What is the Fat Protocol Thesis?
What are Appchains?
Benefits of Sovereignty
Inter-Blockchain Communication (IBC)
Let’s Talk About Security
Interchain Network Effect
It takes “the future will be multichain” one step further; the future will be interchain. A future where app developers can build their own blockchain with specific infrastructure for their app’s needs. Where apps can retain all the value they create rather than “paying rent” to a layer 1. Where blockchains can be connected but stay self-securing, sovereign networks.
There’s so many facets to Cosmos, so this piece will break them all down and explore specific cases of apps warming up to the Appchain Thesis.
NOTE: Ethereum is used throughout as the easiest example of a general purpose blockchain. No hate, the love is mutual.
First, let’s talk about the customizability problem. To build the most powerful apps, developers need to be able to customize at every level including the network level. This isn’t possible on general purpose chains like Ethereum, so how can apps reach their full potential? Most apps would benefit from a chain built for their explicit purpose, built for the application itself. Not a “one-size-fits-all” blockchain.
This leads to the next point, the rent problem. If applications create value, they should be able to retain it. All of it. On general purpose chains, applications leak a lot of value they could be keeping. Whether it be in MEV or gas fees paid by users, apps basically “pay rent” to networks they build on.
Another issue is the scalability problem. General purpose chains offer no possible way for some apps to scale for their purpose. Orderbook DEXes are a good example, which generally require a network that handles thousands of TPS and extremely low gas fees. This is only achievable on existing networks by reworking their core infrastructure.
Last is the interoperability problem, which most blockchains today suffer from. Apps cannot just deploy on one blockchain and be usable from anywhere. They have to deploy copies of themselves on any chain they want to be usable on. This fragments liquidity for the same app across several chains and leaves users with two choices: wait for apps to deploy on their chain, or use a third party bridge to transfer assets to a new network. Bridges are an infamous point of failure for general purpose chains.
With Cosmos, general purpose chains aren’t the only option anymore. App developers get limitless customizability granted by the Cosmos SDK. They can create their own chains that scale to their exact standards and support any feature on their application.
Rather than one blockchain, Cosmos is a network of many app-specific blockchains or “appchains”. Appchains do not have to pay rent and can retain all of the value they create. Apps can choose which token(s) gas fees are paid in, and MEV can be captured as protocol revenue.
Features like customizability, sovereignty, and Tendermint consensus greatly improve scalability for Cosmos blockchains. Being able to create custom networks plays a huge role in scaling apps, especially when virtually any feature can be added.
The Inter-Blockchain Communication protocol (IBC) connects different Cosmos blockchains together. IBC-enabled chains can share users and liquidity in the same way apps on a single blockchain do. This largely solves the fragmented liquidity issue as users can seamlessly move across the interchain. Cosmos tech and IBC plan to eventually abstract away what it means to build an app vs. build a chain--they will be one and the same.
One way to view a blockchain is splitting it into two layers: the base protocol and the application layer on top. The Fat Protocol Thesis is the idea that the base layer protocol is where most value should accrue, contrary to how something like the Web works.
Ethereum is the easiest example of something built around this thesis. It provides the most decentralized base layer for building applications, and users pay $ETH gas fees to use apps on the network.
The Fat Protocol Thesis sees general purpose blockchains as the final settlement layers in the long run, with some believers even suggesting “there will only be one winner”. But this seems really far-fetched.
The Appchain Thesis argues that the userbase of a network is the collective of app users on that network. That apps are largely the source of value creation and attract users to the network, not the other way around.
Value in Cosmos goes to the apps, where it accrues in the first place.
Cosmos is built on the belief there will be many, many blockchains each customized to scale and operate around a certain application’s needs. Not the World Computer, but interconnected “Community Computers”.
Application-specific blockchains (appchains) are networks built for some specific purpose. They are the Community Computers. Developers can create custom networks that precisely meet all of their their app’s needs. Examples of appchains are DEX appchains, lending appchains, infrastructure appchains, NFT appchains, gaming appchains, etc.
