Adapted from two Facebook posts:
A recent article points out that Coinbase users might lose their holdings in the case of bankruptcy, since they would be viewed as unsecured creditors with respect to the assets held by Coinbase, namely its crypto and fiat holdings. (“Coinbase earnings were bad. Worse still, the crypto exchange is now warning that bankruptcy could wipe out user funds.”)
The article tries to push the tedious bitcoin-activist line that you should always hold your own keys, and that the failure of Coinbase customers to hold their own keys is at the root of the problem:
In the event the crypto exchange goes bankrupt, Coinbase says, its users might lose all the cryptocurrency stored in their accounts too.
Coinbase said in its earnings report Tuesday that it holds $256 billion in both fiat currencies and cryptocurrencies on behalf of its customers. Yet the exchange noted that in the event it ever declared bankruptcy, “the crypto assets we hold in custody on behalf of our customers could be subject to bankruptcy proceedings.” Coinbase users would become “general unsecured creditors,” meaning they have no right to claim any specific property from the exchange in proceedings. Their funds would become inaccessible.1
That shouldn’t happen.
An individual’s ownership of cryptocurrency is supposed to be immutable and absolute; that’s one of the key selling points touted by blockchain evangelists everywhere. But when a user creates a Coinbase account, they often end up storing their cryptocurrency in a wallet controlled by Coinbase, which means the individual is giving away at least part of their control over their own funds.
But this analysis is confusing. As I pointed out in a previous post [reprinted below], it appears that Coinbase does actually hold 100% reserves on behalf of its customers. It can’t afford to insure all of this huge $260B pool of assets, so it keeps most of its crypto in cold storage—about 98%; and insures the 2% that is online. See https://help.coinbase.com/en/coinbase/other-topics/legal-policies/how-is-coinbase-insured.
The best way to view this arrangement legally is as an irregular deposit as explained by Huerta de Soto in his book ( https://mises.org/library/money-bank-credit-and-economic-cycles ; see ch. 1). We have deviated from this kind of legal arrangement due to fractional reserve banking, which corrupted money (and which is one reason for the emergence of Bitcoin as a 100% reserve, non-inflatable, non-censorable form of money).
If the law were rational and just, it would view a Coinbase customer as having made an irregular deposit. This means that the crypto deposited may be commingled with that of other irregular depositors (for efficiency and convenience), but the depositors remain the owners of these funds. So they would not be “unsecured creditors” in the case of bankruptcy; they would not be “creditors” at all. They would be the owners of the mass of funds, each depositor having a pro-rata ownership right in the stored crypto.
For example if there are 100,000 BTC on deposit, and you deposited 10BTC, then Coinbase does not own these funds at all. They are just a custodian. So if Coinbase goes bankrupt its own creditors can only pursue its assets which do NOT include the 100,000 BTC. In a sense, you could view the irregular depositors (the customers) as secured creditors, but a better way to view it would be to view them as the owners of the BTC. That is why Coinbase cannot spend or pledge these assets, and it is why Coinbase apparently does, at present, keep 100% reserves (again: 98% in cold storage; 2% online and insured).
But everyone is confused by the idea that customers and depositors of a “bank” are “creditors,” due to the confusion and propaganda and corrupt law behind the practice of fractional-reserve banking (including fractional-reserve free banking, or FRFB) and the corrupt economics underlying the theory that there is any need or use whatsoever for any form of FRB or FRFB.
So the customers of Coinbase, even though they are treated practical as if they are irregular depositors, appear to be legally views as “creditors” and thus, they are at risk of losing their funds, even though Coinbase has 100% reserves, since treating them as creditors means that other creditors of Coinbase could have a claim on these assets “of Coinbase” in a bankruptcy.
I can think of a few solutions to this problem.
