Moonkelp filters news by the theses you believe in — so it has to know what those theses are, how much capital backs them, and how proven each one is. This study is that map: fifteen crypto market theses, placed by maturity and by how strongly they drive investment.
What the map shows
Three-legged triangulation
Each thesis is ranked as a driver of investment by combining three independent measurements — investor surveys (Bitwise / VettaFi and others), capital allocation (top-100 market cap mapped to theses), and search interest (Semrush, US + DE). An ordinal rank, not a score.
Capital concentrates
The dominant, best-proven theses sit top-right — stablecoins, institutional adoption, store of value, RWA. The speculative end — Crypto × AI, prediction markets — sits left: strong survey interest, little capital behind it yet, real break conditions. This is why onboarding curates a short list by driver rank, not fifteen equal options.
Threats ranked up, never smoothed
A “speculative” thesis looks different in the app from a proven one — honesty by design. A story that threatens a thesis is promoted, never filtered out. That’s the anti-hopium core: Moonkelp tells you when a thesis is breaking, not only when it’s winning.
How to read the map
Horizontal is maturity (speculative → partial → proven); vertical is driver rank (top = dominant). Bubble size is rough category market cap; colour is maturity.
Tap any point for the thesis, its counter-thesis (“breaks if …”) and the assets that carry it. The capital layer below shows where the top-100 market cap actually sits per thesis.
Method & limits
Adversarially cross-checked research — 22 of 25 claims confirmed against primary sources. The capital layer maps 95 of the top-100 tokens to theses.
The 15 theses are the investment narratives that recur across institutional research and crypto-native discourse — chosen to cover where top-100 capital actually sits, not for our agreement with them. Evidence as of Jul 19, 2026; capital shares move with price.
Ranks are ordinal triangulation, not a precise index. Perspectives, not investment advice. This study is reviewed quarterly; corrections normally ship within days. Corrections: contact@moonkelp.com.
The map · 15 theses
Hover any thesis for its maturity and its counter-thesis (“breaks if …”).
The indicator, live
Coverage sentiment, running now
This map is static research — but the app turns it into a live signal. Each day the agent rolls up how the coverage leans for every thesis; the result is the running series below (the app’s signature KPI).
Across the theses with tagged coverage, the day’s reporting leans supportive — 573 thesis mentions (30d).
- Store of value51167
- Institutional adoption7691
- Stablecoins7961
- RWA tokenization7749
- Ethereum settlement7141
- On-chain finance7138
- Prediction markets5330
- Privacy6926
- High-performance chains7022
- Real yield8318
- Attention / culture5211
- Crypto × AI918
- Restaking86
- Modular infrastructure305
Coverage sentiment · from the daily thesis tagging · not market positioning.Open the Moonkelp Index
Where the capital sits
The concentration, in numbers
Share of the top-100 market cap per thesis — where capital actually sits. Store-of-value (Bitcoin) alone is more than half; the speculative theses are rounding errors.
Measured differently (no clean market-cap share):Institutional adoption — >$175B in ETPs · Real yield — >$1.4B in buybacks (2025) · Prediction markets — Market cap unverified · Restaking — TVL ~$15B → $7B; EIGEN −86%
All 15 theses · the full dossier
Every thesis, and what breaks it
The belief, its maturity, the capital behind it, the mindshare trend, whether the token actually accrues value, its regulatory exposure, the assets that carry it — and the honest counter-thesis for each.
Stablecoins
ProvenBreaks if Bank / CBDC stablecoins on closed rails win instead
RWA tokenization
PartialBreaks if Tokenization stays on private, permissioned ledgers
Institutional adoption
ProvenBreaks if A run of ETF outflows or a regulatory reversal
Store of value
ProvenBreaks if Positive real rates; BTC trades as a risk asset
Crypto × AI
SpeculativeBreaks if Centralized AI providers need no chain
Real yield
PartialBreaks if ~70% of buybacks underperformed BTC
On-chain finance
PartialBreaks if Regulation; TradFi builds its own closed rails
Ethereum settlement
ProvenBreaks if L2s siphon fees; activity migrates to Solana
High-performance chains
PartialBreaks if Outages; activity was bot / memecoin-driven
Prediction markets
SpeculativeBreaks if Regulated as gambling; token value unclear
DePIN
SpeculativeBreaks if Revenue is only token subsidy
Modular infrastructure
SpeculativeBreaks if "Monolithic wins" (Solana); no value accrual
Restaking
SpeculativeBreaks if Slashing is real; yields keep shrinking
Privacy
PartialBreaks if Delistings and outright bans
Attention / culture
SpeculativeBreaks if Activity collapses — demand is hollow
The theses, their evidence and counter-cases are our own research; the primary sources are listed at the foot of this page.
