SIP - Periodic Rebalancing of sDEFI

The sDEFI index synth is re-composed quarterly. Every three months community members vote on asset inclusion and relative weightage of projects to be included in the sDEFI index synth. The DeFi space moves fast.

Rebalancing the index enables the index synth to innovate as fast as the market. In the first rebalancing of sDEFI with SIP-69, UMA, LEND, COMP, and BAL joined the index. Future sDEFI versions may incorporate new projects such as YFI, MTA, Serum or other projects not yet released to the wild.

Some conditions:

  • Token must have existed for at least 3 months in the market prior to inclusion in the index.
  • A single token cannot comprise more than 20% of the index.
  • There will be a 2 week grace period before the rebalance takes effect.

Changes in sDEFI composition by asset will change via SIP, while changes in the weights of assets will be configured by SCCP. Will attach SIP shortly.

In order to align to with market standards and the quarterly rebalance I suggest also Tokens have to be in market at least 3 months.
Also the fact we should give 1-2 weeks grace period before the rebalance comes into effect.

Is the grace period defined as the time after the new index is finally voted on and before when the index synth actually goes live?

Correct. In terms of governance, and taking into consideration special dates like new years in this schedule (allows for market to come back to normal after EOY while new weights are decided).

Every Quarter (Jan,Apr,July,Oct):

  • First Week of the Q: SCCP is open, decided and approved
  • Second Week of the Q: Grace Period
  • Third Week: implemented on the first day
1 Like

I like that time line and the rules suggested just wanted to add my own comments. I think we’d only need to discuss this in a couple months though if we go for a Quarterly Rebalancing.

Given how DeFi is expanding we also need to decide on if we want to have a global encompassing DeFi index or do we want to limit the number of assets included in it.

Some have floated the idea of having a Large Cap DeFi Index and a Small Cap DeFi Index as well as having more allocation of low caps in the DeFi Index to potentially capture more price appreciation.

We’d also need to determine new assets to be included, and the criterias for their inclusion. Great start here;

Some conditions:

  • Token must have existed for at least 3 months in the market prior to inclusion in the index.
  • A single token cannot comprise more than 20% of the index.
  • There will be a 2 week grace period before the rebalance takes effect.

What we’ve done at CoinGecko as a start for our DeFi list is

  1. Must have a working Product
  2. Must be related to DeFi (Banking, Finance, Lending etc)
  3. Must be composable and able to interact with other DeFi Protocols and not closed walled gardens
  4. Open Source and Auditable code are an added bonus

Some additional points for consideration
a) Should the assets be ERC20 only?
b) How do we derive the prices for these? Are they listed on enough exchanges to derive a median price/avoid market manipulation
c) Have the tokens been distributed widely enough or is the circulating supply float still too little to matter.

I would love to see the inclusion of a few more assets but not now as I think they’d need more time in the market. These include; $MTA , $YFI , $NXM , $KAVA, $CRV

1 Like

I don’t think a hard limit on the number of assets to include is doing us many favors at this point of iteration. As the DeFi market develops and matures it makes sense to explore adding a small-cap DeFi index synth and a large-cap DeFi index synth.

One consideration is that a relatively larger number and correspondingly lower weights are probably needed for a small-cap DeFi index synth to assuage concerns about manipulating one of the underlying index components on spot markets to attack Synthetix and gain relatively riskless profit by taking a position on the small-cap DeFi index synth in advance. Many small-caps are not very liquid and therefore more susceptible to price manipulation.

These criteria are great from the community perspective of deciding what should and shouldn’t be put up for debate. We should not want just any project claiming themselves to be “DeFi” to be a candidate for the index, like “Long Island DeFi Tea Company” or some such gimmick.

Finally:

  1. DeFi is still a developing segment that is probably too young to say will only ever happen on Ethereum with ERC20 tokens. The Serum project from FTX, Alameda Research and Sam Bankman-Fried will run on Solana for example. This will be a project to watch under the DeFi umbrella that’s not on Ethereum.
  1. Decentralized oracles from Chainlink are great when possible. CoinGecko has a strong read on the market price and a public API with real-time price data but as far as centralized oracles go I know these are not publicized in order to deny front-runners any advantage.

  2. The three-month time period in the market is set to give projects time in the wild before including them as index components. While there is risk to adding a token too soon before a wide variety of market participants have expressed their views on it via buying and selling, the tradeoff is waiting too long to add new index components such that the index is not reflecting the dynamic state of the DeFi sector. Index components should be evaluated on a case-by-case basis. There’s a strong case that limited floating supply relatively to total supply could be a criteria that leads the community to assess a token does not have sufficient maturity and liquidity to justify inclusion.

1 Like

Great ideas here, adding my 2gwei:

  • I think rebalancing quarterly is the right interval, and perhaps there’s added value in aligning it to the dates of other major index rebalancings (S&P rebalancing schedule comes to mind)

  • I would like to introduce as little ‘subjective’ reasoning into index-inclusion as it creates governance overhead and is not the intention of the protocol to be giving any guidance on ‘quality’ of projects, IMO.

    • having ‘hard’ numbers. i.e. 30mm marketcap, minimum 2 ‘approved’ DEX/CEX books with 2% depth threshold met
    • must propose for inclusion during a quarterly rebalance, and then ratified for inclusion by governance on the next quarterly rebalance

the first category is intended to prevent tokens that don’t meet the minimum ‘fundamental’ requirements, the second is meant to weed out the low interest as well as the short-lived projects, as there is a significant time delay for inclusion.

1 Like

I like the idea of introducing an asset in the quarter before it can be listed, this is about on pace with the 3 month period an asset should be in the market before it can joint the index. I guess what’s really being proposed is 3-6 months on the market before inclusion in the index.

On the first point, I think it’s okay to have some subjectivity. Arguing for an asset to become a part of the DeFi index is like trying to hit a moving target, which requires multiple critical assessments cross-referenced or corroborated against each other to reach the truest picture of what should be in a DeFi index. “What is DeFi?” is surely a different question to me now than it was 6 months ago. It’s not that there’s no place for hard numbers, just that they shouldn’t be a principal guide. Market cap figures can also change rather quickly due to a volatile market. Like, were Curve and Balancer even around 6 months ago? I don’t think so yet I cannot think of DeFi now without thinking about these AMMs.

Continuing along with the precedent to rebalance sDEFI as frequently as the S&P 500 Index, that is quarterly, there would be 2 rebalancing dates during the remainder of 2020. On 30 July 2020 the sDEFI was rebalanced. The proposal here is to rebalance sDEFI on 18 September 2020 and 18 December 2020. This resource is consulted to reach these dates: https://www.nasdaq.net/PublicPages/assets/MyMID/Trade_Calendar_2020.pdf.