Some unique features of apps as appchains are…
fully customizable, especially at the network level
totally sovereign, with optional shared security
apps are able to retain all value, including gas fees and MEV
share users and liquidity with all other IBC chains (interoperable)
Bitcoin is often perceived as the first appchain ever created, and it definitely has all the features. Bitcoin is purpose built as a decentralized payment network, and its infrastructure (e.g. Lightning Network) is constantly developing to improve this function.
The Cosmos SDK is a toolkit for building custom Proof-of-Stake (PoS) blockchains that can either be general purpose or app-specific. All SDK chains run the Tendermint PoS consensus engine. The SDK and pre-made consensus engine make it so developers don’t have to build a chain completely from scratch.
They can customize this stack to their own standards, and even run sibling PoS consensus mechanisms like Proof of Authority (PoA). This can be customized further too, like BNB Chain’s Proof of Staked Authority (PoSA) or Berachain’s novel “Proof of Liquidity” (PoL).
A hybrid of a general purpose and app-specific chain can exist too, such as “sector-specific” blockchains designed for entire sectors of applications. Sei Network is a nice example with its “orderbook-specific” chain, where any orderbook-based DeFi apps can build on Sei and plug into its universal orderbook shared by all other Sei apps. Sei validators also process transactions in a way that prevents front-running and ensures sub-second finality for traders. This new type of blockchain is only feasible thanks to the SDK.
Last year in September Vitalik explained the buzz around the “layer 3” vision for Ethereum. L3s are are still very loosely defined, but they are basically customizable app-specific rollups built on top of Ethereum L2s. These rollups can be infrastructure for the L2 like privacy layers, or they can be a custom rollup built specifically for some dApp.
Depending on the architecture, L3s assume the security of L2s, all of which today rely on centralized sequencers. Since they are not sovereign networks like Cosmos appchains, the uptime of these L3s will still rely on the main chain.
It’s easy to see the benefits of the SDK to app developers. The SDK = customizability, and this customizability lets developers do anything to make more scalable, secure, and efficient applications. The stars are the limit.
Limited customizability means a limit on an app’s full potential. Developers on general purpose chains are often forced to accept they can’t add certain features or can’t scale their app to a certain extent. To do so would basically require a new blockchain.
In short, the speed, security, costs, and innovative features of an app all rely on customizability. These things impact user experience and an app’s ability to capture value, something developers have to compromise with on chains like Ethereum. With the Cosmos SDK, developers can build their own app-specific network to meet their app’s exact needs; purpose-built networks.
Being able to customize network-level features plays a huge role in scalability. For example, orderbook and lending appchains can modify transaction ordering to make their app more efficient. Orderbook appchains can protect users from frontrunning, and lending appchains can prioritize processing liquidations first.
dYdX is an Ethereum L2 orderbook DEX that’s migrating to its own Cosmos appchain this year. Speaking at Messari Mainnet 2022, the founder mentions they explicitly chose Cosmos for the customizability. dYdX can change network features to improve scalability, such as the need for consensus on certain parts of the protocol. dYdX validators themselves act as an “offchain but decentralized orderbook” where trading fees paid by users are the gas fees.
“You see this parallel the evolution of a lot of tech companies in general… You want to vertically own everything, every sort of piece of the technology, and customize it exactly for your use case.” -dYdX Founder Antonio Juliano
Antonio explains that dYdX is going for a “decentralized Binance”, so this improved scalability is more necessary for dYdX than it is for a standard AMM like Uniswap. General purpose blockchains don’t work for every application, and Cosmos lets dYdX adapt to this. They can create their ideal orderbook appchain that supports 1,000+ TPS with zero gas fees, all while retaining 100% of the value dYdX creates.
The Injective network is no different. The co-founder of Injective explains that pursuing their sector-specific orderbook ambitions would be too costly on an L1 or even an L2, and fixing this scalability issue is only possible with a new chain.
