Share to linkedin Bitcoin rehypothecation and chain forks won’t mix well. Photo credit: Shutterstock Several astute readers asked follow-up questions about the risks facing Wall Street amid a Bitcoin chain fork , a topic that I introduced in a recent Forbes.com piece ( here ).
The chain-fork issue is confusing. Chain splits, for example, seem so simple—if you own bitcoin before a hard fork, you own the post-fork coins too, right?! Nope, not when uncovered bitcoin exposures are involved—these are positions in a bitcoin-substitute that are not 100% collateralized with real, on-chain bitcoins. The uncovered amount is the amount not backed by real bitcoins.
For example, if the financial system has promised 5 bitcoins to customers but has only 3 bitcoins in its custody, the system’s uncovered exposure is 2 bitcoins.
Here’s the problem: how does the financial system come up with the 2 new post-fork coins to meet its obligation? It promised 5 bitcoins to customers but holds only 3 in its custody, so it received only 3 of the post-fork coins. Consequently, its total uncovered exposure after the fork is now the 2 pre-fork coins plus 2 post-fork coins. The system must immediately credit its customers’ accounts with all 5 post-fork coins, but it has only 3.
Can the financial system solve this by conjuring 2 post-fork coins out of thin air? Nope. No one in the financial system has power to create new coins.
Can it just replicate the process by which it originally created the uncovered position in the 2 pre-fork coins? Nope, because it takes time to build uncovered exposure—it builds gradually within the system as institutions lend and re-lend the same bitcoin to each other.
So, the financial system has no easy out.
It must cover the shortfall immediately .
How will it immediately come up with the 2 post-fork coins?
Answer: it would need to purchase them into a what could be a very illiquid secondary market. Why would it be so illiquid? Because forked coins could take several days to start trading in large volume as few holders begin to trade them immediately, and as engineers at crypto exchanges typically take time to determine the feasibility and security implications of adding the new coin. But the financial system owes those post-fork coins to its customers immediately .
Could the financial system delay delivering the 5 post-fork coins to its customers, or never deliver them? This is how many crypto exchanges have handled previous forks. But crypto exchanges serve mostly retail customers, while Wall Street is in a whole different zip code. The fiduciaries for institutional investors will expect all forks to be delivered as soon as they are available, just as they expect all stock dividends and other corporate actions to be credited to customers when paid.
Giving full discretion to any custodian or counterparty regarding whether to deliver a fork might fly for retail crypto investors, but the fiduciaries of institutional investors are unlikely to agree. More on the fiduciary issues below.
Net-net, the financial system would need to source those post-fork coins somewhere.
Fast. And at any cost.
Many folks are surprised to learn that uncovered exposures would ever be allowed to develop, such as the 2 pre-fork bitcoins in the example. Actually, it’s standard operating procedure on Wall Street.
Uncovered exposures build gradually within the traditional financial system, mostly owing to commingling and rehypothecation of securities. This can take many forms, including securities lending, repurchase agreements, derivatives and prime brokerage.
But the end effect is the same—uncovered, fractionally-reserved exposures build within the system slowly and mostly undetectably.
Rehypothecation is nearly impossible to measure at a systemic level, which is why I’ve referred to the practice as insidious and subtle.
Why? Because US GAAP accounting enables multiple parties to report on their financial statements that they own the very same asset. No one really knows how many times that same asset has been double/triple/quadruple/quintuple counted, which means that no one really knows how solvent the financial system is if all such double-counting were backed out.
Returning to the bitcoin hypothetical example, it’s very possible that the bitcoin fork itself causes the system’s uncovered exposure to become measurable for the first time. It’s likely no one (including regulators) knew how big the uncovered position was before the fork.
The magnitude of uncovered exposure usually isn’t discoverable until a triggering event requires an accurate accounting, which is when the game of musical chairs stops. Such “reckoning” events could be a merger (such as the Dole Food example, where brokerage statements showed 33% more shares outstanding than were legally issued), a contested proxy battle (where the true winner in any contest whose vote is closer than 55-45% is unverifiable, according to a prominent Delaware judge), or a pension plan settlement—or, in the case of bitcoin, a hard fork.
Again, any exposed institution could face high loss severity on that uncovered exposure. Not even intra-day (because large intra-day moves can create uncovered “gaps” in margin posting). Every financial institution should simply follow these rules: Never commingle bitcoin. Never rehypothecate bitcoin.
Always require 100% collateralization of any bitcoin-substitute, even intra-day.
It’s really not that complicated! I realize, however, that this is not how Wall Street works. Rehypothecation is as integral to Wall Street as it is antithetical to bitcoin.
Along those lines, I’m grateful to the IMF ’s Dr. Manmohan Singh, whose work I discussed in the prior article , for reaching out to let me know he has published 2017 data on collateral reuse (see Table 1 below). Dr.
