SILK - Stablecoin Trillema & Operation Touch-&-Go (Security, Efficiency, Stability)

Greetings community,

Today I want to open an important topic that intimately impacts the future of SILK. In this post, I want to outline the trade-offs that stablecoins take, while pointing out various potential paths that SILK can take within a tri-attribute paradigm. The balance of ethos and pragmatism can be a slippery slope - and I want to acknowledge the risks that are absorbed by leaning too heavily one way or another during the lifespan of a stablecoin, as well as some general observations about existing stablecoin projects.

The Stablecoin Trilemma
The three key attributes of any stablecoin are as follows:

  • Stability
  • Efficiency
  • Security

Stability: the ability of a stablecoin to maintain its target peg.

Efficiency: the amount of capital required to mint or burn a stablecoin in relation to changes in demand for the stablecoin.

Security: the degree to which underlying collateral is decentralized - the more centralized the collateral, the greater the risk that the underlying collateral could be controlled in a way that negatively impacts the stability of the peg.

A stablecoin that is maximized for stability would have 1000000000x backing per stablecoin issued. Unfortunately, the amount of capital required to mint the extremely backed stablecoin is extremely capital inefficient.

A stablecoin maximized for efficiency would mint 1 stablecoin for $0 of value. This means the ability to expand the supply is not backed by an assets - allowing for rapid expansion of the supply in relation to changes in demand. This, of couse, violates the principle of stability as users would rapidly find that redeeming these stablecoins have not asset backing.

A stablecoin maximized for security would mint 1 stablecoin while demanding 1000000s of different micro-pieces of collateral tied to permisionless and truly decentralized assets. This violates components of efficiency (as expanding supply introduces collateral minting fraction) as well as violation of the stability principle (the more types of collateral included in the peg, the greater the chance one of them will fail).

edit: it was noted that the more collateral types involved, the more risk is hedged (which is correct). Observation made above was about likelihood of a single component of the collateral failing, not about the distribution of risk from said hedging. I’ll also note in this hypothetical example that including arbitrary #s of more collateral doesn’t automatically make things safer as each collateral has its own set of risks. It’s better to have 5 really really good pieces of collateral, than 5 really good pieces of collateral + 5 terrible pieces of collateral.

Naturally, stablecoins exist within a set of sliders between all three variables of the stablecoin trilemma. These sliders ultimately determining the fundamental attributes of any stablecoin, leaving it up to the market to determine what it values and demands.

I want to express that where SILK falls on the trilemma spectrum should be optimized to match long-term product-market demand, our collective ethos, as well as the need to bring SILK to said late game in relation to adoption & stability.

Before we proceed, let’s walkthrough some existing stablecoins:


USDC

Stability: strong - issuance is 1:1 backed by highly liquid financial instruments that can be converted easily to USD, which is what USDC is pegged to.

Efficiency: strong - users can deposit at a 1:1 ratio

Security: weak - only a single issuer is behind this that must comply with external actors, introducing risk to holders byh nature of the permissioned nature of USDC (accounts can be frozen)


DAI

Stability: strong - issuance is ~150% backed by collateral, albeit with risks tied to the liquidity of the underlying collateral. Part of the stability comes from the fact that nearly 60% of all assets backing DAI are centralized assets like USDC.

Efficiency: meh - users have to deposit at a 150% ratio in order to mint DAI, but some of this has been resolved by using other stables as accepted collateral, lowering the LTV required to mint DAI closer to a 1:1 ratio which is considered capital efficient.

Security: okay - roughly 60% of collateral could be impacted by centralied issuers, introducing risk to holders byh nature of the permissioned nature of USDC (accounts can be frozen). Its difficult to quantify the probability of this collateral being frozen as the market impact would be an almost immediate death spiral of DAI that would cause a massive amount of contagion on the entire domain. Some of my thoughts on the Tornado Cash / USDC situation.


FRAX

Stability: alright - issuance is ~90% backed by collateral, aiming to be a fractionalized reserve. This is the closest thing crypto has to a “safe” algorithmic stablecoin.

Efficiency: excellent - users have to deposit stables at a 90% stable + 10% FRAX-Shares rate (burned). Targets the optimal 1:1 mint ratio target.

Security: at risk - FRAX was really built on the back of LP tokens (in the form of centralized stables) + USDC + ETH. FRAX Dune Analytics. Interesting to note is the infographic below - this was at the very beginning of the FRAX journey in January of 2021 when FRAX was 90.75% collateralized by USDC. Note that this number has decreased overtime, pointing towards a trend towards more decentralization.


