. Put differently, hating new ₹10 coins doesn't impose any costs on the hater as long as an equivalent banknote can be used. If there was no such thing as the ₹10 banknote, then anyone who refused to use the ₹10 coin would face much higher costs for being unreasonable. After all, holding two ₹5 coins or five ₹2 coins in the place of a ₹10 coin is inconvenient.

The denomination at which a monetary system switches from coins to notes is referred to by Rocheteau and Lotz (pdf) as the coin-note frontier. In Canada, for instance, the frontier lies between the $2 coin and $5 note, while in Switzerland it lies between the 5 franc coin and 10 franc note. Most frontiers (like Canada's and Switzerland's) are staggered—the largest coin is smaller than the smallest note. This staggering makes a lot of sense. Why should both the nation's mint and its printing presses incur the fixed costs of producing the same unit when one will suffice? Consider too the waste incurred in the doubling-up of the tasks of distributing, sorting and handling a coin and note of the same denomination.

Unlike most countries, India has an even coin-note frontier. For some reason, the Indian monetary authorities have decided to have both the mints and the presses replicate the same task of producing the ₹10. Interestingly, India isn't alone. The U.S.'s largest coin is $1 while the smallest note is $1.

The US's $1 coin, introduced in 1979 and referred to as the Susan B. Anthony dollar, is commonly considered to be a major monetary failure. I wrote about it here. $1 coins have proven to be unpopular with the American public, huge amounts of them accumulating in vaults at various Federal Reserve banks. Because the US monetary authorities decided to introduce the $1 coin without removing the $1 bill, the public was given a choice between a perceived "good" currency, the existing and comfortable note, and a "bad" currency, an unfamiliar coin. They took the less costly route and stuck with the "good" notes. My guess is that the very same forces that doomed the $1 coin could end up killing off the ₹10 coin.

The failure of the $1 and ₹10 coins is unfortunate. As Rocheteau and Lotz point out, replacing low denomination notes with coins is a good idea because the the cost of keeping bills in circulation is greater than the cost of servicing coins. While coins are more expensive to produce, they last much longer than bills.

So not only are the US and India doubling up their costs by having both the mint and printing presses produce the same instrument, but at the same time the decision to keep the note in circulation means that the more efficient instrument—the coin—is destined to fail. The Reserve Bank of India blames the public's "ill-informed notions" for the ₹10 coin's loss of currency. But perhaps it should be blaming itself for providing the right conditions that allow for the spread of these ill-informed notions. Remove the ₹10 note and the problem will be fixed.

---

Amol Agarwal has some comments here.

No rupees left behind

Data on the world's biggest monetary event of the 21st century—Narendra Modi's demonetization—continues to trickle in. The Reserve Bank of India's just published its annual report (pdf) and I'll just say this straight out—I'm genuinely surprised. Out of the ₹15.44 trillion in paper currency that was demonetized by Modi last November, ₹15.28 trillion, or 98.96%, have been returned. My guess would have been that a much smaller proportion of India's monetary stock made it back to the RBI, maybe 92-95%.

For those not familiar with rupee prices, I'll translate the above numbers into US dollar terms. Back on November 9, 2016, $240 billion worth of rupee notes were declared to be invalid. By June 2017, only a small 1.04% sliver of this—$2.5 billion—was still unredeemed, far less than the $15-30 billion many of us though would have been left stranded. (I'll use dollars from here on since it is easier for the international community to understand.)

A glance at the stranded notes counts from a the euro changeover provides some context. When the Euro was introduced in January 2002, people were given fixed windows of time to redeem existing national banknotes before their money status was revoked. In the case of the Italian lira and French franc, individuals had till December 2011 and February 2012 respectively to bring in their notes for euros. By the time this ten-year exchange period was over, 99.15% of Italian lira had been returned while 98.77% of French francs made it back. Much of the unreturned 1% would have been withheld by the collecting community, the rest either being lost, buried, burnt in fires, or destroyed in floods.

Whereas the French and Italians had years to diligently round up old notes before the window closed, Indians only had few weeks, the final day for exchanging being December 31, 2016. Yet even with this much smaller window, Indians were able to bring in greater portion of the outstanding money supply than the French did in ten years. Impressive. Are Indians just great at locating things? Or is something else to explain? The skeptic in me wonders how many counterfeit rupees managed to make it passed the RBI's gatekeepers. No central bank can perfectly screen for fakes—so if the RBI mistakenly accepted a greater proportion of counterfeits than the Bank of France did, then the return rate on rupees would have been artificially improved relative to that of francs. But I'm just speculating.

What makes the final 98.96% return rate even more incredible is that, unlike the European demonetizations—which allowed unlimited, no questions-answered redemption—the Indian demonetization imposed per-person ceilings on the amount of rupees that could be freely converted into new paper rupees. Anything above the ceiling had to be deposited into a bank account i.e. individuals would be de-anonymized and potentially investigated. Yet even with the imposition of such a severe threat, a greater portion of rupees were redeemed in the waning days of 2016 than French francs during the entire 2002-2012 period.

Jugaad, or Indian ingenuity, is one of the explaining factors. Even though they had just a few weeks, Indians who had large quantities of illicit cash were able to contract with those who had room below their ceiling to convert illicit rupees on their behalf, thus evading Modi's blockade. This was money laundering on a grand scale.

