Tuesday, June 25, 2019

Esperanto, money's interval of certainty, and how this applies to Facebook's Libra


Facebook recently announced a new cryptocurrency, Libra. I had earlier speculated about what a Facebook cryptocurrency might look like here for Breakermag.

I think this is great news. MasterCard, Visa, and the various national banking systems (many of which are oligopolies) need more competition. With a big player like Facebook entering the market, prices should fall and service improve, making consumers better off.

The most interesting thing to me about Facebook's move into payments is that rather than indexing Libras to an existing unit of account, the system will be based on an entirely new unit of account. When you owe your friend 5 Libras, or ≋5, that will be different from owing her $5 or ¥5 or £5.  Here is what the white paper has to say:
"As the value of Libra is effectively linked to a basket of fiat currencies, from the point of view of any specific currency, there will be fluctuations in the value of Libra."
So Libra will not just be a new way to pay, but also a new monetary measurement. Given how Facebook describes it in the brief quotation provided, the Libra unit will be similar to other unit of account baskets like the IMF's special drawing right (SDR), the Asian Monetary Unit (AMU), or the European Currency Unit (ECU), the predecessor to the euro. Each of these units is a "cocktail" of other currency units.

Facebook's decision to build its payments network on top of a new unit of account is very ambitious, perhaps overly so. When fintechs or banks introduce new media of exchange or payments systems, they invariably piggy back off of the existing national units of account. For instance, when PayPal debuted in 2001, it didn't set up a new unit called PayPalios. It used the dollar (and for the other nations in which is is active, it used the local unit of account). M-Pesa didn't set up a new unit of account called Pesas. It indexed M-Pesa to the Kenyan shilling.

I couldn't find a good explanation for why Facebook wants to take its own route. But I suspect it might have something to do with the goal of providing a universal monetary unit, one that allows Facebook users around the globe to avoid all the hassles of exchange fluctuations and conversions.

Global monetary harmony an old dream. In the mid 1800s, a bunch of economists, including William Stanley Jevons, tried to get the world to adopt the French 5-franc coin as a universal coinage standard. Jevons pointed out that the world already had international copyright, extradition, maritime codes of signals, postal conventions—so why not international money too? He wrote of the "immense good" that would arise when people could understand all "statements of accounts, prices, and statistics." It would no longer be necessary to employ a skilled class of foreign exchange specialists to take on the "perplexing" task of converting from one money to the other.

But the plan to introduce international money never worked out. (I wrote about this episode for Bullionstar).

Global money like Libra might seem like a great idea. But ultimately, I suspect that the decision to introduce a new unit of account will prevent Libra from ever reaching its full potential. Units of account are a bit like languages. If you are an English speakers, not only do you communicate to everyone around you in English, but you also think in English. Likewise with the dollar or yen or pound or euro. If you live in France, you're used to describing prices and values to friends and family in euros. You also plan and conceptualize in terms of them.

It's hard to get people to voluntarily switch to another language or unit of account once they are locked into it. For instance, in the 1800s L.L. Zamenhof attempted to get the world to adopt Esperanto as a language in order to promote communication across borders. To help facilitate adoption, Zamenhof designed it to be easy to learn. But while around 2 million speak Esperanto, it never succeeded in becoming a real linguistic standard. The core problem is this: Why bother learning a new language, even an easy one, if everyone is using the existing language? 

Facebook's Libra project reminds me of Zamenhof's Esperanto project. Nigerians already talk and compute in naira, Canadians in dollars, Indonesians in rupiahs, and Russians in rubles. Why would any of us want to invest time and effort in learning a second language of prices?

Let me put it more concretely. I do most of my families grocery shopping. Which means I keep track of an evolving array of maybe 30 or 40 food prices in my head. When something is cheap relative to my memory of it, I will buy it—sometimes multiple versions of it. And when it is expensive, I avoid it. But this array is entirely made up of Canadian dollar prices. I don't want to have to re-memorize that full array of prices in Libra terms, or keep two arrays of prices in my head, a dollar one and a Libra one. I'm already fluent in the Canadian dollar ones.

Nor will retailers like Amazon or the local corner store relish the prospect of having to advertise prices in both the local unit of account and Libra, plus whatever unit Google and Netflix choose to impose on us. 