SILK
This of course, leaves us with SILK. Let’s start with some observations:

It is important to note that the market is demanding stability as the #1 attribute it prioritizes above anything else. SILK could make the decision to deviate away from market demand, but I want to acknowledge that there are adoption risks and potentially stability risks by having no interactions with centralized stablecoins in any capacity.

  • The Cosmos ecosystem has not chosen its primary decentralized stablecoins yet - it has however signalled demand for USDC (81% of stablecoin trading volume is coming from USDC on Osmosis)

There is a scenario here where SILK ignores any sort of centralized stablecoin inflows in the Cosmos. It’s important to note that some other stablecoin in the Cosmos will more than likely take heavy advantage of USDC as a proxy for stability within the Cosmos, which in turn can accelerate growth and adoption of said stablecoin at a faster rate than stables that don’t rely on centralized stablecoin inflows.

Some counter-points to this fear: SILK is already positioned as a non-dollar stablecoin, so by default we are already attracting a different audience that cares about the decentralization of the peg, privacy/sovereignty, and more than likely also the decentralization of the underlying collateral.

Counter-point to the above counter-point:

  • We don’t know the probability that USDC or any centralized stable will be a threat vector that gets used against SILK within the next Y years.

That is to say, if the ShadeDAO accepts USDC into its treasury for SILK minting in order to bootstrap growth, what are the odds we encounter a Tornado Cash scenario within 1 year? 2 years? 5 years? 10 years?


Operation Touch-&-Go
Within the balancing act of understanding the target market for SILK, the real competitive pressure to fill the TVL void in Cosmos, understanding the utility of stability proxies (such as USDC), while understanding our goal of having both a decentralized peg and decentralized collateral, I would propose the following strategy.

Y1 SILK

  • 40% of collateral minted against centralized stables
  • 60% of collateral against all other collateral types (decentralized stables, LP tokens, decentralized collateral such as $SCRT & $ATOM)

Y2 SILK (update fees to push towards this)

  • 30% of collateral minted against centralized stables
  • 70% of collateral against all other collateral types (decentralized stables, LP tokens, decentralized collateral)

Y3 SILK

  • 20% of collateral minted against centralized stables
  • 80% of collateral against all other collateral types (decentralized stables, LP tokens, decentralized collateral)

Y4 SILK

  • 10% of collateral minted against centralized stables
  • 90% of collateral against all other collateral types (decentralized stables, LP tokens, decentralized collateral)

Y5 SILK

  • 100% of collateral against all other collateral types (decentralized stables, LP tokens, decentralized collateral)

Operation Touch-&-Go gives SILK half a decade to achieve full decentralization of underlying collateral - giving time for various Treasury Operations & primitive revenue streams be able to partake in this long-term transition of quality of decentralized collateral. With this strategy, the importance of year-1 bootstrapping SILK adoption is acknowledged, with a clear long term plan to achieve indepedence from the risks of the collateral used.


Conclusion
I would love to hear others thoughts and opinions on this topic. My opinion is that SILK should optimize for growth & adoption in the short-term, with a lazer sharp plan to head towards decentralized collateral. Other projects such as FRAX & DAI never had a clear long-term plan for this from day 1, and as such were sucked into the growth-over-decentralization paradigm that have put them into the current situation of incurring centralized collateral risk.

I think we can balance the need for growth with the wisdom to clearly migrate away overtime - using governance and the community to hold ourselves accountability for the vision of SILK:

Private. Global. Stable.

Decentralized.

-Carter Woetzel (Shade Protocol Researcher)

7 Likes

I agree that we should prioritize adoption to an extent, especially early on. It is unfortunate that we will have to rely on centralized stablecoins as collateral, even in part, but it is a practical response to the reality that we are currently in, at least for the short term. Doubly so since we need to maintain fidelity to the peg that we are setting above all else if we want any kind of real adoption imo, whether that is a trading pair on a DEX or in payment platform usage.

I do want to point out that this will likely cause FUD in the form of “Shade Protocol isn’t decentralized, they are backed by USDC” type talk, so the team should be prepared for that from a marketing standpoint.

2 questions immediately arise on my end:

  1. There was talk post-Terra collapse but prior to this post about SILK eventually having some sort of algorithmic component to it down the line. Does this plan discount that as a planned action moving forward? Or is there still some plan around that possibility? If so, is there anything that you can elucidate on that at this time?

  2. What, if any, is the currently plan if USDC were to freeze the collateral before the 5 year rollout plan is full implemented?

