The Libra reserve, discussion of background documents

Here is a 4-pp. appendix of sorts to the core Libra white paper, and it has some of the details that will be of most interest to monetary economists.  I have learned:

1. The Libra will be backed by a bundle of pretty safe, pretty mainstream assets (I don’t know which ones).  It is presented as one hundred percent reserve, though no system with fluctuating prices and also float really will be pure one hundred percent.  And the reserve is in “low-risk” assets, attention all critics of the Basel capital standards.

1b. The paper has a chance to say that the custodians will be separately capitalized, with no cross-collateralization, for purposes of Libra protection, but it does not do so.  I would recommend that change.

2. The assets in the reserve fund will come from users of Libra (how will they be charged?) and from “investors in the separate Investment Token.”  Furthermore “The funds for the coins that will be distributed as incentives will come from a private placement to investors.”

3. What about the public choice issues?  Won’t banks insist — correctly or not — that this represents competition and part of the payments system, and thus it should be brought under deposit insurance control and taxation, Fed regulation, various bank holding company acts, Monetary Control Act of 1980, and so on?  Have banks ever lost a political battle of this kind?

4. We are told “The association does not set monetary policy.  It mints and burns coins only in response to demand from authorized resellers.”  Maybe, of course there are hundreds of years of debate on that one, google “real bills doctrine,” noting that here we have a semi-dominant private issuer rather than a perfectly competitive banking system.  The association policy on interest rate spreads, floats, and credit, of course, can end up being a monetary policy de facto.  I don’t want to prejudge this one against Libra, since to me the validity of the real bills doctrine is a genuinely open question, but it is worth noting that most economists would not agree with the doctrine in most settings.

4b. Won’t some margins arise where there are fractional reserves, even if Facebook/association/Libra are not the ones doing it?  Imagine that a new class of intermediaries arises, offering some intermediate services between the core system and retail use, but not adhering to the 100% reserve provisions.  The logic behind this tendency seems pretty strong, for better or worse, and it can reintroduce risk into the system.  Someone wants to be holding higher yielding assets and then be making claims on them be liquid through the Libra system.  But Facebook/Libra would not seem to have the power to regulate the surrounding system of intermediaries, or is that somehow to be done through covenant (“you can’t use Libra unless you promise not to pile your intermediaries on top of it”)?

5. The crypto angle does seem like a sideshow, for me that is not a problem.

6. Imagine a private payment company issuing SDRs, or some other similar basket, based on 100% backing.  They would offer you new transactions technologies for greater convenience (WhatsApp?), in return receiving access to your transactions data and sharing some of the float and spread all around, to merchants and customers too.  Perhaps that is one way of thinking about how the plan works and where the gains from trade come from?

7. Is there a provision in the system for zero or low-interest loans?  Can I send small amounts of “libras,” say to pay my water bill, without first having them in my account?  Might sellers sign up to participate in such a system, sharing part of the credit risk with Libra?  And is there a way to do it, with crypto and layered assets and float and implicit positions, so that all this is not subject to the usual consumer credit regulations?  Is that part of how the system will make money and attract interest?  This is just speculation, my question marks here are literally question marks, not tricks to make you think that is how it will be.

8. “Who holds intraday credit risk?” is always a question worth asking.

9. Does any of this try to arbitrage away the fees earned by credit card companies for their intermediation?

10. What if the market for the underlying currencies and assets is (for a while?) more liquid than the market for Libras?  Say the basket values adjust before Libra values do.  What kind of arbitrage opportunities does that create?  If we know Libras are due to depreciate, is there a higher nominal rate of interest on them, as with traditional currencies in an international multi-currency setting?  What are the equivalents of covered and uncovered interest parity in this setting?  Does a kind of “program trading” arise to perform the arbitrage?  Can perfect redemption be offered credibly while the prices are still out of whack?

I still don’t feel I have a great handle on the plan, but those are my immediate reactions.  You should take them with a grain of salt, as they may be based on misunderstandings or perhaps even plan incompleteness.  I look forward to learning more.