1. Coinbase could find a way to treat its customer-depositors as “secured creditors” who have a priority claim on the stored assets. This would prevent unsecured creditors from stepping first in line. Coinbase is presumably already engaging in what is in effect 100% reserves, so they should financially be able to do this, since all it would mean is Coinbase could not spend the funds of these secured assets. Nor should they be. When creditors extend credit to Coinbase they should know that they only have resort to assets of Coinbase other than those that back the claims of the customer-depositors.
2. Coinbase could adopt and announce an explicit policy where any creditor of Coinbase agrees not to pursue any of the assets corresponding to the claims of the customer-depositors. This might be tricky but I see no reason it could not be done.
3. Coinbase could investigate whether they could actually have the customers treated, under the law, as “irregular depositors” having actual ownership claims to the crypto and associated fiat on hand—on “deposit” so to speak. I am not sure if the current law would permit this but some creative lawyers might be able to achieve this with the appropriate paperwork and contracts etc.
4. Change the law to be more just and rational — a group of us proposed amending UK law along similar grounds this a while back: UK Proposal for Banking Reform: Fractional-Reserve Banking versus Deposits and Loans.
I’ve got what I think is an interesting question about Coinbase (and related crytpo/bitcoin exchanges) and fractional-reserve banking. The question is: does Coinbase have 100% bitcoin reserves, or fractional reserves? And how do we know?
A few preliminary comments and assumptions as to how this question arose, not all of which you necessarily need to agree with to consider this question meaningful and/or to weigh in.
First, I regard fractional-reserve banking as economic nonsense. I don’t agree that it’s inherently fraudulent, as some of the Rothbardians have argued, but it seems to always in fact have been fraudulent and today exists only because of state deposit insurance (e.g. the FDIC in the US), which is just legalized theft. I think it’s nonsense because it is based on the idea of market failure–sometimes we “need” to expand the supply of money and credit to stop a crash, or because wages are “sticky downwards,” all this nonsense. I think, like Mises, that money is a sui generis good — its not a capital good or consumer good, and increasing its stock does not increase wealth. Because money is not wealth. Unlike capital and consumer goods.
One central premise of FRFB (fractional-reserve freebanking) is that if you deposit money with a bank, then if they lend out 90% of it, then they are issuing you an IOU–a promissory note–and that if they “manage their assets properly” and have judicious use of “suspension clauses” (to stop runs) then the IOUs can circulate as money substitutes–just as paper “warehouse receipts” or title to an amount of deposited gold can. I think this is ridiculous because a promissory note to deliver a future good is going to be worth less than the good itself, and thus, the IOU can’t be a “substitute” for it. It’s inherently riskier. The reason is the future is uncertain and some of the people the bank lends money to will see their projects go bust and they’ll be unable to repay the loans. This will result in the “assets” of the bank (which includes lots of outstanding loans to people, some of whom are deadbeats) being worth less than its “debts” (to its “depositors”) and a run becomes inevitable. And a contagion effect will then happen. Exercising suspension clauses can only delay the inevitable (and their presence is another reason IOUs can’t be money substitutes–the suspension clause makes them also less valuable than the underlying good that is promised).
Some might argue that the banks can simply take out insurance policies to make IOUs as good as gold, but the problem here is this would in effect be insuring the entrepreneurial projects of the bank’s creditors, and this is impossible–it’s why Mises distinguishes case from class probability (see e.g. https://mises.org/library/human-action-0/html/pp/678 and Hoppe on probability). A bank can insure against its gold deposits being ruined by disaster or theft, but a business can’t insure its speculative future profits.
The point is FRB is nonsense. IMHO. This is just minor background.
Now many Austrians have been slow to understand the recent Bitcoin phenomenon. Some claiming it to be “impossible” because of Mises’s “regression theorem,” that any money “must be” based on a physical commodity, etc. These arguments are all half-baked and it’s becoming clearer that bitcoin is not “impossible.” But other Austrians like the freebankers seek leery of bitcoin precisely because its supply is basically limited and fixed, i..e it’s deflationary. And that’s one reason they support FRFB–they want inflation. They think it’s necessary. So some of them think bitcoin is either unsuitable as money or only if you somehow rig up some fractional-reserve scheme on top of it. This is one reason I am eagerly anticipating the debate on related matters between Selgin and Ammous.