Thanks for your work team, and appreciate the discussions as always.

2 Likes

kinda confused by this paragraph, why would adding collateral types make us more vulnerable to one collateralization piece failing? If the solution you described were easy for a user to do it seems like a good solution to me. maybe I’m missing something.

crypto etf for collateralization?

Great questions DC.

(1) The modularity to have SILK be algorithmic is still on the table → the idea is that SILK would head in the direction of FRAX where minting is backed by a decentralized or centralized stable + a burn of SHD. Additionally, its planned for the ShadeDAO to have access to SILK & SHD minting, empowering permissioned algorithmic operations that improve the stability & liquidity of SILK (more details on this in the works)
(2) The plan would be for the DAO to have a trigger looking for this event, and liquidate as large of a position as rapidly as possible. SHD governance would then have to vote to be part of re-couping the costs of the unaccounted for SILK using SHD as collateral or potentially a claim on future revenue streams. Those are some of the existing ideas. It would be a pretty terrible situation all around, but the best you can do is recoup & cover. The opportunity cost of incurring this risk is of course faster growth and adoption.

1 Like

So, the idea is that finding quality collateral is very difficult to do. You’d rather have 5 quality pieces of collateral, than 5 quality pieces of collateral + 95 shitcoins. If you picture quality of collateral as a hiearchy, the more pieces of collateral are added the further you diverge from the most quality kinds of collateral which then increases the risk of the newly added collateral as something that may fail.

The best state to target would be a wide-array of decentralized ETF collateral - but there are limits to how big that basket would be based on liquidity and decentralization.

2 Likes

This is something I have been thinking about since SILK shifted to an overcollateralized model.
Your analogy of the 3 sliders describing the attributes of stablecoins is great, and really helps me visualize the problem and potential solutions.

If USDC is the preferred stable currently, and its best attributes are stability and efficiency, I feel as though we can find somewhere in between the stability and efficiency of DAI/FRAX, even from the start. Most of the questions I have could be answered if you detailed what you think the ideal attributes we should prioritize for SILK most out of the gate?

A few follow up questions:

If SILK were overcollateralized to the extent DAI is, could you shift to a larger percentage of decentralized assets early on? You sacrifice efficiency for greater decentralization, while maintaining high level of Stability.

Could SILK slowly increase its efficiency (lowering collateralization rate) less than 100% based upon revenue generation by the protocol? (kind of like a subsidy) This questions makes me think of the game theory involved with increasing fees/lowering collateralization rate involved with minting SILK.

What degree/purity of decentralization should the protocol seek in its collateral backing? If we accept decentralized assets that are backed by centralized assets, are we REALLY decentralized? Obviously there are only so many “credible” assets that can be selected across all blockchains, and this list becomes drastically smaller if we only consider non-bridged assets, so this question becomes difficult to confront from a purist perspective.

Addendum: The beauty of solving for number and type of decentralized assets can be achieved through governance, over time, as we take note of evolving market and regulatory environments.

2 Likes

Glad to see there is a plan. Fail to plan…you get the point. I think 5 years plans are an excellent way to map out a strategy. Reasoning is clear and concise. Day traders and institutional types I’m sure both understand the logic here. I’m excited to see where the road to transactional privacy leads to in the next 5 years.

2 Likes

I would like to add to this conversation that on the stability/efficiency front one way to increase stability, and increase efficiency is asset liability matching. This seems a bit hard to implement at the moment given the lack of options on blockchain, but something we should keep in mind.

For instance if you were making a stablecoin that was 10% Gold, 10% Bitcoin, 50% USD, 30% EUR you could hold 10% PAXG, 10% BTC, 50% USDC and 30% EEUR. This would mean that you would only need to put 1 Silk of Value (if you put those exact assets in) to mint 1 Silk since it’s perfectly matched. The underlying collateral would directly track the movements of SILK. I know that isn’t fully possible at the moment, but it’s something that we should keep in mind as we design the peg - can we prioritize secure, decentralized assets into our peg that would allow the basket to reflect SILK.

2 Likes

using the peg stability research for the backing as well! very efficient!

