I understand the use case for a stablecoin that needs to link with the crypto world—whatever one may think of crypto, if stablecoins exist, there’s a good argument that the organizer should lack a profit motive and the profits should go to something socially useful.
I am struggling to understand the other use cases (e.g., “savers” on the Glo website). For such users, this seems akin to a money market fund owned by a charity that pays 0% interest and instead donates the interest to charity. I am not sure what the added advantage having a crypto element adds to that basic idea (which is probably a good one, by the way). Right now, I suspect the answer is lower operating costs—but I think there is a high probability that much more extensive regulation is coming for crypto, and that operating costs of companies like Glo will begin to approach costs for banks and other financial institutions.
Indeed, this seems for non-crypto use cases potentially inferior to opening the “GiveDirectly Bank”—i.e., a US bank owned by a non-profit that donates its profits to GiveDirectly. Because Glo isn’t backstopped by the US FDIC (or equivalent), it has to purchase ultrasafe, highly liquid (thus: low-yield) assets to imperfectly simulate that backstop. In contrast, the GD Bank would have a freer hand in investing user capital for higher yield because (up to FDIC insurance limits, which can be creatively stretched to several million dollars per depositor) the risk of loss is ultimately borne by the US Government.
Four separate points here: 1) the stablecoin use case makes sense to you; 2) operating expenses for Glo are low today but likely to be higher in the future after regulation; 3) why crypto versus any other way of implementing the core idea, which is turning some portion of a person’s “cash on hand” into an automatically-donating vehicle; and 4) in particular why not an EA-aligned bank (re: credit union) that just donates to GiveDirectly?
RE: 2: We’re gonna have to wait and see what specifics shake out. Our overall hope is that by registering as a non-profit, and establishing formal oversight and governance within the established category of 501(c)3, we’ll be compliant with whatever comes by default. But this is speculative.
RE: 3-4 We’ve had this idea and I even wrote up a version of it in a doc called “The case for an effective altruist credit union” (which I would be happy to share with you if you’d like to DM me an email address). The basic consensus is: stablecoin makes a lot of sense as a first product, because the DeFi market is already huge and, frankly, in need of some better infrastructure (ultimately, stablecoins are infrastructure), so that is a fine beachhead for the overall idea. Second, launching a bank that is broadly available across the world is going to be very, very challenging. Glo the stablecoin is available just about everywhere the moment it’s on exchanges.
But yes, the overall idea of “passive philanthropy”—generating positive externalities automatically—has broad implications. We dream of ‘philanthriopizing’ other sectors of the economy as well (though I’m not sure we’ll stick wth the term philanthropize’.)
I should add that a good bit of my unease with non-DeFi use cases comes from the additional risk to which users “bringing” funds into crypto will be exposed. Even assuming Glo is 100% risk-free itself, there is still either exchange risk (FTX was not the only one to go under) or self-custody risk (e.g., losing one’s keys) plus risk from compromise of the user’s security environment (e.g., a trojan). For people who needed a stablecoin anyway, these risks are no different than for any other stablecoin; they have made the choice to take the risk for their own purposes. But these risks don’t really have any counterpart for someone whose counterfactual was keeping the money in a high-yield savings account at an insured bank. The FDIC will cover you if your bank goes under, and your bank will cover you in most instances of fraud.
So I am not convinced that putting your money in something like Glo is essentially as safe as putting it in the bank. I don’t think those risks are trivial, and I think there are concerns with exposing altruistic users to them—even if they are very clearly disclosed as meaningful risks. We do not, for instance, allow altruistic blood donors to expose themselves to anything but the mildest or most unlikely of risks. As the folks at 1DaySooner could likely attest, we do allow altruists to take somewhat greater risks when the social goal is important enough and there is very strong informed consent. In those situations, we usually also require a plan to provide care for the altruist if something goes south.
One of the pleasures of liaising with EAs on Glo’s behalf is how quickly the conversations advance to the “frontier” of issues that we are currently debating or have debated extensively. This is a good example.
Section 3 of the previously linked-to Glo EA post outlines the intended stages of Glo’s growth, starting with stablecoin use cases and then moving to savings accounts. I wrote that section and I meant it as a chronological roadmap, meaning the DeFi use case would come first in a sequence. As you observed, that nuance has been lost on the Glo homepage. We’re in the middle of redesigning the website to be more CTA (call to action)-focused in light of Glo’s impending release. If I had my druthers we would focus more on established stablecoin use cases and less on getting new people into crypto for the reasons you outlined. I didn’t feel that way three months ago—in fact I wrote a Twitter thread on getting started with crypto that we published like two weeks before FTX collapsed. Shows what I know.