There is a second explaining factor for the high return rate. Two weeks after the initial demonetization announcement, the government introduced a formal amnesty for demonetized banknote holders. Any deposit of cash above the ceiling would only be taxed at 50%, assuming it was declared. If not declared, the funds might still get through the note blockade undetected, although if apprehended an 85% penalty was to be levied. These new options were better than throwing away one's stash altogether and suffering a sure 100% loss, so previously reticent citizens would have flocked to bring their notes in, even if they had been amassed illicitly.

---

Demonetization was designed to provide a "national dividend" of sorts. If just 90% of the demonetized rupees had made it back to the central bank, the remaining 10% would have been written off, the one-time profit providing a massive $24 billion dividend to all Indians (participants in the underground economy being the folks who funded this subsidy). Does the higher-than-expected 98.96% return rate for rupees mean that Modi's demonetization has failed as a mechanism for redistributing funds from large participants in the underground economy to the rest of the Indian population? Is the national dividend void? Not at all.


As I wrote above, many of the demonetized rupees that have made their way back into the system were deposited into bank accounts. Some depositors will have sought shelter under the 50% amnesty. For the remainder, authorities will be following the paper trails left by deposited currency over the next few years and, if warranted, levy a tax on these deposits. This is a more cumbersome way to collect a tax than stranding banknotes because it requires investigating each suspicious deposit and potentially prosecuting the depositor. It remains to be seen how successful the Indian authorities will be in collecting this tax. Jagdish Bhagwati explains this all in far more detail here.

Remember the ingenuity that Indians used to escape the ceilings that were imposed on them? This was also a form of redistribution. Rich Indians would have paid poor Indians—who had plenty of room under the ceiling—a fee to deposit notes on their behalf, say 25%, or ₹250 on a ₹1000 note. Together, all of these fees would be very much like the national dividend described in the above paragraphs. Except rather than the tax being collected by the authorities and then paid out as a dividend, it would have been collected directly by Indians themselves. If $25 billion was laundered in this way, and an average fee of 25% charged, a $6.25 billion dividend would have been collected by poor Indians from rich ones.

In summary, a large chunk of stranded rupee notes would have provided the 'cleanest' way of taxing to fund the national dividend. However, the fact that very few notes were stranded doesn't mean there will be no national dividend whatsoever. 

You call that counterfeiting? This is counterfeiting!

One of the world's most notorious cases of counterfeiting was the 1925 Portuguese banknote crisis, when Artur Virgilio Alves Reis managed to pass off 200,000 fake five-hundred escudo notes, worth around £56 million in 2016 terms. When the Bank of Portugal discovered the counterfeits in December 1925, it announced an aggressive demonetization of the five-hundred escudo note, giving citizens just twenty days to bring in old notes for redemption. By then, the damage had been done. Reis had managed to increase Portugal's supply of banknotes by 5.9%, spending the equivalent of 0.88% of Portugal's nominal GDP into circulation! (More here.)

An aggressive note demonetization in the face of large-scale counterfeiting is a thoroughly justified response as it immediately puts a halt to the problem. In this context, it's worth revisiting the world's most recent attempt to combat counterfeiting with an aggressive demonetization—India PM Narendra Modi's forced recall of the Rs1000 and 500 notes on November 9, 2016, some 290 days ago. How does it compare to Portugal in 1925?

If we go back to Modi's initial speech, India's demonetization was targeted at the three "festering sores" of corruption, black money, and terrorism. On the latter, Modi said:

Terrorism is a frightening threat. So many have lost their lives because of it. But have you ever thought about how these terrorists get their money? Enemies from across the border run their operations using fake currency notes. This has been going on for years. Many times, those using fake 500 and 1,000 rupee notes have been caught and many such notes have been seized.

Given Modi's fighting words, we'd expect the demonetization to have caught quite a bit of bad currency. Last month, finance minister Arun Jaitley revealed how many counterfeit notes had been detected:

Source

Translating Rs11.23 crore into dollar terms, the monetary authorities have found just US$1.7 million worth of counterfeit notes. Step back for a moment and compare $1.7 million to the full breadth and width of the demonetization. In declaring that all Rs1000 and 500 notes were to be useless, Modi  demonetized US$240 billion worth of paper rupees, around 85% of India's stock of banknotes. So only 0.001% of the entire face value that has been brought in for conversion was fake! That's an incredibly low value of counterfeits, especially when you consider that in 1925, Reis's counterfeits accounted for around 5% of the entire Portuguese paper money supply. 

Central bankers usually measure counterfeiting in parts per million, or PPM, the number of fakes detected in one year for every one million genuine notes in circulation. I've inserted a chart below, which comes from this Reserve Bank of Australia document. Canada, for instance, once had a much higher PPM (it even hit 450 back in 2004), but the Bank of Canada managed to bring this down to the low double digits by introducing new security features and, later, polymer notes.

Australia, the first nation to introduce polymer notes in the early 1990s, used to boast one of the lowest incidences of counterfeiting (below 5 PPM). Criminals are finally starting to make polymer fakes, perhaps because the RBA hasn't bothered to update the series for over two decades, giving plenty of time for counterfeiters to catch up.

In India's case, the recent numbers show a total of 158,000 fake notes detected (out of a 24 billion banknotes demonetized) between November 8, 2016 and July 14, 2017, this totaling up to a minuscule 6.6 PPM—in the same range of countries like Canada or Australia. And certainly not in the same category as Portugal in 1925.

