IIP-7 CoinShares Gold and Cryptoassets Index Lite Version (CGCI-LV)

[mcgpetch]

iip: 7
title: CoinShares Gold and Cryptoassets Index Lite Version (CGCI-LV)
status: Proposed
author: Michael Petch (@mcgpetch) , CoinShares
created: 2020-10-27

Simple Summary

The CoinShares Gold and Cryptoassets Index Lite Version (CGCI-LV) is an adaptation of the original EU Benchmark Regulated index launched by the CoinShares Group in 2020 (see here: Index Strategies) that only includes ERC-20 tokens in order to be compatible in a tokenized index smart contract.

Since the original CGCI is constructed using physical gold and 5 individual cryptoassets some re-engineering of the initial structure was required, whilst ensuring that the core theory and value proposition was preserved. The CGCI-LV therefore deploys the same theory but instead limits the index to 3 ERC-20 token constituents; a gold stablecoin and 2 cryptoassets.

For the gold basket, wrapped-DGLD (wDGLD) was chosen as the ERC-20 gold stablecoin. For the cryptoasset basket wrapped-bitcoin (wBTC) and wrapped-Ether (wETH) were selected based on being the largest and most established cryptoassets.

Abstract

Academic research conducted in partnership with Imperial College London found that a pairing of gold and cryptoassets in a way that accounts for their risk contribution delivers a risk-adjusted return profile that is superior to holding gold or cryptoassets alone. The CGCI-LV is designed with the aim of providing diversified exposure to the alternative asset space in a way that yields a superior risk-return profile when compared to holding such assets in isolation while being orthogonal to traditional financial markets. Accordingly, the Index must:

1. Be comprised of a small number of liquid, investable constituent assets
2. Exhibit a relatively stable composition of asset weights that do not vary dramatically between rebalancing periods, leading to low turnover
3. Utilise some means of principled risk control leading to lower volatility
4. Be specified in a clear and unambiguous manner to facilitate validation and reproducibility
5. Hold constituent assets on a long basis only
6. Not make use of leverage.

Motivation

Two noteworthy characteristics of the returns of non-stablecoin cryptoassets are their high volatility, which brings with it a high level of risk, and their high intraclass correlation, which limits the benefits that can be had by diversifying across multiple cryptoassets. Yet cryptoassets exhibit no correlation with gold, a highly-liquid yet scarce asset which has proved to function as a safe haven during crises affecting traditional financial systems.

Although volatility poses challenges in terms of increased uncertainty, there are also benefits to be had from its proper management through diversification and regular rebalancing (Bouchey et al., 2012, The Journal of Wealth Management. Volatility harvesting: Why does diversifying and rebalancing create portfolio growth?). This is exemplified by the so-called Shannon’s Demon approach in which two, ideally uncorrelated, assets – at least one of which is highly volatile – are periodically rebalanced to maintain an ideal target allocation. The resulting expected growth rate is greater than the arithmetic mean of the individual expected growth rates, while the variance of the returns is less than the mean of the individual variances (Poundstone, 2005. Fortune’s Formula.).

The CGCI-LV is intended to be a low-volatility index that utilises the concept of volatility harvesting through (a) forming a basket of cryptoassets and (b) combining it with gold using weighted-risk contribution as a rebalancing mechanism. By decreasing volatility levels, it seeks to yield superior risk-adjusted returns when compared to a number of alternative strategies, including holding cryptoassets or gold alone. Further, it presents a moderate turnover, which should translate into moderate operating costs.

Size of opportunity

The CGCI has already gained a lot of demand and CoinShares is currently investigating opportunities to deploy the index as an investable benchmark in its passive products business. Additionally, it is anticipated that the CGCI-LV would likely be well received by the native crypto community and is expected to be picked up by CEXs to offer it to their users.

Liquidity commitment

Liquidity and assets will be committed. The gold stablecoin issuer will also commit assets and liquidity.

