IIP-###: Launch the Target Retirement Indices (TRI10, TRI20, ..., TRI80)

Governance Proposal
1

IIP: ##
Title: Launch Target Retirement Indices (TRI10, TRI20, . . . , TRI80)
Status: Proposed
Authors: @JosephKnecht (Article21), @mringz (Index Coop)
Requires: IIP-58: Launching Pulse Aggregate Yield (PAY), IIP-89: Launch Polygon Diversified Index (PDI)
Created: 06 Oct 2021

Simple Summary

The Target Retirement Indices (TRI) are portfolios of crypto indices which risk-adjust as the investor approaches retirement age. The TRI series is based on Target Retirement ETFs and mutual funds which are among the largest and most popular passive investment vehicles by AUM.

Abstract

Target retirement funds are the cornerstone of investment planning. The basic principle is to begin with a growth-oriented portfolio early in an investorā€™s life and then follow a ā€˜glide pathā€™ to a more conservative, income-generating portfolio as the investor approaches retirement.

Target retirement funds currently have $2.8T AUM and are growing rapidly. However, in spite of their enormous size and popularity in TradFi, there are no target retirement indices for cryptocurrency. Index Coop has an opportunity to reinvent one of the largest investment product categories for crypto, as well as to position its indices as the cornerstones of retirement planning.

These would all be Set Protocol ā€œsimpleā€ indices and so require the least amount of engineering effort.

Motivation

Rationale

Broadly diversified, passive portfolios have been shown to optimize risk-adjusted returns. However, investorsā€™ risk-reward preferences and requirements change throughout their life as their personal liabilities change and retirement approaches. For example, a younger investor may be better-served with a higher-risk, growth-oriented portfolio whereas an investor in retirement needs a more stable, income-bearing portfolio.

Target retirement indices are designed to optimize risk-adjusted returns based on the investorsā€™ risk-reward preferences and needs at the current point in their investment cycle. Given their simplicity and high risk-adjusted performance, target date indices have proven enormously popular.

The largest examples include:

Specification

Overview

The TRI indices would contain a broad representation of the crypto market including Index Coopā€™s major thematic indices, mirrored equity indices, and yield-bearing stablecoins. The allocations would follow a glide path from growth- to income-oriented assets as the investor approached retirement age. Additionally, the investor could opt for a more aggressive or more conservative strategy by choosing, respectively, a later or earlier effective retirement age.

Differentiation

Index Coop currently has no retirement-focused indices in its current product line. Based on our research there is no retirement-focused, target retirement, or time-adjusted crypto products on the market.

Example composition

image
image2054Ɨ1292 197 KB

The Target Retirement Index series will consist of the following 8 index tokens:

  • TRI20 (Target Retirement Index 2010) - 10 years post-retirement as of today
  • TRI20 (Target Retirement Index 2020) - At retirement
  • TRI30 (Target Retirement Index 2030) - 10 years away from retirement
  • TRI40 (Target Retirement Index 2040) - 20 years away from retirement
  • TRI50 (Target Retirement Index 2050) - 30 years away from retirement
  • TRI60 (Target Retirement Index 2060) - 40 years away from retirement
  • TRI70 (Target Retirement Index 2070) - 50 years away from retirement
  • TRI80 (Target Retirement Index 2080) - 60 years away from retirement

Each product reflects different allocations for the respective target retirement date. Draft allocations are shown above. These particular index components were chosen purely for illustration and may change depending on their approval status and community input. mSPY is mirrored SPDR S&P 500 ETF Trust (SPY). Note that ideally there would be a total crypto market cap index like VTI to serve as the core growth component. In its absence, we use BED which is the closest we have to a total crypto market cap index. We strongly recommend that IC create a total crypto market cap index as these are by far the largest index in traditional equity markets. The investment literature is highly contradictory on what the optimal balanced crypto-equity portfolio should look like and focuses primarily on Bitcoin as the only crypto asset.

Performance

TBD after DG1 once the composition is finalized barring consultation with the PWG & EWG,

Size of Opportunity

Across all of the retirement indices, we predict the total AUM of the whole series to be the following:

Across all of the retirement indices, we predict the total AUM of the whole series to be the following:

Year AUM
1 $500 000 000
2 $1 000 000 000
3 $2 000 000 000

Most importantly the retirement funds would create another marketing and growth opportunity for the IC products.

Market & Customer Research

Target Customers

  1. Retail investors - From TradFi looking for a similar product to a Vanguard or Fidelity target retirement ETF.

  2. Passive investors - Looking for a broadly diversified crypto index fund, but not necessarily interested in retirement planning per se.

