[Discussion]Tokenised Gold Pool

XAUT should obviously be ignored because Tether is shady to say the least, and should be considered un-backed until proven otherwise.

At least the economics surrounding the sXAU derivative and the backing of its valuation are clear, while the 2 other tokens have credible physical gold backings.
So i’m in favor of a sXAU/PAXG/PGMT pool.

A pool being only as strong as its weakest token, I can say that if a gold pool is made including XAUT, i’ll personnally choose not to participate in it.

As for PGMT I find it funny to see it voted the least despite being what i think is the strongest Gold based token of the 4 by being directly insured by the Australian national mint (which is one of the major gold mints in the world) and having absolutely 0 fees.

Regarding the argument of “staying focused” , i’d say, if anything, it should apply first to all the existing -and incoming- USD pools…
I’m personally not interested in losing money with exposure to a depreciating USD and right now the only choice of non-USD pool is BTC.

So I think a gold pool would be a good thing, even it if it remains a marginal pool. As i understand it, being an easy bridge between similar stablecoins is the reason for this platform and there’s none for gold tokens right now… so its the perfect usecase for Curve.
Let’s just keep Tether out of it, no need for any additional risk on top of everything else.


Agreed on PGMT being severely undervalued here. We are already taking regulatory/institutional risk on USDC & wBTC in every pool we have. PGMT is the most credible gold token. Having sXAU offers a useful gateway into the Synth market and will make the pool have more functionality.

As for XAUT, if the ‘safe’ gold pool shows enough interest, we could consider a second pool that includes it at a later date. Especially when yield/collateral opportunities begin to open up for these tokens to a greater degree.

Who is going to bootstrap the pool? Each of us probably has a stack of one of the other, but i doubt anyone has all three or four in equal proportion and with 500 gas and poor liquidity on chain, somebody is going to take a beating being the first unless we agree to maybe trade privately and securely amongst ourselves beforehand to all equal up.

1 Like

Continuing the discussion from [Discussion]Tokenised Gold Pool:

This is Dan from Universal Protocol; we’re thrilled to see a gold pool under discussion for Curve and would be honored to be included.

We’re extremely excited to support Curve’s gold pool, and UP would be happy to incentivize increased liquidity. We’ll kick around some numbers and put up a post with details, but we’d both incentivize the pool and likely a 95/5 Balancer pool for our ecosystem reward tokens to help with their liquidity.

We’re excited to see this happen; we’re big believers in gold-backed tokens and see the merits of a variety of approaches. Gold tokens are a great fit for Curve, given the relative stability of the asset suits the Curve formula, and there’s many market participants likely willing to warehouse gold risk in a similar manner as stablecoins. Looking forward to being involved, and happy to support and promote as we can. Cheers.


Can get in touch with us on Telegram.

We expect a soft vote giving a chance to governance participants to choose which tokens they would like to see in this pool to go ahead in the coming days.


We would like to suggest that you also include CACHE Gold (CGT) in the list above.

We have submitted a proposal to create a liquidity pool with CGT and also a Maker Collateral Onboarding Application for more information.


Without a storage fee and a transfer there are significant questions as to how a gold-backed token would be scalable and sustainable. Insured, audited storage will always have a cost. If gold-backed tokens don’t charge fees it raises questions about how they will pay these costs and what incentives they have to provide full backing. A transfer fee alone does not solve this problem because there would be no revenue if investors hold tokens for long periods of time without transferring them. A transfer fee could also be avoided by locking up the tokens inside a smart contract and then trading the ownership of the smart contract instead of transferring the tokens.

If Curve and other similar platforms want to deal with tokenized physicals assets, this is an important issue to consider.

@TwoCell, We may be willing to bootstrap a pool.



Agree with @Julian’s reply generally. However, large market caps do not make gold-backed tokens track the price of gold any better. Liquid markets do that and redemption provides an alternative if markets become illiquid or otherwise dysfunctional. Although PAX Gold clearly has the most listings and adoption so far, their markets tend to be illiquid and/or have asymmetric liquidity (more buys than sells or more sells than buys) and there is a huge range in price between different markets - currently $1,830 to $2,070. Redemption should provide an arbitrage opportunity but due to the high minimum redemption of 430 troy ounces, however it doesn’t seem possible to purchase 430 PAXG (~$832,000 USD at current gold spot) on any exchange at anywhere near the spot price.

There are also transparency issues with both PAXG and XAUT.


IMO, sXAU should definitely be added to this hypothetical gold pool, even if the total market cap of sXAU is much smaller than the other gold-backed tokens. I suspect the reason why there is so few sXAU in circulation is because of the trouble in minting it (i.e.: one must acquire SNX, stake that SNX on the Synthetix platform to mint sUSD and finally exchange that sUSD for sXAU) and not because of the stability of the Synthetix platform.

Moreover, I’m personally inclined to trust in the value of sXAU more than others like XAUT, as the redemption process is on-chain (i.e.: one can exchange sXAU for sUSD on Synthetix with 0 slippage at the IRL rate) and the collateral backing sXAU is verifiable and transparent.