I currently own about $30 of crypto. I am lucky to live in a country with trusted (trustworthy?) third parties, which, as you point out in another comment, is not something everyone can say. But it does mean that stablecoins solve no obvious problem for me. So once Glo is released, my plan is to buy up to $100 a month. I’ll try to spend on groceries and such, with the intent to not hold more than $100 at a time until I think the risk profile is totally negligible and the utility approaches that of regular dollars. If something like that level of commitment makes sense to you, great. If not, no worries.
As you point out, Glo’s risk profile is partly a function of things Glo (the company) does, but also of how/where Glo is available and how people interact with it. So a different way to approach your question is to ask what safety/risk benchmarks you’d have to see both from Glo and the crypto ecosystem as a whole to feel that the risk profiles of Glo and a “high-yield savings account at an insured bank” were broadly similar?
I/we have some speculative answers to this but so long as I’ve got you on the line, I’m curious to hear your thoughts.
For Glo: I don’t fully know what means are available to show how much one has in the bank and in Treasuries. I guess the ideal would be some sort of continuous real-time verification, possibly by a network of the world’s largest banks in a custodial role? I am assuming you could prove how much Glo had been issued and how much you held (e.g., from redemptions); from that, it should be possible to show near-full capitalization (perhaps with some lag during heavy redemptions).
I also don’t know enough about running a cryptocurrency to express what I’d need to know Glo was secure (e.g., that someone could not steal the metaphorical printing press and start printing new Glo without asset backing, or hack into a wallet holding redeemed or not-yet-issued Glo). I think some of that would involve proving that there was a technological solution whereby several independent authorities (e.g., your auditors, a partner at a mega law firm, whoever) would need to provide authorization to generate more Glo or to move significant quantities out of cold storage.
For exchange risk: Ideally, there would be something like the SIPC (which insurers customers of broker/dealers from losses in the custody function—not from losses in the value of the underlying securities). I do not think that is particularly realistic—SIPC is actually funded by a levy on broker/dealers, but the exchange market is too concentrated for that to be a viable model to cover losses. It has a credit line with Treasury, but not full faith and credit.
I don’t see a universe with the FDIC/NCUA model for a while—we would need long proof that crypto exchange regulation was working well before I think putting the USG’s full faith and credit behind an exchange insurer would be viable.And to be frank, at least in the mid-term future, crypto exchanges just aren’t socially necessary in the way banks and securities dealers are, and so justifying taxpayer backing would be a tough sell.
So the question is whether something like Japanese regulations would be sufficient. I’m not yet convinced that would be sufficient for the standard you set, although they seem to have protected Japanese users at FTX Japan. Maybe I would change my mind after a few more years and more failures that affect other countries but not Japan.
For consumer-equipment risk: I don’t think you can get to essentially-insured-bank safety for consumer users without either Glo or the exchange eating fraud losses that a bank would have to eat. Consumers on average are just not savvy enough to keep that risk at a minimal level—and although you and I could probably do so, it would come at a steep cost for useability.
That’s a tough issue to figure out, because an entity like Glo isn’t really in a position to verify or cover those sorts of losses. And for me to be convinced of their safety, exchanges need to play it safe and largely stick to providing safe custody and exchange services . . . nothing that exposes the exchange to meaningful risk. Unfortunately, that approach also limits how much money exchanges can make and thus how much they could be expected to spend on fraud reimbursements.
Note that at least US law gives much less protection to business users—if your business computer gets a trojan and makes unauthorized transactions on your account, you’re often screwed. So if you could fix the exchange-risk problem, this might be an acceptable problem for business users.
Also, current US law does not generally protect customers who are fraudulently induced into losses—e.g., you get a fake text from the IRS saying to Zelle $5000 to this account or you’re going to jail, and you comply. Since these are risks borne by users in the banking system, they can be borne by Glo users as well.
All that is to say that I think truly safe crypto is a ways off. Of course, I am willing to accept a somewhat higher level of risk for people holding crypto to accrue benefits to themselves than for those who hold Glo for altruistic purposes.
For the most part. For (2), I primarily meant stuff like KYC and anti-money laundering measures.
If you haven’t already, it’s worth thinking about the tax implications as well. My gut feeling under US law is that all of this interest might be taxable as unrelated business income (UBIT) because running a stablecoin might not be substantially related to the charitable, educational, or other purpose that would be the basis of the organization’s exemption. It would be a means of generating profit for the organization’s charitable mission, which is the point of the UBI tax. The point of the UBI tax is to prevent non-profits from having an unfair advantage over for-profit businesses in commercial activity, which seems to fit here.
Generally, interest is excluded from UBI, but that may only apply to interest from ordinary and routine investments. I can’t give you legal advice as to an answer even if I were qualified; I can only say that it is a question that crossed my mind.