Nor is this data anomalous. Back in January, James Wilson—an Indian civil engineer-turned-demonetization expert—perused official banknote statistics only to find that over the last few years India has not had a high incidence of counterfeit rupees, leading him to describe the demonetization effort as a cannon shooting at sparrows. Using some of the data from James' post, it's possible to calculate that India had a PPM of 7.1 in 2015/16. (I get this from 90.266 billion units of currency in circulation, and 632,926 counterfeits detected that year). This is consistent with earlier data from the Reserve Bank of India showing a PPM from 2007-2012 ranging between 4.4 to 8.1. So India's rate of counterfeits detected is fairly reasonable relative to other countries.

This low PPM could be due to two factors. Either India doesn't have a counterfeiting problem, or it does have one but the authorities are just really bad at detecting counterfeits. If neither Indian banks nor its central bank are particularly good at identifying fake notes, then a large stock of unidentified fakes may be permanently circulating along with legitimate banknotes with no one capable of draining out the fakes. But even so, why instigate a massive note recall to catch counterfeits if the institutions that do the catching are so leaky to start with? If it is the case that India has a counterfeiting problem, then Modi's go-to fixes should have been to improve note security features and the banking system's ability to detect fakes, not implement a massive Portuguese-style recall.

So the data certainly destroys one of the pillars on which Modi's demonetization was based; terrorism and fake currency. You need something like a Portugal in 1925 to justify an aggressive demonetization, and India wasn't even close. Over the next few months more data will pour in, expect Modi's scheme to be further tested.

---

P.S. The title is cribbed from here:

The forking of the Indian rupee

This post is about the dismantling of the rupee-zone between 1947-49, an historical event that is especially topical in light of two modern monetary projects: Narendra Modi's poorly-executed 2016 demonetization and a potential eurozone breakup.

Thanks to a recommendation by Amol Agrawal, who blogs at the excellent Mostly Economics, I've been pecking away at the 900-page history of the Reserve Bank of India, although I have to confess that I've spent most of my time on the chapter on the partition period. For those who don't know, India and Pakistan weren't always independent countries. Up until partition in August 1947, each was part of British India, a British colony. The rupee, which was issued by the Reserve Bank of India (RBI), was the sole medium of exchange in British India. By mid-1949, less than two years after partition, usage of RBI-issued rupees had been successfully limited to the newly-created state of India. As for Pakistan, it had managed to erect its own central bank, the State Bank of Pakistan, as well as introduce a new currency, the Pakistani rupee.

At the time of partition, Pakistan's architects faced a daunting challenge; given that the Brits had announced in early 1947 that the partition of British India was to occur that August, there remained only a few months to create a central bank and issue a new currency. Because printing enough new currency for an entire nation would take far more than a few months to achieve, a temporary solution was arrived at: to use the RBI as an interim agent for issuing currency until the new Pakistani central bank had its own printing presses up and running.

This "bridge" involved using a combination of regular RBI-issued rupees circulating within Pakistan at the time and "overprinted" notes issued by the RBI. To ensure that the purchasing power of the two rupees stayed locked, the overprints were to be accepted by the RBI at par with regular notes. When enough Pakistani rupees had been printed by the newly-created State Bank of Pakistan, or the SBP, the mix of India rupees and overprinted notes was to be demonetized and replaced by Pakistani rupees on a 1:1 basis.

Here is what the Pakistani overprints looked like.

Note that they have the text "GOVERNMENT OF PAKISTAN" inscribed on them. Otherwise, overprints were just like regular rupees of the time.

---

Let's pause and bring this to the present. Like the rupee breakup of 1947-49, Modi's recent demonetization involved the cancellation of a large proportion of existing currency followed by an issue of new banknotes to replace them. (The 500 and 1000 rupee notes represented some 85% of India's paper money.) This is where the similarities between the two projects end. The architects of partition were wise enough to realize that they did not have enough time to print sufficient quantities of Pakistani rupees to replace Indian rupees, and so to avoid burdening the public with a shortage of cash they decided to use existing RBI-issued currency as a bridging mechanism.

Modi and his team of monetary architects evidently did not bother to familiarize themselves with RBI history. If they had, not only would they have realized what a huge task it is to replace the majority of a nation's currency, but they would also have learnt some tricks—like overprinting—to make the project easier. (Overstamping, a technique similar to overprinting, was successfully used in the break-up of the Austro Hungarian krona in 1991 as well as the Czechoslovak koruna in 1993.) This refusal to draw on the RBI's institutional memory means that some eight months after demonetization, Indians are still suffering from cash shortages.

---

Let's return back to the partition and explore the forking of the rupee more closely. The SBP, which was established July 1, 1948, formally took over the RBI-issued overprints as their liability that same day. As fresh Pakistan rupees came off the printing presses over the next months, SBP officials would steadily replace these overprints. That same day, the RBI subtracted the entire stock of overprinted notes from its total banknote liability. (The RBI had been issuing these notes since April.)

In taking over a large percent of the RBI's banknote liabilities, the SBP would need an equivalent set of assets to act as backing. These assets were to come from the RBI. More specifically, a fixed portion of the RBI's gold, coin, sterling-denominated securities, and rupee-denominated securities was to be transferred to the SBP, the rest remaining in India to serve a backing for RBI-issued rupee banknote.