  3. Registered investment advisors - Looking to offer a comprehensive investment solution for their clients that encompasses both crypto and synthetic equity.

User stories

Jane is an experienced passive equity investor. She knows that a target retirement ETF will maximize her risk-adjusted returns over the course of her total investment horizon. She does not have the time, knowledge, or risk-appetite to purchase individual cryptos or even multiple crypto indices. She wants to hold a single token that will accommodate her long-term ā€œset it and forget itā€ approach. Sheā€™s 45 years old and plans to retire in 2040 but has a slightly higher risk tolerance so she buys TRI50 (Target Retirement Index 2050).

John is 18 years old and does not plan to retire until 2070. After suffering multiple rugs, exploits, MetaMask spoofs, and dust attacks, he has become very risk-averse. Consequently, he prefers to invest in stable yield coins but still wants some upside to remind him of his more carefree degen days. He chooses TRI20 (Target Retirement Index 2020) to match his conservative investment style.

Methodology

Initial Composition & Token Inclusion Criteria

Selection criteria for tokens

  • ERC-20 index tokens representing large-cap thematic sectors i.e DPI, MVI, BED, DATA, PAY, JPG, PDI
  • Mirrored equity indices i.e mSPY

Weightings

Weightings would be done according to a glide path that begins with risk-on crypto indices furthest from retirement and then progresses to primarily yield-bearing stablecoin indices and mirrored equity at retirement.

On-Chain liquidity analysis of underlying tokens

To be determined after DG1.

Maintenance

The index will be rebalanced monthly to rebalance as well as to update the allocations according to the glide-path formula.

Costs

Cost to customer

To be determined after DG1.

We anticipate the TRI indices to have a 2-3% management fee given the uniqueness and innovation of the project.

Rebalance frequency

Monthly

Manual Rebalance magnitude

We anticipate trading all of the positions each month due to the glide path.

Fee split

To be determined after DG1.

Meta / intrinsic productivity

Not applicable

Liquidity

To be determined.

Author Background and Commitment

@JosephKnecht is a methodologist at Index Coop and Founder of Article21

Michael Mtenga (@mringz) is co-lead BDWG and has been consistently contributing to the Index Coop since February 2021, contributions that have led to him being awarded Gold Owl within the DAO. His contributions include initiatives with PoolTogether, Perpetual Protocol & Rabbithole. Furthermore, he has been a key member within the business development working group (BDWG) creating and leading the DeFi Partnerships vertical. He also serves as the protocol ambassador for Maker DAO.

Article21 is committed to developing the next generation of crypto index strategies.

Marketing support / distribution / partnerships

The methodologist will promote the token/s. The methodologist will also be available for AMAs and interviews.

Revision history

Copyright

Copyright and related rights waived via CC0.

11 Likes
2

The DAO does not absolve us of duty of care. Because of our stated and percieved expertise any kind of guiding questionnaire crosses the line into (negligent) financial advisory. Ditch the guide in favour of a DYOR statement.

Def an opportunity worth exploring! be great to see some quantitative or even qualitative data to show in market demand.

5 Likes
3

That makes a lot of sense. Iā€™ll remove the questionnaire from the proposal tout suite.

I agree some market research would be very useful. Iā€™ll discuss with GMWG.

2 Likes
4

Love this, cool ideas. Thanks you two!

5 Likes
5

Update: The TRI series is on a temporary pause until more components pass DG2.

1 Like
6

I love the idea! With that said, here is some constructive feedback that could enhance things further. Target date funds started their journey in the early 2000s and really started to takeoff after the pension protection act of 2006. The intent was to follow a glide path that became more conservative over time as you approach your assumed retirement date. Hereā€™s how asset allocation models have been enhanced since then.

Risk based portfolios are designed to achieve a given level of return within a range of acceptable risk. They are designed to identify an investors targeted risk / return balance and engineer a constant portfolio mix to achieve this balance on a constant basis.

Suggestions:

-Consider risk-based models instead of target date funds - aggressive, moderately aggressive, moderate, moderately conservative, conservative. Five models create more simplification while also allowing individuals to look at risk from a different perspective. Target date funds put a strong emphasis on age assuming that you should be aggressive when youā€™re young and conservative when youā€™re oldā€¦ but that shouldnā€™t be the case. Risk is highly personalized and different for everyone. We could be the same age with different needs, expectations, and level of assets which makes risk based models more appropriate.

-Consider a tool to better assist those with picking a risk level - Riskalyze is a tool used within the financial advisor industry that can potentially be a great partner for allowing people to get a more accurate reading of their level of risk so they know which model to choose. I know the intent isnā€™t to provide advice. With that said, if we think of the larger target market for the IC products, we should build in a way that aligns for advisors and their clients.

