Deep tech IP roll ups & investment consolidation: A New Hope?

When a climate hard-tech start-up winds down, CTRL-S pays its founders, with consulting fees and equity in CTRL-S, to walk through every decision they made on a blackboard. Silvan Aeschlimann, its founding head of due diligence, said those debriefs produce what buyers want most, a list of red flags. CTRL-S buys the IP of such companies so that years of R&D do not vanish with them. After three deals it has learned to value raw data and documented failure above the patents most people would expect it to chase.

Silvan has spent most of the past seven years on direct air capture. Of about 150 DAC start-ups, more than a third have pivoted, been acquired or gone out of business. The consolidation is happening to the sector, not by design, but Silvan saw three upsides. DAC companies are usually funded for an edge in one component, so the best parts could be combined into a better system. Pooled results would show how each approach really performs. And companies would stop repeating each other's dead ends.

A participant from green hydrogen equipment doubted the first point. An electrolyser built to run flexibly is engineered as a whole, so a patent lifted from another company may not fit, and acquired IP might suit a large corporate entering the field better. Silvan saw no blueprint answer, but said industrials starting from scratch value a map of what has not worked because it saves them R&D.

The IP never failed, but it was just the wrong moment in the funding cycle.

Patents as defensive islands

A serial deep-tech founder and investor said IP has to be there, because without it a company is not deep tech, but it is overvalued. Its value is the right to stop someone else, and pricing usually starts from how much was spent on it.

Ben Murphy was more sceptical still. Of the patents that fail, a third fall on entry, a third later and a third in litigation, and an owner may not know whether one holds until late-stage litigation. Start-ups often lack the money to enforce a patent, so it becomes a defensive island. Buyers still expect a portfolio. Large chemical companies cannot assess IP in depth, the serial founder said, but they want to see one.

CTRL-S steers away from betting on single patents, whose worth is very hard to judge in DAC. "We don't want to make our money from litigation," Silvan said, and even a hundred thousand is a big bet for a young company. It concentrates on know-how that was never patented.

Data, failure and a blackboard

The first lesson was that experimental data is worth more than expected once pooled. One dataset is just a dataset. At 50, 100 or 150, patterns appear that help in discovering new materials. The second was the red flags, which industrials, consultants, venture studios and accelerators all wanted. The third was that the technology is often sound. "The IP never failed, but it was just the wrong moment in the funding cycle," Silvan said. In one case a corporate partner wanted six months to decide when the start-up had one month left.

For the first deal, CTRL-S compared the IP with rivals', ran the performance claims through techno-economic and life-cycle models and looked for adjacent markets. Anything that works for direct air capture is almost overqualified for point-source capture. Chain of title was the hardest step, because a distressed company has to move fast and a board with different stakes makes handing over data a headache. The process could decide whether to buy, but not where the value lay.

A company rarely folds simply because it ran out of money, Silvan said, and the debrief has to find the specific misfit. CTRL-S now splits each package into business history, technical history, patents, experimental data, engineering know-how and pilot data, since each may suit a different buyer. Some deals come through insolvency bids. In others, founders who know they will not make it come early and negotiate a stake in what follows.

Rescues and roll-ups

Ben works at IP Group, an evergreen deep-tech investor that has tried both. An AIM-listed solid oxide fuel cell company it backed was worth about 320 million before 2008 and almost nothing by early 2012. IP Group had days to decide, put 20 million into a recovery round and eventually took out about 120 million as the company grew to around a billion pounds. On the way the company picked up IP from a failed fuel cell business in Australia. An IP roll-up of European airborne wind companies, which had different answers to the problems of the US pioneer Makani Power, never fully came together.

A founder making a low-carbon ingredient for cement described a newer idea from their investors. Buy a distributor or a concrete producer, with its cash flow, customers and certifications, as a hedge in case the core technology fails. Concrete businesses run on margins of perhaps 3–5%. Because end users pay a premium for the founder's ingredient and sign long-term offtakes for it, owning one could raise its pricing power. The company is not buying anyone yet, but is looking much sooner than expected.

Investors see many companies that worked on something for ten years and then filed for insolvency, the serial founder said, mostly because the timing was wrong. For most of them the timing will come right. CTRL-S is betting that someone should keep their knowledge intact until it does.

This Ripple was hosted by Silvan Aeschlimann (CTRL-S) and Ben Murphy (Kiko Ventures) at The Drop 2026 on 16 September.