To ensure fairness, a formula was settled on ahead of time to determine how the assets were to be apportioned. On a fixed date, the RBI would tally up how many notes were in circulation in Pakistan and how many in India, and divvy up the underlying assets according to the distribution of notes. That's fair way to do things, at least in theory. For the overprints, the RBI would record how many it had issued by July 1, 1948, and for each rupee overprint in existence it would transfer an equivalent quantity of assets to the SPB. When July 1 came, some 9.9% of the RBI's assets were dispatched to Pakistan. Thus one half of the mix of notes circulating in Pakistan, the overprints, had been demonetized.

There still remained the second half of the mix—regular Indian rupees. Dealing with these was more complicated. Unlike overprints, the RBI could have no firm measure for how many regular rupees were still being used in Pakistan, and thus had no way of knowing how many backing assets to transfer to the SBP. Intead, a mechanism for tallying up notes was established such that all Indian rupees circulating in Pakistan had to be brought to SBP offices for conversion into Pakistani rupees before July 1, 1949, one year after the central bank's founding. As Indian rupees flowed into the SBP over the course of the next twelve months, SBP officials remitted them to the RBI. The RBI then transferred an equivalent asset to the SBP for each Indian note it had received up until the expiry of the conversion period on July 1, 1949, after which the RBI ceased to accept remitted Indian rupees.

At this point, the RBI and the SBP were officially divorced. All liabilities and assets had been distributed to each respective party.

---

In a potential breakup of the euro, a formula like the one devised by the RBI will have to be used. The results, however, are likely to be messy. Because as I'll show, the divvying up of the RBI's assets wasn't without controversy.

If you look at the SBP's 2016 financial statements, you'll see an interesting line item called "Assets Held With the Reserve Bank of India":

Source

Go to note 14, and you'll see that:

"These assets were allocated to the Government of Pakistan as its share of the assets of the Reserve Bank of India under the provisions of Pakistan (Monetary System and Reserve Bank) Order, 1947. The transfer of these assets to the Group is subject to final settlement between the Governments of Pakistan and India"

So it seem that Pakistan never received what it believes to be its fair portion of the RBI's assets. For almost 70-years now it has carried these IOUs on its balance sheet (see historical date here). That's a long time to hold an asset that is unlikely to be collected! The reason for this odd balance sheet item can be found on page 568 or the aforementioned 900-page RBI tome. Between the founding of the SBP in July 1948 and the July 1949 cutoff date for note remittances to the RBI, more Indian rupees had filtered over the border into Pakistan than expected. As such, Pakistan was able to stake a larger claim on the RBI's assets than initially estimated.

Indian officials, who were not happy with the amount of assets they were sending over to Pakistan, now claimed that only those notes already in circulation in Pakistan as of July 1948 could legitimately be remitted for underlying assets. Any notes that were imported into Pakistan from India after that date simply would not count to the final tally. Pakistani disagreed. In March 1949 the Indian government informed the bank that "pending negotiations with the Pakistan Government further releases to them should be withheld." This was unfortunate news for the SBP. It had effectively issued Pakistani rupees without a reciprocating asset to back them.* That's where the two parties stand to this day—the SBP grudgingly holds the RBI's IOU on its books as reminder that it never got its perceived fair share of the RBI's assets.

---

This sort of havoc is inevitable when a monetary union breaks up. If existing notes are to be converted into new national notes at a one-to-one basis over a fixed period of time, then everyone has an incentive to export their notes to the region that is expected to have the strongest national currency. I am speculating here, but in September 1949—just three months after the RBI had been successfully divided—India devalued its rupee by 30.5%. Pakistan didn't. So all thoe holding Pakistani rupees were suddenly 30.5% richer than those holding rupees. Maybe the mass banknote exodus into Pakistan during the conversion period was an attempt to avoid this impending devaluation.

This same sort of dynamic would surely characterize a euro break-up. If Europeans are given 6-months to convert their euros into new national currencies like the German mark or the Greek drachma, you can bet that everyone will ship their euros to Germany. Drachmas, like the Indian rupee, are sure to be devalued. And if the final distribution of banknotes is to serve as the marker for divvying up the European Central Bank's assets, then Germany would get most of them. Were it to be prevented from getting its share, then Germany would end up in the same situation as Pakistan, with a shortage of good assets to back up all the fresh marks it has issued.

*If you're interested in specific amounts owed, here's an old World Bank document on the issue.

Demonetization by serial number

This continues a series of posts (1, 2, 3) I've been writing that tries to improve on Indian PM Narendra Modi's clumsy demonetization, or what I prefer to call a policy of surprise note swaps.

The main goal of Modi's demonetization (i.e. note swapping) is to attack holdings of so-called "black money," or unaccounted cash. The problem here is that to have a genuine long-run effect on the behavior of illicit cash users, a policy of demonetization needs to be more than a one-off game. It needs to be a repeatable one. A credible threat of a repeat swap a few months down the road ensures that stocks of licit money don't get rebuilt after the most recent swap. If that threat isn't credible, then people will simply go back to old patterns of cash usage.

It's worth pointing out that the idea of behind demonetization precedes Modi by many decades. In a 1976 article entitled Calling in the Big Bills and a 1980 a follow-up How to Make the Mob Miserable, James S. Henry, who is on Twitter, described what he called "surprise currency recalls."

