Frontier Alternatives: A Three-Pillar Research Practice Across AI, Pre-IPO Secondaries, and Warrant Recoveries

A three-pillar framework for thinking about the next decade of private-market dislocation.

We just finished a long internal exercise: writing down — slide by slide — how we think about frontier alternative investments. The three lanes we research, why we think they matter right now, and what discipline keeps the research honest. The document is private, but the thesis behind it isn’t. We publish on every piece of it, every week, at gp.advalorem.io.

What follows is the public-facing version of the framework.

If you read AdValorem Research, none of the individual pieces will be new. What is new is putting them next to each other and saying: these three things are happening at the same time, and they’re uncorrelated enough to actually compose into a research practice.

Three independent tailwinds, all converging in 2026

There is a version of “alternative investing” that means “anything that isn’t large-cap public equity.” We don’t find that useful. The version we find useful is much narrower: situations where the structural market mechanics have shifted recently enough that the historical playbook hasn’t caught up.

We see three of those right now.

One — the frontier-tech inflection. Physical AI is on the factory floor. Humanoid platforms are shipping production parts at BMW. Quantum is moving from research labs into production stacks at PsiQuantum and IonQ. The window to enter the next decade of compute, physical automation, and post-classical computing at seed and early stage is genuinely narrow and genuinely asymmetric. The companies are private, the rounds are concentrated, and the access path is not “buy the ETF.”

Two — the pre-IPO supply imbalance. Median time-to-IPO has stretched to over eleven years. SpaceX, OpenAI, xAI, Perplexity, Cerebras, Neuralink — all still private. The institutional money already invested at primary needs liquidity, the secondary-market platforms are processing it, and the resulting tape prices at meaningful discounts to last primary rounds. Late-stage frontier secondaries are now a durable access path, not an opportunistic trade.

Three — the judicial unlock on accelerator warrants. In April 2024, Judge Shad Robinson in the Western District of Texas ruled that warrants issued by the Newchip / Astralabs accelerator were severable from any service obligations and freely transferable. That single ruling created the first liquid secondary market for accelerator-issued equity warrants — 81+ warrants across 30+ industries that were previously stranded. Three years later, the doctrine has held up through four subsequent Robinson rulings, and the framework is now being applied to parallel cases (Rhodium Encore is currently on appeal).

Three tailwinds. Each one is large enough on its own to be a full-time research subject. None of them depends on the other being true. That’s the point.

Pillar one — Frontier AI, Robotics & Quantum

The first pillar is single-name early-stage exposure to the companies building the physical and computational layers of the next decade. AI and Physical AI — foundation models, agentic systems, humanoid and industrial robotics. Quantum computing — qubit architectures, error correction, quantum software. Deep tech — semiconductors, novel materials, energy systems. Blockchain infrastructure — base-layer protocols and rails (we are explicit that this means the infrastructure layer, not retail token speculation).

Two disciplines anchor the way we look at these names. First, single-name positions over blind-pool index exposure. We pick the company. We never spray. We have a written underwriting framework that weights founder execution and large scalable markets above narrative momentum. Second, what we call Crash Investing philosophy — stage-flexible entry, valuation discipline, and a willingness to wait for the right price. We publish the underwriting framework as one of our education products.

You do not need to be in a fund to learn this. You can read our weekly pieces on every name in this universe — Figure, Apptronik, PsiQuantum, IonQ, Cerebras, Anthropic — at the insights archive.

Pillar two — Pre-IPO frontier growth (late-stage secondaries)

The second pillar is the other end of the same companies’ lifecycle. Late-stage frontier names approaching public markets, accessed through single-name positions and negotiated secondary structures rather than primary rounds. The thesis is straightforward: post-product-market-fit, pre-public, before institutional repricing is one of the most asymmetric windows that exists in any market — and the secondary platforms (Forge, EquityZen, Hiive, Caplight, direct broker-dealer relationships) have made it transactable in a way they were not five years ago.

The themes we cover most closely:

• Frontier AI labs — OpenAI, xAI, Perplexity

• AI compute and silicon — Cerebras-class hardware and the adjacent infrastructure

• Space and aerospace — SpaceX and the new defense stack

• Brain-computer interfaces — Neuralink-class neurotech

• Quantum (late-stage) — PsiQuantum and quantum-adjacent names

• Marketplaces and mobility — Turo, Flexport, Consensys, Binance.US

Liquidity-gap pricing is real. We wrote about it in this week’s Pre-IPO Secondaries piece.

Pillar three — Newchip warrant portfolio

The third pillar is the strangest of the three, and the one we think about the most. After the Newchip / Astralabs bankruptcy, AdValorem became the operational home for the resulting warrant portfolio — 120+ instruments across 30+ industries and 12+ countries, each carrying up to a $250K initial exercise right plus a $500K follow-on, on a 10-year exercise window severable from any service obligations.

The infrastructure to actually transact those warrants — bidding, custody, severability documentation, exit pathways — is now live as the AdValorem Warrant Exchange at market.advalorem.io. Premier listings (estate-rated names like Charge Zone, Ace Green, Qdeck, Crypto Dispensers) sit alongside active listings with $100K minimum bids and another 86 warrants currently in review.

This pillar is the most operational of the three. It is the pillar where the research desk and the legal posture and the operating infrastructure compound the hardest, because the moat is the documentation. The legal anchor is In re Astralabs / Newchip, W.D. Texas, Judge Shad Robinson, April 2024. We publish the warrant case notes and severability documentation as part of our education library.

Why three pillars and not one

Single-pillar concentration is a strategy. We respect it. It is not what we do.

We run three pillars because the failure modes don’t correlate. A 2026 macro shock that compresses the frontier-tech inflection does not impair a warrant that was issued in 2019 with a 10-year window. A regulatory change at the SEC does not change whether SpaceX is worth more or less in the secondary market. A judicial reversal on warrant severability does not affect whether a humanoid robotics company is shipping production parts.

The compounding shows up at the operating-platform layer. The same research desk that underwrites a Figure-class robotics position is also reading SEC filings on Neuralink secondaries and parsing accelerator bankruptcy dockets. The same Discord community that debates AI compute spend is also tracking warrant exchange listings. The same Substack and podcast (the Mon/Thu market-analysis show, The Syndicate Room) cover all three pillars in rotation. None of that distribution gets cheaper if you only research one pillar. All of it gets cheaper if you research three.

What we publish — and how to follow along

Everything we research lives on the public side:

• The daily insights archive — most weeks, four new pieces (Mon / Tue / Thu / Sat), spanning all three pillars

• The Syndicate Room — Monday and Thursday market-analysis podcast, ten-minute format, available on YouTube and major podcast platforms

• The AdValorem education library — long-form ebooks on pre-IPO investing, litigation finance, warrant enforcement, due diligence, and most recently a six-book Crypto Series covering OPSEC, DeFi tax, MEV, and the agentic-commerce x402 protocol

• The Discord — daily community signal and dinner events

• This Substack — weekly highlights and longer essays like this one

If this framework resonates

We are intentionally not asking anything of this post. If the three-pillar framing matches the way you already think about frontier markets, read more of it at the insights archive. If you would like to discuss any of the three pillars with us directly — not as a sales conversation, but as a thirty-minute working session on the research — you can grab time at tidycal.com/advalorem/syndicate.

The deck we just finished is private. The thesis behind it is not. We publish on it every week. That is the whole point.

—

AdValorem Research publishes daily market analysis on alternative investments, pre-IPO markets, accelerator warrant enforcement, and frontier-tech inflection points. This post is educational and reflects forward-looking views; it is not investment advice and not an offer or solicitation to buy or sell any security.