Thank you for bringing to light an idea that has utility in the DeFi space!
I traded hybrid bank products for a living for many years and now that I am getting into cryptocurrency and learning as much as I can about DeFi, I can see a lot of potential in fixed income and hybrid products in this ecosystem. It's good to see others are interested too.
Late comment to his post, but just wanted to say this was a really cool idea, especially the debt vaults - would love to see this pave the way for more protocols borrowing against future revenue streams.
One quick note: Do you think deposit insurance might cause a fundamental issue with the comparison between capital structures in this post? As you've pointed out, all deposits at most of these banks are backed by the government, so governments get to dictate the amount of risk banks can take. However, IMO these risk numbers (even though they got a lot better post financial crisis stress-testing) don't really reflect a "market pricing" of risk taken on by banks when giving out credit; mostly just numbers that governments have decided reflect safety.
"For this reason depositors would prefer the largest equity buffer possible" -> the crux of my point is that this is actually very relevant in DeFi (especially post Terra) but pretty much irrelevant to retail bank depositors because of government deposit insurance. In my view, banks just prioritize the interests of equity-holders as much as permitted under current regulations (depositors care about things other than risk in my experience) -> If Maker is able to execute on this plan, it would be really interesting to see the kind of "buffer" that the market wants, and how Maker actually balances the interests of depositors and tokenholders. I'd predict that eventually, depositors might expect a higher buffer than is given by current retail banks (assuming the same level of risk on the lending side).
Thank you for your comment Sanat, I do not necessarily agree entirely. Will try to bring more elements to the debate, hoping to be helpful.
When we refer to depositors we refer to liquidity providers to banks. Those do not include only (the FDIC insured portion of) retail depositors, but also wholesale depositors and bond holders. Given that retail depositors are a senior form of lending to the bank, this category includes also senior bond holders.
It is therefore easy today to look at the rate the market is asking to banks to provide funding, at various points of the term and risk structure. I know it's complex, but banking IS complex and extremely developed. A bank has several types, or tiers, of bonds floating, and there is a lot of research on the impact on the funding rate of both the equity buffer AND the liability profile. As you might remember, during the financial crisis several banks were borrowing short (on the money market) and lending long (on mortgages) and this was fine until it wasn't fine anymore. Current DeFi systems do not have a sophisticated analysis of the duration profile of neither their assets nor their liabilities, but I am assuming this will happen as use cases will become more developed.
On the equity cushion for retail banks, it is actually massive. So high that banking has become a very low return on equity business, and banks have ultimately ended up doing mainly non productive lending, which is typically real estate (hugely collateralised and not necessarily the most productive or innovative investment) or even worse government bonds (because...surprise surprise...sovereign regulators assign a 0% risk-weighting to those activities that therefore do not absorb capital). For this reason banks became a kind carry trading conduits that kept the sovereign markets liquid but very much useless to distribute risk capital in the economy.
On your final point, I have no idea how we will end up on risk buffer. But my guess is that probably should be lower than tradfi, because DeFi has better visibility, better enforcement of collateral, better management of payments, and more liquid markets to price assets on balance. But unobservable risk has not disappeared, so we will need a buffer. Trustless doesn't mean we can blindly trust anyone, it means that we can for some stuff operate without the need of struss. But not for everything.
Sorry for the long reply but your question really stimulated me. I am sorry if I wasn't entirely helpful in this, and thank you so much for your comments and contribution.
Really appreciate the thoughtful response - I was definitely focusing too much on the retail deposit side, and will admit that I'm nowhere as familiar with the details of the bond and wholesale deposit side of banking capital raises. Generally super excited to see Maker move in this direction!
When you talk about straight token capital are you refering to MKR (an example) or this applies to any governance token that required this type of capital funding?
great stuff, as always. Question: Couldn't Maker operate with zero capital and just sell MKR when it needed to? Like an over-collateralized version of Terra?
That’s the current ‘burn’ model, but few points: 1) why being in a position to issue shares in the worst possible time with further dilution, and 2) why keep buying them back instead when profits are flowing, things are good and tokens are expansive, and 3) why instead nor reuse funds on internal growth given the startup nature of the protocol? Thanks so much for the kind words.
Thank you for bringing to light an idea that has utility in the DeFi space!
