USD Tez — Collateral² (sq) Model

USD Tez (USDtz) stable-coin 4:1 [over]collateralization Model

Kevin Mehrabi
Oct 28, 2019 · 17 min read

This article

  1. Background
  2. ‘The 4-Silos’ of the USD Tez Reserve
  3. Surplus Collateral (‘The Harvest’)
  4. Filling the Reserve: from 1:1 to 2:1 to 4:1


There are 4 parts to a stablecoin’s design: Collateral, Mechanism, Price Info, Peg. This article focuses on the Collateral model. Basically, The USD Tez Reserve (which harbors the collateral) is composed of 4-silos of collateral (2 are FIAT, and 2 are XTZ), which (when full) comprise a 4:1 backing of all USDtz in existence. The first FIAT silo and first XTZ silo comprise the “Base Layer” ensuring redeemability either form (FIAT or XTZ). The second layer (2nd FIAT silo and 2nd XTZ silo) is the Growth Layer, which, effectively collateralizes the Base Layer. Whatsoever exceeds the 4:1 collateral maximum is surplus called the ‘Harvest’.

The Harvest is distributed into 3 main areas: dividends, recapitalization (scaling the reserve), and ecosystem projects. Harvest distribution plans are decided upon by the USD Tez ‘Trustees’, which will eventually become a token-based on-chain DAO, reflecting Tezos’ own on-chain-governance. Trustees also have the power to amend/evolve the USD Tez system over time. The reserve is grown through baking and banking-like activities, which can be explored in more detail in our other articles.

1. Background

Why does a collateralization model matter for a stablecoin?

What is a ‘stablecoin collateralization model’?

How a stablecoin is collateralized defines the structure of the stablecoin and consequently its total value both as a measurable store of value (which should be stable at a 1–1 parity with the USD), as well as value in a broader sense — that is, for users, backers, merchants, and anyone affected by the coin. From stability to liquidity to governance — both in the short-term and in the long-term — the collateralization model demonstrates the backbone and the spirit of the stablecoin.

The diagram below (Cornell) taxonomically presents an influential context for the role of the collateralization model in the overall design of a stablecoin. Effectively, the collateral model defines, determines, or is dependent on every other aspect of a stablecoin’s design.

Source: “A Classification Framework for Stablecoin Designs”

In this article, I’ll explain the “Collateral” (type and amount) portion — which will imply aspects of “Mechanism,” “Price Info,” on which I will expound in subsequent articles (as necessary). “Peg” is obviously the United States Dollar (USD).


  • Collateral
    Type: FIAT and XTZ Combination
    Amount: Overcollateralized 4:1 (2x FIAT + 2x XTZ); ‘Collateral-squared’
  • Mechanism
    Reserve of Backed Asset
  • Price Info
    Oracles (source: external + distributed on-chain from DEXs)
  • Peg
    FIAT (USD-value)

(Also in this article, I will be using the term “Trustees” to describe those charged with managing the USD Tez reserve. Who the “Trustees” will be will change over time — at first it will be the USD Tez Foundation, and then it will be the DAO which will ultimately be a decentralized on-chain community.)

Values: What matters in stablecoin’s collateralization model?

Designing USD Tez, we’re ascertaining specific mission goals; a stablecoin in its foundation must reconcile the needs of each and every party interacting with the stablecoin (from backer to user to developer).USD Tez’s Values dictate that a stablecoin must always:

  1. Be Redeemable (to the peg) — To maintain 1–1 (USDtz–USD) stability 100% of the tokens in circulation must be backed and redeemable for all users, for their USD-value and under all market conditions
  2. Meet Demand — the amount of liquidity provided by the stablecoin’s minted supply must continue to meet market demand as the serviceable market grows
  3. Balance Incentives — align investor/trustee/DAO financial incentives to that which leads to the betterment of stablecoin and to the public good which it serves, so that what’s good for the coin is good for profits, and what’s good for profits is what’s good for the coin–inseparably

Regarding Balance Incentives, what we don’t want is for the stablecoin to start strong and then degenerate into a workaround hedge fund for the backers who would parasitically strain users and the reserve for their own financial gain and nothing else. Rather, the financial success of the backers should be dependent and proportional to the strength of the service provided.

In essence, the strength of a collateralization model’s long-term prosperity outlook is all about: maintaining balance in all things.