1 Like

One thing I want to add is just that at the end of the day, the only value Silk has is in its utility. In the absence of utility, none of the above matters. The reason I want to emphasize that is because stability mechanisms aren’t a silver bullet and are just meant to create mechanisms for the system to respond to demand changes. Efficiency to me is not really a concern because in the event that silk supply is below demand to the point that the price is above peg, borrowing and reserve minting is literally free money, which is why you never see CDP coins go above peg for any significant period of time. Below peg prices are the result of lack of utility and excess supply, and there’s no secret trick to address that without giving up growth other than making the token more useful. That leaves security, although i don’t necessarily agree that decentralization is synonymous with security. Ultimately security is just one aspect of stability, and i don’t think there’s any strong proof that decentralized stablecoins are extremely secure and there’s plenty of evidence that the opposite is true - not to suggest centralization is the answer, just that we shouldn’t define security by decentralization because they’re not the same thing.

tl;dr the only thing we need to focus on is creating utility for silk and ensuring that the backing is strong, the system as designed has all the tools necessary to maintain a healthy peg as long as silk is actually useful.

3 Likes

The ideal attributes for SILK out of the gate should be heavily slated towards stability + capital efficiency with a migration towards security overtime. If we fail to maintain the promise of stability or make it too difficult for SILK to be introduced into a market that demands it, than we will be in a tough spot. End game attributes of SILK:

  • Private
  • Sovereignty
  • Permissionless
  • Stable
  • Decentralized Peg
  • Decentralized Collateral
  • Flexible peg
  • Governable
  • Globally accessible
  • Censorship resistant
  • Fast

Best way for SILK to increase its efficiency is by having enough stable demand for SILK + arb mechanics available to be able to reliably handle changes in the price of SILK without needing to tap deeply into the capital backing of the peg. This is in essence how FRAX handles it - if there is ever a deep tapping into capital backing, FRAX shares are inflated (walking back the amount that was originally burned) to incentivize the arb. The difference is because FRAX is much much safer with liability issuance, a “walkback” is not protocol breaking, and the unwind effect is much more guarded and planned out with the fee curve to prevent bankruns. I digress. To your point, the more revenue that goes to Shade Protocol, the more flexibility we have with issuance because SHD is essentially a claim on protocol revenue - the more powerful SHD is, the stronger the backing for SILK is with respect to the dynamic growth mechanisms we introduce (like Bounded Protocol Minting & FRAX model minting).

I think the best answer is to migrate away from the most centralized collateral using a score card similar to this:


We could also use something like the Nakamoto Coefficient which roughly measures decentralization.

1 Like

birds eye view, really nice addition, again. I think we can agree that any plan we make will need to adapt continually and It seems that the tools are built out to enable that. onto utility!

Completely agree with this statement, just wanted to add that it would be interesting if you could include a Beta coefficient as part of the peg. @Ranger_Ranger is basically stating something similar but I wanted to explicitly state that if you had 100 pieces of collateral with a Beta coefficient of 1 that is worse than a select set of collateral with varying degrees of a Beta coefficient. Otherwise your backing will be directly proportional to market changes.

This is very well said, while it is vital to think about the stablecoin trilemma as presented wonderfully by @CarterWoetzel the core reason the stable coins listed are being talked about at all is due to the fact that they have been willing to go where the demand goes. USDC can be directly exchanged on many exchanges in a 1:1 swap with USD deposited onto the platform. DAI is pretty much the grandfather of decentralized currencies and started as purely decentralized vaults backed by ETH and pivoted to more centralization risk by accepting more and more types of collateral. Additionally both DAI and USDC are included in pretty much every DeFi and NFT application. There is a reason OpenSea has only these two stable coins as an option to pay with.

I would be interested in looking at LUSD as a “pure” decentralized play to add to this list since it is similar to the first DAI model of ETH CDP backing. It would probably score a bit better than DAI since there is no USDC risk and the efficiency coefficient is ~120%. But the reason it is not mentioned much is because it doesn’t have sufficient utility. So focus on making SILK a cornerstone of Secret/Cosmos/Crypto applications is critical.

Another potential centralized play is BUSD based on the chart Carter shared from Dune, if it captures 18% of the market that should at least be considered.

Depends on adoption and what it is used for, if it is wildly successful then the likelihood increases exponentially, if it stays small then the risk is minute.

Are these percentages of SILK minted and not SILK circulating? I am curious how the treasury operations will be transparent (or algorithmic) to show that the risks above are being pushed out of the system or hedged?

Additionally, it seems like the year 1 expectation should at a minimum be flipped with more aggressive burn down each year OR instead of year-by-year it is a metric of adoption and utilization. As adoption increases it would trigger a change in strategy to become more decentralized. It is vital early on to find product market fit and until that is known having restrictions like this in year 1 may hamper the ability to pivot as quickly as is necessary.

1 Like