Not financial advice — and not a list to copy. These funds have enormous information advantages, longer time horizons and different risk tolerances than individual investors. Quotes are from public sources; portfolio data from public disclosures. The value is understanding their reasoning, not mirroring their bets.
01 // THE FUNDS
Who They Are & What They Watch
a16z Crypto
ANDREESSEN HOROWITZ · CHRIS DIXON, ALI YAHYA · $7.6B AUM
The largest dedicated crypto VC. Thesis: "Read Write Own" — the internet's third era, where users own their data and value through token-based networks. Longest-horizon outlook of any fund on this page; willing to hold through multiple cycles.
DeFi
Consumer crypto
RWA
Stablecoins
Gaming
KNOWN PORTFOLIO
Uniswap · Compound · Solana (early) · Avalanche · Optimism · Base ecosystem · Farcaster · Ondo Finance · Polymarket · OpenSea · Dapper Labs
Multicoin Capital
KYLE SAMANI, TUSHAR JAIN · AUSTIN TX · ~$1B AUM
High-conviction macro thematic bets. Known for going all-in on Solana before it was consensus and building the DePIN (Decentralized Physical Infrastructure) thesis before most knew the acronym. Critical of Ethereum L2 fragmentation — argues one coherent chain beats many shards.
Solana
DePIN
AI × crypto
Cross-chain
KNOWN PORTFOLIO
Solana (SOL) · Helium (HNT) · Pyth Network · Render Network (RNDR) · Hivemapper · IoTeX · Bittensor (TAO) · Serum (now OpenBook)
Paradigm
MATT HUANG, CHARLIE NOYES, DAN ROBINSON · SF · ~$2.5B AUM
The most technical VC in crypto. Known for publishing pre-investment research that shifts the entire field, then backing the protocols it writes about. Thesis is about protocol primitives — MEV, AMM design, restaking, and the long-term infrastructure stack underneath everything else.
Protocol research
MEV / AMMs
Restaking
DeFi primitives
KNOWN PORTFOLIO
Uniswap · Maker (MKR) · Blur · EigenLayer · Matrixport · dYdX · Cosmos (ATOM) · Ethereum (ETH) · Arbitrum
Pantera Capital
DAN MOREHEAD, PAUL VERADITTAKIT · MENLO PARK · ~$5B AUM
Oldest institutional crypto fund (2013). Multi-strategy across Bitcoin, liquid tokens and venture. Thesis has evolved from pure BTC-as-digital-gold to a broader institutional adoption narrative — Bitcoin as reserve asset first, then DeFi rails, then emerging-market financial inclusion as the long arc.
Bitcoin
Institutional
L2s
Emerging markets
KNOWN PORTFOLIO
Bitcoin · Ethereum · Ripple (XRP) · Polkadot · Bitstamp · Alchemy · Ripple Labs · Bakkt · 0x Protocol · Circle
02 // THE CYCLE SPECTRUM
Where They Stand on the Current Crypto Cycle
BTC remains in the mid-$60s — well below 2026 highs — while equities hover near records. How each fund is reading this divergence.
REDUCING RISKNEUTRAL / WATCHINGADDING AGGRESSIVELY
Pantera CapitalWATCHING BTC PATIENTLY
a16z CryptoLONG-HORIZON DEPLOYMENT
ParadigmPROTOCOL-LAYER BUILDER
Multicoin CapitalHIGH-CONVICTION LONG
Pantera sits furthest left not because it's bearish — it's simply more macro-rate-sensitive than the others. With BTC lagging equities and the Fed's Jackson Hole meeting (Aug 27–29) potentially signalling another hold or cut, Morehead has historically argued the 12–18 months after a Bitcoin halving (April 2024) deliver the cycle's best returns — but that window is now well advanced. The other three are less cycle-dependent: a16z, Paradigm and Multicoin deploy on 7–10 year fund horizons regardless of BTC price.