In 1976, @submergingmkt wrote the original call for an aggressive demonetization: https://t.co/tDro2hnvAZ Does he approve of Modi's effort? pic.twitter.com/wM2EzLjkWB

— JP Koning (@jp_koning) January 10, 2017

Specifically, Henry advocated a sudden cancellation and reissuance of all US$50 and $100 notes as a way to hurt "tax cheats, Mafiosi, and other pillars of the criminal community." Rather than a one-shot action, which would only annoy criminals, the idea was that "the recalls could be repeated, at random, every few years or so, raising the 'transaction costs' of doing illegal business."

In order to credibly threaten a series of repeat note swaps, I'd argue that the quantity of notes recalled in each demonetization must be small. Small batches of notes can be quickly cancelled and replaced without disturbing people's lives. This keeps the economic and political costs of withdrawing demonetized cash (lineups, cash shortages, etc) manageable. If these costs are too high, the threat of repeats isn't credible.

Instead of going small, Narendra Modi decided to go big by having the Reserve Bank of India demonetize both of India's highest value banknotes, the ₹500 and ₹1000 note, which together comprised some 86% of India's cash. This has caused all sorts of problems. For instance, almost four months after the November 8 announcement the amount of cash in circulation is still far below the required levels of ₹17-19 trillion, the RBI unable to run its printing presses fast enough to keep up. The RBI's inability to fill the vacuum left by demonetized notes has been ably explained by James Wilson and is illustrated in the chart below:

Because of the enormous disruption it has caused, Modi's massive demonetization departs from Henry's script—it cannot be repeated, not for decades (a point that Russell A. Green makes here as well). Were Modi to begin discussing another demonetization, say for 2018, Indians would probably rise up in anger at the possibility of more lineups, empty ATMs, and hurdles to making basic purchases. Which means that post-Modi demonetization, it's entirely safe for India's illicit users of cash to wade back into the waters. If cash usage patterns return to normal, it seems to me that the entire demonetization project was an exercise in futility.

In India's case, demonetizing entire note denominations is too powerful a tool to ensure repeatability. Even if Modi had demonetized the ₹1000 note and not the ₹500, for instance, the exercise would still have involved some 30-40% of the nation's cash supply. This would have been an arduous affair for all involved, certainly not one that could be repeated for many years.

Weeding out rupee banknotes according to serial number rather than denomination would have allowed for a more refined policy along the lines advocated by Henry. Here's how it would work. The government begins by declaring that all ₹1000 notes ending with the number 9 are henceforth illegal. Anyone owning an offending note can bring it to a bank to be swapped for a legitimate ₹1000 note (one that doesn't end in 9). However, the government sets a limit on the number of demonetized notes that can be exchanged directly for legitimate notes, say no more than three. Anything above that can only be exchanged in person at a bank teller for deposits, which requires that they have an account (i.e. their anonymity will be lifted). Once an individual has deposited five notes in their account, all subsequent deposits of demonetized notes would require a good explanation for the notes' provenance. Should the requisite paper trail be missing, the depositor gives up the entire amount.

The process begins anew a few months hence, the specific timing and banknote target being randomly chosen. So maybe thirteen months after the first swap, the government demonetizes all ₹500 notes ending in 6. Randomness prevents people from anticipating the move and hiding their illicit wealth in a different high denomination note. 

Too understand how this affects black money owners, consider someone who owns a large quantity of illicit ₹1000 banknotes, say ₹70 million (US$1 million, or 70,000 banknotes). This person faces the threat of losing 10% to the note swap. After all, when the 9s are called, odds are that he or she will have around 7,000 of them, of which only eight can be returned without requiring a paper trail. The owner can simply accept a continuing string of 10% losses each year as a cost of doing business.

Alternatively, they might protect themselves ahead of time by converting their hoard into a competing store of value, say gold, bitcoin or low denomination rupee notes like ₹100s (which are not subject to the policy of ongoing swaps). If they flee high denomination notes to avoid subsequent demonetizations, illicit cash users in a worse position than before the adoption of the policy of note swapping. Gold and small denomination notes have far higher storage and handling costs than ₹1000 banknote. And unlike gold and bitcoin, a banknote is both supremely liquid and stable. So even if large-scale owners of banknotes manage to avoid painful note swaps, they still endure higher costs.

As for licit users of high denomination notes, the fact that the 10% clawback would not apply to them means they needn't change their behavior. Nor would the poor--who are unlikely to be able to provide a paper trail--have to worry about the policy. Demonetizations would only occur in high denominations, in India's case ₹500 and 1000s, and the poor are less likely to own these in quantities above the three note limit.

Incidentally, readers may recognize a policy of repeat demonetizations as akin to a Gesell stamp tax, named after Silvio Gesell, who in 1916 proposed the idea of taxing currency holdings in order to increase the velocity of circulation. Greg Mankiw famously updated Gesell's idea during the 2008 credit crisis to remove the zero lower bound. He did so by using serial numbers as the device for imposing a negative return rather than stamps. This post updates Mankiw's idea, except rather than applying the tax to all cash it strikes only at illicit cash holdings, and does so in the name of an entirely different policy goal—attacking the underground economy, not removal of the zero lower bound.

A series of small serial number-based swaps seems like a better policy than Modi's ham-handed demonetization of all ₹1000 and ₹500s. It would certainly do a better job of promoting a long-term decline in undocumented cash holdings and would do so by imposing a much smaller blast radius on the Indian public. There would be no currency shortages, huge lineups at banks, empty ATMs, or trades going unconsummated due to lack of paper money.