Collateral Types

The simplest way to collateralize a stablecoins (and our first step) is a dollar-for-dollar approach:

FIAT-collateral: The logic in using FIAT-collateral can be best summarized as, You give me a FIAT $1 dollar, then I give you a $1-token stablecoin of the same value, and I hold on to your dollar. If you give me back the stablecoin, I’ll redeem it by giving you back the dollar.” That way no matter what happens to the price of crypto (any crypto), the stablecoin will always 100% collateralized. Sadly, there are too many drawbacks to this method of collateralization alone. For one thing, by keeping FIAT-only collateral, the stablecoin cannot grow with the market, whereas crypto-collateral has the potential for inestimable growth (but also for inestimable loss).

Crypto-collateral: Collateralizing the stablecoin in crypto-collateral (XTZ) alone would enable us to grow the USD Tez reserve with the growth of the market. For example, (assume XTZ is priced at $1) if we sell 100 USDtz for100 XTZ and the price of XTZ grows to $10, then suddenly we are 10x (1000%) collateralized overnight. However, if the price of XTZ sinks to $0.90, then we are under collateralized and cannot redeem the purchasers, so the stablecoin would destabilize becoming worth $.90/piece.

This underlies the quintessential challenge that the designers of stablecoin collateralization models must delineate for their own respective purposes.

Our purposes are defined by the values stated above. By starting our reserve with FIAT-collateral accepted in the initial transactions, we avoid the possibility of under-collateralization ever no matter what happens to the XTZ market; USD Tez will always be redeemable for its USD equivalent. This is the safest, smartest, strongest way for any stablecoin to begin collateralization, even if they are to move off of FIAT later and embrace a completely algorithmically-maintained crypto-pegged approach.

Having said that, FIAT collateral is not enough, and neither is ‘double collateralization’ model of FIAT and Crypto for a total of a mere 200% backing (as stated in the Version 1.0 article). While a 2:1 over-collateralization ratio sounds nice, in practice it’s not enough. This article supersedes the previous USD Tez collateralization article.

2. ‘The 4-Silos’ of the USD Tez Reserve

Rather, if the 2:1 over-collateralization sounds nice but isn’t sufficient, a more robust system is needed—one that will assure all goals are met, despite extreme market scenarios. This system can be defined by a method we’re calling the “4-silos” of the reserve.

  1. Silo-1: USD Base Reserve (Base FIAT)
  2. Silo-2: XTZ Base Reserve (Base Crypto)
  3. Silo-3: USD Growth Reserve (Growth FIAT)
  4. Silo-4: XTZ Growth Reserve (Growth Crypto)

Silos 1 and 2 comprise the ‘Base’ layer previously discussed as ‘double collateralization, while Silos 3 and 4 comprise the ‘growth layer’ which protects the reserve from the aforementioned pitfalls.

Is there a Silo-0?

Technically, the supply of USD Tez, both in-and-out of circulation, would comprise ‘Silo-0’, but that’s more obvious so we’ll keep that out of sight going forward. For rhetorical purposes, Silo-0 isn’t quite part of the reserve, as the reserve and its respective 4 silos are the things that collateralize Silo-0. So when referring to “the Silos” we’re referring to the Reserve Silos (Silos 1, 2, 3, and 4). Capiche?

The relationship between the USD Tez supply and Silo-1 — Base FIAT

The reserve silos (1, 2, 3, and 4) are such that each individual silo alone is capable of backing—100% of the minted supply of USD Tez. This means that when all 4-silos are at full-capacity the minted supply of USD Tez will be backed 400% (i.e. a 4:1 ratio).

Silo-1 — “Base FIAT”

The first collateralization method follows the simple model described above; $1 USDtz for $1 USD. FIAT collateralization enables 100% redeemability (unlike crypto) because no matter when happens to the price of XTZ, the price of $1 FIAT will always remain $1.


  • The FIAT serves as a safety fallback so that no matter what happens to the price of crypto, the coin will always be stable and fully redeemable.
  • The price of XTZ could drop 99%, and still, all USDtz issued will be 100% redeemable; the price of XTZ could grow 100x, and still all USDtz issued will be 100% redeemable

This is the easiest, fastest, and the most rock-solid way to begin collateralizing a stablecoin. We could just stop there and have a 100% backed stablecoin, growing with the growth of the reserves. However, if the price of XTZ grows say from $1 to $10 and the trade volume would grow with it, and the serviceable liquidity that USD Tez could provide would be very little compared to the demand. Effectively, a FIAT collateral reserve alone for USD Tez is like shorting XTZ (not a good idea in the long run).