Addendum: If anyone connected to Libra would wish to send more information or address these questions, I would gladly run that material on MR.


" hundreds of years of debate on that one, google “real bills doctrine,” - as TC says, this is a 19th century construct that I feel is 100% valid today. Basically it assumes money is neutral, short and long term. Very sound.

I prefer gold.

Except for the speculative gain potential, nothing about crypto-currency makes sense.

Crypto-currency really is not new. The technology is. What value does Bitcoin have outside Bitcoin users and its complex 'mining' systems? What value will Libra outside Facebook?

"Real bill" theory - bank notes were to be backed by commodities (gold) or by equally-valued assets. It would not contribute to inflation or an elastic money supply. The problem was often the assets were not there or were uncollectable due to credit problems, illiquidity, or whatever. When established in 1913, the law required Fed Reserve Banks to keep gold to back issued Federal Reserve Notes at one-third of the outstanding amounts. For example, each $20 FR Note had emblazoned on it, "Will Pay To The Bearer Twenty Dollars." That was far more valuable than "In God We Trust."

In 19th century, America, state-chartered banks issued bank notes (legal currency), which theoretically were backed by/redeemable in gold coins (money) and to a lesser extent by state bonds.

Some "Wildcat" banks issued notes up to 25x reserves. Only large NY bank notes were accepted at par. Other bank notes were either refused or discounted, especially far out of territory. Many bank failures were due to over-issuances of private bank notes without sufficient gold reserves. con men/counterfeiting/forgeries, "Counterfeit Detectors" published lists.

In 1836, President Jackson issued the "Species Circulars" wherein US land sales could only be paid in gold coinage. It burst speculative land bubble from too many bank notes (some were even collectable) in circulation.

Would "real bills" theory money expand and contract as commerce ebbs and flows? I think not.

Also, there was theoretical issuance of bank notes backed by the bank's loans, not government bonds, which would expand and contract with business activity.

The original 1913 Fed could elasticize the currency by lending, @ the discount rate, reserves to banks in need - usually liquidity shortages from defaulted loans or customers' deposit runs based on impressions that the banks could not pay their deposits. However, Fed discount loans must be secured by high-quality collateral, financial paper owned by the 'troubled' bank.

@Dick The Butcher: ""Real bill" theory - bank notes were to be backed by commodities (gold) or by equally-valued assets. It would not contribute to inflation or an elastic money supply. " - that's not quite really real bills theory. It's more like you cannot over issue bank notes ("money") in a fractional reserve system (that is, each bank note is fractionally backed by gold and the money supply can increase by 1/(1-r), where 0 < r < 1) so long as the borrowing pledges some assets, that is, the borrowing is collateralized. The Real Bills Doctrine says such asset-backed borrowing will never damage the economy. RBD is "pro-cyclical" to use monetary theory language. And I think it's empirically true. Business people expand their ambition during expansions and contract their ambition during contractions, what can be more logical? The idea that a central banker can 'dampen' animal spirits by somehow expanding or contracting the money supply is a quantity theory of money conceit dating back to the days of Copernicus.

Bonus trivia: it took 60 years after Copernicus died for his heliocentric theory to be scientifically PROVED. Even Tycho Brache, born after Copernicus, did not believe in the heliocentric model. Copernicus did not scientifically prove the earth goes round the sun anymore than the ancient Greek Aristarchus of Samos did.

Undeniably believe that that you said. Your favorite justification seemed to be on the net the easiest factor to bear in mind of.
I say to you, I certainly get irked while other folks think about worries that they plainly
don't realize about. You managed to hit the nail upon the top as well as
defined out the entire thing without having side effect , other folks could take a signal.
Will likely be again to get more. Thanks

Thank you for your great post if you want to get informed about startup and technology news you can look our web site. Startuplogy shares startup news, innovations and updates all over the planet, and provides useful information

The arbitrage angle interests me. I'm thinking of arbitrage between FB stock price and the Libra price. And is this a hedge by FB just in case its brand or its business (digital advertising) suffers a fall. With all that data from users and with a banking franchise, it's a short leap. I agree with Cowen, the crypto angle is a sideshow (or what I would describe as a diversion). But I am still amazed at the chutzpah: a company that has achieved huge success by mining its users' data and selling it to the highest bidder offering a crypto whose users are most interested in the privacy of their data and whose value (the crypto value) is dependent on the issuer's credibility - of which FB has none. The sad thing is that it may work. Well, Cowen and his libertarian friends are excited. Be careful what you ask for.