 That being said, while superior to Modi's shock & awe approach, a policy of repeat note swaps certainly has its flaws. In principle, the idea of surprising citizens every few months—i.e. forcing them to keep on guessing—does not seem entirely consistent with the rule of law. Another problem is that once the policy has been ongoing for several years, the list of demonetized serial numbers will be quite long. The process of buying stuff with notes will become evermore difficult given the necessity that the merchant consult this list prior to each deal to ensure that bad notes aren't being fobbed off. Finally, commenting recently on Henry's plan, Ken Rogoff notes that "there is a fine line between a snap currency exchange and a debt default, especially for a highly developed economy in peacetime." Since debt defaults hurt a countries credit standing, serial demonetizations might lead the investment community to be more leery about the nation's other liabilities, say its bonds.

Italian exit and the problem of lira shortages

The topic of euro breakup has slowly been trickling back in the news, especially with the potential for Italy leaving the currency union, a so-called Italexit. In this post I want to explore one of the major conundrums that would-be exiteers must face; the problem of banknotes.

Almost all euro exit scenarios begin with the departing country announcing a shot-gun redenomination of bank deposits into a new currency, in Italy's case the lira. The effort must be sudden—if redenomination is anticipated ahead of time, depositors will preemptively withdraw funds from the exiting country's banking system, say Italy, and put them in the banking systems of the remaining members, say Germany, doing irreparable damage to Italian banks. After all, why risk holding soon-to-be lira when they are likely to be worth far less than euros?

Once the surprise redenomination has been carried out, the next step is to quickly introduce new lira banknotes into the economy. Lira deposits, after all, should probably be convertible into lira cash. This is much tougher than it sounds. Consider the recent Indian debacle. On November 8, 2016 Indian PM Narendra Modi demonetized around 85% of India's banknotes. Ever since then the Reserve Bank of India, the nation's central bank, has been furiously trying to replace the legacy issue with new currency. Because a nation's printing presses will typically only have the capacity to augment the stock of already-existing currency by a few percent each year, a rapid replacement of the entire stock simply isn't possible. India, which has been plagued with a chronic shortage of banknotes since the November announcement, is unlikely to meet its citizens' demand for cash until well into 2017. This in turn has hurt the Indian economy.

Like India, any Italian effort to print enough new lira paper to meet domestic demand could take months to complete. Without sufficient paper lira, existing euro banknotes would have to meet Italians' demand for paper money. Under this scenario, Italians would have to endure a messy mixture of electronic lira circulating with paper euros, reminiscent of the old bimetallic, or silver & gold standards, of yore. I say messy because everything in Italy would have two prices, a lira price and a euro price. In some ways this would be similar to the euro changeover period in 2002 when European shops displayed both euro and legacy currency prices on their shelves (lira, deutschemarks, francs, etc). The difference is that in the 2002 changeover the exchange rate was fixed, so the amount of mental arithmetic devoted to calculating exchange rates was minimal. In the case of Italexit and a new lira, the price of the lira would likely float relative to the euro, so the mental gymnastics required of Italians would be much more onerous.

If the Italian government were to attempt to fix this messiness by forcing retailers to accept euros and lira at a fixed rate, then Gresham's law would probably kick in. Because the government's chosen ratio is likely to overvalue the lira and undervalue the euro, Italians would hoard their paper euros (preferring to use them in Germany and elsewhere) while relying solely on electronic lira to buy things. This hoarding of paper euros, and the ongoing lack of paper lira, would likely lead to a severe banknote shortage, much like the shortages that India and Zimbabwe are currently enduring.

Some readers are probably wondering why Italy wouldn't try printing new banknotes ahead of the redenomination date. That way it could engineer a rapid lira changeover rather than a slow one. The problem here is that if Italian authorities take a preemptive approach, odds are that word will leak out that new lira are being printed, and depositors—spotting an impending redenomination coming down the road—will flee the Italian banking system en masse. So we are right back where we started. A successful Italexit requires that new lira banknotes be printed only after the redenomination has been announced, not before.

One technique that Italy might try in order to get lira paper into circulation as rapidly as possible is to use overstamping (described here). Once redenomination had been announced, Italian authorities quickly produce a large quantity of special stickers or stamps. They would then require Italians to bring in their euro banknotes to banks in order to be stamped, upon which those overstamped euros would no longer be recognized as euros, but lira. The window for getting euros stamped would last a week or two, after which the Italian government would declare that all remaining euro notes are no longer fit to circulate in Italy. Stamped notes would function as Italy's paper currency until the nation's printing presses have had the time to print genuine lira paper currency, at which point Italians would be required to bring stamped notes in for final conversion.

But even here Italy runs into a problem. An Italian with a stash of euro banknotes can either bring this stash in to be overstamped, and eventually converted into lira, or they can break the law and hoard said euros under their mattress. Hoarded euro note will still be valuable because they can be used to buy stuff in Germany, France, and in other remaining eurozone members. An overstamped euro, however, which has effectively been rebranded as lira, will be worth much less than before. Many Italians will therefore avoid getting their money stamped, preferring to get more value for their euro banknotes than less. And this means that Italy is likely to suffer a severe cash crunch, with euros being hoarded and new lira unable to fill the void fast enough, yet another manifestation of Gresham's law.

So any would-be euro exiteer faces several ugly possibilities, including a messy period of multiple prices to massive cash crunches.

