Friday, July 8, 2016

Hyperinflation 2.0?

If you haven't heard, protests are breaking out in Zimbabwe and unpaid civil servants are going on strike. This sort of thing hasn't happened in many years.

It's possible to trace at least some of the motivation for these developments to monetary mischief. Over the last twenty years, no nation has suffered more problems with its money than Zimbabwe has. Everyone remembers the hyperinflation and subsequent dollarization in late 2008. The most recent episode has seen a nation-wide bank run break out as Zimbabweans queue at ATMs to withdraw U.S. dollars, the local currency.

Remember last year's Greek bank run? I'd argue that Zimbabwe's bank run is similar. If you recall, Greek depositors were worried that—in the event of a Greek exit from the Eurozone—their deposits would be redenominated from euros to a Greek version of the euro or even a new drachma. Better to cash out in good euros before getting stuck with something worse. Line-ups grew outside Greek banks until authorities had no choice but to shut the system down.

Like Greece, there is a decent chance that Zimbabwean bank deposits might be made payable in funny money, namely a Zimbabwean version of the U.S. dollar rather than the actual U.S. banknotes. This may explain why Zimbabweans have been desperately queuing up at bank machines—they want to cash out before the worst case scenario happens.

To understand what I mean by 'Zimbabwean version of the U.S. dollar', we need to take a quick tour of the Zimbabwean banking system. A nation's central bank usually runs the plumbing that connects local banks. These banks keep accounts at the central bank—in Zimbabwe's case the Reserve Bank of Zimbabwe (RBZ)—and use balances held in these accounts to clear and settle among each other. These accounts, along with central bank-issued banknotes, constitute a nation's supply of base money, the quantity of which determines its price level. When Zimbabweans spontaneously stopped using the local currency, the Zimbabwean dollar, in late 2008, RBZ accounts (and cash) became worthless. The RBZ-managed plumbing system had imploded.

Zimbabweans still needed to bank, however, so local banks soon began offering U.S. dollar accounts to clients. A new plumbing system was re-erected overseas; instead of maintaining clearing accounts at the now defunct Reserve Bank of Zimbabwe, local banks held U.S dollar accounts at banks in Europe and the U.S., otherwise known as nostro accounts. They used these offshore accounts to settle interbank Zimbabwe payments. [1]

To understand how this offshore plumbing system worked, say Joseph (who lives in the capital Harare) writes a cheque to Robert (who lives in Bulawayo). Joseph's local bank might settle the cheque thousands of miles away by having its New York bank wire funds to the nostro account of Robert's bank, which might be based in London. Circuitous, right?

As for cash, say Joseph wants to withdraw $100,000 in U.S. banknotes from his Zimbabwe bank account. His bank would request its New York bank to debit its nostro account by $100,000 and then ship the $100,000 in banknotes to Zimbabwe. Joseph now has a suitcase full of Ben Franklins.

This offshore plumbing system worked pretty well. However, it didn't take long for the RBZ to re-insinuate itself into the works by offering local banks U.S. dollar accounts. These accounts allowed the local banks to use the RBZ's re-christened real-time gross settlement system (RTGS) to settle interbank payments rather than using the offshore plumbing system. After having lost its printing press, the RBZ had got back into the monetary printing game. It had created a Zimbabwean version of the U.S. dollar.

My understanding is that as time passed the RBZ forced local banks to "repatriate" their clearing accounts from the overseas system and deposit them at the RBZ. In effect, local banks were told to wire U.S. funds from their foreign-based nostro accounts into an RBZ account held at a European/American bank. In turn, the local bank was credited with an equal quantity of U.S. dollar deposits on the RBZ's own books. Voila, local banks had gone from holding U.S. dollars in relatively safe foreign banks located in places like London to holding the domestic RBZ version of the dollar. I can't imagine that bank managers were terribly fond of this forced switch given the RBZ role in igniting the 21st century's first hyperinflation.

Let's see how this new system works. Now when Joseph wants $100,000 in cash, Joseph's bank—call it the Commonwealth Bank of Zimbabwe—has two choices. Use its foreign nostro account as before. Or it can ask the RBZ to debit the Commonwealth Bank RTGS account and provide the proper number of U.S. banknotes. The RBZ in turn sources the cash by requesting its foreign bank to debit the RBZ account—now plump with confiscated dollars—and send the cash to Zimbabwe by plane. The RBZ's overseas dollar accounts in effect "back" the dollar deposits that the RBZ has issued to local banks.