Silo-2 — “Base XTZ”

The base collateral needs to be in FIAT as well as XTZ—double-collateralizing the stablecoin. Silo-2 (the XTZ silo) starts empty, but with the help of lending activities from Silo-1, it can start to fill. Silo-2’s baking activities will commence as soon as some seed XTZ is deposited. Eventually, Silo-2 will be independent and be able to meet its own growth needs strictly from baking activities.

Imagine just 2 silos— Silo 1 and Silo 2. The value of each bucket must independently match the value of all minted USDtz in circulation. In other words, for every $1 USDtz-token issued, there must be $1 USD in a FIAT bucket as well as (assume the price of XTZ is $1) 1 XTZ coin in a crypto bucket; 1–1–1.

Why XTZ?

  • The XTZ reserve silo enables the stablecoin’s collateral to grow with the market since keeping just FIAT collateral while XTZ grows perhaps 10x would not enable USDtz to keep up with liquidity demands.
  • Secondary reason includes keeping the collateral “on-chain” so that we can stake the XTZ held and thereby contribute to the ecosystem on which USD Tez thrives

Together Silos 1 and 2 comprise the Base Layer of the USD Tez reserve. Although we’re starting with Silo-1 alone, and although Silo-1 alone is sufficient, upon the point in which Silo-2 is definitively filled, neither Silo-1 nor Silo-2 should be considered better than the other and at that point, the reserves should never fall under 200% collateralization.

This complete-hedge enables actually enables us to maintain the full-confidence either polarity of the market confidence spectrum:

“I don’t care for crypto-backing. Are you fully-collateralized by FIAT?”

Yes, 100%


“I don’t care for FIAT-backing. Are you fully collateralized by Crypto?”

Yes, 100

Though this is still not the perfect system.

Why 200%+ [over]collateralization is not-good-enough: As the XTZ silo (Silo-2) grows in value beyond the value of the FIAT silo (Silo-1), the USD Tez Foundation can mint more USDtz tokens to meet the new—greater—value of the XTZ in Silo-2). However, there cannot be more USDtz than there is FIAT to back it. This means that some XTZ would have to be liquidated to feed Silo-1 if it were to mint more USDtz.

Despite actually ‘double-collateralizing’ all USDtz by a combined 200% collateralization ratio, these 2 silos alone will hinder growth prospects. In other words, if our goal with Silo 2 is to grow our serviceable liquidity offering over time, then the double-collateralization method (namely, the servicing Silo 1) would hurt that effort.

Even if XTZ were to grow again and again over time, the XTZ silo’s size as a proportion to the overall XTZ market would otherwise be virtually cut in half with every new minting period.

A similar problem emerges if the price of XTZ were to crash. Then part of Silo 1 would need to be used to purchase more XTZ. Since FIAT is the first Silo, the actual supply of USDtz would need to be reduced in order to even make that purchase so that the token isn’t under-collateralized (first rule).

These are the reasons for Silo-3 and Silo-4, which together comprise the ‘Growth Layer’ of the USD Tez reserve.

Silo-3 — ‘Growth FIAT’

Silo-3 is a FIAT reserve designed to match Silo-1 and act as its backup, incase Silo-2 needs replenishing.

If the price of XTZ crashes, then the difference will need to be made up for by a FIAT reserve to purchase more XTZ. We cannot use FIAT funds from Silo-1 without ending up under-collateralizing the USD Tez supply unless we were to slash the amount of USD Tez, which is something to avoid. After all, our goal is to grow our liquidity offering, not shrink it. In a sense, if Silo-2 is the base that collateralizes USD Tez (just as Silo-1 does), then Silo-3 is what provides cover for Silo-2.

It is critical that Silo-1 and Silo-2 become full as soon as possible and to remain that way in perpetuity, whereas, Silo-3 can be more flexible. After all, the purpose of Silo-3 is to be an active reserve that can be used to replenish Silo-2. It’s a “growth” reserve which will have times where it’s at capacity and times in which it is less than capacity. Ideally, it need not break from full-capacity at all (how that would work will be described later) but it’s not a cause for alarm if it does.

For Silo-3’s reserves to go down from 100% to a fraction of that if need be, should just be seen as Silo-3 doing its job, and that it will rise up to 100% over time with further growth activities.

