Proposed sDEFI Index Synth weighting

[garthtravers]

Here is the proposed sDEFI Index Synth weighting, determined according to current (31/10/2019) market caps. Please indicate support or rejection of this weighting using the emoji reactions.

[Arthur0x]

Thanks for sharing the table @garthtravers

I am against giving Chainlink such a high weightage for a DEFI index since it's debatable whether they should be considered as DeFi in the first place.

Chainlink belongs to Oracle category which is a much broader and cover use-cases beyond just DeFi although the focal point is on DeFi at the moment.

Chainlink also have its own idiosyncratic catalyst (4chan meme and LINK Marines)and price movement that are not related to DeFi and will likely distort the value of the DEFI index.

This kind of defeat the purpose of giving people an opportunity to have exposure to DeFi sector in the first place, as 40% of the index is an oracle project which is not conclusively a DeFi project.

I am very critical of the DEFI index created by NASDAQ which include Amoveo with a 20% weightage! This is a perfect example of poor index construction. We should strive to create an index that's much more representative of the DeFi sector.

I propose to reduce the weightage of LINK to be on par with MKR and bump up the weightage of other DeFi projects token proportionally.

Proposed Weightage

Project | Weightage LINK | 25.00% MKR | 25.00% ZRX | 13.00% SNX | 13.00% REN | 6.00% LRC | 6.00% KNC | 6.00% BNT | 3.50% MLN | 2.50%

[degenspartan]

Mmm, counterpoint to @Arthur0x is that: is the sDEFI index that we are constructing supposed to simply be a representative of investible DeFi projects, or should it be tracking the success of projects involved the DeFi space?

I would argue that LINK should be included in the sDEFI index since its first 2 public use cases would be to provide oracle services to Compound and Synthetix, which are both clearly DeFi.

While I do not really like mcap weightage allocation method for indices as a sector play, any alternative method to determine weightage seems very arbitrary.

I am also looking at it from an inverse synth holder point of view. Allocation methods other than mcap weightage is usually constructed with the intention to optimize the upside returns. Having it plain and simple and standard to be mcap weighted is fair to both longs and shorts.

[Arthur0x]

I am fine for LINK to be included in sDEFI index since they do have a lot of touch point with DeFi but oracle cover much broader use cases than DeFi.

My main issue is over-representation of LINK in the DEFI token therefore making its value proposition moot. We have the chance to construct the first robust DeFi index and we shouldn't squander this opportunity.

I disagree with the final statement as for people who are bearish on DeFi as a whole (Bitcoin Maximalists, Marmot Boi etc), They should gain more for owning the inverse DEFI token since DeFi projects with lower market cap usually exhibit higher volatility on downside and go down in value faster and more furious.

Mmm, counterpoint to @Arthur0x is that: is the sDEFI index that we are constructing supposed to simply be a representative of investible DeFi projects, or should it be tracking the success of projects involved the DeFi space?

I would argue that LINK should be included in the sDEFI index since its first 2 public use cases would be to provide oracle services to Compound and Synthetix, which are both clearly DeFi.

While I do not really like mcap weightage allocation method for indices as a sector play, any alternative method to determine weightage seems very arbitrary.

I am also looking at it from an inverse synth holder point of view. Allocation methods other than mcap weightage is usually constructed with the intention to optimize the upside returns. Having it plain and simple and standard to be mcap weighted is fair to both longs and shorts.

[hughkarp]

I'm clearly biased but no love for NXM (Nexus Mutual)? You can even use our bonding curve as the price "oracle" mcap is in the order of $10m

[kaiynne]

The number of coins needs to be much greater than 9. From a statistical perspective, 30 components are needed for representative sampling. I'm not an expert on the DeFi space, but looking at the following lists makes me think that constructing an index using 30 coins should be easy.

https://defipulse.com/defi-list/ https://defirate.com https://media.consensys.net/the-100-projects-pioneering-decentralized-finance-717478ddcdf2 https://www.block123.com/en/feature/awesome-ethereum-defi-decentralized-finance-list

I don't want to be overly critical but it is hard to take this view seriously when there are not even 30 defi projects with tokens. I understand the argument for a larger sample size, but this is what we have to work with. It was unfortunate that REP did not make it in. But the point of the polls was to ideally see which assets people were interested in having exposure to and leveraging those communities.