It is because of these difficulties (and others) that I see euro exit as an incredibly unlikely proposition. The euro isn't a glove, it's a Chinese finger trap—once you've got it on, it's almost impossible to get out.

 ---

If there is to be an exit, the most likely one will be the euros without the Eurozone approach. Rather than announcing a new lira, Italy simply says that it will officially leave the Eurozone while continuing to use the euro unofficially. This means that Italian banks would continue to denominate deposits in euros and keep euro banknotes in reserve to meet redemption requests. The euro would still be used by Italian merchants to price goods, and euro banknotes would continue circulating across the nation. The difference now would be that the Bank of Italy would no longer have the authority to print euro banknotes. Instead, Italy would have to import banknotes from the rest of Europe, much like how Panama—a dollarized nation—imports U.S. banknotes from the U.S., as do Zimbabwe and Ecuador. By employing this sort of strategy, all the hassles I wrote about in this post (multiple prices & cash crunches) are cleanly avoided while at the same time an exit of sorts is achieved.

The shrinking rupee

Earlier this month I criticized the architects of India's recent note demonetization for not using the traditional overstamping technique for replacing large quantities of banknotes.

This week I want to examine another feature of Modi's demonetization: the concurrent change in note sizing. The new series of ₹500 and ₹2000 notes are smaller in size than the ₹500 and ₹1000 series that they have since replaced. This has caused huge logistical problems. Since each cartridge in an ATM must be manually configured to handle a certain note size, ATMs were not equipped hold the newly issued ₹500s, ₹2000s, or additional ₹100s for that matter. Instead, they were forced to operate at a fraction of their capacity. Indians, desperate to replace their demonetized notes with good cash, were left on the lurch.

Let's explore the reduction in banknotes size. I'd argue that independent of the decision to crack down on black money, the decision to go smaller makes a lot of sense. But twinning a banknote size reduction with a demonetization was a recipe for disaster.

Consider that the length of the current issue of rupee banknotes grows as the denomination increases, like this:

Denomination: width x length

₹100: 73mm x 157mm

₹500: 73mm x 167mm

₹1000: 73mm x 177mm

To Americans and Canadians, this may seem odd since all our money is the same size. However, a pattern of progressively longer notes is quite common in other countries. Euro banknotes, for instance, also increase in size as denomination rises as do Swiss francs and Japanese yen. Presumably this format is chosen to to make manual sorting easier.

Now if the Reserve Bank of India, the nation's central bank, had continued to follow its traditional size progression, the newly issued 2000 rupee note would have had these measurements:

₹2000: 73mm x 187 mm

This would have been an awfully big note, one of the largest in the world by surface area. It would have clocked in 32% larger than a US$20 bill, for instance, and 43% larger than a 20 euro note. Not only would a note of this size have been expensive to print, but the combined costs of storage and handling incurred by hundreds of millions of Indians over time would have been quite large. Reducing the size would cut down on both expenses.*

The trend among central banks is to reduce the dimensions of banknotes. For instance, euro banknotes are quite a bit smaller than the francs, deutschmarks, and other notes that they replaced. The five euro note is one of the smallest notes in the world (see this pdf). When the Swiss began to introduce the ninth generation of Swiss banknotes in 2016, they lopped around 11 mm off the length of the 50 franc note and 4mm off its height (it now clocks in at 70 x 137 mm, down from 74 x 148).  By doing so, the Swiss National Bank will be lowering manufacturing and handling costs of the currency. In the chart below, you can see the evolution of the dimensions of Swiss cash over time.

Data source: Wikipedia

So India's decision to reduce the size of the new notes is very much modern practice. 17mm has been removed from the length of the ₹500 note; it measures 150 mm rather than 167mm. As for its height, it has gone from 73mm to 66mm. The new ₹2000 note measures 66mm x166mm, a 20% reduction from what it would have measured had the RBI continued with its old progression. Presumably the RBI will eventually do the same with the smaller denomination like the ₹100 as well.

While a note size reduction makes sense, twinning it with an aggressive demonetization was a bad decision. To reduce the odds of damaging the economy, the void left by demonetized notes must be filled as rapidly as possible. In India's case, the discontinuity in banknote size interfered with this re-cashification process. The authorities should have split the two policies apart, say by enacting a gentle two or three-year conversion of existing notes to a new and smaller series, and only announcing a surprise aggressive demonetization of the two highest denomination notes four or five years from now, say in 2021.

Alternatively, the authorities could have proceeded with their November 8 aggressive demonetization, but without enacting a note size reduction. The RBI should have taken incoming demonetized 500 and 1000 rupee notes and stamped them for re-circulation to ensure the banknote supply was sufficient, as I went into here. By using existing banknotes, ATMs cartridges would not have required adjustment. As for the new ₹500 note, it should have been the same size as the old series to ensure that cash distribution worked smoothly. Then—say five years from now—the Reserve Bank of India could have gradually phased in reduced note sizes for the entire range of denominations from the ₹10 to the ₹2000. This would have cut down on the chaos of the last few months.

The RBI seems not have been involved much in the planning stage of demonetization. According to recent press reports, the Board was "asked" to consider the demonetization just a day before it was enacted and had not discussed the matter before then. This is a shame. While an aggressive demonetization needs to be planned in secret, as I pointed out here, having at least a few closed-mouth central bank types involved from the outset seems like a basic requirement. They might have been able to fix the mistake of combining a demonetization with a note reduction.

