Silo-02, Milestone 1 - Evaluate Treasury Investment Strategy and Recommendations

**There is a vote currently active to redeem a portion of the treasury for YAM tokens that if passed will effect the evaluation and recommendations. Updates will occur when necessary.

Evaluating Treasury Investment Strategy and Recommendations

The Yam Re-Org that Ross and Designer have been working on for the last few months is a change to the operational structure of the DAO. Yam’s new operational structure is to allow anyone to request funding from the treasury to add value to Yam DAO and/or YAM Tokenholders.

Based on past behavior, managing an on-chain treasury is difficult and subject to many personal interpreations of best course. Considering the restrictions and limitations for Yam’s Treasury, this document hopes to define best course rules and guidelines in order to minimize friction for long term maintainace of Yam’s Treasury. There is NO PERFECT SOLUTION to invest and manage a treasury.

How should the treasury be organized to support the Yam DAO Organization?

There are 2 directives of the treasury:

  1. To support funding requests for approved Yam projects/silos
  2. To grow the value of the treasury to support directive #1.

Short history of the treasury

Understanding how we got here to determine how we move forward.
The treasury initially was created from the rebasing function of YAM token which has since been turned off.
Yam DAO and its treasury is entirely governed by YAM Tokenholders. The treasury has been utilized in a number of different ways:

  1. Investment Treasury
  • Current investment strategy of a moderate portfolio targeting < 0.75 Beta and < 0.5 Correlation.
  1. UMA Synthetic Liquidity providing, earning fees and UMA rewards
  2. Sushi DEX liquidity providing, earning fees and SUSHI rewards
  3. Index Coop DPI / ETH liquidity providing, earning fees and INDEX rewards
  4. Proof of concept for Yam DAO House treasury management solution
  5. Expenses for Yam DAO operations and contributors

Yam’s Treasury has a history of effectively utilizing assets to maintain the DAO but also to grow the treasury.

Current State of the Treasury

The current composition and size of the treasury can be seen here: YAM Finance

Numbers as of 7.15.2022
Current Value: $3.2 million

  • USDC (deposited in Yearn) - ~50%
  • ETH (some deposited in Yearn, some not) - ~30%
  • DPI (some LPing and some in treasury) - ~12%
  • UMA, INDEX, SUSHI, GTC - ~8%

Current inflows into the treasury:

  • Yearn Vault USDC Interest: $1.6m @ 1-2% APY
  • Yearn Vault ETH/stETH Interest: $165k @ 4-7% APY
  • xSushi: $62k @ 1.89% APY
  • Total Estimated Inflows: ~ $24k per year

Current outflows from treasury:

  • Yam Design Studios for June: $10200 70% USDC / 30% YAM
  • Specific Architectures Ross for June: $12,250 70% USDC / 30% YAM
  • Mona Community Management for June: $1,666 100% YAM
  • Designer Silo for June: $10,200 70% USDC / 30% YAM
  • 0xE Developer Compensation for June: $14,167 70% USDC / 30% YAM
  • Subscriptions, backend costs and gas fees for June:
  • Total outflows for June: ~$33k USDC / ~16k YAM

We have a healthy treasury utilizing ~50% of its assets earning interest.
Currently there is no active or in-development project to earn revenues from.

Current investment metrics and strategy. Current investment strategy of a moderate portfolio targeting < 0.75 Beta and < 0.5 Correlation.

  • As of 7.13.22, Beta = 0.45 and Correlation = 0.38
  • Beta / Correlation fall well below targets
  • For complete list of assets refer to yam.finance dashboard

Review of current investment strategy

Investment strategy targeting a specific beta and correlation initially was created is generally a good metric to assess risk for non-crypto portfolios. Since the creation of Yam’s treasury and investment strategy it has become apparent that this method of assessing risk is not the most optimal. A few reasons why:

  1. Crypto assets in general are highly correlated therefor using beta and correlation with respect to ETH as a metric is not as helpful as originally intended. For example, since most assets that an on-chain treasury can acquire either very close in beta/correlation to ETH or a stablecoin, so it does not provide any quality information on which assets to hold in the treasury.
  2. Beta/Correlation does not account for liquidity, inflation, potential revenue and etc of tokens.
  3. There are crypto primitives that allow for hedging risk but the hedging space has not fully matured and carry additional risks beyond the financial instrument (liquidity risk, smart contract risk, ect)
  4. Rebalancing assets becomes a hard question to answer. If the treasury needs to rebalance and there are multiple ways to do it, which assets do we sell, which do we buy?
  5. Original investment strategy relied on yield farming which many of the farms have ended rewards. Original forum post: YIP: Investment Portfolio Strategy Polling / Voting

In addition, there are several limitations of an on-chain governed treasury:

  1. Any transactions that involve the treasury require someone familiar with the governance module to write the code to interact with the treasury.
  2. Swaps require a contract that can interface the treasury with a DEX or Aggregator DEX. Currently we are limited to Uniswap V2 DEX because the contract has not been created to interface with others.
  3. The time lock on interacting with the treasury is 5 days after a 2 day on-chain vote approval. There is no immediate asset to treasury assets, therefor smart contract risk is of greater concern than normal.

Recommendation on Investment Strategy

My recommendation is to simplify investment strategy, allocation model and use rebalancing as a tool to manage risk. Using sushiHOUSE investment strategy and allocation model from sushiHOUSE - A Treasury Management Proposal - Proposals - SushiSwap

Using the same investment strategy risk level of 3, the Investment Strategy would target allocating 55% ETH 20% BTC and 20% Stablecoin 5% DPI.

VOTE on Investment Strategy

Current model targeting beta and correlation. In addition:

  1. Create framework on which assets to be kept in treasury and how to rebalance to maintain targets.

vs

Use sushiHOUSE investment strategy and allocation model (above) with risk level 3 - Medium. Strategy would target allocating 55% ETH / 20% BTC / 20% Stablecoin / 5% DPI. Additional rules and guidelines to maintain efficiency:

  1. Investment strategy does not include operations reserve which would be a 12 times monthly expenses.
  2. Assets that do not fit this model will be disposed of as long as slippage doing so does not exceed xxx% otherwise it will be kept in treasury but ignored.
  3. Rebalancing trigger limits will be set at ± 10% allocation. All assets will be sold/bought back to stated allocations when triggered. Rebalancing can be done at most once every 30 days.

!!VOTE!!

  • Current Model = Beta / Correlation Target
  • Allocation Model = 55% ETH / 20% BTC / 20% Stable / 5% DPI
  • Seek another alternative
  • Don’t do anything with treasury

0 voters

Operations Reserve

The Yam treasury should be split into an operations reserve and the rest into the investment fund. Operations reserve should accommodate a runway of 12 months (subject to vote) in a low risk yielding stablecoin. Recommended general guidelines to easily maintain reserve:

  • If the reserves fall under 3 months of expenses, the operations reserve will be refilled to 12 months runway and then a treasury rebalance will occur.
  • Any project or silos that require funding in excess of 3 months of expenses will require additional approval and funding will come from the investment fund. Anything funding requests under 3 months of expenses can come from the operations reserve.

This process is similar to many corporate accounts where payroll and expenses are kept out of the at risk investment accounts.

Links and Additional Referrence Materials

1 Like

Hey Feddas. Lots of good stuff here. I have some comments and questions:

in #1, it is unclear what “investment treasury” means in this context.

#3 isn’t accurate. The DAO does not provide liquidity or earn fees. The DAO earns Sushi Tokens in return for paying depositors YAM to LP.

It would be good to specify which of these are still active (only 1, 3, and 6).


It would be very helpful if you could give a quick rundown of what beta and correlation measure. Why are they used and what do different levels mean? what do the current beta and correlation numbers mean for the treasury? I.e. a higher beta and correlation would mean that our treasury would perform more similarly to ETH in general, and a low beta would mean it would perform more differently (although more similar to what is not always clear).