“In order for Injective as a network to even be viable, you would have to really optimize, customize a lot of the base layer primitives within a blockchain itself, and only Cosmos can enable you to do that.” -InjectiveLabs Co-Founder Eric Chen
He also mentions that the Injective appchain can act as “common infrastructure” for other DeFi applications that require its high throughput, low gas fees, and unique transaction ordering.
Osmosis has a similar story for scaling UX. Instead of users having only one token to pay gas fees in, Osmosis users can pay gas fees with any asset of their choice. This decision is only possible thanks to the customization granted by the SDK. Osmosis can also integrate network infrastructure like Skip Protocol to capture MEV as revenue.
Compound Chain was an initiative by the Compound Finance team to launch a custom parachain in the Polkadot ecosystem. The Compound founder agrees that appchains are a superior way to scale applications, and maybe Compound will revisit the idea with enough enthusiasm from Cosmonauts. As its own appchain, Compound can set $COMP as the gas token, keep low fees, innovate network features, retain all MEV, share a common IBC userbase, and be secured the Cosmos Hub if it wants.
Tendermint (now known as “Ignite”) also plays a huge role in scaling Cosmos appchains as the default consensus engine of SDK chains. It’s like Ethereum’s Gasper. Tendermint PoS is simply a default protocol that lets appchains keep low fees and quick transactions, while still giving developers room to customize on top.
Tendermint’s high throughput was tested with the collapse of Terra’s $UST stablecoin, as the Terra network flawlessly handled transactions quickly and without failure (Terra did halt, but due to the fall in $LUNA price, not due to scalability issues).
General purpose blockchains may not be the most ideal solution for enterprise adoption. Some entities want control and are looking to leverage blockchain solely for its data availability, transparency, and tamper-proof infrastructure. The customizability of Cosmos appchains is just the solution.
“Indeed, enterprise-grade appchains operate in a vastly different way to Web3-native appchains. Whereas Web3-native appchains often focus on the tokenomics and use the blockchain structure to financialize their products, enterprise-grade appchains focus on using the blockchain as an efficient way of record and maintain data.” -0xFishylosopher
Similarly, appchains can be designed for issuance of Real World Assets (RWAs). Noble is a brand new Cosmos chain for generic asset issuance, just recently announcing the native issuance of $USDC on Noble for the broader IBC ecosystem. Institutions can deploy on Noble and issue tokenized RWAs using a totally compliant framework.
On chains like Ethereum, application and network governance are separate. Apps are built on an infrastructure they do not own, so they’re forced to inherit changes to the network passed by network governance. Apps also rely on the uptime of their network (a Solana dApp and an appchain walk into a bar…).
Similarly, if developers tried getting Ethereum governance to make changes to the network just to accommodate their app, they probably wouldn’t be very successful.
This is why sovereignty is important. Even while connected, Cosmos appchains are sovereign networks with their own governance and validator sets. They answer to no one but themselves. If you think about it, sovereignty is the heart of the Appchain Thesis.
Customizability is ultimately a product of sovereignty. This one is sort of self-explanatory. I mean, appchains being able to make changes to the application and network is only possible with sovereignty.
Scalability is ultimately a product of sovereignty. The appchain universe isn’t a bunch of apps competing for blockspace under a single state tree. Each appchain has its own blockspace demand, building its own state tree, handling its own transaction volume.
Security is ultimately a product of sovereignty. Apps do not have to rely on any network’s uptime but their own. Even if one appchain goes down, the rest stay standing, and this was well-demonstrated when Terra emergency halted.
There can be no compromise when it comes to sovereignty. Polkadot, another blockchain interoperability solution, tried to compromise on sovereignty with its shared security model. Like appchains, Polkadot has “parachains” where app developers can build their own app-specific blockchain.