3 Likes
7

Thanks for the feedback.

Iā€™ve tried to accommodate both a target date and a risk preference by suggesting customers can shift their effective target date. For example, if a customer plans to retire in 2050 but wants a higher risk portfolio she could invest in TRI60. One alternative could be to have the target date indices and then 3 additional static portfolios, eg, TRI Growth, TRI Balanced, and TRI Conservative.

My original post proposed exactly that: a risk-tolerance questionnaire to guide customers to the appropriate product. I was reminded though that weā€™re not allowed to give financial advice in any way, shape, or form, and so the questionnaire proposal was removed.

8

Excellent proposal! I have spent 10+ years structuring and investing in traditional index funds and my suggestion would be to include a mechanism for consistent dollar cost averaging, where a user has recurring investments on a defined cadence (optimally biweekly or monthly). This creates a long-term mindset and manages the psychology of market volatility.

9

Thanks. DCAā€™ing can only be done on the customer end and would be cost-prohibitive on L1 except for whales. There is an interesting related question of how to re-invest the yield.

10

No prob! Love what you came up with. Itā€™s definitely a needed product. Just had minor tweaks or enhancements to make it better for the masses. Target date funds are good as they are used in many 401k plans. But over the last 10 plus years, there had been a shift to risk based modelsā€¦. Aggressive to conservative. This is what I wanted to get across in my post.

Cheers,
Eric

1 Like
11

I agree. We plan to do market research for this project and we can ask if thereā€™s appetite for static portfolios as well as time-varying ones. Thanks for the useful feedback.

12

Thanks for this submission. Iā€™ve listed some constructive feedback below to help strengthen the proposal. We need more information to determine the attractiveness of this particular product (suite) and de-risk some key assumptions being made:

1. TDF interest outside of the 401(k) market
Target date funds (TDFs) are, as a few here have pointed out, very popular with 401(k) plans in the U.S. Iā€™ve personally invested in TDFs beforeā€¦but have never considered (nor would ever consider) purchasing one outside of a retirement plan. What evidence do you have that retail investors meaningfully purchase these products outside of 401(k) plans? (I have no clue but curious to hear what youā€™ve learned because this seems to be an important assumption underpinning success).

2. TDF interest specifically within the crypto market
There is an implicit assumption here is that because itā€™s a popular product outside of crypto, it will be popular within crypto. Are on-chain buyers - known for flipping JPGs, apeing, yield farming, and buying meme coins - actually looking en masse for retirement options on-chain? How do we know this?

3. Validating the need for several target date funds
The differences among these eight proposed index options is an assumption that there are a lot of users on-chain at varying ages that will be eyeing retirement at various points in the future. And that users would become more conservative over time. My hunch would be that the current market skews significantly towards younger users and also institutional funds, but maybe Iā€™m wrong and itā€™s more evenly split. What evidence do you have that the age distribution for on-chain retail buyers is more of a normal distribution to justify all these options?

4. Operational considerations
Having eight products seems like an operational nightmare. Weā€™d need sufficient critical mass from an AUM standpoint for eight products. There would be eight monthly rebalancings. This approach appears to fragment potential AUM, increases gas costs, and assumes that there is a critical mass of buyers of various ages that would be attracted to this concept. Iā€™d argue that requires a higher degree of validation.

5. De-risking the projections

How did you arrive at these figures? For comparison, our most popular product (DPI), which targets the promising, emerging theme of decentralized finance had around $200mm at the one-year mark - less than half of the AUM youā€™re projecting. As another data point, MVI currently has less than $50mm now seven months in - less than 10% what youā€™re projecting. These projections seem extremely aggressive.

6. The expected users

How many ā€œJohnā€'s do you think exist? The (likely capital-lite) 18-year-old once-degen-turned-stablecoin-only-investor? At age 18? And there are a lot of these Johnā€™s? I donā€™t buy it. This seems like an edge case.

Jane seems more reasonable, but I also donā€™t buy that there are many Janeā€™s that have moved on-chain (which indicates sheā€™s pretty sophisticated) but on the other hand needs retirement help through TDFs.

Recommendations

  1. To Ericā€™s point, move off of products centered on age and focus instead on the degree of risk. I was on the Investment strategy team at J.P. Morgan, and our model portfolios did exactly that. We had three high-level portfolios for new money: a) conservative, b) balanced, and c) aggressive. Thatā€™s it. Let people decide what level of risk makes sense for them, regardless of age.

