Magpie started long ago as a project built on the then-new Wombat Exchange, similar to the Curve/Convex combo but on Binance Chain. Back then, I chose Wombex, Magpie's main competitor at the time. I remember participating in the ICO on PancakeSwap and holding onto my tokens like a true long-term investor, even though I could have cashed out with huge profits. DeFi was even more inefficient back then.

Recently, Magpie evolved from just a layer on top of Wombat into a Mega DAO. For those not in the loop, Magpie keeps a portion of tokens when launching new sub-DAOs to ensure they have the deciding vote in their sub-DAO decisions. Now, the token represents a basket of various projects, and each project generates income for vlMGP holders, which is the staked version of the token.

I love these kinds of projects—investing in a token that lets you earn income from one or multiple projects. With MGP, you're literally investing in various DeFi sectors.

Determining the exact value of the MGP token is tricky, but let's give it a shot.

Is MGP overvalued ?

Currently, the MGP token offers just 10% APR, which is quite low for a highly volatile token. This return comes from the yields provided by PNP, CKP, and RDP.

Penpie is the biggest income contributor right now, with Cakepie's returns growing significantly, while Radpie adds just a bit.

As a potential investor, I'd want at least 30-40% APR for taking on such risks. So, does this mean the token is overvalued from my point of view right now ?

The image shows that Penpie provides the most income, but Cakepie is currently the most efficient project. In May, Cakepie almost matched Penpie's returns, even though only 16% of CKP tokens are vested. As more tokens vest, the income will increase—assuming Pancake keeps up its performance.

By the way, the image doesn't show Wombat. It's unclear where the returns from this project are. Either they haven't added it to the statistics yet, or there are some complicated relationships between the projects, and we shouldn't count on this income.

Additionally, we see three projects in their very early stages that haven't fully launched yet. While Campie and Eigenpie are built on well-known and serious projects, I have no idea what Listapie is or how it was chosen by the team. As a potential MGP investor, I’d like to understand the criteria for selecting projects under subDAOs, but right now, I don't have that clarity at all.

Will these three projects add enough returns to bring MGP's APR to my desired 30-40% ? I highly doubt it. So, does this mean that the MGP token is overvalued ?

Classic researchers might ask, where are the serious studies on FDV, P/E, market cap, supply, and other boring ratios and figures ? Let me save you some time by mentioning that only 38% of the tokens have been issued so far, of which 40% are staked as vlMGP. Yes, we are in for a long period of MGP token inflation, but this will be offset by the increase in vested tokens from subDAOs, which will start generating income.

I believe we should look in a completely different direction. Maybe it's not that MGP is overvalued, but rather that the tokens of its subDAOs not generating enough income? If so, why is that ?

Inefficiency of MGP Votes

And here we come to what I believe is the main inefficiency of the Magpie project as a MegaDAO.

If we look at the current valuation of the PNP token, we'll see that the vePENDLE/vlPNP ratio is 3.55. This means that one vlPNP token contains the value of 3.55 vePENDLE tokens, which doesn't impact the PNP token price significantly. This is normal because it’s offset by the higher yields in the bribe market. I held growing PENDLE tokens for a long time but only earned 20% annually. Then, I switched to a dividend income strategy and moved to the more undervalued PNP compared to PENDLE.

As a PNP-maxi, I'm not particularly worried about the PNP token price as long as I'm getting 70-100% APR on the bribe market, and sometimes even more, because I can choose the most profitable bribes.

This is exactly what Magpie can't do.

Or rather, it can but doesn’t, because it needs to vote for its subDAO pools. They can't just allocate all votes for a week to a pool like Renzo ezETH, which offers 100% APY. They have to vote for Penpie mPENDLE, which currently gives 41% and usually not much more than 25%.

I can't definitively say that they only vote with vlPNP tokens for their own pool, as I couldn’t find a way to see the exact distribution. However, based on the rewards they receive in their treasury wallet, most of the rewards come from the Penpie mPENDLE pool. The same applies to Cakepie.

So there's a dilemma: you could potentially earn almost twice as much with your votes, but you can't because a lot is tied to the mPENDLE pool. This includes the mechanism for maintaining the mPENDLE/PENDLE peg and the PNP buyback, which subsequently becomes the bribe rewards from this pool that you earn income from.

The Future of Magpie MegaDAO

I believe that addressing this inefficiency must be the project's main goal to increase its income.

Simply put, the project should vote for the 5-7 most profitable pools each week. Then, they should convert a large portion of the profits from various tokens into PNP, CKP, RDN, and other subDAO tokens to pay rewards to vlMGP holders. The remaining profits should be converted and distributed for various needs like maintaining the peg, and so on. Sounds simple, but technically, my God, I don't envy the developers.

There's another option: paying vlMGP holders in the tokens and chains as they were received from voting rewards, similar to what Radiant does. For example, "We've earned 24 USDC, 13 USDT, 0.01 ETH, 0.007 wstETH for you. Take them as they are, or you can use a one-click compound button to allocate the tokens to the respective pools in the corresponding subDAOs." This is technically easier but might have mixed reactions from users.

Redirecting votes to more profitable pools opens up another strategy: Buybacks!

Imagine Magpie starts using part of its bribe income to buy back subDAO tokens and store them in its treasury, generating even more income for vlMGP stakers. They could create a bribe market where vlMGP holders vote on which tokens to buy back from the market.

Even I would start buying MGP to vote for PNP buybacks. Imagine, Eigenpie is earning profits, and you have the power to vote so that these earnings are used to buy back PNP. Now, imagine if Pancake, for example, decides to buy MGP to vote for CKP buybacks. What would happen then ?

This is what Magpie would look like in the perfect world, or but maybe I'm overthinking it and getting too carried away.

After all, there's a more likely scenario.

There’s also the possibility that Magpie might ignore this optimization and focus solely on launching new subDAOs. In that case, the project would need to offer something else to increase MGP token returns.

After all, why would I buy and lock MGP for a 10% return in PNP, CKP, and RDP tokens when I can manually buy those tokens in the same or different proportions and get 5-7 times more return by voting, just like I do now with PNP ?

Only time will tell which path Magpie MegaDAO will choose. For now, the project has finally broken free from Wombat, which was holding back its growth, and it has an interesting future ahead.