However, this only covers the balance if the price of XTZ were to crash. If the price of XTZ were to skyrocket, say 10x, then Silo-3 would not be very helpful.

For that reason, there’s Silo-4

Silo-4 — ‘Growth Crypto’

In a state of XTZ growth, the dollar value of Silo-2 will have outgrown the value of all FIAT in Silo-1 and of course, all USDtz in circulation. At the same time, it will likely be of great interest to all stakeholders that the occasion is used to increase the liquid supply of USDtz available.

To validly increase the USD Tez supply after an XTZ growth spurt, at the very least, the amount of FIAT in Silo-1 must be increased making Silo-1’s liquid dollar value match that of Silo-2’s liquid dollar value. This cannot be done by liquidating any XTZ since that practice would lead to a halving of its proportion to the overall market supply of XTZ in the market (very bad in the long run).

Instead, Silo-4 will provide the solution. Silo-4’s reserve is designed to match that of Silo-2 to act as a backup for Silo-1. If Silo-1 collateralizes all USDtz in circulation (as does Silo-2), then Silo-4 is a source of coverage for Silo-1.

Just like 2-keys needing to be turned to launch missiles from a submarine, so too should Silo-1 and Silo-2 be of equal size before new minting of USDtz tokens is to occur. However, if the Trustees decide that USD Tez’s liquidity offering must grow at full speed, then they can decide to spend all available funds to recapitalize Silo-1 alone. This is because Silo-1 backing all USD Tez in circulation is the absolute minimum requirement.

Together at full-capacity, all 4 silos will be collateralizing USDtz 400%. This is the maximum reserve ratio allowable. After that, all other funds earned from baking and lending activities enter the surplus chamber, which we call the “Harvest.”

III. Surplus Collateral — ‘The Harvest’

At the end of each designated fiscal cycle — any growth earnings from the reserve management beyond 400% collateralization of the USDtz supply (i.e. when the 4 silos are already filled and we have extra earnings left-over) will be considered a surplus, which we call “the Harvest”

A surplus is what trustees intend on achieving. Without a surplus, the trustees cannot collect a profit (issued as a dividend through a smart-contract).

The trustees can decide to divvy the harvest in 3 main ways:

Trustees” refer to the USD Tez Foundation at first, until the USD Tez Foundation revolves this and other its other responsibilities to the on-chain DAO
  1. Mint more USDtz-tokens — equally distributing the surplus amongst the 4 different silos, maintaining equilibrium and growing the reserve size overall
  2. Transact a dividend payment — issuing the surplus as a reward to the Trustees as early backers of the project
  3. Earmark ecosystem projects — research and development should be conducted in a manner that does not disrupt the 4-silos of the Reactor Core, and thus should come from the surplus

If the Trustees are the ones to decide when to take a dividend for themselves, and when to grow USDtz-supply and thereby serve the USD Tez ecosystem, wouldn’t they just ultimately act out of greed and take a dividend always?

USD Tez has a trade-off measure installed that would force (an otherwise parasitic) greedy person to reconcile the broad-ecosystem interests before acting in a parasitic manner. This is the “400% rule.”

The 400% rule (or 4:1-ratio rule) is what balances incentives for the USD Tez trustees (Foundation and DAO)

Balancing the Trade-offs — the 400% rule in action

Here are some case examples which will illustrate the psycho-social economic games at work. These decisions would follow the governance logic of the Tezos itself as a model.

Assume the most extreme possibility: If the Trustees use 100% of the surplus to take a dividend and 0% for recapitalization (and 0% for ecosystem development), they will benefit financially at the moment, but the cost of that action is that they will preclude themselves from greater earnings in the future. The dividend they take is ultimately a function of the size of the reserve as a whole since it’s growth earnings from lending and baking that are producing the surplus.

Alternative: If the trustees who are tempted to take a dividend payment instead decide to use the surplus to entirely recapitalize the Reserve, enabling expansion of the USD Tez supply, they are opening the door for greater gains in the next fiscal cycle since the larger reserve would effectively mean a larger ‘principal investment’ size for future surpluses. Furthermore, to keep USD Tez competitive, the liquidity supply must be robust enough to meet demand, putting a further incentive on the Trustees to reinvest the surplus in areas outside of their own dividend payments.