Can you elaborate on this? I’m not sure I agree with the premise as written. What smart contract risk?


It is unclear what the differences between the 2 options are. What would the allocation of the “current model” be? What is the beta and correlation of the sushihouse strategy? As I go to make a decision, I don’t know how the information in the first part of the post helps inform my decision in the vote.

Why is the sushiHouse model being proposed? Why the “3” risk / exposure level? Given that we aren’t using an auto-rebalancing portfolio manager, is a 5% allocation of different DeFi assets really doing anything for us, or the treasury health? Why add WBTC instead of adjusting ETH and stables to account for similar metrics given the significant overlap in correlation between ETH and BTC?


I agree that we need to think about the runway in all of this but I don’t think this analysis is complete.

Earlier you described the directives of the treasury:

If we agree that 1 is the main directive, then we should be designing our strategy, and runway numbers, to account for it. This support should be assumed to be what we expect if the DAO is growing and building projects. The current runway is 400k/yr in stables (according to your numbers above). This is with 4 people being paid (the fewest the DAO has ever had). This falls below the 20% of the treasury in stables as listed in your sushiHouse metrics ($~650k). But if the DAO does find a product that needs development work, the runway figure above is going to grow, potentially significantly (2-3x+) This would mean that the stablecoin allocation would also have to grow, exceeding the 20% target and breaking the strategy.

If “supporting building” is the purpose of the treasury, then I would prefer to see a more conservative portfolio strategy to cap downside risk.

Yes you are correct.

There’s a time delay (at min 6 days) when interacting with any assets inside the treasury, which magnifies any additional smart contract risk. ie. If funds needed to be withdrawn from Yearn vault due to a smart contract bug it would take us much longer due to the time delay which increases risk.

Risk 3 because that’s the most similar risk model to our current approved risk model. The current risk model has worked well, it’s participated in the upside of the market and protected the downside. Now it needs to be rebalanced and maintained which hasn’t been done.

While there are an infinite possibilities on what to do / what assets we should hold / what risk we should take / how we should define risk / why DPI vs WBTC vs ETH / why 10% variance ect. The goal of this is to create a framework that specifically fits for Yam DAO that can easily be used to manage the current risk profile of the treasury and NOT to figure out the infinite number of possibilities on portfolio management.

I offer a simple and reasonable solution that will allow for very easy and low cost maintenance of Yam’s treasury given it’s limitations.

Currently we have no active project aside from things that are to “repair” Yam. Do we have any in the foreseeable future? No. While I would like to plan for a significant need of the treasury to build out fantastic projects at Yam, it isn’t currently necessary to design the strategy around something non-existent. It is smarter and more prudent to build a flexible system that can account for a change in the future. The runway is calculated in a multiple of monthly expenses, therefor if the expenses spike due to a new project, when the rebalance is triggered, it will account for the new higher expenses. Viola, it works both with lower and higher monthly expense. Ross before we go back and forth on this, let’s take into account that I worked professionally as a financial planner.

Feddas

Additional Recommendations

Yield Farming

Currently the treasury is utilizing USDC and stETH vault.
USDC ~$1,500,000 earning ~ 0-2%
stETH ~$200,000 earning ~ 4-7%

Total APY revenue approx: $8,000 to $44,000

Adding Yearn Vault 3 Crypto

I recommend including the Yearn Vault 3 Crypto (BTC/ETH/USDT) earning 5-9% APY.
This strategy uses a convex reinvest including the LP swap fees rewards between the 3 assets. No leverage, uses well established protocols to earn fees and yield.