The other consideration is price feeds and liquidity for any assets that are included, we need to be able to get a price feed with multiple sources which means we need a number of different exchanges. As you go further down the long tail of assets it becomes progressively harder, and their weighting would need to be so tiny as to not materially impact the index itself.

[kaiynne]

So if you don't see you have a problem here, it's really a waste of my time trying to explain it.

I think you are coming at this from a strange angle to be honest. You want some some kind rigid framework for inclusion, but that is just not possible in the current market. There are only 10-15 tokens that represent DeFi projects, many don't have tokens, Compound and dYdX being just two examples.

The reality is that any construction is going to be 'somewhat' arbitrary, but that does not mean there is no value in creating an index. The idea that it must be perfect or is worthless is simply wrong. There is value even in an imperfectly representative basket of DeFi projects, especially given that everyone is able to exercise their own discretion as to whether they want to hold it.

I suppose I most take issue with the fact that you are criticising the construction without putting forward any viable suggestions for improvement. But I definitely curious as to what your ideal index would look like? What assets would be included?

What is the minimum market cap? Minimum liquidity? Minimum number of exchanges it's listed on? Geographic coverage? Number of trading pairs? Minimum age?

Again it appears you want a rigid framework for exclusion here, and that is simply not possible for a number of reasons. But to just highlight one critical one. We need to have multiple feeds to ensure that we can medianise the prices, so we need at least three at a minimum exchanges with reasonable liquidity. This unfortunately puts us in a position where we need to exercise judgement, because if you look at something like say Electroneum you might think it is actually fine to include, it has multiple exchanges and 100k USD of volume per day, but the issue is that the majority of that volume comes from exchanges that are not that reputable. So you if you used a rigid framework you could easily end up with a false positive.

Final point, why would we include stablecoins in the basket? I'm really confused by that suggestion.

[Arthur0x]

I'm clearly biased but no love for NXM (Nexus Mutual)? You can even use our bonding curve as the price "oracle" mcap is in the order of $10m

Unfortunately NXM was not included in the voting round for DeFi Synths last time, wasn't aware there's a market price for NXM due to lack of exchange listing.

[mikedemarais]

adding UNI-V1:DAI to index should be considered

[CryptoToit]

1) I don't "mind" $LINK being included as they are focused on helping build and support the decentralised crypto space. But the main question is are they themselves not a fully centralised project?

2) I like the idea from Mike to consider UNI:-V1:DAI :)

_PS - The current s/iDEFI Index for me is a version 1 we are experimenting with and should not get caught up in trying to get it perfect or worry too much about adhering to some weird industry standard. Lets experiment and then create a better version next time._

[sdchetty]

Honest questions;

1. Why is LINK being included and given such a strong weighting when it's use case extends much further than just DeFi? It feels like the odd one out, I get that they're supporting the platform but that should be divorced from this decision.

2. Why are there more coins included in sDEFI compared to iCEX? I don't see the point of including a long tail of liquidity platforms for an allocation of less than 5% that barely have adoption. If you go on DeFipulse/State of the dApps it clearly shows the leading DeFi protocols/platforms - why not just include those that are actually being used by people? It would make more sense to just simplify it and go for a Top 5 of MKR, SNX, REN, KNC, BNT. Once things like REP & MLN start gaining traction then yeah revise the composition by all means but at this stage it's still too early to include them.

Not a criticism at all, I'm really looking forward to this index, just want to make sure it reflects current state.

[adivated]

RUNE should prob be added

[jjgonecrypto]

Sorry @Rjrunner44 but this conversation is now closed. At some future point we can vote on a new sDEFI index (or have multiple).