A Decade of Change: The Demise of Crypto Startups

Foresight NewsPublicado a 2026-07-06Actualizado a 2026-07-06

Resumen

"The Decade-Long Transformation: The Demise of Crypto Startups" The article traces the dramatic evolution of the cryptocurrency industry from its anarchic beginnings to its current highly regulated and institutionalized state. In the early days (circa 2017), launching a crypto startup was remarkably simple: a whitepaper, a GitHub repository, and a Telegram group could attract thousands of retail investors via an Initial Coin Offering (ICO). Founders operated anonymously with near-zero regulatory and financial barriers, enabling rapid, global innovation but also widespread fraud. By 2026, the landscape is fundamentally different. To operate in major markets like the US, EU, and Asia, crypto businesses must now navigate a complex web of regulations akin to traditional finance. Compliance costs are prohibitive: estimates for a US multi-state operation range from $750,000 to $1.2 million in the first three years, with annual costs exceeding $2 million thereafter. Regulations like MiCA in the EU and New York's BitLicense have created high capital and operational hurdles that act as barriers to entry. Simultaneously, venture capital investment has shifted dramatically. Following the collapses of Terra and FTX, funding has concentrated in later-stage, established companies, creating a "barbell market." Early-stage and seed funding has shrunk significantly, while mega-funds like Andreessen Horowitz's $15 billion pool dominate. Most capital now flows to trading platforms, lending ...


By: Andjela Radmilac

Compiled by: Saoirse, Foresight News


In 2017, a few developers could launch a token and start a crypto company within just a few days, armed with only a whitepaper and a GitHub repository. Back then, the capital barrier was extremely low, and licensing and regulation were either non-existent or trivial afterthoughts. With a sufficiently compelling idea, even a product that hadn't materialized could attract tens of thousands of retail investors through an Initial Coin Offering (ICO).


But by 2026, crypto firms operating in compliant markets in Europe, America, and Asia must assemble legal and compliance teams, secure banking partnerships, establish full anti-money laundering systems, and allocate substantial funds to meet licensing and operational requirements before they can serve customers at scale.


The crypto industry, originally built by anonymous founders coding in their bedrooms, is now dominated by large enterprises with full balance sheets, regulatory licenses, and institutional sales teams. Crypto startups haven't vanished entirely, but the barriers to entry for launching a company have become comparable to those in the traditional financial sector, which has long been insulated from new entrants.


The Crypto Startup Ecosystem of Yesteryear


The first decade of crypto entrepreneurship was characterized by low capital requirements, minimal regulatory friction, and open collaboration among numerous anonymous developers worldwide. Exchanges, wallets, and blockchain protocols could be built by small, globally distributed teams primarily coordinating via Discord and GitHub.


Ethereum itself was born in 2015 through a public crowdsale, raising about $18 million from thousands of ordinary individuals, not venture capital syndicates. The ICO boom of 2017-2018 pushed this lightweight startup model to its extreme: any team with a website, a token contract, and a Telegram community could fundraise directly from the public, bypassing venture capital due diligence and equity lock-up periods.


Among these startups, some grew into foundational industry infrastructure, but many more failed or were exposed as fraudulent. This resulted in massive investor losses, providing the core justification for stringent regulatory scrutiny worldwide.


There were no institutional gatekeepers back then: developers didn't need to partner with banks, as all transactions could be conducted using cryptocurrency; they didn't need to apply for state money transmitter licenses, as regulators often didn't understand the tokens they issued; and they didn't need to actively acquire customers, as early users found them through social media rather than corporate procurement channels.


The barriers to entry, both in terms of capital and compliance costs, were almost zero. While this led to significant industry chaos, it also gave rise to a multitude of highly innovative experiments in finance and social organization.


The New Industry Reality


Today, the industry's operating logic has fundamentally changed. A crypto company serving clients in the US, EU, or Asia must adhere to a licensing and regulatory framework strikingly similar to that of traditional banks.