I understand the use case for a stablecoin that needs to link with the crypto world—whatever one may think of crypto, if stablecoins exist, there’s a good argument that the organizer should lack a profit motive and the profits should go to something socially useful.
I am struggling to understand the other use cases (e.g., “savers” on the Glo website). For such users, this seems akin to a money market fund owned by a charity that pays 0% interest and instead donates the interest to charity. I am not sure what the added advantage having a crypto element adds to that basic idea (which is probably a good one, by the way). Right now, I suspect the answer is lower operating costs—but I think there is a high probability that much more extensive regulation is coming for crypto, and that operating costs of companies like Glo will begin to approach costs for banks and other financial institutions.
Indeed, this seems for non-crypto use cases potentially inferior to opening the “GiveDirectly Bank”—i.e., a US bank owned by a non-profit that donates its profits to GiveDirectly. Because Glo isn’t backstopped by the US FDIC (or equivalent), it has to purchase ultrasafe, highly liquid (thus: low-yield) assets to imperfectly simulate that backstop. In contrast, the GD Bank would have a freer hand in investing user capital for higher yield because (up to FDIC insurance limits, which can be creatively stretched to several million dollars per depositor) the risk of loss is ultimately borne by the US Government.
Hi Jason, thanks for the kind words!
Four separate points here: 1) the stablecoin use case makes sense to you; 2) operating expenses for Glo are low today but likely to be higher in the future after regulation; 3) why crypto versus any other way of implementing the core idea, which is turning some portion of a person’s “cash on hand” into an automatically-donating vehicle; and 4) in particular why not an EA-aligned bank (re: credit union) that just donates to GiveDirectly?
RE: 2: We’re gonna have to wait and see what specifics shake out. Our overall hope is that by registering as a non-profit, and establishing formal oversight and governance within the established category of 501(c)3, we’ll be compliant with whatever comes by default. But this is speculative.
RE: 3-4 We’ve had this idea and I even wrote up a version of it in a doc called “The case for an effective altruist credit union” (which I would be happy to share with you if you’d like to DM me an email address). The basic consensus is: stablecoin makes a lot of sense as a first product, because the DeFi market is already huge and, frankly, in need of some better infrastructure (ultimately, stablecoins are infrastructure), so that is a fine beachhead for the overall idea. Second, launching a bank that is broadly available across the world is going to be very, very challenging. Glo the stablecoin is available just about everywhere the moment it’s on exchanges.
But yes, the overall idea of “passive philanthropy”—generating positive externalities automatically—has broad implications. We dream of ‘philanthriopizing’ other sectors of the economy as well (though I’m not sure we’ll stick wth the term philanthropize’.)
Does that make sense?
I should add that a good bit of my unease with non-DeFi use cases comes from the additional risk to which users “bringing” funds into crypto will be exposed. Even assuming Glo is 100% risk-free itself, there is still either exchange risk (FTX was not the only one to go under) or self-custody risk (e.g., losing one’s keys) plus risk from compromise of the user’s security environment (e.g., a trojan). For people who needed a stablecoin anyway, these risks are no different than for any other stablecoin; they have made the choice to take the risk for their own purposes. But these risks don’t really have any counterpart for someone whose counterfactual was keeping the money in a high-yield savings account at an insured bank. The FDIC will cover you if your bank goes under, and your bank will cover you in most instances of fraud.
So I am not convinced that putting your money in something like Glo is essentially as safe as putting it in the bank. I don’t think those risks are trivial, and I think there are concerns with exposing altruistic users to them—even if they are very clearly disclosed as meaningful risks. We do not, for instance, allow altruistic blood donors to expose themselves to anything but the mildest or most unlikely of risks. As the folks at 1DaySooner could likely attest, we do allow altruists to take somewhat greater risks when the social goal is important enough and there is very strong informed consent. In those situations, we usually also require a plan to provide care for the altruist if something goes south.
One of the pleasures of liaising with EAs on Glo’s behalf is how quickly the conversations advance to the “frontier” of issues that we are currently debating or have debated extensively. This is a good example.
Section 3 of the previously linked-to Glo EA post outlines the intended stages of Glo’s growth, starting with stablecoin use cases and then moving to savings accounts. I wrote that section and I meant it as a chronological roadmap, meaning the DeFi use case would come first in a sequence. As you observed, that nuance has been lost on the Glo homepage. We’re in the middle of redesigning the website to be more CTA (call to action)-focused in light of Glo’s impending release. If I had my druthers we would focus more on established stablecoin use cases and less on getting new people into crypto for the reasons you outlined. I didn’t feel that way three months ago—in fact I wrote a Twitter thread on getting started with crypto that we published like two weeks before FTX collapsed. Shows what I know.
