Crypto-assets and decentralized finance through a financial stability lens
federalreserve.gov
federalreserve.gov
"... same risks that are all too familiar from traditional finance, such as leverage, settlement, opacity, and maturity and liquidity transformation."
Yes. I've been making this point for years. It's mostly problems that are well known. It's just that the crypto crowd doesn't study financial history, so they get to repeat it. Although crypto has its own unique risks - it's really hard to secure crypto assets and still use them, and fast anonymous remote irrevocable transfers make it hard to catch crooks.
"... there is a risk that a single dominant stablecoin might emerge, given the winner-takes-all dynamics in such activities. Indeed, the market is currently highly concentrated among three dominant stablecoins, and it risks becoming even more concentrated in the future. The top three stablecoins account for almost 90 percent of transactions, and the top two of these account for 80 percent of market capitalization."
Network effects tend to concentrate stablecoins. The Fed points out that they're already very concentrated. It's a single point of failure. If USDT crashes, a lot of people will be hurting. Many of them had no idea they had a big exposure to risk.
*"...decentralized lending, which relies on overcollateralization to substitute for intermediation, can serve as a stress amplifier by creating waves of liquidations as prices fall."
Right. This is the concept that if you bundle enough high risk investments together, you get a low risk investment. The trouble is that the risks are usually correlated; multiple investments of the same class crash together. Add in leverage, and the crash speeds up.
The GSE loan guarantee program is largely backstopped by the Federal Reserve. It underwrites $6 trillion worth of US mortgages, which is approximately 50% of the entire market.
In 1999, the GSEs set out to increase liquidity in so-called under-served residential mortgage markets, and thereby reduce the risk for private lenders to lend in that market, as this article describes:
https://www.nytimes.com/1999/09/30/business/fannie-mae-eases...
>>''Fannie Mae has expanded home ownership for millions of families in the 1990's by reducing down payment requirements,'' said Franklin D. Raines, Fannie Mae's chairman and chief executive officer.
>>'Yet there remain too many borrowers whose credit is just a notch below what our underwriting has required who have been relegated to paying significantly higher mortgage rates in the so-called subprime market.''
In terms of the Fed's direct role, its actions in 2003 - 2005 largely followed the playbook prescribed by Fed mouthpiece, Paul Krugman in 2002:
https://www.nytimes.com/1999/09/30/business/fannie-mae-eases...
>>To fight this recession the Fed needs more than a snapback; it needs soaring household spending to offset moribund business investment. And to do that, as Paul McCulley of Pimco put it, Alan Greenspan needs to create a housing bubble to replace the Nasdaq bubble.
The market's natural behavior, which is to bankrupt irresponsible lenders, borrowers and speculators, seems to work in preventing bubbles from growing enough to cause financial instability. The collapse of Terra/Luna imposes more financial discipline in the crypto markets than any set of regulations could.
Attempts to artificially reduce instability can often have the opposite effect, by socializing the harmful effects of irresponsible risk taking, and thus creating moral hazard. This is the case with the Fed's role as the publicly funded lender of last resort, and the FDIC's role as publicly funded deposit insurer.
With respect to the latter, here is a study drawing that conclusion:
Can we put this in less technical jargon that describes what happens based on the structural system of how the feds actions make its way to economy as a whole?
Attempts to reduce instability create stability for wealthy class, like wall Street and bankers, while socializing the cost of maintaining stability by increasing the cost of living for everyone else, which most disproportionately hurts people on fixed incomes, and makes social welfare harder to provide by government and private charity alike.
Yes. The short term stabilization brought about by the socialization of losses and due diligence costs also degrades the weeding out process that instills long term discipline on financial institutions.
- The Fed decides who can use US dollars and what you can do with them; stable coins provide ways to evade such mechanisms.
- It wants to avoid spillover in case the cryptocurrency market collapses entirely. Shadow banking and eurodollars played a role during the 2008 crisis.
I would caution against trying to forestall financial crises through imposed centralization. The mundane benefit of a permissionless/dynamic financial system - which is to facilitate investments that produce economic returns - accrue daily, usually in imperceptible increments, while financial crises occur once every few years, with large and sudden drops in market values.
The latter gets the attention, due to their suddenness, but the benefits of the former compound to enormous proportions, and go largely unseen, due to how gradual the process is.
> Innovation has the potential to make financial services faster, cheaper, and more inclusive and to do so in ways that are native to the digital ecosystem. Enabling responsible innovation to flourish will require that the regulatory perimeter encompass the crypto financial system according to the principle of like risk, like regulatory outcome, and that novel risks associated with the new technologies be appropriately addressed.
The Fed recognizes this is a new tech stack with potentially disruptive consequences. Reconcile that with this anachronistic tulip mania point of view you're parroting.