Something I have seen zero comments on in either the white paper or elsewhere is mining. Is there mining involved in this blockchain? If so, will it be more or less efficient than the mining we see in the leading existing cryptocurrencies, which seems to be massively inefficient?

"...which seems to be massively inefficient?"

Bit coin mining is intentionally inefficient. That's the point. It's an artificial barrier that creates an incentive for miners to invest significant effort in the process to create value from scarcity.

From a quick glance, Libra doesn't use the energy intensive approach of "proof of work" style mining to drive consensus but uses a voting mechanism based on validators chosen from the founding members of Libra. Crypto fans call this "proof of stake." This makes their consensus mechanism significantly less decentralized than say Bitcoin or Ethereum but it has the advantage of being much more energy efficient. See pg 2 of the following document under "Sustainability":

The front page NY Times story on this says they use a blockchain. Do not XRP and Ehterium already use proof of stake? My understanding is they are more efficient than bitcoin, but still use lotsof energy. What is going on here? Is Libra based on a blockchain or not, and if it is, is it the new program that makes it more efficient than Etherium or XRP? I mean Etherium offers a lot of the fancier stuff Libra has, although not backing by a basket of currencies and cash from a bunch of companies. The latter Libra has, maybe, but anything else?

The only question I had when hearing of the plan was why would anyone trust FB with their money or payment activities? Given the news that keeps coming out about their near absence for respecting the privacy of their user that though has not changed in the past few years.

I think this is the big point here. Not so much perhaps about whether Facebook is going to run algorithms on your data (which it almost certainly will), but note that this is very much an identifiable and infinitely traceable currency. This is not cash or bitcoin: if you spend money on something, it's forever recorded and stored somewhere, and can be tagged to you. Even if Calibra doesn't hand over your data to a government which is looking to trace someone, they're still the target for a hacker who wants to know. Note too that your FB account is almost certainly tied to your mobile phone GPS as well.

Bear this in mind when you're looking at the HK street protests, where everyone is buying train tickets using cash and wearing face masks. So long as you have enough cash in your pocket and those machines still take cash, this is potentially workable.

I don't necessarily think that adding another perfectly traceable currency is a killer, but reducing cash transactions to only include suspicious or illegal transactions would be a massive problem. We need a mechanism to keep some portion of transactions kept untraceable if we want to keep some level of privacy in society.

"why would anyone trust FB..."

VS what, Wells Fargo, PayPal or BOA? :) The privacy uproar around FB is kind of like the Standard Oil controversy. Every other company is trying to and would be thrilled to do the exact same thing. Most other companies just aren't very good at it. Clearly FB has top talent. Buy the stock.

Save a few prog mandarins, most of the public couldn't care less about data privacy. They're not voting with their feet at any rate.

9. It will seriously dent banks profitability with credit cards. My chain of retail stores pays 3.9% intermediation fee with visa, more with American, in Costa Rica, even if it is one of the top 10 generators of credit card fees in the country. We are definitively interested in getting paid with this thing, at least in that country. Not to mention western Union and other remittance companies, this can be their death.

Politically it is a giant step forward for seriously backed (gold) stable coins, and in the future real cryptos.

I dislike how credit cards are a 3-5% tax on almost everything. Even if you pay cash, you are subjected to the same price! Reducing these fees is similar to moving down one or two tax brackets.

Is it possible that credit cards offer a protection against counterfeit cash? The transaction fee acts as a form of insurance?

Also, credit cards allow people to make purchases with money that they don't yet have, which stimulates consumption.