In any case I have mused that one benefit of Bitcoin as money would be that it would be harder for FRB to emerge. In my view what happened was gold was deposited in banks for obvious reasons, and paper deposit certificates issued and used as substitutes. This made it easier for people to get confused over time and to equate the paper thing with money, and this made it easier for FRB to emerge. the banks were engaged in two functions: storing money, and also, credit intermediation: A loans money to the bank (and gets and IOU) and then the bank lends it out to borrowers. This way the bank shares some of the interest payments with its “depositors” and so if you deposit money you get paid interest. This leads to the possibility of bank runs, and also over time confuses people about the difference between title to money and IOUs (or promissory notes) — and then bank runs happen and the state steps in and while it’s cutting ties to gold because people have forgotten the connection between paper and gold, the state insures the deposits. This leads to central banking the whole point of which is to allow the state to inflate: to print money: without being restricted by physical gold supplies, and without making them have to engage in unpopular taxation. So then they can finance their wars, feed money to their friends in the military industrial complex, yada yada.
Anyway so with bitcoin you don’t really need a bank. You can easily keep all your bitcoin on a thumb drive. So if you give bitcoin to a bank for credit intermediation, you are under no illusions that you own the bitcoin: you are lending it to the bank, and they givey ou an IOU in return; and then you hope to make interest, but you take the risk that you might lose it all. And so on. But the idea that these bitcoin IOUs would circulate “on par” with actual bitcoins is absurd. No one would accept a note saying “I promise to pay A 1 bitcoin on demand” if the price for the good A wants to buy is 1 bitcoin. The promissory note is riskier than an actual bitcoin.
I like the approach of Huerta de Soto in his book, where he says that a deposit would be an “irregular” deposit: a bunch of people put their gold into the bank. Instead of each one keeping his particular coins in a safety deposit box—a “regular deposit”—because this is too expensive, and because gold is fungible, all the customers coins are pooled an each depositor has a pro-rata claim to that amount of gold. But the customers own it, not the bank, and the bank may not lend it out. If you want to earn interest on your gold, then instead of depositing it, you would “lend” it to the bank (and lose title to it) and receive an IOU instead. But then this IOU would not circulate as a money substitute because it is riskier than the gold (or title to the gold in the irregular deposit). So banking would have two functions: storage, and credit intermediation. In fact in the draft Bill I helped draft for the UK we proposed just this: to distinguish the two. (UK Proposal for Banking Reform: Fractional-Reserve Banking versus Deposits and Loans)
But it occurs to me that bitcoins are individually traceable and not quite as “fungible” as gold coins are, so that even if people might use a “Coinbase” type service to “hold” their BTC, there would be no reason to commingle them—i.e. no reason to use irregular deposits. Instead, all deposits would be regular deposits, akin to safety-deposit boxes. (Although my understanding is that Coinbase actually does commingle them—another reason not to use Coinbase and instead to have your own wallet.) In any case, it still seems to me that in a BTC system the risk that actual bitcoins (or claims to actual bitcoins, or to actual amounts of bitcoin in a storage service) would be conflated with bitcoin IOUs from the credit intermediation side, would be radically minimized. Plus, in an imagined BTC world (of the future) the state would either not exist or its role in money would be radically limited (otherwise BTC can’t emerge). So the state’s role that ie played in helping cut the tie from paper notes to gold and thus establish the Ponzi scheme we have now—would not exist. I.e., it’s hard to imagine FRB or FRFB emerging, in a BTC world. In fact, this is in my view one fear the freelancers have: they are worried Bitcoin can’t support FRB (or, to put it the way we Rothbardians think: there is little chance the huxters could get away with foisting a bullshit, unnecessary FRB scheme on a BTC world).