However, Polkadot limits the total number of parachains to 100. Developers have to pay to get their parachain slot, and parachains must pay rent to the Polkadot main chain to keep their slot and inherit the main chain’s security. By introducing an upfront and long-term cost for apps to build their own parachain, demand was snubbed with few willing to pay.
Parachains on Polkadot are still their “own chains”, but they’re not sovereign at all. This is also true for Avalanche subnets and Ethereum layer 3s. Customizability is only so beneficial on its own, and this is why sovereignty is the heart of the Appchain Thesis.
“Everyone who uses Ethereum… essentially is paying some sort of rent to this service provider. But the fact remains that people don’t want to pay rent to landlords, they want to become their own landlords, and they want value to accrue to their own token.” -Celestia Co-Found John Adler
Cosmos appchains are totally sovereign, fully custom, self-securing, and pay rent to no one. They can choose to be secured by the Cosmos Hub. They need no permission to launch and there is no limit on the number of appchains that can exist. They also use IBC to interoperate, which is a smooth experience for developers (so much so that Polkadot’s Syntropy is migrating to Cosmos after frustration with Substrate).
IBC is the skeleton and pulse of Cosmos, connecting 50+ chains under one infrastructure. With the SDK, developers can enable IBC on their chain and connect to any other IBC chain in the ecosystem. IBC lets networks share data, so users can transfer assets from chain to chain in seconds with no bridge risk involved. Appchains remain sovereign networks, but can share users like apps on a general purpose L1 do.
There is no fragmented liquidity with IBC. Apps don’t have to deploy several instances on multiple chains to reach new users. Instead, they can exist as one instance (their own chain) that users can seamlessly deposit to and withdraw from via IBC. Since appchains can share IBC users, they can share IBC-native liquidity.
“I think very soon you won’t even be able to know what chain you’re on when you’re using these applications, and that’s how it should be.” -dYdX Founder Antonio Juliano
The reach of IBC isn’t just limited to the Cosmos ecosystem either. Protocols like Polymer and Electron are building IBC-compatible ZK solutions for Ethereum. Composable and Landslide are building infrastructure so other non-SDK ecosystems like Polkadot, NEAR, and Avalanche can connect to IBC too.
IBC enables composability of blockchains. For example, appchains can share staking and slashing information via IBC to create new security models like Interchain Security or Mesh Security. Infrastructure chains are another way networks compose via IBC, examples being the Band oracle appchain, Axelar bridge appchain, and Babylon security appchain.
Cosmos chains can also have their own layer 2s, like Sei Network’s “Nitro” and Injective’s “Cascade”. The Cascade L2 leverages the SVM for execution but is secured by the Injective main chain. It’s the first smart contract layer in Cosmos that Solana apps can easily deploy on, and apps on Cascade get access to the broader IBC userbase/liquidity by default.
Cosmos has long been criticized for its lack of application-level interoperability. On general purpose chains, apps can easily compose with each other and create new collaborative products for their users. On Cosmos, composability has been limited to the network-level. After all, apps are their own blockchains.
Interchain Accounts enables app-level interoperability in Cosmos. Appchains can connect and compose with each other via IBC just like smart contracts on Ethereum. Composability is way cooler in Cosmos because it’s an a bunch of sovereign, customizable blockchains, the possibilities are endless:
cross-chain stablecoin minting
interchain DEX aggregators
liquid staking many assets from one chain
privacy infrastructure and much more
Ethereum’s security is what makes its blockspace so valuable (I think we all can agree Ethereum isn’t popular for its scalability). This is Ethereum’s “moat” against the Appchain Thesis. Apps share the security of Ethereum, this attracts more apps, which attracts more users, which increases blockspace demand, which makes $ETH more valuable.
But as the network bloats, transactions get slower and more expensive. Apps could turn to L2s for better scalability, but still today no L2 is actually secured by Ethereum yet (centralized sequencer problem). Ethereum also has other points of failure for applications, like bridges, oracles, and RPC nodes. There are definitely tradeoffs that apps make for this top-tier security.
So sure, single appchains have nowhere near as many validators or as much stake as general purpose L1s, making them “less secure” by default. But there’s a lot more nuance. With IBC and the Cosmos SDK, apps can explore this nuance and take a new perspective on security for their app.
With Interchain Security (ICS), shared security is optional. Appchains can rent security from the Cosmos Hub rather than building their own validator set. These “consumer chains” use Hub validators to make up as much of their validator set as they want, and Hub validators accept the chain’s native asset as gas. ICS validators slashed on consumer chains are slashed on the Hub too, so this keeps incentives balanced for validators to behave.
Soon, ICS will allow appchains to rent security from any IBC-enabled chain, not just the Cosmos Hub. Consumer chains will be able to choose validators from multiple chains to make up their validator set, so they can inherit as much security from as many different networks as they’d like.
With Mesh Security, appchains can form “mutual defense agreements like NATO” as Osmosis founder Sunny Aggarwal likes to describe it. This combines the stake and security of two or more chains together, and this joining of forces could provide stronger security than any layer 1 today including Ethereum.
Unlike Interchain Security, appchains are not inheriting security. They’re combining it. They still maintain their own validator sets, but get the additional protection of every other Mesh Security chain’s validators/stake. There are over 2,600 active validators throughout the IBC ecosystem today, holding a diverse total stake of over $5 billion. This is more than Fantom’s 64, Polygon’s 100, Avalanche’s 1,200, Solana’s 1,600, and getting close to Cardano’s 3,100. At scale, Mesh Security could be very powerful.
Appchains like Babylon are infrastructure that other IBC appchains can connect to. Babylon offers improved security for appchains using Bitcoin timestamps. Appchains can use these secure timestamps on critical consensus data, which helps protect against long-range network attacks. Stakers on these appchains can also unstake much faster (as short as a few hours) without compromising network security.
Berachain is a custom general purpose L1 built with the Cosmos SDK. It supports an EVM-compatible execution layer, so Ethereum and other EVM-based developers can easily port their apps onto Berachain.
Berachain has a unique consensus mechanism called “Proof of Liquidity” (PoL). This mechanism lets Berachain validators accept any asset as stake that is approved by the network’s governance ($wETH, $ATOM, $AVAX, $USDC, $USDT, etc.). These assets are liquid staked by default, so stakers can use their Berachain-staked assets to earn yield and unstake whenever by selling the liquid derivative. One use of the LSD could be minting the $HONEY stablecoin, so users can take out debt against their staked assets.
This custom consensus infrastructure is only possible with the Cosmos SDK. It doesn’t just enable customizability for app features, but also the ability to customize new consensus mechanisms like Proof of Liquidity.
As their own networks, appchains can add features or make certain decisions that aren’t possible on general purpose chains. The dYdX founder outlines the app-specific security tradeoffs on chains like Ethereum, and why Cosmos works for them even with “less security”.
“We’re not going to get to a point on the dYdX chain where there’s even close to as many validators as there are on Ethereum, but at the same time you get sovereignty... and that lets you deal with issues, things like hacks, things like bugs, or upgrades much more seamlessly.” -dYdX Founder Antonio Juliano
For apps, security isn’t just about the number of validators. It could be certain controls or scalability improvements that make the app more resilient, and this is exactly what Cosmos offers.
Superfluid staking is a novel creation by the Osmosis team, basically backwards liquid staking. It allows liquidity providers to stake a portion of their LPed $OSMO to earn additional staking yield on top of their LP.
With interchain accounts, Osmosis can extend support to native assets of any IBC appchain. This means $ATOM, $JUNO, $SCRT, $STARS, and other LPs alike can superfluid stake their LPed assets from Osmosis. As such, IBC networks can secure themselves with a portion of their token liquidity on Osmosis.
Celestia is a modular blockchain solution that allows developers to build blockchains that use separate consensus, execution, and application layers. Appchains can get shared security from Celestia, but still keep their own network governance.