I traded hybrid bank products for a living for many years and now that I am getting into cryptocurrency and learning as much as I can about DeFi, I can see a lot of potential in fixed income and hybrid products in this ecosystem. It's good to see others are interested too.
Late comment to his post, but just wanted to say this was a really cool idea, especially the debt vaults - would love to see this pave the way for more protocols borrowing against future revenue streams.
One quick note: Do you think deposit insurance might cause a fundamental issue with the comparison between capital structures in this post? As you've pointed out, all deposits at most of these banks are backed by the government, so governments get to dictate the amount of risk banks can take. However, IMO these risk numbers (even though they got a lot better post financial crisis stress-testing) don't really reflect a "market pricing" of risk taken on by banks when giving out credit; mostly just numbers that governments have decided reflect safety.
"For this reason depositors would prefer the largest equity buffer possible" -> the crux of my point is that this is actually very relevant in DeFi (especially post Terra) but pretty much irrelevant to retail bank depositors because of government deposit insurance. In my view, banks just prioritize the interests of equity-holders as much as permitted under current regulations (depositors care about things other than risk in my experience) -> If Maker is able to execute on this plan, it would be really interesting to see the kind of "buffer" that the market wants, and how Maker actually balances the interests of depositors and tokenholders. I'd predict that eventually, depositors might expect a higher buffer than is given by current retail banks (assuming the same level of risk on the lending side).
Thank you for your comment Sanat, I do not necessarily agree entirely. Will try to bring more elements to the debate, hoping to be helpful.
When we refer to depositors we refer to liquidity providers to banks. Those do not include only (the FDIC insured portion of) retail depositors, but also wholesale depositors and bond holders. Given that retail depositors are a senior form of lending to the bank, this category includes also senior bond holders.
It is therefore easy today to look at the rate the market is asking to banks to provide funding, at various points of the term and risk structure. I know it's complex, but banking IS complex and extremely developed. A bank has several types, or tiers, of bonds floating, and there is a lot of research on the impact on the funding rate of both the equity buffer AND the liability profile. As you might remember, during the financial crisis several banks were borrowing short (on the money market) and lending long (on mortgages) and this was fine until it wasn't fine anymore. Current DeFi systems do not have a sophisticated analysis of the duration profile of neither their assets nor their liabilities, but I am assuming this will happen as use cases will become more developed.
On the equity cushion for retail banks, it is actually massive. So high that banking has become a very low return on equity business, and banks have ultimately ended up doing mainly non productive lending, which is typically real estate (hugely collateralised and not necessarily the most productive or innovative investment) or even worse government bonds (because...surprise surprise...sovereign regulators assign a 0% risk-weighting to those activities that therefore do not absorb capital). For this reason banks became a kind carry trading conduits that kept the sovereign markets liquid but very much useless to distribute risk capital in the economy.
On your final point, I have no idea how we will end up on risk buffer. But my guess is that probably should be lower than tradfi, because DeFi has better visibility, better enforcement of collateral, better management of payments, and more liquid markets to price assets on balance. But unobservable risk has not disappeared, so we will need a buffer. Trustless doesn't mean we can blindly trust anyone, it means that we can for some stuff operate without the need of struss. But not for everything.
Sorry for the long reply but your question really stimulated me. I am sorry if I wasn't entirely helpful in this, and thank you so much for your comments and contribution.
Really appreciate the thoughtful response - I was definitely focusing too much on the retail deposit side, and will admit that I'm nowhere as familiar with the details of the bond and wholesale deposit side of banking capital raises. Generally super excited to see Maker move in this direction!
sir may i ask what software you use to create these "handdrawn" charts? thank you!
Excalidraw: simple, amazing, and free.
thanks so much!
When you talk about straight token capital are you refering to MKR (an example) or this applies to any governance token that required this type of capital funding?
I’m referring to issue tokens of some sort (with governance right most probably) in exchange for cash at the protocol level.
Got it. Thanks! Great article.
great stuff, as always. Question: Couldn't Maker operate with zero capital and just sell MKR when it needed to? Like an over-collateralized version of Terra?
That’s the current ‘burn’ model, but few points: 1) why being in a position to issue shares in the worst possible time with further dilution, and 2) why keep buying them back instead when profits are flowing, things are good and tokens are expansive, and 3) why instead nor reuse funds on internal growth given the startup nature of the protocol? Thanks so much for the kind words.
makes sense .... thanks!