So Facebook is inflicting an inconvenience on its users by forcing us to adopt a new unit of account. To make for a better user experience, it should probably index the Libra payments network to the units of account that we're all used to. 

If not, here is what is likely to happen. We'll all continue to think and communicate in terms of local currency. But at the last-minute we will have to make a foreign exchange calculation in order to determine out how much of our Libra to pay at the check-out counter. To do this calculation, we'll have to use that moment's Libra-to-local currency exchange rate. This is already how bitcoin transactions occur, for instance.

But this means that Libra users will lose one of the greatest services provided by money: money's interval of certainty. This is one of society's best free lunches around. It emerges from a combination of two fact. First, most of us don't live in a Libra world in which we must make some sort of last-minute foreign exchange calculation before paying. Rather, we live in a world in which the instruments we hold in our wallet are indexed to the same unit of account in which shops set prices.

Monetary economists call this a wedding of the medium-of-exchange and unit-of-account functions of money. This fusion is really quite convenient. It means that we don't have to make constant foreign exchange conversions every time we pay for something. A bill with a dollar on it is equal to the dollars emblazoned on sticker prices.

Secondly, shops generally choose to keep sticker prices fixed for long periods of time. Even with the growth of Amazon and other online retailers, Alberto Cavallo (who co-founded the Billion Prices Project) finds that the average price in the U.S. has a duration of around 3.65 months between 2014-2017. So for example, an IKEA chair that is priced at $15 will probably have this same price for around 3.65 months. This is down from 6.48 month between 2008-10. But 3.65 months is still a pretty long time.

Why do businesses provide sticky pricing? In the early 1990s Alan Blinder asked businesses this very question. He found that the most common reason was the desire to avoid "antagonizing" customers or "causing them difficulties." Blinder's findings were similar to Arthur Okun's earlier explanation for sticky prices whereby business owners maintain an implicit contract, or invisible handshake, with customers. If buyers view a price increase as being unfair, they might take revenge on the retailer by looking for alternatives. (I explore these ideas more here).

Anyways, the combination of these two factors—sticky prices and a wedding of the unit of account and medium of exchange—provides all of us with an interval of certainty (or what I once called money's 'home advantage'). We know exactly how many items we can buy for the next few weeks or months using the banknotes in our wallet or funds in our account. And so we can make very precise spending plans. In an uncertain world, this sort of clarity is quite special.

Given Libra's current design, the interval of certainty disappears. Store keepers will still keep prices sticky in terms of the local unit of account, but Libra users do not benefit from this stickiness because Libras aren't indexed to the same unit as sticker prices are. Anyone who has ≋100 in their account won't know whether they can afford to buy a given item two weeks from now. But if they hold $100, they'll still have that certainty, since dollar prices are still sticky.

If money's interval of certainty is important, it is particularly important to the poor. The rich have plenty of savings that they can rely on to ride out price fluctuations. The fewer resources that a family has, the more it must carefully map out the next few day's of spending.  The combination of sticky prices and a wedding of the unit-of-account and medium-of-exchange affords a vital planning window to those who are just barely getting by.

This clashes with one of Libra's founding principles: to help the world's 1.7 billion unbanked. Here is David Marcus, Libra's project lead:

Most of the world's unbanked people are poor. But Libra won't be doing the poor much of a favor by choosing to void the interval of certainty that they rely on. If Facebook and David Marcus truly wants to help the unbanked, it seems to me that it would better to index Libras to the various local units of account.

I suppose there is an argument to be made that Libras could provide poor people in nations with bad currencies a haven of sorts. Better Libras than Venezuelan bolivars, right? But the nations with the world's largest unbanked populations—places like India, Nigeria, Mexico, Ethiopia, Bangladesh, and Indonesia—all have single digit inflation, or close to it. Extremely high inflation is really just a problem in a few outliers, like Zimbabwe and Venezuela.

Besides, providing those who endure high inflation with a better unit of account isn't the only way to help them. Offering locally-denominated Libras that offer a compensating high rate of interest would probably be more useful. Not only would these types of Libra offer inflation protection, but they would preserve the interval of certainty.