I currently own about $30 of crypto. I am lucky to live in a country with trusted (trustworthy?) third parties, which, as you point out in another comment, is not something everyone can say. But it does mean that stablecoins solve no obvious problem for me. So once Glo is released, my plan is to buy up to $100 a month. I’ll try to spend on groceries and such, with the intent to not hold more than $100 at a time until I think the risk profile is totally negligible and the utility approaches that of regular dollars. If something like that level of commitment makes sense to you, great. If not, no worries.
As you point out, Glo’s risk profile is partly a function of things Glo (the company) does, but also of how/where Glo is available and how people interact with it. So a different way to approach your question is to ask what safety/risk benchmarks you’d have to see both from Glo and the crypto ecosystem as a whole to feel that the risk profiles of Glo and a “high-yield savings account at an insured bank” were broadly similar?
I/we have some speculative answers to this but so long as I’ve got you on the line, I’m curious to hear your thoughts.
As always, thanks for engaging.
For Glo: I don’t fully know what means are available to show how much one has in the bank and in Treasuries. I guess the ideal would be some sort of continuous real-time verification, possibly by a network of the world’s largest banks in a custodial role? I am assuming you could prove how much Glo had been issued and how much you held (e.g., from redemptions); from that, it should be possible to show near-full capitalization (perhaps with some lag during heavy redemptions).
I also don’t know enough about running a cryptocurrency to express what I’d need to know Glo was secure (e.g., that someone could not steal the metaphorical printing press and start printing new Glo without asset backing, or hack into a wallet holding redeemed or not-yet-issued Glo). I think some of that would involve proving that there was a technological solution whereby several independent authorities (e.g., your auditors, a partner at a mega law firm, whoever) would need to provide authorization to generate more Glo or to move significant quantities out of cold storage.
For exchange risk: Ideally, there would be something like the SIPC (which insurers customers of broker/dealers from losses in the custody function—not from losses in the value of the underlying securities). I do not think that is particularly realistic—SIPC is actually funded by a levy on broker/dealers, but the exchange market is too concentrated for that to be a viable model to cover losses. It has a credit line with Treasury, but not full faith and credit.
I don’t see a universe with the FDIC/NCUA model for a while—we would need long proof that crypto exchange regulation was working well before I think putting the USG’s full faith and credit behind an exchange insurer would be viable.And to be frank, at least in the mid-term future, crypto exchanges just aren’t socially necessary in the way banks and securities dealers are, and so justifying taxpayer backing would be a tough sell.
So the question is whether something like Japanese regulations would be sufficient. I’m not yet convinced that would be sufficient for the standard you set, although they seem to have protected Japanese users at FTX Japan. Maybe I would change my mind after a few more years and more failures that affect other countries but not Japan.
For consumer-equipment risk: I don’t think you can get to essentially-insured-bank safety for consumer users without either Glo or the exchange eating fraud losses that a bank would have to eat. Consumers on average are just not savvy enough to keep that risk at a minimal level—and although you and I could probably do so, it would come at a steep cost for useability.
That’s a tough issue to figure out, because an entity like Glo isn’t really in a position to verify or cover those sorts of losses. And for me to be convinced of their safety, exchanges need to play it safe and largely stick to providing safe custody and exchange services . . . nothing that exposes the exchange to meaningful risk. Unfortunately, that approach also limits how much money exchanges can make and thus how much they could be expected to spend on fraud reimbursements.
Note that at least US law gives much less protection to business users—if your business computer gets a trojan and makes unauthorized transactions on your account, you’re often screwed. So if you could fix the exchange-risk problem, this might be an acceptable problem for business users.
Also, current US law does not generally protect customers who are fraudulently induced into losses—e.g., you get a fake text from the IRS saying to Zelle $5000 to this account or you’re going to jail, and you comply. Since these are risks borne by users in the banking system, they can be borne by Glo users as well.
All that is to say that I think truly safe crypto is a ways off. Of course, I am willing to accept a somewhat higher level of risk for people holding crypto to accrue benefits to themselves than for those who hold Glo for altruistic purposes.
For the most part. For (2), I primarily meant stuff like KYC and anti-money laundering measures.
If you haven’t already, it’s worth thinking about the tax implications as well. My gut feeling under US law is that all of this interest might be taxable as unrelated business income (UBIT) because running a stablecoin might not be substantially related to the charitable, educational, or other purpose that would be the basis of the organization’s exemption. It would be a means of generating profit for the organization’s charitable mission, which is the point of the UBI tax. The point of the UBI tax is to prevent non-profits from having an unfair advantage over for-profit businesses in commercial activity, which seems to fit here.
Generally, interest is excluded from UBI, but that may only apply to interest from ordinary and routine investments. I can’t give you legal advice as to an answer even if I were qualified; I can only say that it is a question that crossed my mind.
I’ll DM you an e-mail address.