Singh is the foremost expert on the topic and I highly recommend this paper for readers interested in learning more about how rehypothecation works, as well as more insightful writings on this topic by Dr. Singh here , here , here , here and here .
Again, Dr. Singh has recommended that regulators’ financial stability assessments be adjusted to back out “pledged collateral, or the associated reuse of such assets,” but policymakers have not heeded his wise advice.
Dr. Singh estimates that collateral is re-used 2.
0 times (as of year-end 2017). This means only one of the three people who think they own a U.
S. Treasury bond, for example, actually does own it— namely, the original owner of the bond and the two parties that reused the bond. The brokerage statements of all three people show that they own the bond, so it’s impossible for them to detect that behind the scenes only one bond exists! Source: IMF working paper # 17/113 “Collateral reuse and balance sheet space”-updated. IMF
Dr.
Singh’s data show an improvement since the financial crisis, when four parties reported in 2007 that they owned the same asset (namely, the original owner plus three reusers of it). By “improvement” since the financial crisis, I mean that the 3.0 number from 2007 declined to 2.0 in 2017.
Rehypothecation is a major source of the financial system’s instability, so declining rehypothecation means less instability. But there’s more.
Only once a year can we catch a glimpse into the magnitude of rehypothecation, because many banks only disclose collateral statistics once a year in their annual reports. The true magnitude of rehypothecation is likely much higher because financial institutions “window-dress” their year-end balance sheets to bring leverage numbers down for financial reporting dates —indeed, this year’s BIS Annual Report found: “The data indicate that window-dressing in repo markets is material .
” On any given day, no one really knows.
Some commentators defend rehypothecation—a form of fractional-reserve banking—as a necessary element in the financial system to ensure liquidity and encourage price discovery. These commentators would rather see the 2.0 number in Table 1 go back up, not down, because that would lubricate the financial system with liquidity as the $7.
5 trillion volume—$7.5 trillion(!)—of pledged collateral goes forth and multiplies.
That is a debate for another day.
But, when talking about bitcoin, there’s a game-changing twist to this debate about fractional-reserve banking. There is no lender-of-last-resort for bitcoin. For cash, there is a lender-of-last-resort—the Federal Reserve—and it has bailed out the financial system’s uncovered exposures many times before. But bitcoin is different. No more than 21 million bitcoins will ever be created.
Again, in the bitcoin example, the financial system would need to acquire those 2 post-fork coins somewhere, fast, and potentially into a massive short squeeze. The bigger the uncovered exposure in the financial system, the bigger the pain —all of which, incidentally, is self-inflicted.
Finally, there are two more prongs of the fork topic to discuss.
The first is how legal agreements pertaining to forks (soft or hard) will be written.
The attorneys tasked with drafting Wall Street’s legal contracts have an impossible task, since no two forks are the same and it is impossible to foresee all potential forks.
Here’s how LedgerX , a crypto company regulated as a SEF (swap execution facility), is handling it: “Our management and risk committees will evaluate each hard fork on a case-by-case basis.”
Think about that statement. Does it promise anything?
Open-ended discretion regarding a potentially high-value issue will likely be rejected by the fiduciaries of most large institutional investors, such as mutual funds and pension funds.
When investors don’t receive the full value of forks, someone’s got to pay. It shouldn’t be Mom and Pop. It’s likely to be either the custodian/central counterparty, or the fiduciary that delegated open-ended discretion to them.
And ERISA fiduciaries have personal liability. Litigators will be busy…
Second, some bitcoin forks may raise compliance issues for Wall Street. Hat tip to Andreas Antonopolous, an engineer who is a highly respected bitcoin educator, for raising this scenario : what happens if (when?) bitcoin adopts confidential transactions? These would likely make it impossible for regulated financial institutions to comply with know-your-customer and anti-money laundering laws. Would Wall Street’s compliance officers allow adoption of such a fork?
In that scenario, Wall Street itself could fork into its own “ corpocoin ,” to use Andreas’s phrase.
But then “corpocoin” would be just another financial asset, just another IOU issued by a centralized intermediary that is leveraged and might default, just another asset tracked in non-transparent ledgers that insidiously create the kind of systemic leverage detailed by the IMF.
Nope, bitcoin rehypothecation and chain forks do not mix well, and the simple reason is that the two systems are fundamentally not compatible. One is decentralized, the other is centralized.
One is natively-digital, owned directly and allocated to individuals. The other is natively-paper, owned indirectly, commingled at inception and then rehypothecated. One has a ledger that prevents multiple people from owning the same asset, the other creates multiple owners for the same asset on a systemic level.
I’ll leave you with a question to ponder: which is the bigger risk—Wall Street to bitcoin, or bitcoin to Wall Street?? I’m a 22-year Wall Street veteran who has been active in bitcoin since 2012, and whose passion is a fair and stable financial system. I saw inaccuracies in Wall Street’s ledger systems while running Morgan Stanley’s pension solutions business (2007-2016), holding senior role..
.
.