As if being able to transact with anyone else on the planet sans middlemen isn't enough utility on its own. I would instead ask people who make comments like the one that started this thread to substantiate their affirmative claims about how crypto is only used for these shady things. For some reason, no such evidence is ever forthcoming.
1. International settlements of any sum of money (small to large) in 10-20 minutes instead of days
2. Markedly lower money transfer fees compared wire transfer or ACH
3. Ability to be one’s own bank if so desired, avoiding government bank account pillaging (Cyprus, Argentina, many more)
4. Ability to send money anywhere (try paying your staff in Russia with USD these days… we have two; they both get paid in Bitcoin since following Russia’s removal from SWIFT cryptocurrency is now the only way)
5. When comparing Layer 2 transactions, orders of magnitude more transactions processed (global credit card network: 19,000 transactions per second; Bitcoin’s Lightning network: several million transactions per second), for significantly lower fees (very appealing if you are a merchant); payment settlement in hours instead of days
6. Deflationary savings account over medium- and long-term timescales
2) Debatable because of volatility.
3) Yes but being your own bank also means losing everything is a forgotten/misplaced password/seed phrase, clicking the wrong link, etc away. I don't know how many password resets a big bank in the US does everyday but I'm sure it's a mindbogglingly large number. The vast majority of the population isn't ready for and won't tolerate this. The fees you describe for payments in the financial system have anti-fraud measure costs (reversals, etc) baked in. Yet another feature of the traditional financial system that has been developed (in reality) after decades of real-world experience. It's the equivalent of every single individual building their own hardened vault and hiring armed private security.
4) Fair enough but there are realities in potentially "skirting" the "law" like this. Banks have significant processes to make sure you're not (for example) "funding terrorism" or whatever which is a serious crime in the US and very easy to do with cryptocurrencies. I, for one, don't want to risk the Feds showing up at my door because my funds ended up with someone on my government's enemies list. Just because you can do it with Bitcoin doesn't magically mean the people with guns and prisons will just say "Oh Bitcoin - nevermind - that's ok".
5) Lightning (and for that L2s) are mostly bolt-on hacks that sacrifice one or more features/properties of cryptocurrencies as originally intended when it became clear they fundamentally don't work for anything beyond toy-level. Again, from a users perspective as long as the payment network allows you to swipe a card and walk out with your purchase ASAP transaction rates are invisible to the user. If traditional payment systems needed higher transaction rates they would magically appear.
6) This is a very dangerous supposition.
5 confuses theoretical maximum throughput with actual throughput. The theoretical maximum throughput of credit card networks is far higher than the actual, and the actual throughput of lightning is far lower than theoretical. Since only three wallets can join the lightning network per second (if all Bitcoin transactions were setting up lightning channels), its utility for actual transactions between arbitrary wallets remains theoretical as well.
6 confuses the value of Bitcoin with its supply. The value quickly approaches 0 when regulation takes away advantages 1-4.
Its utility is here. To say otherwise is just sticking heads in sand.
As long as people can interact with both easily, they will, and a lot of the risk/rewards will get tied together. Things move up together and down together.
I agree with the beanie baby analogy when it comes to the various tokens that people "invest" in. But I wouldn't be so quick to say there was no utility, at least conceptually. It didn't really "click" for me until I implemented a smart contract. I realized that I could implement the entire financial system relatively easily, e.g. I could create an analog for stocks, options, margin lending, etc. Building a similar system without the blockchain would be much more difficult. I certainly couldn't do it alone. And even if I did do it with the blockchain, there would be an additional step of convincing people to trust that my system would stay up. Smart contracts run "forever" (in a sense).
Name the specific parts of mortgages that you believe should not exist and that DeFi eliminates.
Should we not do income checks on homebuyers to ensure that they can reasonably be able to pay the debt?
If they can’t pay their debt, what happens to the home? Is it collectively owned by the smart contract? How does that work? Where does the deed go?
Speaking of the deed, what if there’s a bug in the contract? Are we really going to pretend that these contracts have the force of law and a house can be stolen because someone wrote a bug? Or are we accepting that we can override the contract and none of the benefits of “smart contracts run[ning] forever” are real?
How should a smart contract value a home, to determine if the loan is adequately backed? If the buyer wants a $1m mortgage on what is actually a $10k plot of land, they can default and the contract (?) gets (??) an asset two orders of magnitude less than the check it wrote.
Hand-waving “well every real estate transaction will also be on the blockchain with perfect metadata so AI can make that decision” is not an answer.
How does a DeFi platform prevent money laundering? Should it not? This requires people! And regulations!
This is a tiny slice.
But we cannot pretend that any of this makes sense without thinking more than one layer deep. Which is really the problem.
Let's use your example of mortgages. When you get a mortgage, your lender wires money to the seller. How do they do that? They use Fedwire, which is a system maintained by the Federal Reserve for banking transfers. Such a system can be totally replaced by a blockchain. If you replaced Fedwire with a blockchain, you wouldn't remove the regulatory and legal requirements, you would just replace a legacy technical system with something that is more powerful. A blockchain can do everything Fedwire can do, and a blockchain implements this functionality in a more generic fashion that allows for additional constructs, like smart contracts, to be added.