If I'm a retailer, credit cards increase sales while acting as an insurance against counterfeits.

Ah, someone fighting the good fight. We can have a legitimate debate about whether credit card interchange is too high (should some consumers really get 1% cash back rewards?). But you are on the right path.
Benefits to a merchant include:
a) protection from counterfeit
b) less cash on site, less risk of robbery
c) less cost of handling cash (someone has to pay for those armored cars and drivers)
d) faster checkout times
e) keeping your card using customers happy
Then there are the indirect benefits, namely that credit cards allow people to spend money they don't have, which means you get the sale!

Benefits to the consumer:
a) free float (average 30 days to pay for the txn)
b) protection from merchant misbehavior (ability to chargeback)
c) protection from robbery (generally if someone steals your card you're out at most $50, usually $0. if they steal your cash, you're never going to get it back.
d) insurance on many transactions
e) rewards
f) centralized transaction records (e.g., for tax filings)
g) faster checkout

We could reduce interchange (and card cost) by legislating it down, other countries have. This will just reduce rewards and make it uneconomical to offer credit cards to lower income folks. You may be fine with that. But usually some point of sale lending system just pops up to replace that role for such folks, "hello Klarna".

Let’s see. My retail chain in Costa Rica sells about 100m$ per year. We sell 70% with credit card with an average commission of 4% (3.9% visa and MasterCard, 4.5% Amex). That’s 2.8m$ that the banks get with 25 PoS, which they come to check maybe 1 time per year. About 20% of that goes to merchant (the bank that put the PoS), 80% the acquirer (the bank that has the clients and take the risk of default). On the other hand, we have a total 15% gross margin, and our own expenses outside credit card are about 10%, so we make 2.2m$ before taxes, 1.5m$ after taxes. Credit card fees siphon double the amount I make running the business.

I would never ask for the State to come and force them to decrease the commission. I would just use market forces.

I assure you that I could give a 4% discount to people paying in cash, and the banks would not try to enforce the contract clause that forces me not to discriminate among clients paying in cash or in card, because it knows some other bank would take it and close an eye (they do not compete on percentage because Visa and Co. set a floor by category of merchandise).

So, they bought the State. By law, in Costa Rica (and other countries as well) I cannot discriminate. That’s a typical baptist and bootleggers coalition, Bootleggers being the banks and baptists the ignorant mob, that do not even know that there are commissions charged to the stores.

I can deal with market forces, competing is what we entrepreneurs do every day. I can’t deal with the bastards with the guns and the plaque.

There is no mining in Libra. There isn't actually a chain of blocks, either - the current state of the system is just stored in a database. They call it a "blockchain" more as an analogy to Bitcoin than as a reference to the traditional blockchain data structure. is a good summary of many of the technical details of Libra. I recommend reading that for anyone interested in these questions.


Really? Tyler says the crypto angle "is a sideshow." If there is in fact no blockchain at all, then this is not a cruptocurrency but something else. Why are the crypto markets jumping up on reports of this? Well, they did. Some of them are down today.

Crypto is the whole reason we are having this discussion. Ownership of the private key is proof of Libra ownership. They use a simpler "blockchain" structure in the form of a distributed database that all validators have a copy of. It serves the exact same function as a regular blockchain just a different implementation plus/minus a few features. For example you get faster queries and authentication compared to traditional linked-list blockchains.

So, is it that much faster than XRP or Etherium? They do not use proof of work, from my understanding. What is the key to this supposed greater speed for Libra? I see people claiming it is not a blockchain at all, but just a fancy version of PayPal, whose former president is the guy running this operation, David Marcus.

It sounds like we can hedge the listed assets privately within their net by holding more or less libra and less or more of some safe asset in their list. The game becomes one of predicting the composition of the list. The libra is stock in a index traded fund. The libra association is competing with JPM, but they don't have the assets, yet. How in the world are hey going to pick the proper class of assets without getting their ass eat.