Differentiation

While there are several existing cryptoasset indices that offer broad exposure to crypto assets through market capitalization weighting, they have proven to be limited in risk diversification due to the high correlation among cryptoassets and are characterised by a volatility close to that of a single cryptoasset. CoinShares developed the index to bring more effective risk control to cryptoasset index products.

Statistics

Annualised Return: 0.4604

Annualised Std Dev: 0.2498

Annualised Sharpe (Rf=0%): 1.8431

Annualised Sortino (Rf=0%): 3.0473

Annualised Avg Turnover: 1.0233

Year-to-Year Return (%): 60.9701

30-day High: 7355

30-day Low: 6635.82

Return Since Inception (%): 632.248

Source: CoinShares. The CGCI-LV was launched on October 28th, 2020. Figures for the CGCI-LV prior to this date have been derived by applying the methodology described on this page. The crypto-basket price base level was set to 100 on July 1st, 2015. The Index base level was set to 1,000 on January 1st, 2016. Calculations do not include fees and expenses. Past performance is no indication of future performance.

On-chain liquidity analysis

All assets are supported by on-chain liquidity providers and committed resources will be mandated to maintain liquidity for the CGCI-LV.

Fee Split

CGCI-LV will have an annual streaming fee of 0.60% (60 basis points). Revenue from the annual streaming fee will be split with 60% to CoinShares, 40% to the Index Coop.

Methodology

Constituent Weighting:

For the constituent weighting, we examine each constituent’s native asset (i.e. for wrapped-bitcoin we investigate bitcoin, for wrapped-DGLD we investigate DGLD, etc.). We follow a bi-level approach that involves studying the historical volatilities of a crypto-basket (containing BTC and ETH) and gold (DGLD), separately in order to inform the crypto-gold asset allocation decision.The Index is calculated following a two-stage allocation scheme that involves:

1. Computation of the historical volatility of (a) the Equally weighted crypto-basket, and (b) DGLD;
2. Asset allocation among the crypto-basket and DGLD expressed as the bivariate weighted risk contribution problem.

Rebalancing schedule:

In order to fully capture the diversification benefits of the time varying correlations between gold and the cryptobasket, a monthly rebalancing frequency is employed. Constituent weights are announced on the last business day of the running month. The rebalancing is employed on the first business day of the following month.

Index Calculation:

In the CGCI-LV, the crypto-basket is formulated as an Equally Weighted basket of 2 defined cryptoassets, each with a weight of 0.5. The crypto-basket price base level is set on 100 on July 1st, 2015. The Index base level is set on 1 000 on January 1st, 2016.

Author background and commitment

The CoinShares Group is a pioneer in digital asset investing and manages more than $1 billion in assets on behalf of a global investor base, with offices in Jersey, Stockholm, London, and New York. Over the last seven years, the firm has built innovative, regulated, institutional-grade products and services backed by robust research. This legacy started with the first regulated Bitcoin fund in 2014, and continues today via the XBT Provider ETP offerings and other asset management and trading services built by the Group. For more information on CoinShares, visit: https://coinshares.com/.

The Group originally launched the CGCI in May 2020 and was the first EU Benchmark Regulations (EU BMR) compliant index for the digital asset industry that combines digital assets and gold. The index is now live on Bloomberg Terminals (ticker: COINCGCI) and Refinitiv (ticker: .COINCGCI).

Reference

Index Whitepaper: https://coinshares.com/assets/index-strategies/koutsouri-poli-alfieri-petch-distaso-knottenbelt-marble-2019.pdf

Original CGCI Methodology: https://coinshares.com/assets/index-strategies/CGCI-Methodology.pdf

Stress Testing the original CGCI: Index Strategies

Original CGCI Benchmark Statement: https://coinshares.com/assets/index-strategies/CGCI-Benchmark-Statement.pdf

[Etienne]

Hey. I really like the idea and we definitely need larger underlying markets if we want to grow considerably the AUV. Some questions / remarks:

• wBTC is centralized, but reliable so far - what do you think about adding some RenBTC to the index ? I like the idea of market cap weight between BTCs versions to reduce the risks - and there is a really low slippage between these assets on Curve notably

• what is the current liquidity depth of the gold token ? Could you incentivize liquidity on Uniswap ? I’m a little bit worried about the spread compared to the real price of gold.