03 // THIS WEEK
What They're Saying & Doing
Chris Dixon A16Z CRYPTO BULL ON PROTOCOL LAYER
AUG 2026
"The killer apps for crypto aren't gambling apps or memecoins — they're the apps that don't know they're crypto apps."
Dixon's framing of why he's still deploying aggressively: the most durable crypto applications will be the ones where the user never sees a wallet address or a gas fee — stablecoin payments in emerging markets, tokenized real-world assets (RWA) settling on public rails, prediction markets like Polymarket that simply work better than their Web2 equivalents. a16z's current deployment focus tracks this: Ondo Finance (tokenized US Treasuries), Polymarket (prediction markets), and Farcaster (decentralized social) are all applications where the crypto infrastructure is invisible to the end user. The argument against: "infrastructure before applications" has been the bull thesis in every cycle and the applications always arrive later than expected.
Kyle Samani MULTICOIN CAPITAL BULL ON SOLANA + DePIN
AUG 2026
"DePIN isn't a narrative — it's a business model. The unit economics of decentralized wireless work, which is something you couldn't say about Web3 gaming."
Samani continues to argue that Helium's migration to Solana and subsequent recovery is the proof-of-concept for the DePIN category: a protocol where token incentives coordinate real physical infrastructure (cell towers, storage nodes, mapping data) that generates actual revenue. Multicoin's current positioning reinforces this — Render Network (RNDR) has become a key holding given its position at the intersection of AI compute demand and decentralized GPU supply, while Bittensor (TAO) represents their highest-conviction AI × crypto bet. The risk: Solana's own token is still trading well below its 2021 peak despite the ecosystem's fundamental strength — a reminder that great protocols and great tokens are not the same thing at any given moment.
Paradigm Research PARADIGM BUILDING THROUGH CYCLE
AUG 2026
Published new research on EigenLayer restaking risks — arguing that while restaking dramatically increases capital efficiency, it introduces correlated slashing risk that the market is underpricing.
Paradigm rarely makes public macro calls, preferring to publish technical research that simultaneously signals where they're thinking about deploying. The EigenLayer restaking paper is the clearest public tell of the current period: they're long on the primitive (restaking as a concept), skeptical of the current implementation's risk disclosure, and implicitly signalling they'll back whatever comes next that solves the correlated-slashing problem. Separately, Paradigm's Blur NFT marketplace continues to dominate professional trader volume even as the broader NFT market remains subdued — a reminder that their application bets are made for protocol-level usage, not sentiment cycles.
Dan Morehead PANTERA CAPITAL WATCHING THE MACRO
AUG 2026
"Bitcoin's post-halving cycle historically delivers its best 12–18 month returns. We're now 16 months past the April 2024 halving. The window isn't closed but it's narrowing."
Morehead's macro framework is the simplest of the four: Bitcoin follows halving cycles with a lag, and the Fed's rate trajectory is the second variable. With Jackson Hole on Aug 27–29 and September's Fed decision still in play, Pantera is watching the rate environment more closely than the other funds — all of which have 7–10 year horizons that make a single rate decision less relevant. Morehead's framing implies BTC at mid-$60s should be in recovery territory 16 months post-halving, which makes the current lag vs. equity markets either a buying opportunity or a signal that something has structurally changed. He hasn't specified which.