Industry licensing guide data shows that if a startup aims to operate comprehensively across multiple US states, total compliance-related expenditures in the first three years can reach $750,000 to $1.2 million. As the company scales, ongoing annual compliance costs can exceed $2 million.


The New York BitLicense is widely regarded as the most stringent state-level crypto license in the US. Licensing consultants typically advise applicants to allocate over a year for preparation and budget for substantial legal, compliance, and operational expenses.


The EU's Markets in Crypto-Assets (MiCA) regulation sets tiered minimum capital requirements: €50,000 for advisory services and €150,000 for trading platforms. However, these figures represent only the foundational costs. The real major expenditures lie in the mandated governance structures, dedicated compliance personnel, and continuous high-frequency reporting mechanisms. Analysts note that this compliance framework has significantly increased the operational costs of European crypto businesses compared to 18 months ago.


US regulatory rules are gradually becoming clearer, but at a cost: The GENIUS Act establishes a federal framework for regulating payment stablecoins, but its implementation details and effective timeline still require accompanying regulatory guidance, which must be finalized no later than 18 months after the Act's passage. Meanwhile, the CLARITY Act, which aims to standardize market structure, is still under Senate review and has not yet been enacted into law.


While regulatory standardization benefits the industry's long-term development, it also significantly raises the hard barriers to legal operation. Licensing practitioners state that these high compliance costs create a natural moat, protecting early entrants from low-cost new competitors.


The collapse of Terra and the implosion of FTX fundamentally altered venture capital's investment logic in the crypto space. Gate Ventures data shows that annual VC investment in crypto plummeted from a peak of over $44 billion in 2022 to around $9 billion in 2024, before recovering to over $20 billion in 2025.


According to Galaxy Digital, in Q1 2026, venture firms completed 355 crypto-related investments, deploying approximately $4 billion in total. The median deal size exceeded $4.5 million, hitting a record high. Later-stage, mature companies captured 57% of all invested capital, while the proportion of pre-seed deals shrank to 19%.


Concurrent analysis by CryptoRank shows an even more extreme concentration of capital: just 9 Series C and later deals, surging 1020% year-over-year, accounted for 28.4% of all venture funding. Combined, seed and pre-seed rounds represented only 5.2% of total funding raised. Analysts term this a "barbell market": substantial capital at the very early and very late stages, with a sustained contraction in mid-stage growth funding—the very stage that traditionally supported companies in expanding their institutional client base and scaling operations.


Funding for early-stage startups continues to shrink: in Q1 2026, investors committed just under $1.1 billion to 8 new crypto-focused venture funds, the lowest quarterly fundraising figure since 2020.


Today, funding is highly concentrated in the hands of a few, unprecedentedly large institutions. Andreessen Horowitz announced in January 2026 the closing of several multi-strategy venture funds totaling over $15 billion. The firm stated this sum represented more than 18% of all VC capital deployed in the US in 2025.


Dragonfly completed fundraising for its fourth fund of $650 million in February. Its managing partner, Robbie Hadick, bluntly stated that the entire crypto venture ecosystem is undergoing a "mass extinction event."


VC preferences have also shifted. Galaxy data shows that nearly 60% of Q1 2026 funding flowed into trading platforms and lending infrastructure. Payments and prediction markets—infrastructure projects serving institutional clients—secured the quarter's single largest funding rounds, with Kalshi raising approximately $1 billion.


Merger and acquisition (M&A) activity is filling the market gap left by native startups and VC incubation. PitchBook data indicates 267 disclosed crypto M&A deals in 2025, totaling $8.6 billion, nearly four times the 2024 total. The pace is accelerating: from $272 million deployed in crypto M&A in Q4 2025, the figure skyrocketed to $7.23 billion in Q2 2026—a more than 26-fold increase in six months. Coinbase's $2.9 billion acquisition of Deribit remains the largest crypto M&A deal in history. Ripple spent $1.25 billion acquiring prime broker Hidden Road, choosing to build its institutional service capabilities through acquisition rather than in-house development.


