One could say that it’s kind of condescending name for a blog post. I guess. My late father used to lovingly refer to my brother and I as bozos, usually after we did something dumb, or pretty much any day that ended in “y”, so you should know that I consider it a term of great endearment. Anyway, if you’re reading this, the title worked.
Don’t like to read more than a paragraph, and still want to tweet your response? TL;DR the Libra Reserve poses all kinds of questions that will take years for properly clued-up economists to iron out (of which Libra currently appears to have none on hand) and will probably end up being economically unfeasible. In all cases, getting these things right are more important than getting things right now.
Having read many a hot takes on Libra, from supporters and straight up haters, I decided to craft a post to consider some of the lesser appreciated aspects around Libra, and more specifically the reserve mechanism that is envisioned to underpin its stablecoin. This post was written for people who are already more familiar with cryptocurrencies and stablecoins, if these are new to you then you’ll have some background reading to do. Look, I never promised to be nice.
Also, if you’ve not already picked up on this, my takes are usually quite snarky, this post is absolutely, 100%, in no way a departure from that. I make no apologies. Happy reading… Bozo!
On 18 June 2019, the Heavenly Divine brought to us the Libra genesis block, an open source code repository and a website… Wait, maybe it wasn’t exactly like that. At least they were a bit more honest about how it was created than a certain Small-wave-shape-like Ledger. Very simply, Facebook via its wholly-owned subsidiary, Calibra, along with their 27 partners announced the launch of the Libra association.
The Libra Association states that it is a Geneva based, independent, non-profit membership organisation. Either the Canton of Geneva hasn’t gotten the memo, or someone setup a SARL (Société à responsabilité limitée, a Limited Liability Company). Somewhat removed from Calibra CEO, David Marcus’s claims that while “[i]t’s easy to assume from the headlines that Libra is only associated with Facebook, but that is not the case”, other than a few Switzerland and Ireland based directors (whom, by the way are also directors for Facebook’s Swiss operations), Ernst and Young, and Facebook (Facebook Global Holdings II). Facebook is, at the time of writing, the only listed shareholder. Maybe the other 27 partners just haven’t gotten their paperwork together though.
Moving right along, it is important to stress that Libra isn’t live. They released some code in a Github repository, and a test network. You can download this code, connect to other computers on the network and create and move some test Libra tokens. Like the future Libra tokens, the test Libra token are stablecoins. Test Libra tokens are precisely stable at zero Dollars, zero Pounds, zero Yen, zero Euros and zero bitcoin.
So concretely, we got a website, some questionable Swiss companies, a whitepaper and a Github. Sounds like one of the most productive ICO projects ever! Naturally, it was met with little fanfare and everyone objectively assessed exactly what had happened. Psych! Everyone went overboard.
Three of the biggest areas of public attack were, 1) privacy consideration, 2) censorship concerns, and 3) cartel-like behaviour. Others questioned semantics, is this a blockchain? And is it fair to call Libra a cryptocurrency?
Whilst all of these are more-or-less worthwhile topics to discuss, to me Libra (as I’ve understood it) poses two more important considerations. The first, which has garnered more of the debate amongst the Bank for International Settlements (BIS), G7 (Specifically German and French delegations) and the US Congressional Committee on Financial services is the potential systemic risk, should Libra ever become large enough. The second, and very much related to it, which has astonishingly (at least to me) been greatly overlooked, is the business model for Libra Association members/Libra nodes. I believe that these two aspects are inherently inter-linked, and will form the core discussion of this post.
Enough talk, financial geek-out time!
Turning our attention to systemic risk issues, let’s first outline what a systemic risk is. The most widely shared definition that I came across comes from Professors George Kaufman, and Kenneth Scott, “What is systemic risk, and do bank regulators retard or contribute to it?”:
Systemic risk refers to the risk or probability of breakdowns in an entire system, as opposed to breakdowns in individual parts or components, and is evidenced by comovements (correlation) among most or all the parts.
Translating that to crypto bozo, systemic risk is the difference between everyone getting rekt, versus smaller contained blow-ups. One of the most recent demonstrations of this concept is the Great Financial Crisis of 2007 & 2008, rather than just Bear Stearns and Lehman going bust, the entire global financial system was on the verge of collapse.