‘Bob’ Example’— Bob is an individual USD Tez Trustee, about to vote on surplus distribution. The growth of XTZ was especially great that period and the surplus is 100% above the value of the full-reserve. If the full amount of the surplus is issued as a dividend, Bob alone would take a $1,000,000 dividend for that period.

Alternative— Bob considers that if that surplus amount were to recapitalize the reserve, perhaps by pouring all of it into Silo-1, then the amount of minted and issued USD Tez could potentially double. With an expanded market from the growth of XTZ’s price, this may be a critical juncture in which to take such a step. Not to mention, expanding the pot to play with, and taking a dividend next time around can leave Bob with a dividend of $2,000,000 on the flip side — something much less likely to occur if the surplus is accepted as a dividend alone right now.

IV. Filling the Reserve — from 1:1 to 2:1 to 4:1

The way in which USD Tez’s Reserve grows to meet full-capacity (400% collateralization over the amount of USD Tez minted and in circulation) is through a combination of baking, lending, and banking activities. The Trustees are incentivized then to achieve 400% collateralization as fast as possible.

Filling the Base Layer — Silos 1 and 2

Automatically, Silo-1 will always be full from the start, as USD Tez may not be minted without an equivalent USD value in Silo-1. Silo-1 is not allowed to lose value relative to the amount of USD Tez in circulation.

The first goal then is to fill Silo 2, which will be a combination of banking and lending activities. We can fill Silo 1 with another friendly-neighbor stablecoin; USD Coin, and accrue interest on it from our banking partners. The interest accrued will be exchanged for Tezos XTZ and set in Silo-2.

The Tezos XTZ in Silo-2 will then be baked using an equitable syndicate made from technology.

After Silo-2 is full, it should not be left to contain a value amount that is of lower value to Silo 1. Therefore, in the long-run, we should consider Silos 1 and 2 as equally full and keep them in harmony with one another.

The real pursuit then is to fill Silo-3 and Silo-4, respectively.

Filling the Growth Silos (Silos 3 and 4)

After the Base Layer (Silos 1 and 2) are full and assuming the price of XTZ remains the same through the duration of the growth period, the time to achieve full-capacity (4:1 collateral) can be projected by computing “doubling time”. In other words, we are at 200% collateral, and now we need to get to 400%.

N(t) = the number of objects at time t
= doubling period (time it takes for object to double in number)
N0 = initial number of objects
t = time

(Assuming, however, a major growth spurt in Tezos occurs after the Base Silos are full, the doubling time would be accelerated significantly by redistributing the newly over-collateralized Silo-2.)

Silo-3 will grow from a combination of growth-overflow from Silos 1 and 2. Once Silo-3 has been seeded with capital it can also help support its own growth to capacity from its own banking activities.

Once Silo-3 is full, a 3-pronged growth machine emerges to support the growth of Silo-4. Silo-4 will, therefore, be filled quicker than Silos 2 and 3 were filled. Silo-4’s growth will be seeded by a combination of banking activities from the FIAT silos (Silos 1 and 3), as well as from baking activities from the other XTZ Silo (Silo-2).

Additional banking activities employed by the Trustees to grow the USD Tez Reserve will increase ‘r’ thereby helping us achieve a lower doubling time Td.

Banking activities are rooted in “Savings & Loan” discussed in other articles —both equity and debt security structures. Future articles will discuss derivatives securities as well, including futures-smart-contracts on USD Tez lending agreements. This will include TZ-Bills issued by the USD Tez Reserve

Coming soon: USD Tez Whitepaper

To stay up to date on the USD Tez project and to participate in our discussions, you can visit the official USD Tez website:

and join our social media channels:

Telegram — USD Tez

Twitter — USD Tez

USD Tez (USDtz)

USDtz is a Tezos-based stablecoin pegged to the US dollar

Kevin Mehrabi

Written by

I have some thoughts. Founder/CEO, doing creative product things with #Tezos and #DeFi — oh and I like TV…a lot #Bahai #TzFi

USD Tez (USDtz)

USDtz is a Tezos-based stablecoin pegged to the US dollar

More From Medium

More on Stable Coin from USD Tez (USDtz)

More on Stable Coin from USD Tez (USDtz)

On-Chain Lending — with Tezos

More on Tezos from USD Tez (USDtz)

More on Tezos from USD Tez (USDtz)

TZ-bills — from the USD Tez Reserve

More from Kevin Mehrabi

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade