Factor Capital Update - September 2025

In recent conversations with LPs and GPs, it's been noted that the summer marked a noticeable slowdown in the crypto venture markets broadly. While one can be inclined to attribute this to just summer months, I think there is also something notable in terms of the overall shifting landscape of the crypto markets that applies to AI markets in interesting ways as well. I'll try to cover all of this in this month's update.

The End of the Hype Cycle

For years, underwriting tokens felt like guessing. Valuations for Layer 1s and other early protocols often drifted far ahead of reality. A well-hyped launch, backed by the right VCs, could reliably summon a multibillion-dollar market cap without revenue, retention, or even a live product.

That dynamic is fading. We are now watching a clean separation in crypto markets between assets priced on speculation and those priced on utility. For outsiders, the idea that a token could have value without fundamentals seems absurd. But for a long time, it seemed crypto was destined to trade on goodwill and vibes alone. That era is over.

The Market Shifts

Today, fundamentals are the key focus. The market is rewarding protocols that can show durable usage and real economics. I think this is most clearly visualized by looking at this chart showing BTC, ETH and Stablecoins as percentage of overall crypto market cap. One would think that in a time where each month I seem to be highlighting more and more institutional adoption of the asset class and utility, these assets would be losing share relative to some of the other protocols that are gaining traction.

In reality BTC, ETH and SOL (not pictured) are only taking more of the capital in the market. The primary beneficiaries outside of this cohort are a handful of solidly fundamentally valued applications with much smaller market caps. For example:

• Hyperliquid, a decentralized exchange, posted ~$106M in August revenue on nearly $400B in volume, a new monthly record. Its token (ironically tickered $HYPE given it’s fundamental value) is up roughly 4x since April.

• Pump.fun, a platform for launching new tokens, has generated over $775M in cumulative revenue and continues to print seven-figure daily fee lines, leading to its successful token launch at a $4B valuation in July.

• Maple Finance, an on-chain credit marketplace, provides the most compelling before-and-after test. They had a big valuation as DeFi coins peaked in 2022 and then, after plugging along for years, released a new version of its application with real revenue generation that massively accelerated engagement, leading to $2.9B in active loans, $8M in annualized revenue, and a ~500% increase in its market cap.

These data points support the broader story: value is accruing to products people use, not just stories people tell. This shift is causing widespread dissatisfaction among crypto LPs who grew accustomed to quick wins and who are feeling like the market is depressed right now despite the big three trading near all time highs.

For investors, success now will increasingly depend on disciplined, long-term analysis and identifying applications that are creating durable businesses. Both venture and liquid fund managers must shift from the short-term, speculative mindset that defined the last decade and focus on patient, fundamentals-driven underwriting.

I observed this while Factor portfolio company Plural Finance completed their raise this Summer, a $7M round led by Paradigm. They represent a perfect example of a business providing fundamental value with a deep institutional pipeline of assets coming onto their platform. At the start of the raise, this was an uphill sell to most of the crypto investor crowd since it isn’t a hot token launch. But ultimately the market caught up to their narrative and they’re poised to truly accelerate their growth with this funding completed and a deep and exciting pipeline of investors and issuers joining the platform.

So, with the old playbook gone, where does the fast money turn?

Enter DATs…

The biggest story of the summer became the rise of Digital-Asset-Treasury (DAT) vehicles. In this model, groups acquire public companies with no meaningful business, inject capital, and use the funds to buy crypto assets like BTC and ETH for the balance sheet.

This strategy was pioneered by Michael Saylor’s Microstrategy ($MSTR ( ▼ 1.26% ) ) and effectively wraps crypto beta in an equity shell. The most dramatic recent example is $BMNR ( ▲ 7.9% ) , a company with just $8M in assets on July 27th. After announcing a $250M plan to acquire ETH, its stock rallied, allowing it to raise ~$2B in additional capital. The company is now valued at nearly $8B.

The bet is that crypto prices will rise, assisted by their own consistent buying, while they issue new equity at a premium to the value of their crypto holdings (their Net Asset Value, or NAV). As long as the stock trades for more than the tokens it holds, it's a guaranteed way to generate levered upside. While potentially profitable, this is a reflexive trade, and the inevitable unwind is not a game Factor is set up to play. Our edge is in finding non-consensus opportunities that can durably build and compound over time.

The Institutionalization of the Base Layer

Complicating things further for speculators is that the ground is shifting beneath their feet. Building the next "better, faster, cheaper" Layer 1 blockchain is now incredibly difficult because established, regulated giants are entering the space. August was a landmark month in this regard:

• Circle, the issuer of USDC, unveiled Arc, its own L1 for stablecoin finance.

• Stripe, in partnership with venture firm Paradigm, introduced Tempo, a payments-first L1 to move merchant settlement off legacy rails.

• Google Cloud teased Universal Ledger, an institutional L1 now piloting tokenized settlement with partners like CME Group.

These are not crypto-native yield farms. They are distribution, compliance, and treasury plumbing for the enterprise. Five years ago, launching a "new L1" was a default pitch that could command a $200M+ seed valuation. Today, competing at the base layer means going up against teams with immense capital, regulatory muscle, and existing customer distribution. The higher-probability path for startups is now to own a specific workflow at the application layer and slot into these institutional rails rather than trying to replace them.

Factor portfolio company Dinari represents a version of this as well, where in August they announced the launch of their Layer 1 alternative to DTCC for equity settlement. At the start, governance will come from a consortium of institutions including Gemini, custodian BitGo and asset manager VanEck, who will serve as validators and also offer custody services. This combination of specialized value proposition and institutional partnership is a winning approach now.

The AI Counterpoint

While writing about the 2021-23 era of Layer 1 fundraising, I realized it sounded a lot like what’s happening in AI venture markets right now. Top-line revenue for AI businesses is exploding, and seed rounds are happening at $2B valuations for pre-product (arguably pre-team in some cases) companies started by pedigreed researchers from the top labs, as discussed last month. Yet even the market leader, ChatGPT, which is approaching one billion monthly active users, reportedly loses billions as it scales model training and inference. OpenAI posted $12B in annual revenue in August but is still raising tens of billions to fund its data center buildout.

This is the classic "growth first, margins later" phase. As we saw with the rise of search (Google) and social media (Meta), it will likely resolve to a power law with a few dominant platforms capturing nearly all the value and a lot of the investor dollars chasing the rest of the space mostly flowing straight through to the platforms and infrastructure providers.

The Anatomy of a Durable Business

The takeaway from these parallel shifts in crypto and AI is not simply that "fundamentals are back." It’s a more specific lesson about where the most durable value will be created in the next cycle.

The platform layer in both ecosystems is consolidating into a capital-intensive arms race between giants. For most startups, competing head-on is a low-probability, high-burn endeavor. The real opportunity isn't in building the next foundational AI model or Layer 1 blockchain. It lies in using these powerful new rails to build something different: highly-focused, capital-efficient businesses.

This is the core of our thesis for Fund II, and most investors I speak with are aligned with this model. We are looking for founders who leverage these new platforms to create targeted solutions for niche markets—businesses that can achieve profitability quickly with high revenue per employee. They aren't playing the same "growth at all costs" game as the giants. Instead, they are building fundamentally outstanding companies by establishing a durable advantage in a specific market they understand well, benefiting from the rails being supplied by the giants.

I believe this is an entirely new investment category that will emerge over the next few years, differentiated from the traditional venture model that has persisted over the past half-century.

Thanks, as always, for reading.

Jake Dwyer
Founder & Managing Partner
Factor Capital