“Great”, you say, “long bitcoin, short the bankers!” Not so quick buckeroo, systemic risk doesn’t only happen in fiat systems, crypto can experience systemic shocks too. Take for example, Mt. GOX, accounting for some 70% of all bitcoin exchange volumes in 2013, upon its collapse and eventual bankruptcy they claimed that nearly 7% of all bitcoin in existence were missing. No need to tell you that much of the troubles around GOX happened around the run up to nearly $1,300 in 2013, and subsequent collapse down to $200ish in 2015. Some have even stated that the eventual liquidation of GOX’s assets in late 2017, early 2018 contributed to the fall from nearly $20,000.
“That’s just price, bitcoin goes up and down, shut up you FUDster”, you say. Sure it does, and as I write this the price is well above where it was at the peak in 2013. But 2014–2016 were often regarded as a “crypto-winter”, kind of like 2018 & the first half of 2019, and not too unlike 2008–2010 in the global financial world.
Talk about Libra, I want more on Zuckbucks!
Details are still extraordinarily thin on the ground, but what we do know is that every Libra (currency) will be backed by a basket of assets, denominated in major currencies, such as the US Dollar, Euro, Japanese Yen, and British Pound. Believe it or not, this gives a pretty wide range of exactly what they could be holding. Libra elaborates that “[t]he actual assets will be a collection of low-volatility assets, including bank deposits and government securities in currencies from stable and reputable central banks.” This still doesn’t narrow things down too much, but does highlight some of the more obvious ones, namely t-bills (short term debt issued by governments), and commercial bank deposits (not too dissimilar from what’s available to small retail investors at their local bank, but significantly larger in size here). What it doesn’t rule out (possibly on purpose, cynically because they don’t actually understand this stuff) is investing money into commercial paper, repos, and Eurocurrency markets.
Looking at the first two assets we know of, the concern and the ability for regulators to contain them are pretty straight forward. Libra raises some money, they use that money to buy debt with less than 1-year maturity from the US, core Eurozone countries (Germany, France, Netherlands, etc..), UK, and Japan. As things stand now, they probably stay away from southern Eurozone countries like Italy, Spain and Greece. But then again, maybe not, they might need the juice. In addition to those, they call up large banks in the same countries and place large deposits with them. Assuming all this money is pretty much just a replacement for other domestic funds (i.e. for every $1 from the US, it goes into US banks and T-bills, €1 from Germany goes to Eurozone banks and T-bills, etc…), things are pretty ho-hum. Given this is not likely to be the case, as indicative of the global reach of Libra Association members (specifically Facebook), we’ll have to return to this point later on in this post.
Now let’s return to the Global Financial Crisis, and some of the root causes. While we could all speculate and blame your favourite boogeyman, fortunately for us the Financial Crisis Inquiry Commission (FCIC), commissioned by the US Congress, published a ‘light’ 663 page paper on it. This report contains an in-depth blow-by-blow of the causes leading up to the crisis, as well as actions taken during the earlier periods of the crisis. All your favourite “bankster” memes appear there, failures in corporate governance and risk management; deregulation, excessive borrowing, complex over-the-counter derivatives, deteriorating lending standards and lack of transparency; a breakdown in ethics and accountability, from banks to credit agencies; and an ill-prepared government.
The section that really sticks out to me, however, details the shadow banking system. It gently meanders through financial history, discussing bank runs in the 19th and early 20th century, which led to the establishment of the Federal Reserve, to the Savings and Loans crisis of the 1980s that led to “Thrifts” being forced to cap their interest rates paid to depositors at 6%, sparking the boom in shadow banking by way of the money markets. Money markets, and more specifically their commercial paper investments, are short term debt issued by high quality companies, the yields of which are returned to the money market fund investor. These yields were outside of the regulation that kept Thrifts from paying higher return and resulted in the market for commercial paper ballooning from under $125 billion in 1980, to $1.6 trillion in 2000. Money markets funds are of course famous for keeping their share value at $1, fully backed by high quality assets, paying out returns to investors should they go above. The original stablecoin! Well not really, money market funds are significantly better, because investors get some upside.
