What survived the crypto winter — which protocols, use cases, and business models proved durable
Which Crypto Projects Actually Survived — And Why?
Based on analysis of a 95-node, 304-edge knowledge graph mapping the protocols, business models, and events that defined the crypto industry from 2022 through 2025.
What This Is About
Imagine a giant storm wipes out most of the farmers in a region. When the dust settles, you want to know: which farms survived, what made them different from the ones that failed, and what does that tell us about farming?
That is roughly what happened in crypto in 2022. A cascade of failures — exchange collapses, failed “algorithmic” currencies, companies going bankrupt — wiped out a large portion of the industry. This analysis maps out what the wreckage looked like and which structures were still standing afterward.
The knowledge graph behind this is essentially a map of causes, effects, and feedback loops — a diagram of “this caused that, which enabled this other thing, which contradicted that.” Reading it carefully reveals patterns that are not obvious from just watching prices.
The Core Survival Test: Were You Making Real Money?
The single most connected idea in the entire graph is something called the “real yield paradigm shift.” With 42 connections, it functions as the primary test that every surviving protocol either passed or failed.
Here is what it means in plain terms.
Before the crash, many crypto projects paid their users in their own tokens — tokens they printed themselves. This is like a company paying employees in gift cards to their own store. The employees feel paid, the company looks generous, but no actual value is being created. The “yield” (the return on your investment) was fake: it came from new token creation, not from the project doing something genuinely useful and charging for it.
The projects that survived mostly share one thing: they generate fees from real activity. Traders pay fees to use a trading platform. Borrowers pay interest to use a lending protocol. That revenue is real. It does not depend on convincing new people to buy tokens.
In the graph, almost every surviving protocol points toward “real yield” as evidence that it belongs in the survivor category. The graph treats real yield not as a thing that causes other things, but as an endpoint — a badge that says “this project passed the test.” That structural detail matters: the graph is not saying real yield is a new invention that fixes everything. It is saying it became the definition of fitness.
The Collapse Planted the Seeds of Its Own Recovery
One of the most striking structural findings: the 2022 crash did not just destroy things. It directly triggered many of the conditions that later reversed it.
Think of it like a controlled burn in a forest. The fire clears out dead wood and invasive species. It also deposits nutrients in the soil and opens the canopy for new growth. The destruction and the recovery are not separate events — one causes the other.
In the graph, the 2022 collapse directly triggered:
- A shift toward “real yield” protocols (people stopped trusting fake returns)
- Increased trust in decentralized exchanges over centralized ones (because centralized exchanges were the ones that collapsed)
- A regulatory response in the United States that eventually led to Bitcoin ETF approval
- Infrastructure investments in institutional-grade custody of crypto assets
And then those outcomes eventually led to Bitcoin’s recovery via institutional investment — which the graph encodes as partially “inverting” the original collapse. The system turned over and righted itself, but through a specific sequence, not randomly.
Two Stablecoin Systems That Do Not Overlap
A stablecoin is a cryptocurrency designed to always be worth one dollar. The graph shows two completely separate ways of achieving this — and they lead to very different places.
Tether’s path looks like this: Tether issues dollars, collects the interest on the real-world assets backing those dollars (this is called seigniorage), and uses that income to fund distribution across emerging markets — places like Turkey, Argentina, and Nigeria where the local currency is unstable and people want a reliable dollar substitute. That adoption, in turn, generates more transaction volume, which generates more seigniorage income. It is a self-reinforcing loop that does not depend on the regulatory environment in Europe or the United States.
Circle’s path (the company behind USDC, the other major stablecoin) looks very different. Circle operates transparently, complies with regulators, holds segregated reserves, and publishes audits. This sounds like an advantage — and in some ways it is. But the graph encodes something non-obvious: the same regulatory frameworks that were designed to legitimize compliant issuers like Circle also raise Circle’s costs significantly more than Tether’s. Tether, which operates more opaquely, does not face the same compliance overhead. So regulation designed to help Circle appears to hurt Circle’s margins relative to Tether.
These two paths share no meaningful connection in the graph. They are parallel systems solving the same problem through incompatible methods.
A Risk That Keeps Accumulating With Nowhere to Go
One of the graph’s structural gaps is worth understanding, even if you are not technical.
There is a system called EigenLayer that allows people to “restake” their Ethereum — essentially reusing the same collateral to secure multiple systems simultaneously. The risk label for this in the graph is “EigenLayer Restaking Contagion Risk.”