Distribution Channels as Corporate Moat


Technological innovation is no longer the decisive factor for crypto companies. In 2026, the winning companies succeed not through protocol-level innovation, but through banking partnerships, enterprise client relationships, multi-jurisdictional licenses, and the brand trust that allows institutional partners to operate with confidence.


This is why acquisitions have become the optimal path for rapid market entry—theoretically, a company could build all these capabilities internally, but the process is lengthy. The core value of Coinbase's acquisition of Deribit was gaining a compliant derivatives license and years of accumulated institutional trust. Onboarding institutional clients to a new trading platform often requires months of due diligence; the value of that license and client base far exceeds that of Deribit's codebase.


The same logic applies to Ripple's acquisition of Hidden Road. The industry calls such deals "bridge acquisitions": established giants directly acquiring compliance credentials and distribution channels, bypassing the cost of building from scratch.


Banking partnerships are a core bottleneck that cannot be overcome by technology alone. Even if a startup's product is technically flawless, without a willing banking partner to custody fiat reserves, the project cannot launch. For businesses dependent on fiat on/off-ramps, this hurdle can be fatal.


Companies with established banking channels possess a massive competitive advantage unrelated to technical prowess. The logic applies similarly to licensing: companies that already hold BitLicenses or MiCA licenses have crossed the time and capital thresholds that new entrants still face. Regulators also favor applicants with proven compliant operational records overseas, amplifying the first-mover advantage. The industry trust accumulated over years of regulatory scrutiny is an intangible asset that cannot be bought with a single funding round.


While industry maturation brings obvious benefits, it comes at a cost, and opinions on the future trajectory diverge significantly. The optimistic view is clear: higher barriers to entry make it difficult for underfunded, unaudited, and low-quality projects to launch, potentially preventing disasters like vaporware ICOs or the collapse of Terra's algorithmic stablecoin.


A clear regulatory framework attracts institutional capital. Licensed exchanges, compliant custodians, and audited stablecoin issuers form a standardized system, giving pension funds and traditional banks the confidence to participate. This regulated system reduces the flow of underfunded projects into compliant distribution channels, and regulators gain clearer tools to police misconduct.


However, significant concerns persist. Founders lacking capital, industry connections, and institutional resources face far greater challenges compared to five years ago. Even engineers with groundbreaking on-chain infrastructure ideas must first raise substantial funds, partner with licensed entities, or limit their scope to avoid regulated consumer-facing businesses, scaling later.


Venture capital's increasing preference for mature infrastructure reduces funding for frontier exploratory projects. Experimental fields like decentralized social media, novel governance mechanisms, and innovative wallets are seeing dwindling research and development funds.


Industry influence is increasingly concentrated in the hands of a few companies with capital, licenses, and distribution power. Latecomers can only compete for existing market share within the framework established by the incumbents.


This trajectory has played out in other industries before: post-2008 financial crisis, banking consolidation continued, with only large institutions able to bear high compliance costs; the payments industry consolidated around providers with scale advantages in risk management and cross-border settlement; social media giants leveraged ample capital to build safety and trust systems that smaller platforms couldn't match.


Each of these sectors experienced a freewheeling, innovative early stage, until regulation and capital requirements rose to levels only resource-rich incumbents could sustain.


The very purpose behind the crypto industry's creation was to escape the monopolistic structures of traditional finance. Yet, as the data and current reality demonstrate, the crypto industry is replicating the maturation curve of other sectors. Entrepreneurs without capital, licenses, or backing from giants must now assess for themselves: within this development curve, does any space remain to create something entirely new from the ground up?

Preguntas relacionadas

QHow has the barrier to entry for crypto startups changed from 2017 to 2026, according to the article?

AThe barrier to entry has dramatically increased. In 2017, startups could launch with just a whitepaper, a GitHub repository, and raise funds via ICOs with minimal costs and almost no regulatory hurdles. By 2026, crypto firms targeting regulated markets like the US, EU, and Asia must assemble legal and compliance teams, establish banking partnerships, build full AML systems, and secure significant capital to meet licensing requirements, making the entry cost comparable to traditional finance.

QWhat are the key factors driving the high compliance costs for crypto businesses in regulated markets today?

AKey factors include the need to obtain stringent licenses (e.g., New York's BitLicense, EU's MiCA), which require significant legal preparation and ongoing reporting. These regimes mandate specific governance structures, dedicated compliance personnel, and minimum capital reserves (e.g., €50k to €150k under MiCA). Total compliance-related expenditures for a US multi-state operation can reach $750k to $1.2 million in the first three years, with annual costs exceeding $2 million for larger firms.

QHow has venture capital investment in the crypto sector shifted in recent years, as described in the article?

AVC investment has shifted dramatically towards larger, later-stage deals and away from early-stage funding. While total annual investment rebounded to over $20 billion in 2025 from a 2024 low of ~$9 billion, the distribution changed. In Q1 2026, later-stage mature companies captured 57% of all funds, with the median deal size hitting a record high. Seed and pre-seed rounds accounted for only 5.2% of total funding. This creates a 'barbell market' with money concentrated at the very early and very late stages, squeezing mid-stage growth funding.

QWhy are 'bridge M&A' acquisitions like Coinbase buying Deribit or Ripple acquiring Hidden Road becoming a preferred strategy for crypto firms?

AThese 'bridge M&A' deals are preferred because they allow established firms to quickly acquire critical non-technological assets that are now essential for success: compliance licenses, institutional trust, banking partnerships, and pre-existing customer distribution channels. Building these capabilities from scratch is time-consuming and expensive. For example, Deribit's value to Coinbase was primarily its licensed platform and trusted relationships with institutional clients, not just its technical infrastructure.

QWhat is the core dilemma or trade-off presented by the crypto industry's maturation, according to the article's conclusion?

AThe core dilemma is that while maturation brings benefits like reduced fraud, institutional investment, and clearer regulations, it also replicates the monopolistic tendencies of traditional finance. High compliance costs, concentrated VC funding, and the importance of distribution channels create significant barriers for new, under-resourced founders. This centralizes power and market share among a few well-funded, licensed incumbents, potentially stifling the innovative, decentralized experimentation that defined the industry's origins.

Lecturas Relacionadas

El primer filósofo de IA del mundo, en sus 9 años en Google DeepMind: Corriendo por la seguridad de la AGI

**Resumen: El filósofo dentro de Google DeepMind: 9 años abogando por la seguridad de la AGI** Iason Gabriel, filósofo político de Oxford, ha trabajado durante nueve años en Google DeepMind, siendo durante mucho tiempo el único filósofo en un laboratorio de IA de vanguardia. Su tarea era responder a preguntas fundamentales sobre la ética y la naturaleza de la IA. Gabriel desarrolló un marco de alineación que considera cuatro partes: el sistema de IA, el usuario, el desarrollador y la sociedad, reconociendo que sus intereses pueden entrar en conflicto. Este marco, adoptado por el equipo de seguridad de AGI de DeepMind, influyó directamente en las decisiones de entrenamiento de modelos como Gemini. Su investigación temprana sobre los riesgos de la antropomorfización llevó a que los modelos de lenguaje de Google se entrenaran para no fingir ser humanos. Sin embargo, el rápido despliegue tecnológico supera el ritmo de la reflexión ética. El incidente de 2025, donde un hombre estadounidense acabó con su vida tras miles de mensajes con Gemini, mostró las limitaciones de los mecanismos de seguridad. El concepto de "social reward hacking", acuñado por Gabriel, describe cómo una IA entrenada para complacer puede terminar socavando el juicio humano. El panorama actual está dominado por una carrera masiva de inversión (se estiman 670.000 millones de dólares para 2026), lo que presiona a las empresas como DeepMind a cambiar a un "modo de guerra" comercial. Este contexto hace que líneas rojas éticas anteriores, como la prohibición de aplicaciones militares, se vuelvan difíciles de mantener, como demuestra un acuerdo de 2026 entre Google y el ejército de EE.UU. Gabriel y su equipo ahora estudian el impacto sistémico de la AGI en la economía y la política, anticipando una transformación comparable a la Revolución Industrial. Lo que comenzó como una pregunta sobre la IA ("¿Qué es? ¿Es segura?") ha evolucionado hacia una cuestión filosófica más profunda sobre nosotros mismos: ante una máquina que rivaliza con capacidades humanas únicas, **¿qué somos nosotros?**

marsbitHace 40 min(s)

El primer filósofo de IA del mundo, en sus 9 años en Google DeepMind: Corriendo por la seguridad de la AGI

marsbitHace 40 min(s)

Gemini 3.5 Pro se filtra en secreto, superando a Fable 5 en frontend

**Gemini 3.5 Pro: Filtraciones revelan su capacidad superior en generación de código front-end, superando a Fable 5** Tras dos meses de espera, las filtraciones apuntan a que Gemini 3.5 Pro, el modelo estrella de Google, podría lanzarse el 17 de julio. Lo más destacado es su aparente salto de capacidad en la generación visual y de código para front-end, donde según pruebas de desarrolladores superaría a Fable 5. Sus mejoras clave incluyen: un mejor criterio de diseño (colores, espaciado), interfaces de usuario más limpias y con menos código redundante, una generación de SVG significativamente más precisa y capaz, y una mayor completitud en la creación de páginas con una sola instrucción. Ejemplos mostrados generan desde retratos en SVG complejos hasta una isla flotante estilo steampunk con Three.js, todo a partir de descripciones breves. Sin embargo, el modelo no sería superior en todas las áreas. En tareas de razonamiento complejo, ingeniería de software a gran escala (nivel repositorio) o ejecución de agentes de larga duración, las filtraciones indican que aún estaría por detrás de Fable 5 y GPT-5.6. La demora en su lanzamiento se atribuiría a que Google no solo amplió el modelo, sino que realizó un nuevo preentrenamiento desde una base diferente a la de Gemini 3.5 Flash. Este nuevo "cimiento" también serviría para 'Nano Banana Pro', un futuro modelo de generación de imágenes que competiría con GPT-Image 2. La posible llegada de Gemini 3.5 Pro marca un intento de Google por recuperar terreno en la intensa competencia de modelos de IA, centrando su fortaleza en un ámbito específico mientras trabaja para igualar a sus rivales en otros frentes. La comunidad espera la confirmación oficial para el 17 de julio.

marsbitHace 43 min(s)

Gemini 3.5 Pro se filtra en secreto, superando a Fable 5 en frontend

marsbitHace 43 min(s)

¿Ethereum se prepara para aligerar la carga? ¿Qué opinan los guardias E sobre la mejora 'Lean Ethereum' y su impacto futuro?

Autora: Biteye Tras una racha bajista en junio, el precio de Ethereum (ETH) se ha recuperado recientemente, impulsado por las expectativas de recortes de tasas. En este contexto, Vitalik Buterin ha presentado el mapa de ruta de "Lean Ethereum", descrito como la "tercera gran actualización" de la red. El objetivo es rediseñar desde principios fundamentales la capa de consenso (CL), la capa de datos (DL) y la capa de ejecución (EL), buscando hacer a Ethereum **más simple, seguro (resistente a la computación cuántica), verificable y escalable**. Este movimiento coincide con la reestructuración de la Fundación Ethereum (despidos del 20%) y la creación de nuevas organizaciones como EthLabs, sugiriendo un esfuerzo colectivo para que el ecosistema avance de manera más ágil. **Opiniones optimistas:** * **Sassal.eth (The Daily Gwei):** Considera que reducirá drásticamente los requisitos para operar un nodo, aumentando la descentralización. Lo ve como el estado final deseable para una blockchain. * **Ryan Sean Adams (Bankless):** Interpreta Lean Ethereum como un regreso del enfoque hacia los problemas fundamentales del protocolo: seguridad, escalabilidad, privacidad y credibilidad a largo plazo. * **BITWU y otros:** Lo ven como una fase de "depuración" para crear una base mínima y confiable, que responda a la pregunta de por qué Ethereum seguirá siendo relevante en la próxima década. * **gigi发财猪 y Xiyu:** Perciben una reorganización positiva dentro del ecosistema y creen que la nueva narrativa puede ayudar a reparar el sentimiento del mercado. **Opiniones cautelosas:** * **Ignas (investigador DeFi):** Aunque la hoja de ruta es atractiva, señala que no aborda la economía del token y que la competencia en el espacio RWA es feroz. Su éxito depende de una entrega puntual. * **Dankrad Feist (ex-investigador de la Fundación Ethereum):** Coincide en la dirección, pero critica el horizonte temporal de 3-4 años como demasiado lento para la velocidad actual del mercado. **Conclusión:** Mientras los optimistas ven en Lean Ethereum una reafirmación crucial de la visión descentralizada a largo plazo, los cautelosos enfatizan que las buenas ideas deben traducirse en entregas rápidas y valor tangible para sostener el precio de ETH. En cualquier caso, la propuesta ha devuelto a Ethereum una narrativa técnica propia en un momento clave.

marsbitHace 2 hora(s)

¿Ethereum se prepara para aligerar la carga? ¿Qué opinan los guardias E sobre la mejora 'Lean Ethereum' y su impacto futuro?

marsbitHace 2 hora(s)

AI se pliega, Fable 5 y GPT-5.6, convirtiéndose en un privilegio de unos pocos

**Resumen en español (europeo):** Vivimos en un mundo de "IA plegada". Mientras una minoría privilegiada utiliza modelos de élite como **Fable 5 o el próximo GPT-5.6**, la gran mayoría del público solo tiene acceso a modelos básicos o versiones gratuitas y limitadas, como ChatGPT de 8B-30B parámetros. Esta brecha crea dos realidades paralelas: una donde la IA es una herramienta revolucionaria y otra donde parece un juguete poco útil. El problema va más allá de la potencia. Para el usuario común, la IA es un chatbot que responde preguntas. Para la élite (ingenieros, empresas), es un sistema completo de trabajo: diseña arquitecturas complejas, escribe código, gestiona proyectos y hasta ofrece consejo médico preciso (mientras la versión gratuita tiene una alta tasa de error en temas de salud). El coste es una barrera enorme: se habla de gastos de **1000 dólares diarios** en proyectos con Fable. Los flujos de trabajo más avanzados combinan varios modelos, cada uno con una función especializada (creatividad, arquitectura, ejecución, revisión), multiplicando la productividad. Sin embargo, para la mayoría de las tareas empresariales comunes, modelos más asequibles como GPT-5.5 podrían ser suficientes si se integraran mejor con los datos y procesos reales. La conclusión es cruda: **el verdadero poder de la IA se está convirtiendo en un privilegio**, alejándose del público general y acentuando una nueva forma de desigualdad digital. El futuro se está "plegando", y la brecha entre quienes usan herramientas de juguete y quienes usan superinteligencias se hace cada vez más profunda.

marsbitHace 2 hora(s)

AI se pliega, Fable 5 y GPT-5.6, convirtiéndose en un privilegio de unos pocos

marsbitHace 2 hora(s)

Trading

Spot
活动图片