Since I last looked at Alchemix the price has fallen from $312 to $18 and the ETH ratio has fallen from .155 to .01. There’s no way to sugarcoat that. It hurts to see a good project getting hammered by 2020 era tokenomic designs but the price chart we see for ALCX is pretty typical of these inflationary incentive design tokens. My conclusion at the time was use the platform and give ALCX no thought. I think that was fair criticism then and my prediction has played out fairly accurately.
But some things have changed regarding ALCX. The issuance has fallen from 19874 per week to 7270 with a corresponding 95% price drop since I last wrote. That brings the most pessimistic ALCX sell pressure down to $130k per week from $6.2M. It takes almost a year to emit as much dollar denominated ALCX sell pressure as it used to emit in a week.
While the inflation schedule has been playing out those ALCX emissions have been used to buy up a sizable treasury of yield bearing assets that have indirectly turned ALCX into a sort of index fund in its own right. According to this beautiful stats page, if you discount the ALCX holdings their treasury has amassed $15.05M in interest bearing assets that are responsible for about $2.5M in income. This entire category of income did not exist in my last valuation.
Next, they have turned what used to be the idle assets in the Transmuter system into actively managed assets that are being used to yield farm until they are actively needed by the Transmuter. This is called their Elixir system and it resembles the Frax AMO. Again, using that stats dashboard as a source of truth, this brings in another $3.2M in revenue per year. 2/3rds of this revenue goes to the treasury so that’s roughly another $2.2M in revenue controlled by the governance token.
Lastly, they are in the final stages of a governance system revamp which will change the utility and dividend structure of the ALCX token. That’s a lot of changes. All told, I think it’s worth having another look here.
For those unfamiliar with their system I’d suggest starting with my old Transmuter Boost post. If you’d like a good laugh you can also revisit my eponymous Double Logris post while you’re in the archives. TLDR: Alchemix makes revenue from the interest people forfeit when self-repaying their loans, from liquidity farming with the sum of all assets forfeited to date, and by managing its treasury. The treasury skims 10% of the harvests and additionally receives the assets the ALCX inflation is being dumped for. ALCX holders control that treasury. The harvests in total account for a few hundred thousand of protocol revenue per year and therefore only a few tens of thousands of treasury revenue. There are also some Yearn kickbacks for an affiliate program but those are just a fraction of the harvest data I’m already going to be glossing over. Over time the story for Alchemix evaluation has quietly transmuted from one of TVL and harvests to one of a managed fund with a valuation dominated by its holdings.
TLDR: Alchemix makes revenue from the interest people forfeit when self-repaying their loans, from liquidity farming with the sum of all assets forfeited to date, and by managing its treasury. The treasury skims 10% of the harvests and additionally receives the assets the ALCX inflation is being dumped for. ALCX holders control that treasury.
Combined we’re looking at $2.5M liquidity farming with the treasury. $2.2M from liquidity farming with the Elixir funds and some misc revenue from harvests and yearn kickbacks. Putting our governance managed revenue at north of $4.7M. My belief is that a governance token should always be worth more than the assets it manages. If not, someone can just buy controlling interest in the Alchemix DAO and vote for a rage-quit function that burns ALCX for a pro-rata share of the treasury. The implementation of this belief when calculating PE is to subtract the treasury asset value from the market cap. Adjusted market cap: $32.41M-$20.83M=$11.58M
If inflation was 0 you’d have a justified market cap of about $94M (standard PE 20 assumption) based on the $4.7M revenue. Against the adjusted market cap that gives ALCX PE before costs of 2.5. For ALCX, things look promising at face value so far.
Naturally, that’s half the picture. Let’s have a look at those expenses. According to their last financial report the DAO is approved to spend $450k per quarter, $1.8M per year. That’s a good chunk of the profit right there. There’s going to be another 222k ALCX issued in the next 12 months which is another $4M in inflation. Pessimistically, you’d assume that is all going to be sold, but there is some nuance to that. If we look at emission weights we can see that 64.5% of issuance is going to bribes for Curve pools of alAssets. You’d normally assume all of that would be sold but the Elixir AMO is a significant holder of those pools. So, the protocol is just paying itself that ALCX. For example: the AMO has $35.4M in the alUSD/FraxBP pool. The total pool is $59.6M, so 60% of those emissions go back to the protocol. Some napkin math later we can discount the emissions by about 40% because some of it what is emitted isn’t going to market even under pessimistic assumptions.
Bringing us to a treasury revenue of $4.7M, an adjusted market cap of $11.58M, and adjusted costs of $1.8M+1.6M. Given many assumptions about relative asset prices holding that gives them a $1.3M net revenue and a PE of 8.9 after accounting for administration and discounted inflation costs. A remarkably different valuation than when I last wrote about it.
Given many assumptions about relative asset prices holding that gives them a $1.3M net revenue and a PE of 8.9 after accounting for administration and discounted inflation costs. A remarkably different valuation than when I last wrote about it.
Onto the growth story, the long term success of Alchemix should be from their harvests, not just from liquidity farming with their treasury. This income is roughly proportional to their TVL.
- How fast can Alchemix TVL can expand?
- Why is the growth limited at all?
- Wouldn’t more TVL usually be a good thing?
- Why cap it?
Well, the usual flow of a vault user is to deposit assets, then mint alAssets, then sell them. That’s why they are taking a loan after all. Selling the alAsset is the expected use case. So, more users = more sell pressure. This wrecks the price of the alAsset (which is bad). So the TVL of vaults is limited in order to protect the alAsset peg. Fundamentally the biggest headwind of Alchemix is their limited ability to protect the alAsset peg while growing quickly enough. It’s not for a lack of product market fit, they literally have too many customers banging at the gates to fit them all in the vault capacity.
Let’s explore this a bit further. Let’s take ALCX inflation and Curve incentives entirely out of the picture for a moment. How does a sustainable version of this protocol work once all the ALCX has been emitted? Today the Elixir system has about $68M TVL. Those are LP positions owned by the protocol on places like Curve. Let’s use that number for some napkin math. How much TVL can you support with 68M of protocol owned liquidity? Well, that depends on how strongly you want to defend the peg. For the sake of argument let’s say that on some Curve pool there is a 3:1 alAsset/asset ratio at a .97 ratio due to liquidity compression. In that case, their 68M consists of about 17M of underlying. The other 51M is the alAsset they can support. It is literally being held in Curve by the Elixir system and requires no yield.
At a 1:2 loan-to-value ratio that means they can support 102M of TVL. If the underlying is earning 3% yield that means they are earning about $3.06M in underlying asset protocol profit. This profit would allow them to expand the alAsset market cap by $9M and in turn that means they can expand the TVL by 18M. More realistic numbers I’ve run put the YoY growth target of Alchemix at 15-25%. For how low the valuation of Alchemix is, people are expecting it to be compared to moonshots that can explode with an exponential base of 10 (e.g. 10^3). Instead it has an exponential base of 1.15-1.25. That puts it in-line with more mature tech companies like Microsoft. At best, with the current designs, this is going to be a marathon not an explosive success. Very little about this calculation changes due to their governance revamp.
How can they get this number up? Per unit of protocol profit, how can you maximize the amount of alAsset you can support? One uncomfortable way is to allow the peg to slip more. That hurts users and only works a little bit further until you hit the really bad end of the liquidity compression system on Curve. Another possibility is to use leverage. That would involve integrating the Elixir system with something like Gearbox that would allow them to borrow 3x the collateral and make proportionally more LP. I’d have to run the numbers more on the viability of that strategy from a gearbox interest cost vs vault profit. Keep in mind there is like a 6x ratio on Elixir liquidity to interest bearing TVL if the Curve LP ratio is at 3:1 so this could turn out to be net profitable even after paying to borrow the underlying asset. Naturally this comes with doomsday scenarios where they get their entire system liquidated and it all ends in tears. The last option is to flip the system somehow and find another way to protect the peg that doesn’t involve liquidity depth. For example they could copy mechanisms from Liquity or Frax as Cat In a Box has done. But it’s not as though LUSD is a success story in terms of adoption and growth right?
The last thing I’d like to touch on is what the coming governance revamp does change:
1) It removes the cost of ALCX/ETH liquidity from the issuance which frees that issuance up to acquire more strategic assets like CRV. It pushes this cost onto ALCX governance participants.
2) It adds profit mechanisms to ALCX holders. This offsets the costs of 1). Although, speaking candidly, dividends/buybacks aren’t strictly required as long as the treasury continues to grow and is on track to surpass the ALCX market cap. Eventually the price of ALCX would just be floored by hostile takeover risks threatening the treasury like we’ve seen happen to Digix Gold for example.
3) There is this Flux system which monetizes the timelock portion of the ve model. It’s frankly brilliant. Though I wish they went further with it then they did.
Final conclusion, this asset is underpriced compared to traditional tech equities and frankly quite beat up by inflationary pressures. It can and probably will dip a little further on the ETH ratio just because of bear market conditions and the fact that the token and their revenue streams are largely misunderstood. That said, ALCX becomes a better buy each day. Almost every calculation above becomes more favorable with time. The assets in the treasury are up 84.3% in Q1 and by token count are growing about 20% a year. The inflation schedule is approaching its minimum issuance February of next year. By February 2024, using todays pricing data, treasury would grow another 10% which would reduce the treasury-adjusted market cap by about $2M to about $9.5M. Inflation would be at 2200 ALCX per week or $2M a year down from $4M right now. That gives you an cost-adjusted income of $2.1M. If the market cap were to hold rather than grow in line with inflation that means they would be looking at a PE of 4.5 with basically nothing changing except the date. That’s too good to pass up in my mind. As the market cap approaches the treasury value I’ll be DCAing into this starting now.