Here is the unusual part: six high-weight nodes in the graph all feed into this risk node. The core Ethereum infrastructure — the staking system, the layer-2 networks, the fee-burning mechanism — all amplify this risk. Yet the risk node itself has no outgoing edges. The graph shows the risk accumulating but does not model what happens when it discharges.
This is not necessarily an error in the graph. It may reflect that, as of the time of analysis, the consequences of this risk have not yet been observed. But structurally, it represents a pressure that is building in a container with no modeled release valve.
Bitcoin and Ethereum Are Playing Completely Different Games
The graph shows Bitcoin and Ethereum following separate value chains that almost never touch.
Bitcoin’s chain looks like: scarcity from the halving (when new Bitcoin creation slows down) plus ETF approval creates demand from institutional investors, which companies like MicroStrategy amplify by buying large quantities and putting it on their balance sheet, which eventually contributes to a race among governments to hold Bitcoin as a reserve asset. The entire chain is about Bitcoin as a store of value — digital gold.
Ethereum’s chain looks like: a technical upgrade (EIP-4844) reduces costs for layer-2 networks built on top of Ethereum, which allows those networks to capture transaction revenue, but this also means less fee revenue flows back to Ethereum itself. The graph actually encodes this as a paradox: Ethereum made its ecosystem cheaper, which helped its layer-2 networks thrive, but in doing so reduced the value it captures at the base layer.
These two architectures share almost no edges in the graph. The one meaningful connection is that MicroStrategy’s Bitcoin accumulation is inversely correlated with DeFi real yield — meaning when institutional money piles into Bitcoin for store-of-value reasons, it tends not to flow into Ethereum-based yield products.
Things That Are Connected in Non-Obvious Ways
A few connections in the graph are worth highlighting because they are not visible from surface descriptions:
A North Korean hack specifically threatens a synthetic dollar. A cryptocurrency exchange called Bybit was compromised. The graph connects this specifically to Ethena, which issues a synthetic dollar backed by short positions on centralized exchanges. That is not an obvious connection unless you understand Ethena’s mechanism — the hack attacks the specific infrastructure Ethena depends on for its collateral.
Meme coin infrastructure is also serious payment infrastructure. Pump.fun is a platform on Solana that lets anyone launch a meme coin cheaply and instantly. The graph also shows it undermining the traditional venture capital model of launching tokens — because Pump.fun’s bonding curve mechanism makes the high-valuation/low-float token extraction strategy structurally unavailable. A meme coin tool is encoded as an institutional financing reform.
EU regulation raises costs for the compliant player. As described above with Circle — the regulation designed to help the rule-follower amplifies the rule-follower’s cost disadvantage. This is a non-obvious second-order effect.
The Questions the Graph Does Not Answer
The graph is explicit about several things it cannot resolve:
The relationship between layer-2 networks and Ethereum’s base layer remains genuinely ambiguous. Is Ethereum’s success at enabling cheaper layer-2 networks good or bad for Ethereum itself? The graph shows both, without resolving it.
The US legislative framework for stablecoins simultaneously constrains Tether and enables Tether. Two equal-weight edges point in opposite directions. The net effect is not modeled.
Bitcoin’s long-term security relies on transaction fees once block rewards shrink to near zero. Multiple nodes “partially address” or “temporarily solve” this problem. None resolve it. The graph encodes Bitcoin’s security future as an open question with only provisional answers.
The AI-crypto intersection has many structural enablers but remains at minimum weight. The graph treats AI-integrated crypto protocols as anticipated but not yet arrived.
Bottom Line
What the graph’s structure reveals, taken as a whole:
Survival was defined by one criterion. Real revenue from real activity replaced token emissions as the measure of a legitimate protocol. Everything else flows from that shift.
The crash was generative, not just destructive. The mechanisms that eventually reversed 2022 were mostly triggered by 2022 itself — regulation, trust shifts, institutional infrastructure, and the real-yield standard all emerged from the collapse.
Tether’s dominance is a geopolitical phenomenon, not a crypto one. Its feedback loop runs through emerging market currency instability, not through crypto market cycles. Regulatory frameworks designed to constrain it may not reach the markets where it actually operates.
The largest unresolved risk is invisible from the surface. EigenLayer’s restaking risk receives inputs from the entire Ethereum infrastructure stack but has no modeled outputs. The graph shows accumulation without discharge.
Bitcoin and Ethereum are not competing. They are running entirely separate experiments with different hypotheses about what makes something valuable. The graph treats them as parallel rather than competing.
The AI-crypto thesis is structural but premature. The graph’s architecture predicts that durable AI-crypto value will come from payment infrastructure first, and from autonomous financial agents only after that payment layer is established — and that establishment is not yet encoded as complete.