When everything is going well, these instruments work great. Though occasionally things don’t, and the value of the money market funds “break the buck” or drop below $1. The FCIC report talks about the Penn Central Transportation Company, which filed for bankruptcy in 1996, taking with it its $200m in outstanding commercial paper. This froze up commercial paper markets, until banks were allowed to be hired as insurance agents to stand in, if and when a commercial paper issuer defaulted. Though this did open the commercial paper market back up, it also created huge potential liabilities on banks’ balance sheets, as they would have to cover for their clients (massive companies) if they weren’t able to make the payments themselves.
Repos, or repurchase agreements, are another popular money market instrument. They are fully collateralized loans which allow a bond holder to borrow cash secured by their bond holdings for a low interest rate, in agreement that they would return the cash by buying back their bonds from the lender. These tend to be useful for lots of reasons, including helping bond dealers manage the assets in their trading books, and long term bond holders lend out the cash loaned to them at extremely low rates to other investors for a slight “yield pick-up”. Like the commercial paper market, repos tend to be seen as low risk, given their strong collateral backing. However during times of stress these markets have also been known to seize up, leaving businesses that rely on them to face financial ruin.
The third area of potential Libra Reserve backing, are the Eurocurrency markets. No, I haven’t forgotten that I already mentioned the currency that I use to buy a baguette at my local boulangerie here in France.
Eurocurrencies, of which the Eurodollar is the largest and most widely known, are unsecured (uncollateralised) time deposits held by banks outside of their home markets, and thus are subject to different (often less stringent) regulation than deposits issued by banks in their home markets. The prime example of a Eurodollar are Dollar deposits held in a bank in London. If you’ve heard of the London-InterBank-Offer-Rate (LIBOR), that’s what this is measuring. Eurocurrency markets don’t only exist for Dollars, but for other currencies, Yen (“Euroyen”) and Euros (“Euroeuros”) being the other majors. The Eurodollar market grew out of the Marshall plan when Europe was awash with Dollar loans following WW2, and the Soviet Union feared that their US-deposits would be seized by the US government and sought the relative safety of UK banks. See Bitcoiners, even bankers cared about censorship resistance!
One of the basic aspects of the fiat system is that, while a central bank can rescue a commercial bank from failure by lending it the money it creates (e.g., the Bank of England could save a London based bank by loaning it Sterling), they cannot unilaterally decide to create money in another currency. In this way, Eurocurrency markets look a lot more like how Bitcoiners envisage a Bitcoin-based financial system could look.
Given the inability for a central bank to create foreign currency (e.g., BoE creating USD), should one of their banks run into issues, lending rates tend to be slightly higher than the rates that the central bank borrows at. In USD, for example, we look at the difference between LIBOR and T-bills, which is known as the TED spread. Much like we saw in commercial paper and repo markets, Eurocurrency markets are free(ish) markets and are also susceptible to freezing during times of distress. Given their massive size (comparable to the size of the onshore banking system in the case of the Eurodollar) this can create slight headaches, as was seen in 2007–2008.
TED Spread in 2007–2009
One tool used to help alleviate the issue in the financial crisis (as well as at several points since the inception of the Eurocurrency market), are currency swap lines. These are essentially the ability of two central banks to exchange their own currencies. In this case quite useful when the conversation is:
Mark Carney (Governor of the Bank of England): Hey Jerome, long time no speak!
Jerome Powell (Chair of the Federal Reserve Bank): Marky-Mark, good to hear from you, how’s it going?
MC: Not so great actually. Look mate, I’ve got a favour to ask. We let our banks loan out too many Dollars to dodgy construction projects in Asia, and now they are broke. Mind if we borrow some Dollars? I’ll give you some of our stunning Sterling, they’ve got great pictures of the Queen on them.
JP: No worries man, happens to the best of us. I’ll go fire up the printing presses in the back, and get those Greenbacks over to you. Do me a solid and see if you could get Princess Markle on the next set of £10 notes, yeah?
MC: Thank you. I’ll see what I can do.
Obviously the conversation isn’t as formal as the above, in fact it’s pretty automated now as the largest central banks have established permanent currency swap lines. The ECB put together a pretty accessible explanation here. The bottom line here is that currency swap lines can now make another country the bag holder for their country’s Eurocurrency-derived financial crisis.
“Sounds risky, Eurocurrencies must have been reduced after the financial crisis”, you might say.
Well, you’d be wrong.