Apple can do this right, they have Apple privacy.
Create the secure iCoin S/L bank, open to all with apple ID on a congestion price, in iCoin unit. iBank uses congestion fees for collateral, making sure the queues are stable implies risk equalization, so deposits and loans, on demand', are balanced to a quart point market error.

Then Apple convinces CoinBase to put iCoin on the exchange. Deal done, Apple just lets the iBank bot free run, there is low risk. Apple ID means we can carry around iCoins with us, like tucked under the actual screens of our iPods, no need for third party ledger. Thus, the entire world converts suddenly to iCoin, perfect automated universal liquidity, just buy a hundred dollar iPod. I comes with price neutral currency functions, perfectly bettable, all automated. There are a number of bot codes around that can do this optimum currency banker trading pit..

You should write a crypto blog...

Or: let's begin importing EU tech regulation ASAP.

(What would or do EU tech regulators [or EU banking regulators, for that matter] make of FB's financial "service" ambitions?)

This talk does not fear the Fed, their bluster is for traditional institutions at the moment. But the Fed is going to be competitive, they are going to support much of this stuff, especially risk equalized auto-priced S/L currencies, that is what the Fed wants to do. Soon the techies are forcing the issue, and Apple seems to be taking the lead. The Fed is going to need a new contract with Treasury, a version 4 of the current Fed series since 1913, but a big revision. This is a contract coming up, a Bretton Woods kind of meeting of the elders, a Nixon Shock, an FDR repricing of gold, one of these versions is going to happen such the the Fed moves fully into automated portfolio systems. The Fed is getting a Moore's law update.

Timmy and Apple ID is a great thing, Timmy is my new hero. Right now we can use AppleID and sign onto a trading site, and trade securely and anonymously. Just that possibility is enough to drive Apple id through finance, it becomes an entry into JPM net, for example.

The Fed and finance needs a response, a demand the concept become a standard, and Timmy might agree. But we need a bonsue with the NSA, the AppleID standard allows personal bots to hold private keys, with no one looking, even the owner. That is what gets us digital bearer assets. Ask for a low limit, $20 maximum, to start in a trial, the smart technology can enforce that. A personal bot can go spend $20, on our behalf while we are at the bar.

Pretty sure this new asset will create a taxable event every time an American uses it. Not sure how that is useful to Americans. They should have backed it with USD instead

It's a stable coin, right? No gains ... If it works correctly...

@Will it's stable relative to a basket of major currencies+bonds, not the USD. So, for US residents, whose unit of account is the USD there will be some volatility.

@Max the taxes can be handled by software. Nothing very complicated to for a program to inspect each of your transactions and keep track of what you may owe, unless you choose to do it by hand. Backing by USD only makes sense for US residents, rest of the world would face the same problem, so they might as well hedge geographic-risk by choosing the basket of major currencies+bonds.

I'm gonna wait for the service to actually exist to start commenting on it. Old fashioned, I know, in the Hot Take Era.

Nothing wrong with raising concerns or asking questions. Better to do that now than after it gets rolled out, don't you think? Facebook might employ smart, capable people but they haven't thought of everything.

Why did end up with the sensation that they create a new Federal Reserve that eventually has a currency, than properly creating a new currency that incidentally has a clearance agency?

Anyone who runs their finances through Facebook will get exactly what they deserve, total exposure of your finances to anyone willing to pay. That's not even considering the hacking risk.

What do you think about the theory that this inted to sove problems that in Europe were already solved decades ago?

TrustScience Inc., a leading provider of AI-powered credit scoring, and Inovatec Systems Corporation, a new breed of Loan Operating System (LOS) provider, announced today they will partner to release a fully automated lending platform that enables end-to-end loan management across the entire credit spectrum.
Lenders can be up and running on a fully customized LOS and an AI-powered loan underwriting model within weeks, not months (or years).
Trust Science CEO Evan Chrapko comments, “This partnership gives lenders the ability to accurately score and lend to an additional 64 million consumers in the U.S. alone, with unprecedented accuracy and speed. The end-to-end, customizable nature of Inovatec Systems’ LOS makes it a perfect partner for Trust Science and our API-based scoring solution.”
So, they claim, a completely automated S/L tech.