---

*On the other hand, there are arguments for increasing note size too. See Garber.

Modi and the overstamping of demonetized currency

1913 Austro-Hungarian banknote with 1919 Czechoslovak overstamp

When Indian PM Narendra Modi and his small group of would-be monetary architects were putting together their plan to suddenly demonetize the 500 and 1000 rupee note and replace them with new notes (including a new 2000 rupee note), they must have been missing a monetary historian. That's because there is a long history of nations suddenly demonetizing the entire circulating paper issue and introducing a new paper currency. These rapid switches have tended to follow a well-trodden script, one that Modi did not follow. Had he chosen to adopt it, the last two months might have been less chaotic.

One challenge faced by any prospective note switcher is to print the new currency fast enough to replace the legacy notes. When the switch is a slow one that is planned long beforehand, like the euro introduction, this is not an issue. In the case of a rapid switch that cannot be prepared for, however, the printing challenge is overwhelming. In India's case, pre-printing notes was not the answer. Because its goal was to surprise a large number of cash-users with undeclared cash, the rupee switch had to be sudden—printing large batches of notes ahead of time might have tipped off the prey. Without enough currency, however, an economy undergoing a switch is cursed to endure a temporary cash crunch, much like the one India has experienced. To cope with the period between the demonetization of the old notes and the issuance of new ones, nations have resorted to an old monetary trick called overstamping.

The 1993 conversion of old Czechoslovak koruna into new Czech and Slovak money is the best modern example of a successful rapid currency switch, thanks in no small part to its use of the overstamping technique. With the January 1993 dissolution of Czechoslovakia into Slovakia and the Czech Republic, the public also anticipated an ensuing breakup of the still-circulating Czechoslovak koruna into two national currencies. Since Czech was expected to be the more economically robust of the two nations, depositors began to move their paper money and bank accounts to the Czech side of the border to ensure their savings would be held in the stronger of the two soon-to-be currencies.

Anxious to put an end to capital flight before it crippled the monetary systems of the fledgling nations, the Czechs and Slovaks were forced to introduce new monetary units ahead of schedule. Unfortunately, the banknotes hadn't been printed yet. As a temporary expedient, the monetary authorities decided to affix different coloured stickers, or stamps, to existing Czechoslovak banknotes in order to demarcate them as either new Czech koruna or new Slovak koruna.

In executing a currency switch, affixing stamps on existing legacy notes has several advantages over relying entirely on new banknotes. Stamps take far less time to design and print than banknotes, they can be rapidly distributed thanks to their small size, and they are cheaper to store. In the case of Czechoslovakia, for six days in February 1993 all koruna banknotes were brought in so as to have the proper Czech or Slovak stamp affixed to them (below is a legacy koruna note with a Czech stamp on it). After the six day period passed, any unstamped currency was declared worthless. Cross border movements of cash between Czech and Slovakia were made illegal for that period and cash withdrawals from banks suspended.

1985 Czechoslovak koruna with yellow overstamp (top left)

Once the new notes had been printed up, Czechs and Slovaks could bring their stamped legacy notes to the bank to get new ones. All in all, it was a relatively smooth currency switch.

As I said at the outset, there is a long history to currency swaps. In 1919, the existing Austro-Hungarian krona was dismembered, with each of Czechoslovakia, Serbia, Hungary, Austria, and Romania printing unique stamps to be affixed to the krona circulating in their jurisdiction. Once new notes had been printed, stamped krona could be brought in for redemption. Peter Garber and Michael Spencer go into some detail on this episode here. As for the general practice of affixing stamps to money, I'd suggest Arnold Keller and John Sandrock here.

Zooming forward in time, to fill the void vacated by demonetized banknotes Modi's planning team decided to rely entirely on the ability of the Reserve Bank of India (RBI) to print new currency. And that's where its problems began; with 24 billion pieces needing to be printed, there simply wasn't enough time.  Blogger James Wilson calculates the replacement effort might not be completed till as late as September 2017:

Long lineups developed at banks and, without enough cash to go around, trade has been impaired. Had it borrowed from the above precedents, Modi's demonetization team might have been able to avoid at least some of this mess by procuring a supply of several billion worth of stickers or stamps. Some of these could have been printed and hoarded by the government in the months prior to November 8, the rest could have been printed on demand after the announcement date. The trick would be to ensure that the stamps had enough special security features so that, once issued, they succeeded in throwing off counterfeiters, who would have a strong incentive to make fakes for sale to those who had large amounts of undocumented cash.

Under a stamp scheme, come the November 8 demonetization announcement the RBI would have distributed the stamps to banks. Members of the public could then bring their demonetized notes to their local bank in order to have them stamped, up to a certain quota. These stamped notes would quickly reenter circulation, helping close the void left by the sudden demonetization of 80% of India's money supply and the slow trickle of newly-issued 500 and 2000 rupee notes entering the economy. Once the RBI had managed to print enough 500 and 2000 notes, it would set some fixed conversion period for turning in stamped 500 notes and 1000s for new notes.

It could be that Modi's team considered a stamping mechanism and decided against it for some justifiable reason that I'm not aware of. Or perhaps they simply didn't do their homework. If they didn't, overstamping could very well have diminished the shock faced by the Indian economy over the last few weeks. Countries that decide to follow India's path in the future would do well to include it in their plans.