Our good friends from the Bank for International Settlements (BIS) keep track of such things, as of the end of Q3 2016, Eurodollar funding stood at nearly $9Tn, surpassing the $7.85Tn amount in Q1 2008 (n.b. they’ve also looked at how the reforms of money markets in the US have led to a reduction in the size of that market… curious).
Looking at the BIS chart on the right, we can see that one of the biggest gainers has been bank deposits (in USD) outside of the United States. Looking over to the ECB, they have done some of the heavy lifting to highlight comparable measures for the Euro, Yen and “the rest”. As you can see below, the USD is a fairly large market with $4.2Tn deposited with banks outside of the US in Dollars. Euros follow at a distant second, with $1.9Tn deposited with banks outside of the Eurozone, and denominated in Euros. The comparable Japanese Yen market is $205 billion. Sterling is hidden in the other category, presumably smaller than the JPY figure.
Cool, but why is this important?
Much like what we’ve seen in the money market, Eurocurrency markets are also susceptible to market forces, in particular during times of financial shock. The peak in early 2008 and the subsequent trough during 2009 speaks to that fact. Should the Libra Reserve keep a substantial amount in Eurocurrency deposits they may find getting liquidity from these markets quite difficult when they need it most. This could be brought about not only by perceived risk, but also if other sources dry up and they are left needing to access their Libra balances to pay for real world things.
Well, let’s recall what the Libra Association is trying to be: a Swiss-based non-bank entity, with a stash of cash in Dollars, Euros, Sterling and Yen which they need to put someplace.
We’ve highlighted that they can do the domestic thing, placing USD from US clients in US banks, or US T-bills. Also, putting EUR from Eurozone clients in Eurozone banks and EUR T-bills from core Eurozone countries, etc… While this could work to a limited extent, such a view overlooks that Libra is a basket, and that “international” is part of the deal. Meaning that the US client may deposit USD, and the British client may buy Libra with GBP, but both clients hold a risk that is part USD, EUR, JPY and part GBP. So while Libra may be able to hold limited amounts of domestic funds in domestic options for risk management purposes, they almost certainly need to go out a bit in search of other options. The other option would be the money markets (commercial paper and repo) and deposits in banks outside of the direct control of central banks which manages the currency in denomination (Eurocurrency markets).
As we’ve highlighted, none of these things on their own present an insta-rekt situation. We’ve seen that these are extremely large markets, measuring in the billions or trillions, surely adding in whatever underpins the Libra Reserve isn’t going to cause a shock. So, all of these things are true, and when all is well we need not worry, though systemic risks usually only become obvious when it’s too late, like when Libra holders are trying to get their Libra balances out, and into cash or moved to their local banks, denominated into currencies which surely aren’t Libra.
But SDRs bro!
Let’s look at that some more. When Libra first came out announcing that they would be backed by a basket, some people started making comparisons to the SDR. My loyal crypto bozos, however, began googling: “What is SDR?”, shortly before firing off tweet storms about it. No, not you, of course I’m not talking about you. But because we’re all on the same page, I won’t insult your intelligence by explaining what IMF Special Drawing Rights are. Nor will I remind you that SDRs aren’t a currency per se.
What is important is highlighting that the SDR measures the four currencies that we’ve discussed, plus the Chinese Yuan. From what we know so far the Libra Reserve isn’t publicly discussing using Chinese Yuan, possibly because it isn’t freely tradeable outside of China, possibly because Facebook is blocked in China. If we look at the weighting of that basket during the last review in 2015 (valuable for 5 years), and do the math to reassess the weightings without Yuan we see the following:
Let’s make the big assumption (because we otherwise have no idea) that Libra will pick a similar weighting for the Libra Reserve. Close enough for government work, as they say.
Interesting and all that, but we don’t know how much money will be in the Libra Reserve.
True, we don’t and won’t know until the thing goes live. What we can do is give ourselves an idea of how much might be in there from other similar projects. I know, I know, “Colin there isn’t currently a Libra”. Absolutely, but we do have an idea of what electronic payments could look like from China, by virtue of AliPay, TencentPay and the like. From filings at the People’s Bank of China, we know that two largest of these “third-party payments providers” had some 1 Trillion Yuan, or about $150 billion underpinning multiple services as of mid-2018. While China is a large market, surely a service that seeks to operate globally, or at least cater to a large chunk of Facebook’s 2.4 billion clients should be able to match or even exceed such a number. For now though, we’ll just go ahead and use $150 billion to get started.