[Kiba]

This is a very interesting and pragmatic strategy and obviously having industry leaders like CoinShares is great.

On tokens:
I can’t find much about DGLD but it says it’s issued on a private Bitcoin sidechain. If you’re wrapping, is the bridge also centralized? That would be compounding a lot of risk, even if the backing assets are secured and certified by public institutions. Could you issue natively on ETH like Tether and Circle do?

The proposed (not finalized) code for new index smart contract doesn’t allow WETH to be added as a component (source). I think that might only be because of the one tradeRemainingWETH method but I’m sure Set team can comment further.

[overanalyser]

It looks like an interesting proposition and something distinct from $DPI which is a good thing.

Are there any ERC20 based gold tokens?

The fact that coinshares have got the related ticker incorporated into TradFi markets already means that the index might attract a different customer base to crypto natives .

I’ve been using coinshares Ether Tracker Euro for a few years to give crypto exposure in my UK pension https://coinshares.com/etps/xbt-provider.

I would be interested to see the proposed fee / free split (coinshares vs INDEXcoop)

[richard]

Tagging @mcgpetch here

[mcgpetch]

Hey @Etienne, thanks for the comments and questions!

Regarding the ERC-20 bitcoin constituent, we went with wBTC because it is the largest by market cap and the first token to standardize bitcoin to an ERC-20 token. We could certainly explore holding a basket of ERC-20 bitcoin’s, however, a few concerns that immediately come to mind are the following;

1. It would increase turnover for the index
2. It would make the index a bit more complex for users to understand at first. At the moment the index design is simple, bitcoin/ether & gold. If we added multiple bitcoin’s to the index we might confuse users who are not customed to DeFi (note, this token eventually should be listed on CEX’s and target their users).

In regards to wDGLD, the CoinShares Group just launched the token and we haven’t even announced it, or marketed it publicly yet. We’re just gearing up for our official launch in the coming weeks and we will be driving a lot of liquidity from our in-house trading desk and liquidity providers (including Kyber) to ensure spreads are as tight as possible, especially on Uniswap. Our goal is to always have the spread less than 1% but we’re going to get it as minimal as possible. We’ve also chatted with the Curve team about a gold-token yield pool.

[mcgpetch]

Thanks Kiba!

As mentioned above in my reply to @Etienne, we’ll have more information to publicly share in the coming weeks about wDGLD. However, to answer a few of your comments, the wDGLD wrapping bridge is centralized (managed by CoinShares). We went with this approach b/c when the DGLD network was being built in 2017, the project was driven by one major pillar; gold investment is heavily reliant on institutions and does not have a proper digital format to work, and thrive, in the digital world . We aimed to solve this problem by creating a robust digital format for gold, operating on cryptocurrency rails, which gave users full ownership in physical gold stored at the most secure vaults in Switzerland. In order to do this we needed to build our own blockchain network since the Ethereum network would not support the functionalities required to provide direct ownership of the underlying gold to the token holders and the ability to scale in any jurisdiction from a regulatory and technological standpoint. That is why we created the DGLD blockchain which fulfills all of these objectives. Learn more about the network here: https://dgld.ch/assets/documents/dgld-whitepaper.pdf .

However, there is of course a lot of benefits of being on the Ethereum network (like being able to put gold in a tokenized index). But to issue natively on Ethereum would not allow users to have the same regulatory assurance and full legal ownership in gold with the ability to redeem for physical gold, silver, platinum and palladium. This is why a wrapped token, while maintaining the DGLD network, made sense for the DGLD project. The inherent values which drove the project since inception remain intact, while being able to tap into the wider digital economy by being supported on the Ethereum network via wDGLD.