So one of my friends is somewhat in favor of FRFB. I tried to point out that he is a crypto-fan and thus he has to realize that crypto and FRFB don’t mix–and this is one reason the freebankers are leery of bitcoin: because it’s too deflationary. So I pointed out what I mentioned above: that you would not have a money storage “bank” service emerge like we. did with gold, since it’s so easy to store your own bitcoin.
He countered: but people use Coinbase now.
I said yeah but bitcoin is in its infancy, and eventually it’ll be more user friendly. People used AOL at first too instead of their own email and browsers, but now no one uses AOL.
So then he said but Coinbase is now paying interest. I said…. Nooooo that can’t be possible. Because that would mean Coinbase actually doesn’t own all the BTC that its customers have claims on. That would be fraudulent, because I have never heard them disclose this bullshit to me. So he points me to some WSJ article “Coinbase’s New Customer Incentive: Interest Payments, With a Crypto Twist” — but I think the title is misleading. It’s not interest, just some kind of joint staking operation for some ‘proof of stake’ coins, but not Bitcoin: “Because of the difference between proof of stake and proof of work, bitcoin isn’t an option for staking. Coinbase’s staking service will start with only one cryptocurrency, called Tezos, which launched last year. Other cryptos will be added later.”
So it’s not clear to me that “interest” is being paid as a result of FRB.
So I said, Coinbase must have 100% BTC reserves— so that it holds enough BTC to reimburse all the customers. Otherwise it’s fraud. So then he said no, Coinbase and these other exchanges put limits on how much you can withdraw at a time. So there could in principle be a “run” on Coinbase.
So this gave me pause. First, it made me glad I no longer use Coinbase. Second, I’m thinking—even though I don’t quite agree with the Rothbardians that FRB (including FRFB) is “inherently” fraudulent—if Coinbase really doesn’t hold as many BTC as all of its customers have claims to, that is some fraudulent bullshit.
So now I’m wondering. What the hell. Is this true? Does Coinbase have 100% BTC reserves—or not? What I’m thinking is that my friend is confused by the fact that it might take some time delay for customers to get large sums of BTC out—and this is not because they are not 100% backed, but because most of their reserves are in cold storage. As far as I can tell BTC has private insurance but only in case of theft and only against about 2% of their assets—the “online” part. The rest is “offline” or what I assume is “cold storage” and thus not insured, and thus would take longer to retrieve:
SEE: Coinbase Help: How is Coinbase insured?
Coinbase prioritizes the security of our customer’s funds, all digital currency that Coinbase holds online is insured. If Coinbase were to suffer a breach of its online storage, the insurance policy would pay out to cover any customer funds lost as a result. Coinbase holds less than 2% of customer funds online. The rest is held in offline storage.
So I return to my original question: Does Coinbase have 100% BTC reserves, or not? If not, why don’t they clearly disclose this fact so that customers are not deceived into thinking they are “depositors”?
Update: in the aftermath of the FTX implosion, Coinbase is sending out messages to its users reassuring them that they are not engaged in FRB. For example, this email was sent to a friend. It reads, in part:
A more transparent and secure crypto exchange
Hi David Johnathan,
With the recent liquidity issues in the crypto markets this week, we want to reiterate how Coinbase’s business is different and ultimately better protects your account & digital assets.
For the last 10 years, we have built Coinbase in a way that allows us to be transparent about our track record and balance sheet strength, and effectively and prudently manage risk for our 108+ million customers and ourselves.
See below for more information about how we keep your funds safe: …
We hold your Assets 1:1
Coinbase holds customer assets 1:1, and we won’t lend those assets without your consent. This means funds are available to our customers at any time.
Also in response to the FTX implosion, this from Grayscale Bitcoin Trust:
Safety, Security, and Transparency
Due to recent events, investors are understandably inquiring deeper into their crypto investments. Custody of the digital assets underlying Grayscale’s digital asset products is unaffected, and our products’ digital assets remain safe and secure. We take transparency seriously, and have a long history of working constructively with regulators to create and strengthen full and fair risk disclosures for our digital asset products. In that spirit, we affirm the following, as of November 18, 2022:
- The holdings of Grayscale’s digital asset products are safe and secure. Balances are reflected in historical public filings and have been evaluated by our third-party auditors.