Now just attach the Apple ID system, change the name to iCoin. Like all of two weeks worth of work, and a large payment to these auto-lending folks. What do you get? A complete replacement of central banking.

So, in other words, we have some reforming to do with the Fed if we want the Fed to compete. The Fed is losing market share rapidly as this tech enters JPM net.
The Fed can cannot guarantee of your government insurance program and compete against automatic portfolio adjustments. The Fed needs to go full auto-bot trading.

In terms of our abstract algebra tree, the auto-S/L makes the optimum tree trunk, there is no better version. The bots do not check to see if all our precious government goodies are paid, it just matches a queue of loans to a queue of deposits, all risk equalized. Makes a better hologram (velocity equations works fine enough).

The Apple ID makes lonas in advance work, the member account treats is like a credit card. And, more importantly, we keep the concept of cash. I idea that I can carry around my iPod full of cash and just digitally give some to another iPod with no intervention by a ledger service.

Both techologies up anbd running. All we need os for the Apple ID project manager to call up the marketing teams for Trust Science, and have a talk. Or, even simpler, if apple and Trust just read my comment. This is how fragile our precious Fed is, a very precarious situation. It is losing market share, fast.

On June 21, the Financial Action Task Force, or FATF, is expected to adopt a proposal that could have a big effect on the cryptocurrency world. FATF wants cryptocurrency exchanges to act like banks, collecting identifying information not only from its own customers, but also the people to whom these customers send cryptocurrency.
This seems to go against the very nature of electronic cash. Cash provides consumers with an information-light method of transacting. Want to protect your personal data in order to avoid identity theft and hackers? Use cash, either the paper or electronic version. But the moment all parties to a cash transaction must submit identifying information, all of cash’s data-shielding virtues are voided and it stops being cash.
Now this makes Timmy mad, this negates AppleID, makes it illegal. The battle is joined, time is short.

Will there be any consequence for the taxable basis in countries which population largely/considerably adopt libra - or other crypto?
Some countries already have trouble fundind democratica republican institutions as their economies were to informal. Will this intensifies as crypto adoption furthers the govt-independent fraction of the economy?
I mean, some crypto promise to promote underdevelop countries, but they may result in weaken them even more.

I don't see what problem this solves. The only one would be for a country that has an unstable currency and bans the usage of stable alternatives. But in that case, why would they permit the usage of Libra.

If people want the institutions backing their stable fiat deposits to be in the stable country, then it is merely a matter of letting the banks do the same things that the Facebook consortium is planning.

"'It mints and burns coins only in response to demand from authorized resellers.' Maybe, of course there are hundreds of years of debate on that one, google 'real bills doctrine,' noting that here we have a semi-dominant private issuer rather than a perfectly competitive banking system."

The suggestion that any argument to the effect that the quantity of a market-supplied money will automatically adjust to accommodate changes in the demand for same must rest on appeal to the real-bills doctrine shows that he hasn't followed well, or perhaps has forgotten, much of the modern literature on competitively-supplied currencies. For starters, that doctrine plays no part at all in the modern free banking literature. Both Larry White and I expressly disavow it, our theories relies on a completely distinct mechanism from what Fullarton and other banking-school theorists envisioned. The real-bills doctrine also plays no part in any established cryptocurrency scheme. Yet it's certainly possible to design such a scheme so that it allows for demand-accommodating adjustments, at least to some considerable extent.

So let's not mix-up two separate issues: whether the real-bills doctrine has merit (A: not as a means for regulating the money stock) and whether one can have a demand-accommodating private currency (A: yes, but not the real-bills doctrine isn't part of the "how").

Sorry: For "he" I of course mean Tyler.

It's good to hear that the real bills doctrine is a genuinely open question. Apparently, economics has progressed since 1997.

" I am forced to agree with Charles Calomiris’ (1997) assessment that “to my knowledge, the real bills doctrine has no current advocates.”

Comments for this post are closed