Applying to Allocation Strategy

If we were to adopt the new strategy of 55% ETH 20% BTC and 20% Stablecoin 5% DPI and add Yearn Vault 3 Crypto (BTC/ETH/USDT).
The assets could be allocated:
stETH = 35% ETH
3 Crypto = 20% ETH / 20% BTC / 20% Stable USDT

At a $3,000,000 treasury the allocations:
stETH = $1,050,000 earning 4-7% APY
3Crypto = $1,800,000 earning 5-9% APY
Approximate APY revenue $132,000 to $235,500 a year

Yearn Vault Links

3 Crypto: yearn.finance
USDC: yearn.finance
stETH: yearn.finance

Protocol Owned Liquidity

Ross has written a few forum posts on protocol owned liquidity (POL):

This would be valuable to Yam because:

  1. It would allow YAM to stop offering liquidity farming rewards.
  2. Allows the treasury to earn trading fees.

My recommendation for the simplest solution that requires no maintaince:

  1. Currently TVL in YAM liquidity is ~$615,000. We could use $310,000 of the treasury’s ETH and mint an equalivalent amount of YAM to be paired in a Uniswap LP pool. Whatever amount of YAM minted would need to be burned if the Protocol Owned Liquidity were ever unwound.
  2. Discontinue liquidity farming rewards.
  3. The POL would not be spent or utilized in other ways. It is only created for providing a liquid market for YAM.

I think we both agree that a simplified treasury strategy is the way to go. The question is what version of a simplified treasury strategy is the one we should go with. I don’t want to “figure out the infinite number of possibilities on portfolio management”, but I do want to come to consensus about what “specifically fits for Yam DAO that can easily be used to manage the current desired risk profile of the treasury”.

I’m not asking for every option to be considered, but when an option is considered (as you are doing with the 2 options that you propose), it should be explained as to why different assets are included. Because there are infinite options, the choice that we make should tell a story about what we want to achieve, and then show how it does so.

The fact of the matter is you do have an opinion on how the treasury should be allocated. That is fine. We all do. My concern is that this opinion is being obfuscated with language about risk factors, metrics, and rebalancing strategies. As I understand it, the treasury could have a solid rebalancing procedure whether it is aggressive or conservative. Some assets are easy to rebalance, others aren’t. Understanding these parameters helps us limit our options, which is a good thing.

I’m actually ok with a more aggressive strategy if that is what is being proposed. But I would also like to know how that strategy compares to our baseline (what we have). We haven’t rebalanced the treasury in a year, so it would be good to see what the treasury would look like if we had been rebalancing at +/- 10% with our current strategy. And then something similar for the sushi “allocation model”.

Without that it is very hard to know what the differences are between the 2 options. Maybe one is better at one thing but worse at another. right now it still feels like a black box, and the only reassurance that we have gotten is “the current risk model has worked well” even though we essentially haven’t been updating it. And “trust me I’m a financial advisor.”

So long story short: I would like to a story about why choose a particular strategy and some data backing it up. Your answer at the end of your first response to me does this and upon further thought I pretty much agree with it. But I need that story to make a decision and I don’t think I’m alone in that.

Assuming that a 55/20/20/5 strategy is what we go with, this seems solid.

USDT isn’t my favorite, but they seem to have weathered every storm thrown at them. A real problem with USDT would cause all sorts of other problems and would probably be impossible to avoid the fallout even if you didn’t hold any.

That 5% DPI still feels like added complexity for little return. 60/20/20/20 (40% stETH, 20ETH/20WBTC/20USDT) would mean that everything could be in Yearn and the metrics profile should be mostly the same.

I see what you are saying now. I would argue this is a bit of a red herring though since the chance that we notice any exploit and pull funds out faster than yearn does and pauses or closes a pool seems unlikely. Our funds are mostly in their hands whether our timelock is 1 day, 5 days, or 30 days.

The scope of this project isn’t to change the risk profile of the treasury but to create a maintenance framework around Yam’s treasury that accounts for the current limitations and requirements.

Yes, the way that the treasury is allocated is an opinion and subjective. Given our current assets / allocation / risk the model that I’ve proposed is closest to what we currently are aiming for.

This silo isn’t here to create a story, it is here a framework for maintaining an already approved risk strategy by first simplifying it, then creating guides and tools so that anyone would be able to rebalance the treasury in the future.

It is not a red herring if your comparing to a multisig or something else without a timelock. In addition risk is greater the longer the timelock, it is unquantifiable but still true.