- Each of Grayscale’s digital asset products is set up as a separate legal entity: a statutory trust for single asset products, and limited liability company for diversified products.
- All digital assets that underlie Grayscale’s digital asset products are held by Coinbase Custody Trust Company, LLC (Coinbase Custody) as custodian for each product. A letter that verifies the number of digital assets held by Coinbase Custody for each of Grayscale’s digital asset products can be found below.
- The organizational documents governing each of Grayscale’s digital asset products, as well as the custodian agreement with Coinbase Custody, prohibit the digital assets underlying the products from being lent, borrowed, or otherwise encumbered.
Let’s dig a little deeper:
Each Grayscale digital asset product is structured as a separate legal entity – a statutory trust for each single asset product, and a limited liability company for each diversified digital asset product – sponsored or managed by Grayscale Investments, but ultimately owned by the product’s shareholders. For example, this means that Grayscale Bitcoin Trust (OTCQX: GBTC) holds bitcoin — and only bitcoin — and each share is backed by a proportional amount of the trust’s holdings, approximately 0.00091502 BTC per share of GBTC, as of November 18, 2022. To be perfectly clear: these digital assets are owned by GBTC and GBTC alone.
Grayscale’s nine largest digital asset products are SEC reporting companies that file publicly available Forms 10-K, 10-Q, and 8-K. With each Form 10-K and 10-Q, Grayscale’s CEO and CFO certify that such reports meet the requirements of Section 13(a) or 15(d) of the Securities Exchange Act of 1934, and fairly present, in all material respects, the financial condition and results of operations of each product, among other certifications, as required by the Sarbanes-Oxley Act of 2002. Annual financial statements for these products are audited according to U.S. generally accepted accounting principles (GAAP). These products’ public filings are available at the links below:
- Grayscale Bitcoin Trust (OTCQX: GBTC) SEC Documents
- Grayscale Bitcoin Cash Trust (OTCQX: BCHG) SEC Documents
- Grayscale Ethereum Trust (OTCQX: ETHE) SEC Documents
- Grayscale Ethereum Classic Trust (OTCQX: ETCG) SEC Documents
- Grayscale Horizen Trust (OTCQX: HZEN) SEC Documents
- Grayscale Litecoin Trust (OTCQX: LTCN) SEC Documents
- Grayscale Stellar Lumens Trust (OTCQX: GXLM) SEC Documents
- Grayscale Zcash Trust (OTCQX: ZCSH) SEC Documents
- Grayscale Digital Large Cap Fund (OTCQX: GDLC) SEC Documents
Additionally, we provide GAAP audited financial statements in compliance with the highest reporting standards prescribed by OTC Markets, the Alternative Reporting Standards, for five of our newer publicly traded digital asset products:
- Grayscale Basic Attention Token Trust (OTCQB: GBAT) OTC Markets Documents
- Grayscale Decentraland Trust (OTCQX: MANA) OTC Markets Documents
- Grayscale Chainlink Trust (OTCQB: GLNK) OTC Markets Documents
- Grayscale Filecoin Trust (OTCQB: FILG) OTC Markets Documents
- Grayscale Livepeer Trust (OTCQB: GLIV) OTC Markets Documents
We also directly provide investors with GAAP audited financial statements pursuant to each product’s governing documents for our three newest private placement-only products, which are not currently publicly traded. (Grayscale may attempt to have the shares of new products quoted on a secondary market. Although the shares of certain Grayscale products have been approved for trading on a secondary market, such as OTCQX, investors in a new product should not assume that the shares will ever obtain such an approval because of questions the SEC and/or FINRA may have regarding the status of the product’s underlying digital assets under the federal securities laws.)