On paper this sort of system should work fine... as long as the RBZ doesn't abscond with the funds in the foreign bank accounts. Unfortunately, this may be exactly what happened. The RBZ had effectively gone from being bankrupt to having amassed large amounts of U.S. funds overseas. This proved tempting, and according to former finance minister Tendai Biti the regime began dipping into the RBZ's foreign stash to pay for expensive junkets and to finance public sector salaries. The upshot it that there may not be enough U.S. funds in the RBZ's foreign accounts to back its promises to local banks.

This means that now when Zimbabweans go to their banks to get U.S. cash, the banks—which before had no problems meeting these requests via their nostro accounts—are hamstrung. They have U.S. dollar accounts at the RBZ but the RBZ is unable to draw on its depleted overseas accounts to get the requested cash. The lineups that have developed are the public's attempt to squeeze out whatever spare dollars remain in the system, an attempt that is rendered much hard given the withdrawal limits that have been instituted to slow down the run.

Zimbabweans are already starting to see a divergence between the price of an electronic dollar and a paper dollar. Various media reports say the practice of "cash burning" has re-emerged for the first time since the hyperinflation of 2007-08. Anyone who needs to convert deposits into cash, frustrated by long lines at ATMs and withdrawal limits, can instead approach an informal dealer who offers to buy their deposits at a discount of 10-20% of their cash value (see here and here). Think of the 'cash burning' discount as the market value of an RBZ-backed bank deposit. If the regime has indeed wasted all the money in its nostro accounts, this discount will only widen.

The theory that the regime has absconded with the RBZ's overseas funds is consistent with a flurry of official proclamations over the last month or two. If the RBZ is indeed bankrupt, it would make sense for the ruling regime to adopt the same strategy that Greece did last year; implement capital controls to trap as many U.S. dollars in the banking system as possible, thus limiting the damage and buying time for the government to rebuild the balance sheets of both the RBZ and the local banks before reopening for business. This would probably require some sort of loan from China or elsewhere. Under this scenario, Zimbabwean deposit holders could very well have to take a large haircut.

As in Greece, the RBZ has started to ring-fence the system by instituting daily withdrawal limits (of around $100); enough to allow Zimbabweans to get by but not enough to hurt the banking system. To coax people into accepting electronic dollars rather than paper dollars, the central bank has suddenly decreed much lower fees on bank payments and transfers. The government has also invoked the Bank Use Promotion and Suppression of Money Laundering Act, which punishes citizens and business if they refuse to deposit their money in banks. More radically, it has imposed severe import restrictions on a broad variety of goods from furniture to beans to fertilizer, a policy that presumably prevents cash leaking over the border. Together, all these regulations seem designed to help stuff as many U.S. banknotes back into the RBZ as possible.

Alternatively, it's possible the Zimbabwe government cribs from the Argentina play book and sets up a corralito, or coral, followed by a redenomination of dollar accounts into the local unit. Unlike Argentina, which had pesos, Zimbabwe is fully dollarized and doesn't have its own paper currency in which to redenominate deposits. But so-called bond notes (which I wrote about last month), an issue of paper money set to debut this fall in denominations of $2, $5, $10 and $20, may be a step in the Argentinean direction. Rather than meeting conversion requests by providing U.S. dollars, the RBZ will be able to print off any quantity of bond notes it deems necessary. In this way U.S. dollar claims on Zimbabwean banks will cease to be payable in actual dollars but in the RBZ's peculiar brand of U.S. banknotes, probably worth far less than the real thing.

It seems perverse that Zimbabwe could see another hyperinflation while on the very dollar standard that was meant to immunize it from a hyperinflation scenario, but I'm starting to worry this could happen. Consider that Robert John Mangudya, the head of the RBZ, claims that retailers are beginning to put two different price tags on one product, a higher one for electronic payments and a lower one for cash. If the RBZ-issued electronic dollar continues to inflate then electronic dollar sticker price will rise but the U.S. paper dollar price will stay constant. This second set of prices would at least provide some modicum of price stability to the nation.

Not so fast. Mangudya warns that the central bank will prosecute any retailer that sets two prices. If retailers comply and set only one price for their wares, that effectively undervalues U.S. banknotes and overvalues RBZ-issued U.S. electronic dollars. Gresham's law will take hold as shoppers use only bad electronic dollars to pay for things while hoarding their good, and undervalued, paper dollars in their wallets. Unwilling to be the dupes and accumulate overvalued and unwanted electronic dollars, retailers will have no choice but to jack up their prices, essentially adopting the RBZ U.S. e-dollar as the standard unit of account, or unit in which they set prices. With U.S. dollars no longer being used as a medium of exchange and unit of account, price stability in Zimbabwe will cease to exist.