04 // THE DISAGREEMENT
Where These Funds Genuinely Conflict
| Question | Bull side | Cautious side |
|---|---|---|
| Which chain wins long-term? | Multicoin: Solana — one coherent high-throughput chain beats fragmented L2s; DePIN and AI compute demand reward speed | Paradigm: Ethereum's restaking + L2 ecosystem — the protocol layer is where value accrues, not the L1 token |
| Will applications drive the next leg? | a16z: Yes — RWA, stablecoin payments and prediction markets are the first DeFi applications with real PMF and non-crypto users | All four: History says applications always arrive later than expected. The infrastructure-first pitch has been made every cycle. |
| Is BTC's lag vs. equities a buy signal? | Pantera: Yes — halving cycles have historically resolved in BTC's favour with a lag; rate cuts are the catalyst | Multicoin: BTC's cycle narrative is less important than ecosystem fundamentals; Solana at current levels vs. 2021 highs is the better question |
| AI × crypto — real or narrative? | Multicoin: Render Network and Bittensor are real businesses — GPU supply and ML model markets are genuinely better when decentralized | Paradigm: The intersection is real at the protocol layer (verifiable compute) but most "AI crypto" tokens are narrative-only |
| Is DePIN a durable category? | Multicoin: Helium's unit economics work — token incentives can bootstrap physical infra that otherwise requires billions in capex | a16z: DePIN works for some categories (wireless) and failed for others (storage, computing in prior cycles) — execution track record is thin |
05 // THEMES TO TRACK
The Narratives These Funds Are Shaping
RWA — REAL WORLD ASSETS
Tokenised Treasuries, credit and real estate
a16z's Ondo Finance bet, Pantera's Circle position and Paradigm's Maker stake are all bets on the same thesis: that DeFi's killer use case is bringing tradfi instruments on-chain. Total RWA TVL has grown from ~$5B in 2023 to over $25B, led by tokenized US Treasuries. BlackRock's BUIDL fund is the mainstream validation signal.
DePIN
Decentralised physical infrastructure
Multicoin's core thesis. Helium (wireless), Hivemapper (mapping data), Render (GPU compute) and IoTeX (IoT data) are the biggest names. The category's underlying argument: token incentives can outcompete capex-heavy incumbents by turning users into suppliers. Still early and execution-dependent.
RESTAKING
EigenLayer and the capital efficiency trade
Paradigm's current research focus. Restaking lets ETH stakers secure additional protocols with the same capital, dramatically increasing yield but also creating correlated slashing risk. EigenLayer has attracted over $15B in restaked ETH. Paradigm is long on the concept, worried about the current risk disclosure.
PREDICTION MARKETS
Polymarket's breakout as a legitimate data source
a16z-backed Polymarket has become the market of record for election and macro event probabilities — cited in mainstream financial media alongside traditional polls. The user acquisition model for crypto that doesn't feel like crypto. Bull case: it expands into every prediction category. Bear case: regulation risk in US markets.
AI × CRYPTO
Verifiable compute and autonomous AI agents
The emerging intersection: AI agents that hold wallets, verifiable AI inference on-chain, and decentralized GPU markets (Render, Bittensor). Multicoin is the most convicted. Paradigm is skeptical of the narrative tokens but bullish on the protocol primitives underneath. Still mostly pre-revenue.
STABLECOIN RAILS
Payments as the first mass-market crypto use case
Circle (Pantera) and the broader stablecoin ecosystem now processing over $10T in annual volume. USDC on Solana's throughput makes it competitive with Visa for payment rails in LatAm and Southeast Asia. The Circle IPO, when it comes, will be the first real retail-accessible vote on whether stablecoins are infrastructure or commodity.
06 // HOW TO USE THIS
Reading Crypto Fund Signals Without Getting Hurt
Their time horizons are not yours
a16z, Paradigm and Multicoin run 7–10 year venture funds. They can hold a token that drops 90% and not be wrong — they expected it. An individual investor buying the same token on a 12-month horizon is in a completely different position.
Token ≠ protocol performance
Multicoin has been right about Solana's ecosystem growth and wrong about SOL's price at various points simultaneously. The best-positioned protocol can still have an overvalued or undervalued token. Separate the network thesis from the token trade.
Positions are disclosed late or not at all
Unlike 13F filers, crypto funds have minimal disclosure requirements for liquid positions. What they publish on X/podcasts is what they want you to know. Assume their actual book is more nuanced than their public thesis.
BETTER USE
Mine the research, not the ticker
Paradigm's technical papers, Multicoin's thesis memos and Dixon's essays are the most useful output — they force-test your own assumptions about which primitives win. Match those arguments to your own Decision Journal rather than copying trades.