- Grayscale Decentralized Finance (DeFi) Fund
- Grayscale Solana Trust
- Grayscale Smart Contract Platform Ex-Ethereum Fund
All digital assets that underlie Grayscale’s digital asset products are custodied with Coinbase Custody. As background, Coinbase Custody is a wholly-owned subsidiary of Coinbase Global, Inc. (NASDAQ: COIN), and is licensed to custody client digital assets as a New York-chartered limited purpose trust company. Coinbase Custody has been regulated by the New York State Department of Financial Services since 2018, the same regulator that oversees some of the United States’ biggest banks. Coinbase Custody also services as a fiduciary, which means that it is required to always act in its clients’ best interest under the New York Banking Law.
Coinbase Custody frequently performs on-chain validation as part of their custodian operations. Due to security concerns, we do not make such on-chain wallet information and confirmation data publicly available through a cryptographic Proof-of-Reserve, or other advanced cryptographic accounting procedure.
Coinbase Custody is also obligated under the terms of the custodian agreement in place with each Grayscale product to fulfill a number of critical operational practices that ensure that the digital assets underlying each Grayscale product are safe. Coinbase shared more information about our relationship for Grayscale’s investors and indicated the exact quantity of digital assets underlying each of our digital asset products in the letter linked here.
Sanctity of Assets
Grayscale and Coinbase Custody are prohibited from borrowing, lending, rehypothecating, or otherwise encumbering the products’ underlying assets pursuant to the terms of the trust agreement or limited liability company agreement governing each digital asset product. No other entity, including DCG, Genesis, nor any other Grayscale affiliate, has any control over the digital assets underlying the Grayscale products. Each of our products carries identical prohibitions to those outlined in GBTC’s trust agreement :
As custodian, Coinbase Custody is also prohibited from engaging in such activity, as outlined in the products’ corresponding custody agreements. Keeping with GBTC’s custody agreement, as an example:
Crypto captivated many investors not only for the investment opportunity, but because of a shared belief that decentralization matters, financial inclusion matters, and self-sovereign digital money matters. There is a whole community of talented individuals and developers who continue to work together to achieve this vision. As always, Grayscale remains committed to providing investors with access to this ecosystem through a suite of accessible, secure, and future-forward investment products.
- See this language from a recent 10Q:
As of March 31, 2022, we held $256 billion in custodial fiat currencies and cryptocurrencies on behalf of customers. Supported crypto assets are not insured or guaranteed by any government or government agency. We have also entered into partnerships with third parties, such as with the Centre Consortium, as a reseller of USDC, where we or our partners receive and hold funds for the benefit of our customers. Our and our partners’ abilities to manage and accurately safeguard these customer assets requires a high level of internal controls. As our business continues to grow and we expand our product and service offerings, we must continue to strengthen our associated internal controls and ensure that our partners do the same. Our success and the success of our offerings requires significant public confidence in our and our partners’ ability to properly manage customers’ balances and handle large and growing transaction volumes and amounts of customer funds. In addition, we are dependent on our partners’ operations, liquidity, and financial condition for the proper maintenance, use, and safekeeping of these customer assets. Any failure by us or our partners to maintain the necessary controls or to manage customer crypto assets and funds appropriately and in compliance with applicable regulatory requirements could result in reputational harm, litigation, regulatory enforcement actions, significant financial losses, lead customers to discontinue or reduce their use of our and our partners’ products, and result in significant penalties and fines and additional restrictions, which could adversely impact our business, operating results, and financial condition. Moreover, because custodially held crypto assets may be considered to be the property of a bankruptcy estate, in the event of a bankruptcy, the crypto assets we hold in custody on behalf of our customers could be subject to bankruptcy proceedings and such customers could be treated as our general unsecured creditors. This may result in customers finding our custodial services more risky and less attractive and any failure to increase our customer base, discontinuation or reduction in use of our platform and products by existing customers as a result could adversely impact our business, operating results, and financial condition. [↩]