However, to address some of the concerns of centralization, we have independent auditors in place and all assets are signed by the vault operator and viewable on our block explorer here and of course you have the MKS PAMP Group and the CoinShares Group backing the Consortium.

[owl]

This is definitely a good proposal to be considered by the community. I agree that volatility harvesting between gold and cryptoassets would be a great option in this money go brr… era. However, I have one qualm regarding this proposal:

• I understand that CGCI has performed well with markedly lower volatility compared to cryptoassets, suffering less than 25% correction during the Black Thursday early this year. However, considering possibilities of another white swan events in the near future (ETH2 Deposit, Bull Market etc.) what is the risk of wBTC and possibly wDGLD losing their peg? Would this present potential attack vectors for flash loaner?

I am positive that this product could become the index asset for the masses. However, I also think that careful considerations are needed for this to be as safe and stable as possible.

[mcgpetch]

Hey @overanalyser --glad to hear your support of our XBT Provider products over the years. Great to have you with us

Re. ERC-20 gold stablecoins, there are many but the only legitimate projects are; PAXG, XAUt and wDGLD. We support wDGLD since the CoinShares Group launched the ERC-20 token and is part of the DGLD consortium (see here). Also, wDGLD is the most sophisticated in terms of partners (you know who is behind your gold and where), regulatory structuring (DGLD is a payment currency and not a security meaning no risks for index creators and exchanges to support), technological (the holder is the legal owner in the underlying gold that the token maps to on our explorer) and is actually physical redeemable for physical gold, silver, platinum and palladium.

On the note of our existing tickers on BBG and Refinitiv, we certainly believe this index will interest new target markets and users. The index is a great gateway into crypto, especially for internet investment gold users ($7.2bn market), and once they’re in they will definitely be interested in other products, like bitcoin, etc. Many partners we’ve spoken with also share the same opinion.

[mcgpetch]

Thanks @owl for the comment!

Specifically on the topic of how the index would react under different market regimes and movements, you can find here a paper we wrote with Imperial College that you might find interesting.

Re. the white swan events that you mentioned, I think you and @Etienne raise a good point about possible concerns with wBTC. In this case, in our original CGCI methodology the index committee is able to extraordinarily decide if a constituent is not referencing the proper asset price, for which the committee would replace that asset. In this case perhaps the committee should be able to decide whether or not wBTC is always the best wrapped-bitcoin token to use. In relation to wDGLD, our liquidity providers are obligated to keep the spreads low and have the capabilities to keep spreads tight through our trading systems which should never be a problem.

[neozaru]

Hello,

First, good idea. This is something I haven’t seen in the crypto-space yet.
Also timing-wise it makes double-sense :

• No clear answer today about what is a good risk/reward ratio investment. As risk-off still risk-off and risk-on still risk-on ? Having a basket a uncorrelated assets can be a type of response.
• BTC and GOLD tokens (ERC20) are now available with enough liquidity to start thinking of this type of stuff, which was not the case some months ago.

Now, I have no specific bias on a specific “wrapped” BTC implementation (same for tokenized gold) but I do want to point out that Curve.fi (https://curve.fi) pool tokens could address the issue. In short, Curve is an AMM for coins which are supposed to have the same value. Currently there are BTC pools combining WBTC, renBTC, sBTC and other, diversifying the risk. Every pool has its own output token - with a nice and long name - that any project can reuse (Yearn is doing it for example).
They are planning to have a Gold pool in the near future, so that could also be something that interests you, should you not find a consensus about which “flavor” of BTC and Gold is the best.
Ah, and those Curve pool tokens also accrue interest through trading fees

[reganbozman]

I’m a fan of Coinshares and so its great to see ya’ll actively engaging with the community.