Thankfully, I suspect that Libra is very much a work-in-progress. The current whitepaper seems to give only a hint of what the project might become. If so, one of the changes I suspect Facebook will have to make if it wants to get traction is to link the Libra network to already-existing units of account. A new unit of account is just too Utopian.

11 comments:

  1. A lot of the unbanked are in countries where gold is used for protecting themselves against local currency devaluation and as such probably "think" in terms of gold ounces, so Libra could have used gold as the backing unit.

    It seems to me that going with a new unit of account is so obviously wrong as you note that I think it was more about not wanting to have to pick a single currency (most likely USD) at this early stage while they are trying to "shape" the regulators, as that would put other countries offside.

    The fact that they have not defined what is in their new basket is deliberate as they want the major countries to think their money & govt debt will be part of the basket but I guess later (after getting various country regulators onside) they will come out and say on further research/reflection (as per your points in this article) that a basket won't work so we will go with X currency.

    ReplyDelete
    Replies
    1. Good points, Bron. I had attributed their design choice to the age-old dream of creating international money, but it could also be driven by regulatory concerns, as you point out.

      Delete
  2. A very good post. I agree, depositing in a Libra denominated account curtails one’s interval of certainty. But does the historical parallel of European banks that offered their own unit of account, e.g. Hamburger Bank’s Mark Banco, indicate that the benefits from holding a part one’s liquidity in “Esperanto deposits” too is significant? No, I don’t have statistics for the relative size of Mark Banco deposits. “Why bother learning a new language, even an easy one, if everyone is using the existing language?” With others now speaking the language, the advantages of becoming fluent in Esperanto are currently higher than in 1887. Is your concern the cost-benefit calculation of potential pioneer Libra depositors?

    ReplyDelete
    Replies
    1. You're right that we've had other instances where the medium of exchange and unit of account have been separated. You mention the Hamburger Bank, which I don't know much about, but there was also the American colonial experience.

      My feeling is that these sorts of systems were an evolutionary and spontaneous response to early monetary fragmentation and/or monetary disasters. But over time, societies inevitably move towards more standardized systems where the unit of account and medium of exchange are unified. Facebook's effort doesn't seem to fit.

      Delete
  3. Good post. (My thought on your passage about the price array in memory for grocery shopping: it's hard enough for me trying to keep $/kg vs $/lb straight, even though there's a fixed exchange rate between kg and lb.)

    ReplyDelete
    Replies
    1. Thanks Nick. That's my experience with kg-to-lb too. I have to look it up on the internet every dang time. And forget about Celsius-to-Farenheit. I've never understood that one.

      Delete
  4. Correct me if I am wrong, but the difference between a SDR and Libra is that the SDR has some formulaic and transparent link to the basket of currencies it tracks. The Libra does not appear to have that - rather the link appears to be opaque and (worse) at the arbitrary decision of FB?

    It will be interesting to see how the 'on-ramp' in currency-controlled countries works. You give FB your Bolivars, and FB gives you some Libra - but then what does FB do with the Bolivars? Stick in the basket that Libra is linked to?

    ReplyDelete
    Replies
    1. There is very little information on how they are planning to operate the system. I took the following as a sign that it was going to be managed like the IMF's SDR unit of account:

      "Since Libra will be global, the association decided not to develop its own monetary policy but to inherit the policies of the central banks represented in the basket."

      So that would seem to indicate a formulaic approach to determining the Libra's value. But at this point, it's hard to know for sure.

      Delete
    2. Thanks. I guess my (cynical) assumption is that when a seller of a product/service leaves a detail unspecified, the eventual specification of that detail will not normally be in the buyers favour. Caveat Emptor.

      Delete
  5. It might well take off in countries that lack much certainty in their home currency anyway. Venezuela, Zimbabwe, Argentina etc. Two big ifs:
    - If it's easier to access & use than USD, and
    - If the value of the Libra is perceived to be at least as stable as USD,
    then it will catch on in those sorts of markets.

    ReplyDelete
  6. As always an enlightening and high quality article. Question about M-Pesa, which is successful in a lot of African countries beside Kenia. Is M-Pesa always linked to the corresponding national currency ?

    ReplyDelete