I really like the idea of combining of those types of digital gold (backed, synthetics, …). Even the output goldCRV token would be valuable as something you can put as collateral on other platforms (instead of under a mattress).

I also like the idea of a “WPaxG” although I don’t know if it solves all the fees-related issues nor if consulting Paxos first wouldn’t be better.

1 Like

Formalisation of process starts now: CIP#12 - Which tokens should be included in a gold pool?

1 Like

I think it’s a fair point that:

I would ask the audience here if Curve’s algorithm is likely to work better and earn more on:

  • a large deep liquid market or
  • a small shallow chunky one.

How about including Cache Gold? They seem to be the only one that has a real connection to the physical market, nice to know you can take delivery of your gold when fiat currencies/synthetics go to pieces


Any movement on this? I now there were concerns about the tiny market caps of each different type of gold token, but i was looking today and the market cap for paxg for example has increased markedly since this topic initially came up - I presume the others have seen similar growth. It would be nice to have more on chain liquidity in these tokens and be able to swap between them with no slippage since some are convertible to physical and some are not.

1 Like

Hi Twocell,

You can read about the progress of CACHE Gold here: https://cache.gold/assets/docs/CACHE-Growth-Report-2021.pdf

We definitely have a working and efficient redemption system for bars as small as 100 grams.

Please let us know if you have any questions.

Yes, but only use PGMT and sAUX to begin with, then add others who prove to redeemable/liquid in the future.

sAUX is by definition not redeemable because it’s synthetic. PMGT is redeemable by almost nobody. Have you tried? Their app (required to redeem) is not supported in the vast majority of jurisdictions. The redemption process is complicated requiring redemption of tokens for a certificate and then redemption of the certificate for gold.

Do you know anyone that has redeemed it? How much did they redeem? How much did it cost? How did they do it and where are they located?

PMGT is only listed on two exchanges and only one of them has volume (very little) also.

There is a more in depth discussion on the same topic here: CIP#12 - Which tokens should be included in a gold pool?

1 Like

CGT is proven to be redeemable. It is the only gold token that provides full onchain transparency for all redemptions.

When tokens are redeemed they are removed from circulation and locked up in a smart contract address we call the “Unbacked Treasury”. After tokens have been removed from circulation the corresponding bar(s) are unlocked and released to the redeemer. After redemption bars no longer appear in the CACHE Gold Explorer as CACHE Locked Gold backing tokens but the full history of each bar is available up to the point of redemption and removal from the vault. Here is an example bar redeemed on June 3, 2020: https://explorer.cache.gold/r/E2801105200076C59AA10A28

Bars can never be unlocked unless tokens are removed from circulation first and unbacked tokens locked in the smart contract cannot re-enter circulation unless an equivalent amount of gold is deposited and locked first.

You can see by inspecting the CACHE Unbacked Treasury address how many redemptions have occurred to date and for how much gold. You can also audit the supply to ensure full backing quickly and easily with publicly available, real-time data.

1 Like

I don’t know enough about cryptocurrency-based gold tokens to say which ones should be included and based the two tokens on previous posts in the thread. Don’t see those two as fact but my opinion which is easily changed since I still have much to learn about this topic. :slight_smile:

I do however see it as important that gold tokens either be redeemable or in the case of synthetic ones; liquid and pegged harder than my gimp.

1 Like

I like it. It has transparency.

The redemption page says:

The Redemption Process

To redeem your tokens, you must have an approved CACHE account and provide an Ethereum address that holds CACHE tokens you wish to redeem.

You can select any parcel for redemption as long as the pure gram total of the parcel is equal to or less than your CACHE Gold token balance. During the checkout process you can choose to:

  • Pick up your parcel(s) at the vault (only available at some vaults)
  • Ship your parcel(s) to your mailing address (this incurs additional costs)
  • Sell your parcel(s) to a liquidity provider that accepts gold deliveries at the vault

Once you submit your redemption request, CACHE will reserve the parcels for 10 minutes in order to allow time for payment of the balance due. If you fail to send the full balance due within the time limit, the redemption request expires and any partial transfer received will be refunded to the originating address.

Some questions:

  1. What does it involve to create an account on CACHE, KYC I assume?

  2. By your estimate, what would it cost on average to ship: 10g, 100g, and 1000g to different locations around the world? Do you have a table for countries and shipping costs?

  3. What should someone do if it isn’t feasible to ship to gold? I assume selling it to a liquidity provider would be an option. Are there option to do this today, and if so how?

4.(A bit off-topic) Any plans to add Silver, Platinum or Palladium?

I would love to see tokens representing 25% each of Gold/Silver/Platinum/Palladium being exchanged by Curve in the future, but that would require at least two of these to exist…

I think Silver, Platinum and Palladium pools would be much simpler to implement on Curve and would prefer to see it done as fast as possible seeing how inflation will hit the dollar sooner or later in the coming years (probably sooner). Diversification is a must but we can’t just rely on BTC, ETH and other cryptocurrencies since they are in the same basket. The same goes for fiat currencies.

1 Like