I think this index is interesting intellectually but I have open questions on whether this will dramatically increase AUM, which is the primary focus right now.

Can you share some data on the float right now of other tokenized gold indices? My instinct is that the market for these things has not really taken off - probably because people who want gold exposure today have a lot of of good options - but i certainly may be wrong on that.

[owl]

Curve is a great project but I think introducing it for an index product is not the way forward. If we were to use curve tokens as the underlying assets we will expose end users to various risks (associated with smart contracts and underlying assets). I will not support curve tokens for underlying assets of this specific proposed Index product.

[overanalyser]

OK, some more thoughts.

Initially I was twitchy about the centralised nature of the wDGLD, and "2) The sidechain provides enhanced control over KYC and AML: The DGLD sidechain has whitelists and blacklists permissions, providing greater control over**KYC and AML;" on page 8 of the white paper. https://dgld.ch/assets/documents/dgld-whitepaper.pdf

However, I don’t think there is anyway that gold can be incorporated into DeFi without some form of centralisation and I suspect come form of AML / KYC. What I don’t know is if, by the time if has been tokenised on the bitcoin sidechain and then wrapped into ERC20 if it’s still controlled (I suspect so, based on analogy to USDT…).

So for INDEXcoop, if we go with ANY gold containing index, we are likely to have some people say "xxx from INDEXcoop is just centralised [as per wBTC, USDT and many others]. So that is something we need to consider.

I see a huge liquidity problem, my understanding of the set protocol is that for an efficient market we will need DefI native protocols for liquidity:

• Set protocol issue and redemption
• secondary market for the new token (xxx:ETH uniswap pair).
• secondary market for the individual tokens in the index (i.e. ETH, BTC and wDGLD), I would suspect that most arbitrage would prefer ETH pairs.

The current w-DGLD DeFi liquidity looks too low to support large traders (~$38 k https://info.uniswap.org/pair/0x65bc62d64f2cdca65dd8abe56265609524a25804). So there would need to be some market makers for both wDGLD and the new index. Market makers willing to accept divergence loss.

It’s possible that the main market for this INDEX fund will be retail acting through brokers who issue and redeem the tokens as needed so DeFI liquidity isn’t essential to grow AUM. However, I think that there may be some DeFi potential customers who will avoid anything with such low DeFI liquidity.

Looking at the other two gold tokens mentioned:

https://www.coingecko.com/en/coins/pax-gold 72 M market cap, 1.6 M daily volume
https://www.paxos.com/paxgold/ ~23 k volume on Uniswap, 190 k deep pool

https://www.coingecko.com/en/coins/tether-gold 92 M market cap, 70 K volume.
https://gold.tether.to/ %22 k deep pool on Uniswap

So, I’m certainly interested in this proposal as a second INDEXcoop product.

I think at the moment the Coops focus should remaining on promoting DPI and building it’s AUV within our core market of crypto natives. But I don’t see that preventing a second fund being launched alongside CoinShares with there commitment to (on chain) liquidity and a presumed marketing push by coinshares as they promote DGLG and wDGLD.

@mcgpetch Two questions:

1. How quickly do you see you launching the DGLD initiatives?
2. What kind of fee structure would you propose for the $CGCI-LV fund?

[Kiba]

The whitelist/blacklist also concerned me since what is the real value in the token if you are unable to redeem. The fact that it is backed by gold means nothing if you can’t get the gold, might as well use a synth like sXAU. I think he addressed this in his comments above though. I do think there is a ton of value in the regulatory/institutional aspect in contrast to DeFi tokens which opens a whole new market for us considering all the assets in this index are widely accepted in traditional financial markets now (e.g. CME).

I took the fact that DGLD isn’t publicly launched yet and the statement “committed resources will be mandated to maintain liquidity for the CGCI-LV.” in the proposal to mean CoinShares and other DGLD members would provide liquidity before launch.