Check our work · the claim ledger
Every load-bearing claim, and where it comes from
25 claims went through adversarial verification: 22 survived, 3 were killed. The load-bearing ones are listed here with their sources — including the three that failed, which we do not use.
| Claim | Source | Status |
|---|---|---|
| Stablecoin raw transaction volume $46T (+106% YoY), $9T bot-adjusted organic | a16z State of Crypto, Oct 2025 | confirmed |
| Stablecoin market cap ~$310B; USDT $184B (~59%) | DefiLlama Jun 2026 · CoinGecko Q1 2026 | confirmed |
| 83% of institutions use or plan stablecoins for payments; T+0 settlement 88% | Coinbase/EY-Parthenon survey, Jan 2026 (n=351) | confirmed |
| Tokenized RWA ~$33B in Q2 2026 (+45% YTD) | Bitwise Q2 2026 · RWA.xyz | confirmed |
| 64% of asset managers want to tokenize their own assets (up from 40% in 2025) | Coinbase/EY-Parthenon, Jan 2026 | confirmed |
| >$175B in BTC/ETH ETPs, +169% year over year | a16z State of Crypto, Oct 2025 | confirmed |
| Advisor crypto allocation at 32% in 2025, up from 22% — an all-time high | Bitwise/VettaFi 2026 survey (n=299) | confirmed |
| 73% of institutions plan to increase crypto allocation in 2026 (from 62%) | Coinbase/EY-Parthenon, Jan 2026 | confirmed |
| Top survey themes: stablecoins + tokenization 30%, digital gold 22%, crypto x AI 19% | Bitwise/VettaFi 2026, Jan 2026 | confirmed |
| 45% of younger US investors hold crypto vs. 18% of older cohorts | Coinbase/Ipsos Q4 2025 (n=4,350) | confirmed |
| ~40-70M monthly active crypto users, up ~10M year over year | a16z State of Crypto, Oct 2025 | confirmed |
| IBIT was the most-traded BTC ETP launch of all time | a16z, Oct 2025 · public market data | confirmed |
| 19% advisor interest in Crypto x AI | Bitwise/VettaFi 2026 | confirmed |
| TAO search interest spikes in US and DE while broad ai-crypto search declines | Semrush, first-party, Jul 2026 | confirmed |
| DeFi/token buybacks >$1.4B in 2025; Hyperliquid ~$645M (~46%) | CoinGecko · The Block, 2025 | confirmed |
| Hyperliquid routes ~97-99% of trading fees into HYPE buybacks (>$1.3B since Jan 2025) | protocol data, verified 2-of-3 | confirmed |
| ~70% of token-buyback programs underperformed BTC | Bill Hsu analysis via Yahoo Finance, 2025 | confirmed |
| EigenLayer slashing live on mainnet since Apr 17, 2025 | The Defiant, Apr 2025 | confirmed |
| Restaking TVL fell from ~$15B peak to ~$7B by end 2025; EIGEN -86% | PANews et al., late 2025 | confirmed |
| Polymarket search volume ~450k US / ~110k DE monthly, rising | Semrush, first-party, Jul 2026 | confirmed |
| Stablecoins settle 20x PayPal / 3x Visa | attributed to a16z | failed — not used |
| Prediction markets at $1.8B open interest / $43B quarterly volume | Odaily | failed — not used |
| 46% use crypto as an inflation hedge | MEXC press release | failed — not used |
Confidence notes live in the research record: survey percentages are ordinal interest signals from small, self-selected samples, and most survey sponsors have a commercial interest — which is why market-cap and on-chain figures are weighted higher throughout.
Sources & references
What this study is built on
15 crypto market theses, mapped in July 2026 by triangulating investor surveys, capital allocation and search interest, then cross-checked against critical sources — plus our own Semrush mindshare analysis. The key claims were verified against the sources below (22 of 25 confirmed).
Surveys & industry reports
Critical / opposing views
Our own analysis
Sources were cited by name and date in the research; the exact fetched pages aren’t stored, so each link points to the provider’s own site rather than the specific page. Traffic figures are our own Semrush pulls — primarily the US database, so they compare sources rather than count absolute audiences.