Ok, so let’s return to our previous discussions on Eurocurrencies, now that we have some actual numbers to play with. Assuming that Libra is able to raise these sums, which doesn’t seem like a stretch, and places them in line with SDR sans CNY, we’d expect them to find a home for this currency somewhat in line with the above. Though, as we’ve discussed, Eurocurrencies are not their only option, these sums represent nearly 1.7% of Eurodollars, 2.5% of Euroeuros, and nearly 7% of Euroyen. We don’t have numbers for Eurosterling, but it is likely to not exceed the size of Yen, so 7% is probably not a million miles away. Even if we conservatively assume that only 25% of Libra Reserves, the need to move 0.5% of the entire Eurocurrency market would likely not go unnoticed even in the best of times.
If you think that moving that amount of money to or from the markets in good times is tough, let’s ponder what would happen when these time-limited deposits (when they say you can’t use them for 1,3,6 months, you really can’t without borrowing from someone else, which means you actually do care about liquidity), and so if banks with access to central bank money and currency swaps struggle, how does Libra expect to cope? Not a pretty picture. The issue becomes more acute for JPY and GBP markets, which are smaller, less liquid, and of which Libra would comprise a larger portion.
Similar issues arise when we consider Libra Reserve assets being invested in commercial paper and repo markets, which also may be an option for part of their holdings, but certainly not a silver bullet. Another consideration for Libra is that a lot of this data would be pretty public; it’s supposed to be in a blockchain, ain’t it? That means, much like as is the case for Tether, the market sees the intention to create or reduce Libra and anticipates the impact on the market. In the case of Libra Reserve money market rates go down slightly before Libra deposits new money, and increase before it tries to sell out. For EUR and USD holdings these markets might be large enough to allow them to ignore it most of the time, in GBP and JPY this probably isn’t the case. When market liquidity gets tight (it’s harder to move money), Libra might even face a situation where they are unable to sell GBP or JPY in line to maintain the basket, and need to sell more EUR and USD to make up for it, leaving remaining Libra holders overweight these (less liquid) assets and underweight the more liquid assets. This could, in turn lead to more selling pressure, eventually causing the Libra Reserve to sell GBP and JPY assets significantly below market value. Ouch. But anyway, it’s just technology, maybe they’ll sort that part out.
Maybe Libra shouldn’t manage the money so much itself and just leave the banks to it.
Probably not a horrible idea, but here too there are significant considerations, ones that large corporations around the world employ sizeable (and costly) teams to navigate. Which banks do you give it to? We talked about corporate deposits, they are somewhat like retail deposits, but on a larger scale. What we didn’t yet highlight is there is one critical difference, when you deposit money in the United States at a bank you are protected by the FDIC, if your bank goes bust the FDIC gives you money, up to $250,000. Similar things happen in Europe, UK and Japan. For corporate deposits which would likely measure in the tens, hundreds of millions, or billions, this level of backing is a drop in the bucket. This means that the Libra Reserve would need to closely monitor the health of their banks, making sure to reduce the risk of a bank or banks failing and taking client money with it. At least, for their 9% potential allocation, Libra may have a better option, depositing them directly with the Bank of England and earning some interest, though details aren’t out yet, so we’ll have to wait to see if this really is an option (BoE may, for instance, insist on Libra Reserve setting up shop in the UK or conforming to UK anti-money laundering rules, they may also take pause at the realisation that Libra’s GBP holdings alone, under our assumptions, would expand their 2018 balance sheet by 1.7%, or be equivalent to 65% of their deposits from other global central banks). Beyond the unicorns at Threadneedle Street, we highlighted this possibility for Eurocurrency deposits (which may have no deposit guarantees), they are just as worrisome for domestic deposits. Given the sums of money in question, these banks too might want reassurance that they won’t be left with a Libra shaped hole one day to the next in their deposits, and might demand that Libra give them 90 days notice before withdrawing more than a proportionally small amount.
“Yes the planet got destroyed. But for a beautiful moment in time we created a lot of value for shareholders”
Up to here, we consider the implications of what happens if Libra is allowed to earn money off of the money deposited with them.