One hopes that rumors that the regime has absconded with the RBZ's funds are false and that the current bank run and potential inflation is just a temporary spate of animal spirits. But in my experience, most sustained bank runs are underpinned by something real.

[1] I get much of this information from here.


  1. Interesting that retailers are starting to post two price tags. Miles Kimball believes that this won't happen when he breaks par between bank money and central bank notes, because he doesn't expect the two to diverge by much.

    1. Yes, I think the valuation gap between U.S. paper money and the RBZ version of the U.S. dollar is wider than anything Miles Kimball envisions for his crawling peg plan.

    2. You say above that the cash burning is at a rate of 10-20% in Zimbabwe. Perhaps Kimball doesn't envision anything like that in the US, but the proponents of a 2% inflation target also did not expect the ZLB ever to be a problem.

      It's fine to break par between bank money and central bank notes. But there is no point in calling for bank money to be the numeraire (as Kimball does) and then hoping that merchants will comply with this.

    3. "...but the proponents of a 2% inflation target also did not expect the ZLB ever to be a problem."

      Yep, good point.

      "...and then hoping that merchants will comply with this."

      In the section entitled "What If the Government Has Trouble Establishing Electronic Money as the Unit of Account?" in this paper...

      ...Kimball and Agarwal go into some more depth on enforcing bank money as numeraire.

    4. Sure, they can get people to choose bank money as the accounting unit for keeping records for and writing contracts. But Kimball is also hoping that we don't see the appearance of dual price tags at stores. That is a key feature of his proposal, but we really don't know what to expect. I remember low-margin electronics stores here in Toronto using a dual price system, to pass on the cost of merchant fees (less than 3%) to customers. I'm not sure why that stopped - perhaps it conflicted with rules set by the payments networks. But if we have deep negative rates for a couple of years in a row, the accumulated discount on bank notes could be large enough for differentiated prices to reappear, at least for big-ticket items.

    5. Don't you think government can legislate the use of electronic money as the unit of account in stores? ... say by introducing large fines on those who violate the rules? For many years Visa and Mastercard have managed to prevent credit card surcharges--essentially forcing merchants to adopt credit card money as the unit of account. You mention seeing surcharges in Toronto electronic stores but in my experience surcharging in Canada is very rare. If the networks can dictate the unit of account on a retail level, so can the government, no?

    6. Yes, the government has even more power than Visa and Mastercard to dictate the way we conduct transactions. But our government is democratic, and such policies would be very unpopular. This is a question some people raise about Kimball's proposal: whether or not the public would accept it.

    7. Yes, the unit of account switch might be a bit intrusive. Central banks may hesitate to introduce a switch without getting support from government (i.e. by having them alter central bank legislation) and governments may hesitate to enact legislation without getting support of voters, say by a referendum.

      There are a few ways to get further below zero without requiring a unit of account switch, but they don't allow for deeply negative rates:

  2. Cancelling the planned issue of bond notes would remove that element of uncertainty and might calm things down.

  3. JP-

    What do you mean with this line?

    "constitute a nation's supply of base money, the quantity of which determines its price level."

    Since 2009 the US supply of reserves has increased by more than 400% ($900B to $3.7T) and yet the price level (CPI) has gone up just 11% (211 to 236)

    In the equivalent 7 year period before 2009 the supply of reserves in the US dollar system went up 22% ($670B to $870B) and the price level (CPI) had gone up 16% (177 to 211)

    QE has definitively proven that base money does not drive anything, least of all the price level.

    1. That line is a bit sloppy. I should have said something like:

      "...the quantity of which is one of the determinants of its price level"

      As you point out, the supply of dollars has exploded since 2009 but prices didn't rise. We also need to think about the demand for base money, which jumped quite dramatically over the same time period.

  4. Very insightful article, great work! Just one thing though, the Reserve Bank Governor is John Mangudya (not Robert Mangudya)

  5. Zdollars are helicopter money. Backed by thin air. Same as BoE issuing reserves backed by Adair's zero coupon perpetuals.
    Credible money is a receipt issued on Credible collateral.

    Of course you can get away with it for a while. Enron did. Worth 50bn one moment and zero the . What changed?