  2. While I just recommended moving from eight to three products, Iā€™d go further. Start with just a balanced portfolio. If you do, this is exactly what @MrMadila proposed in a meta-index of Index Cooperative Products. If that is successful, we can launch conservative and more aggressive variants in the same way that $GMI followed the product-market fit of a product like $DPI.

  3. Nobody understands these products better than the Index Cooperative community. So as we think through what demand might look like, build a book of interest based on our community members. Create a survey to understand what % of the portfolio a member would (will?) allocate to this and what existing product it would pull from. For example, my portfolio today is INDEX, ETH, DPI, and airdrops. Assuming I agree with the asset allocation of the balanced portfolio, Iā€™d put alllllll my index exposure into this product because Iā€™d like to get exposure to MVI, DATA, PAY, and JPG. That would represent about 20% of my portfolio. Do that across the Coop. Youā€™ll be able to better answer the question ā€œDoes anyone care?ā€ If so, great (and get them to actually commit). If not, why not? And why would you expect anyone to care if the most vested index folks arenā€™t? This helps you better define the target market, assess the cannibalization effects this product will have on the rest of the product suite, and get a sense of market sizing. We should be doing this exercise all the time by the way for products. (cc @catjam @jdcook)

Hope this is helpful!

3 Likes
13

Hi Mike,

Many thanks for your in-depth feedback. Itā€™s like gold dust. Many of these are marketing questions that we donā€™t have the answers to yet but let me give some initial replies below ā€¦

There are no figures on how popular conventional TDFs are outside of retirement accounts (RAs). I presume theyā€™re less popular outside of RAs because of the tax inefficiency. However, in ICā€™s case I understand that the capital gains from rebalancing are tax-free so thereā€™s no disadvantage to having them outside of an RA. In fact it could be an advantage that the rebalancing is CGT-free versus a conventional TDF.

For benchmarking, crypto IRA provider iTrust Capital has 20,000 customers, $1B in AUM, $3.5B in volume. Kingdom Trust which markets self-directed IRAs has 100,000 customers and $12B crypto AUM. I donā€™t know how much of Galaxyā€™s funds sit in RAs but I suspect itā€™s sizable. From my personal investing perspective, investing in a crypto TDF and checking it only 40 years later is a mega-chad move.

Roughly half of DeFi users are over 30 (and so should be thinking about retirement planning) and probably a much higher fraction on a dollar basis. Twenty percent are over 40. These demographic surveys are helpful:

The costs and effort would need to be taken in context of the profitability which has not been modeled yet. As long as the funds meet our profitability hurdle rates, managing 8 funds in a series is just as expensive as managing 8 separate funds.

I would tactfully make the opposite point that developing similar funds in a series is a way to boost product development productivity. All of that said, I agree that the number of funds is too much and will need to be revisited.

For clarification, ~$500M AUM is across 8 funds so $62.5M per fund.

I took a simplistic top-down estimate. TDFs are ~2% of the total equity market. I took 2% of the total non-custodial crypto market (~$1.3T) and then divided that by a further factor of 50 for safe measure.

I understand that our launch AUMs are determined by our ability to raise the seed capital as opposed to the broader market demand. Without insight into our access to seed capital itā€™s hard to estimate the launch AUM other than top-down. Iā€™m open to any suggestions on how to estimate AUM bottom-up.

Sorry, John was a lighthearted example to show how the risk adjustment worked. I agree that ā€˜Janesā€™ and older are the target demographic. I think the DeFi age survey above is very useful

1,2. Iā€™m open to having 3 static portfolios or even starting that with a single static portfolio. Thereā€™s no cannibalism b/c the fees are stacked.
3. I strongly agree that all of our products need more market research and market input earlier. Itā€™s a major gap. @catjam made the point earlier that we should be doing our market research on our customers and not on ourselves, which I agree with. Weā€™re subject to all sorts of bias, both positive and negative, when evaluating our own products. I also suspect the IndexCoop demographic skews much younger and risk-tolerant than the general on-chain customer population which this product is directed towards.

Let me propose the following. This project is on a temporary pause anyway while we wait for the components (PDI, PAY, JPG) to hopefully get approved. Iā€™ll work with @catjam and GMWG to do some initial market research where we can probe the demand for a crypto TDF, static vs time-varying, etc. Does that sound like a sensible plan?

Thanks for the rigorous feedback. I live for this stuff.

Best,
J

3 Likes
14

How could the picture change if it were on an L2?

15

Thatā€™s a great point. The rebalancing costs drop dramatically; likewise, if weā€™ve socialized gas costs by then. The value decay depends on how much liquidity our products have on L2 but assuming itā€™s roughly the same as now on L1 I estimate the decay at 0.2-0.3%/month.

1 Like