Following your mortgage a little further: most mortgages are not held by the bank. The was the main issue in the financial crisis in 2008. Banks don't hold mortgages, so they were very loose in who they lent to. Rather than hold on to the mortgages, the bank sends them to a clearing house that packages up the mortgages into a mortgage backed security. All of this infrastructure could be replaced by a smart contract.
A blockchain can be thought of as a generalization of a financial system (system here meaning the technical system, i.e. the nitty gritty details of how and when money is moved). This is a powerful generalization and can implement existing financial systems in a much more efficient manner. As an individual, I could implement a mortgage-backed security system. That type of productivity is not possible in our current financial system.
This does not mean that I'm advocating for some type of anarchistic hellscape where regulations fall by the wayside. You can still have the same set of regulations, but implement the underlying nuts and bolts financial system in a way that's more efficient, standardized, and democratic.
These pseudo zero-risk loans aren't exactly useful. The hard part of finance is in valuations, risk assessment, and regulatory compliance. And I stress the word pseudo, even DAI is built on a house of cards, although somewhat sturdier than its competition [order of mangitude depreciation, and when the market can't catch a bid it may be game over]
If tulips bulbs were of a limited supply, (relatively) infinitely divisible, (relatively) impossible to forge, and could be transferred anywhere in the world, in any amount, (relatively) instantaneously… Would it really be that crazy to imagine that they may have kept significantly more value? Or be used as a kind of currency?
It seems like anything with those properties could be a good basis for a currency, whether it’s Beanie Babies or Pokémon cards (neither of which actually do have those properties), or cryptographically secure blocks of teleporting cheese.
True until another batch of cryptotulip bulbs is released under a new name with slightly different properties and it joins the cryptotulip ecosystem. Any one species of cryptotulip might be a limited supply, but on the whole it's a very inflationary ecosystem, with trust on which is the "true currency" extremely malleable and divided. Even the "original" BitTulip has as many haters expecting its inevitable dethroning as it has proponents.
Even if the original Dutch tulips had somehow limited supply growth per species - say only selling neutered bulbs incapable of reproducing, while "miner" growers do the actual growing - you still have inflation from new sellers growing new species. Are people just supposed to arbitrarily say "Only orange Tulipa gesneriana tulips are a valid currency"? How does that consensus happen in a decentralized environment, when clearly it's not happening in this crypto bubble?
Is the argument that the limited supply comes from limited farmland? (AKA limited miners to generate the underlying security?) If so, how does that scale past the point where farming/mining is taking a significant chunk of the world's energy/land and just maintaining the currency is costing more than it saves on efficient asset transfer of actually productive useful things (like say - growing more crops, or using GPUs for AI)?
"Proof of Stake solves this!" you say? Well, now we're back to arbitrarily naming a tulip species as the only True Currency, with no physical resources backing it, and no particular reason people will stay loyal to it...
(Granted, the current non-crypto alternative is a heavily-armed centralized tulip farm that can grow its supply as it sees fit, and use its military to stifle other growers and guarantee its continued dominance. It allows a few other varieties (CAD, GBP, Yen, etc) but makes sure those growers back their supplies with a large stock of USDtulips, making them subsidiary growers bound to rise and fall with the central farm.)
Am I missing something?
The earth is a decentralized environment in which USD, EUR, GBP, and JPY float against each other, and their value at any time is determined by market confidence. You shouldn't expect 100% global consensus on a particular cryptocurrency any more than you expect 100% global consensus on (say) USD or EUR.
If I print a stack of ThrowBills, would you buy them for $1? If I mint ThrowCoins, would you buy them for 1 BTC? If you decline my offer, then you understand that value does not come for free (crypto or not), and merely inventing a new tulip does not inflate existing tulips -- it competes with them.
Back to tulips: though inventing a new tulip does not inflate existing tulips, any level of public support for that tulip dilutes everyone's confidence in the original tulip, and creates two more unstable tulip currencies vs just the one. This is because everyone has just a bit more doubt on which one they should hold.
"But that's decentralized then. If people lose confidence in one of those tulips - good thing you diversified in both.". Sure, except for losing confidence in one is likely to create a cascading effect of lack of confidence in the others as the public starts to rightfully doubt their magical belief that a useless commodity is valuable. Same panics happen with banks. And a big plethora of smaller currencies means the initial dominos of the panic are that much easier to push over - leading to the possibility of the entire market wiping. Tulips all the way down.
Again, not that this isn't the same process found in conventional finance. There's just additional confidence that there will be (perhaps violent) interventions to stem any such lack of confidence/panic before it becomes critical. And there's been a long history of independent currencies being shut down with the same force due to not wanting to compete.