Besides the relatively minor issues about tokens/liquidity which can be easily solved, they are more than capable of pushing this forward without taking resources away from DPI. It’s not zero sum either, the news from this could attract more talent and money to Index Coop. The fact that CGCI-LV is far less risky and easier to understand than DPI makes it more likely for exchanges to list which opens the doors for DPI as well.

[overanalyser]

In some ways, i think that there is no avoiding a centralised participant when tokenising gold. While some customers may not want exposure to a centralised asset, there are plenty of people using wBTC and USDT…

[mcgpetch]

Hey @overanalyser, thanks again for the comments.

On the topic of centralization for the gold component, you’re right that no matter what asset-backed token you work with there has to be some form of centralization. At the end of the day your gold has to sit somewhere. It’s a bit of a give and take situation where you have to have a certain element of centralization. In the case of DGLD we did our past to get the minimal amount of centralization and holders have legal ownership in the gold and it is outside of the traditional banking system.

Regarding the liquidity situation --we are definitely aware of the lack of liquidity in DEX’s. As mentioned in another comment, we haven’t launched wDGLD yet hence the lack of liquidity on UniSwap but once we launch we will have plenty of resources committed to unchain liquidity and on DEX’s to support any type of demand.

To answer your q’s:

1. By the end of this month. Keep an eye out over the next few weeks
2. Fee model is still to be decided and would be great to discuss it w/ everyone on the upcoming IIP-7 call EoW but we’re thinking of going for a low fee. Idea here is to have a lower fee but get the index to an economy of scale from large AuM across multiple distributors (e.g. DEX’s, CEX’s and platforms).

[mcgpetch]

Hey @Kiba, thanks for the comment.

I actually disagree with your comment that if you can’t redeem then it doesn’t have value. Average investors holding precious metals ETF’s usually aren’t able to redeem unless they hold a lot of units yet billions of dollars are held in those products. In the case of the w-list/b-list situation that is employed, this is a necessity if you want to operate as a regulatory compliant stablecoin issuer. At the end of the day that should give you more trust in the issuer and the underlying given that we operate to trad-fi standards and your assets are protected. Like I replied to @overanalyser, bringing asset-backed tokens into crypto is a bit of a give/take in terms of centralization. IMO I would much rather know the issuer than trusting an entity like Tether for example.

On the other hand I wholeheartedly agree that this isn’t a zero-sum game. This type of index has a different target market than the DPI as it has different investment objectives. Keep an eye out over the next few weeks as CoinShares starts to share the IIP-7 and our on-chain asset management ideas…

[mcgpetch]

Thanks @reganbozman! I actually wouldn’t limit this index’s reach to only gold investors, although that is a great market itself to be in. For instance, the internet investment gold market is a $8.6bn market that is primarily occupied by out-dated platforms (take Bullion Vault for example that has about $3.2bn on its platform). I don’t think it would take much to push some of these investors into this type of product which is more compelling (+ the avg. client has $29k in holdings which makes the CPC much more appealing). Precious metal asset managers are starting to look for ways to beef up their offering as their investors want more innovative products that involve crypto. I think we’ll start seeing more gold investors start acquiring bitcoin over the next few months and the CGCI-LV is a great place for them to start as it provides better risk and return profile that is superior to holding gold or cryptoassets alone.

I think this product is a great gateway for trad-fi and gold investors coming into crypto. Also, it’s a product that is likely to be well-received and supported by CEX’s and platforms (like Coinbase, for instance). If this product can help with opening up the door to some of these platforms then that’s a good win for future INDEX Coop initiatives and products.

[Kiba]

Yea I totally agree for “average investors”, I’m just talking from my perspective as a crypto-native. If all I want is price exposure I’d much rather use ChainLink oracles and Synthetix contracts rather than have the counter party risk, government surveillance, and private IT systems of DGLD. To other people those things are definitely valuable and I don’t think you’re targeting people like me just wanted to mention other token options.

Excited to see this launch!