I Analyzed 59 DeFi Yield Aggregators. Here Is What I Learned | by Max Yampolsky | Jan, 2023

The features and flaws of top DeFi yield products

The emergence of Decentralized Finance (DeFi) enabled a new generation of financial products to be built on blockchain, such as:

• Yield Aggregators — apps that are indexing protocols/pools across multiple chains and provide a user interface with an overview of all available farms (e.g., Beefy, Alpaca)
• Yield Optimizers — dApps that utilize existing protocol and use a particular strategy to maximize yield farmed in its pools (e.g., Convex, Yearn)
• Decentralized Funds & Vaults — dApps that allow asset managers to host their proprietary strategies and investors to subscribe to them (e.g., Token Sets, dHedge)
• CeDeFi Products — apps built on top of DeFi protocols but with centralized custody (most frequently single-asset staking). For example, Flint, Swissborg, Midas.

Yet, DeFi is still very young (less than 5 years old), and there are many things to improve. Here are some of the apparent flaws I noticed while looking at the industry:

All these popular tools, like DefiYield App, Beefy, Autofarm, Plasma, Alpaca, they provide you with an aggregate view of all the available pools to invest, which is great, no doubt.

But after looking at this overview, the next obvious question that comes to mind for any investor: “Which pool do I invest in?”

How do I decide the most optimal allocation for each pool? What’s the difference between one pool and another? What can go wrong?

These questions still need to be answered as the UX aspect is designed poorly across all apps.

These apps view pools from the context of chains but instead should do from the context of investment strategies.

Most importantly…

Showing the APY and TVL data is just a part of the picture. How about risk?

Risk is the most critical aspect to assess for any investment opportunity, yet most DeFi aggregators don’t communicate it properly.

Focusing simply on APY numbers is no longer a solution. Hopefully, more and more projects will realize this.

One particular project that positively impressed me is Exponential.fi.

They’ve built a comprehensive risk assessment system for each chain, protocol, and asset.

When reviewing a particular investment pool, you can get a much better grip on its risks.

The downside is that assessing and maintaining the databases requires a lot of manual work.

Can this be automated by AI with new tools like GPT-3? An idea for the next unicorn.

Fixed interest rates are still a fantasy in DeFi, and the yields of 95% protocols are flexible. For example, if today you invest in a pool with 40% APY, tomorrow it can be 20%, and in 7 days, drop to 10%.

The apps tend to display the APY as of the current moment, ignoring the variability aspect, giving a UX illusion that you will receive the original interest indefinitely.

Protocols with proprietary yield strategies (not farming-based) can be particularly nasty. They show the expected APY based on the strategy backtests (e.g., Don-Key.Finance), but the real APY turns out to be completely different.

For example, for a T-USDC-P-ETH put-selling strategy on Ribbon.Finance, the projected yield is +22.69%. Yet, when you look at the vault’s cumulative performance, the actual APY is -27.27%.

About 20–30% of all projects are not real DeFi projects but centralized apps built on top of DeFi.

Many popular investment apps like Flint, Swissborg, Umami, Osom Finance, Bincentive, DeFiato, deliver lucrative yields yet require users to sign up for an account on their platform and use in-app internal wallets to deposit and farm cryptocurrency.

But the critical question is: where is the yield coming from? For such apps, it is impossible to verify by design.

Investors in these products expose themselves to additional custodian risk, which may lead to situations like those with FTX, Celsius, Voyager, and, more recently, Midas.

Some apps, like Finoa and Staked.us, even require you to be an accredited investor to use them, which creates another UX barrier toward wider DeFi adoption.

DeFi has a graveyard of failed projects, with more and more joining the ranks, unable to survive the crypto winter.

Some projects were simply rug pulls, some were shut down due to an exploit/hack or mismanaged funds, and some just slowly died out and went off the grid. Some examples:

Execution is still a significant problem in DeFi, with the majority of projects failing to deliver on their promise.

DeFi is young but not new. Since it originated in 2018, many projects have tried to develop The DeFi liquidity aggregator/optimizer, but until now, there is no clear-cut winner on the market.

The problem with most apps is that they don’t deliver the user experience needed for an average joe to open a DeFi position. Hence, they fail to get meaningful traction.

Some projects died for various reasons, with the most common being execution. Many developer teams succeed in raising money but fail to deliver/build a meaningful product. Often, the issue is in the skills/competencies of a founding team and the lack of product development background.

But the main premise is that the opportunity and the market are there. DeFi is here to stay, and whoever can capture the yield market in its infancy — will remain the dominant player in the industry for years.