For those of you that clicked on the AliPay, TencentPay link, well done, you probably know what happens here. And, for those of you who didn’t, the PBOC likely did some of the same calculations that I’ve just done, and decided to tell them to bugger off with their shadow banks. Unlike the plans with Libra, Chinese tech giants went with a simplified approach, mainly attempting to move domestic funds from cash and bank accounts in CNY to digital form then into wherever the managing companies felt that they should best invest. I’ve got no insight on whether their investment decisions were good or bad, but the PBOC, seeing the growth of the money held in these systems to more than 1 Trillion Yuan decided to gradually force the funds into non-interest paying accounts. First 20%, moving to half and then eventually last year, all of them.
The question for the Libra Reserve thus becomes, who has the power to tell them to hold all of the reserve funds in a central bank, and which ones, in what proportion. Assuming that Libra is trying to track our modified SDR you might say, “well that’s simple, 47% to the Fed, 37% to the ECB, 9% for the Bank of England and Bank of Japan”. Again, things are unlikely to be so easy, recall that Libra is meant to be global. How would the Australians feel, should Libra take off, if money starts getting sucked out of their retail bank deposits and into the Fed, ECB, BoE and BoJ? Especially if they are for domestic Australian goods and services?
Take a more striking example, money for remittances paid from the United States to the Philippines, might not ever make it into Philippine Pesos, just getting converted into Libra and used locally in the Philippines to pay for things in Libra. There could be even greater secondary effects if countries with high inflation and falling exchange rates see money and commerce moving to Libra and sitting in the central banks of countries which issue those exchange reserves which they require to access international trade, putting ever greater pressure on their currencies.
Stepping back from doomsday for non-Libra Reserve economies, and the assumption of guaranteed success for Libra. One has to question, if the Libra Reserve isn’t allowed to earn money off the float, how will the companies paying $10m to join the association and managing nodes make a return? The obvious answer is transaction fees, but this may even be limited.
The second thought is, why not just create a Libra based banking system, allowing someone with lots of Libra to lend to people in need of Libra?
On the surface, and in limited quantities, this has obvious appeal. Even with a relatively small size, Libra could offer people buying and selling goods and services on Libra to quickly see a value in borrowing and lending in that currency. In time this might even become a driver for internet only businesses connected to Libra to fund themselves, take investment, get paid by customers and pay their staff. While there are similar concerns for non-Libra-Reserve economies on what this may do, there are other concerns. Let’s come back and again assume that this does take off, and that lending by larger Libra holders to Libra borrowers becomes a big business. What then happens when default rates increase? Maybe Libra-based banks have appeared, allowing causal moderate Libra holders to lend across a portfolio of Libra borrowers, and they in turn leverage out that money across the system. Until one day it doesn’t work. Unlike Bitcoin, new Libra can be created, and this Libra bank could just turn to the underlying fiat markets, borrowing USD, EUR, GBP, and JPY to fill holes in their balance sheet from bad loans. This could create sizeable risks, not easily viewable between the fiat-world and the Libra world.
We must also consider that Libra Reserve might feel pressure from non-Libra-Reserve countries, and not like the idea of not getting interest on assets placed in central banks and attempt to negotiate the ability to lend them in greater quantities to Eurocurrencies markets. At least this way Australia and the Philippines could get investment in exchange. This of course brings back the circular logic of previous discussions on the risks created by investing in these markets versus having always liquid central bank deposits.
Of course none of these scenarios are guaranteed to occur should Libra get the go-ahead. It may very well do, and then become a massive commercial failure. But I still believe that it is important to move beyond jabs at the technology, or trite discussions on money laundering, to look at the wider implications for the financial system and the global economy, should this thing take off. At minimum, I don’t see this Libra, the one which was presented to the world on 19 June, to be ready for primetime in any meaningful way for at least the next 5 years. There may be smaller, more focussed efforts in the meantime that give better datasets on how such a global asset backed digital currency may work. Stablecoins, such as Tether, USDC, PAX, etc… are the obvious candidates, and in some senses have already started to show many of the aforementioned risks. What regulators see is another question.
For us crypto bozos, rather than being quick to shout “innovate or die” or accuse the men and women tasked with keeping the financial system from collapsing on itself, let’s consider some of the wider implications of these ideas, and whether they are desirable, rather than what might make our lives slightly easier when we want to book and pay for a backpacking holiday in South East Asia from our phones.
Thank you to those who helped edit this!