This is like an improved model of Interchain Security, since appchains sharing security on Celestia can interoperate much more securely than ICS chains. Appchains can fully customize their stack even further using Celestia, which has benefits for security as well.
Network effect is the increase in value of a network as more users join. Ethereum and Bitcoin have some of the strongest network effects known today.
The exact same network effect of a standard L1 can be reproduced using an L0 connecting apps as their own blockchains. The interchain network effect is arguably even more powerful than that of a monolithic L1, since users from dozens of blockchains share a common network. The Cosmos IBC network.
In Cosmos, a single instance of an app is connected seamlessly to every new chain that launches with IBC. This is best exemplified by the de facto interchain DEX, Osmosis. More traders and volume come to Osmosis as existing chains grow in size or as new chains launch. Most appchains don’t have a native DEX, so Osmosis is the IBC-native DEX.
Interchain UX is great, and is constantly being improved by organizations like Strangelove. Interchain wallets like Keplr make it so users don’t have to switch between networks as they move across IBC chains, and users can make interchain transfers from their Keplr wallet rather than needing a token bridge.
All of this abstraction makes it so users of one IBC chain can access every other IBC chain with ease. In this, all chains in the IBC network share a universal Cosmonaut userbase. Users aren’t exclusive to a single appchain on Cosmos in the same way they aren’t exclusive to a single app on general purpose chains.
And remember, there are novel general purpose chains in Cosmos like Juno, Berachain, Agoric, Canto, and Evmos. Sector-specific chains like Sei, Injective, and Kujira. All of these chains can share a common userbase and IBC-native liquidity. The network effect of one chain is the network effect of all IBC chains.
dYdX got criticism for moving to an ecosystem with “too few users”. Except the dYdX founder argues apps shouldn’t view blockchains as “a method of distribution”, and Mr. Leshner would probably agree. Apps would be better off as a single instance that can be seamlessly reached by users from any chain. Where apps don’t have to keep deploying on the “hottest chains” to capture new users. IBC solves exactly this.
“Once you don’t have to get shipped off to some third party site to do the bridging and it’s just like ‘okay, let’s deposit into Osmosis or deposit into dYdX', you won’t really even know what chain you’re on. And I think with the interoperability it makes for a much better product experience.” -dYdX Founder Antonio Juliano
The interchain will start to develop its own sort of “economic zones”, or ecosystems of appchains clustered around certain use cases/partnerships. Sort of like the “Curve ecosystem” on Ethereum, or all the protocols building on top of GLP. Appchains can build out their own app-specific ecosystems too, such as Osmosis and Mars Protocol. Similarly, Kujira has an entire suite of products around its orderbook DEX as well as its own over-collateralized $USK stablecoin.
Cosmos is largely community-driven with a tight-knit ecosystem of appchain users. The Interchain Network Effect survives thanks to the customizability from the Cosmos SDK, the scalability from Tendermint, and connectivity from IBC. In their own respects, each core component plays a role in the Interchain Network Effect. The Cosmos Thesis is only “whole” with these three things combined.
Talk about a Big Bang. I feel like a Big Bang went off in my brain just writing this. Anyways, hopefully this piece was able to explain every part of what Cosmos is all about.
Cosmos has been one step ahead since 2014 pioneering Proof of Stake, interoperability solutions, and app-specific blockchains. The Fat Protocol Thesis is dying. It’s inevitable that more apps will follow dYdX and warm up to the Appchain Thesis, and there’s no doubt that Cosmos will be their first choice.
Follow me on Twitter
Come join the Galaxy Box
Follow Cosmos on Twitter
Join the Cosmos Discord
DISCLOSURE: $ATOM is one of my largest holdings next to $ETH. I was not asked to write this article and have not been compensated in any way. The information provided in this article is solely for educational purposes and should not be considered as financial advice. The views expressed in this article are my own and do not necessarily reflect the official policy or position of any company or organization. Readers should always conduct their own research before making any financial decisions.