The Beat that Wasn't Enough

Science before it hits the news

Last issue, we promised a deeper look at two papers that quietly moved the goalposts for excitonic and photonic computing. Neither made a business headline. Both matter more than most things that did.

Excitons just learned to read magnets.

A team at City College of New York, led by postdoctoral researcher Pratap Chandra Adak in Vinod M. Menon's Laboratory for Nano and Micro Photonics, published a review in Nature Materials titled "Excitons in van der Waals magnetic materials" — surveying a body of work with a genuinely new idea at its center: in certain atomically thin magnetic semiconductors, the same electrons that form excitons also carry the material's magnetism. Light and magnetism stop being separate systems and start talking to each other directly.

In plain English: shine light on the right material, and you can read — and potentially write — its magnetic state optically. No wires, no electrodes, just photons. The review covers three material platforms doing this today: chromium triiodide, nickel phosphorus trisulfide, and chromium sulfur bromide.

Why should you care? Because magneto-optical reading and writing is a foundational requirement for next-generation data storage and spintronic computing — the kind that doesn't heat up the way today's electronics do. The work was funded by DARPA and the Gordon and Betty Moore Foundation. When both a defense research agency and a foundation known for backing long-shot science fund the same lab, that's usually a signal the field is about to get crowded with follow-on money.

Light just got a lattice — in time, not space.

The second paper is, if anything, stranger. An international team from École Polytechnique, Collège de France, and Germany's Helmholtz-Zentrum Dresden-Rossendorf built what's being called the world's first all-optical photonic time crystal — a material whose optical properties flicker in a repeating pattern so fast it happens roughly as quickly as the light wave itself oscillates.

Ordinary crystals repeat in space. This one repeats in time. Using HZDR's TELBE terahertz source to drive a gold-and-semiconductor metamaterial, the team modulated the material's reflectivity on a picosecond timescale — one trillionth of a second — and cut terahertz photon losses in half in the process.

By the numbers:

The terahertz band sits in an awkward gap — too fast for electronics, historically too weak for photonics to fully exploit. This result is a small number (a factor of two) attached to a genuinely new capability: on-demand, ultrafast control of light itself. As one researcher not involved in the work put it, this is early and modest — "but so were the first transistors."

Neither of these papers will move a stock price this week. Both are the kind of foundational result that, eighteen months from now, gets cited in the methods section of whatever excitonic device actually reaches a production line.

Stocks, funding, company moves

We told you last issue to watch Coherent's earnings. They landed — and the market's reaction is the most instructive thing that's happened in this sector in weeks.

The numbers were genuinely excellent. On August 12, Coherent reported fiscal Q4 revenue of $2.05 billion, up 34% year-over-year and a company record. Non-GAAP EPS hit $1.74, beating consensus estimates of roughly $1.58–1.65. Full-year revenue crossed $7.12 billion — the first time in Coherent's history it has topped $7 billion — up 22.5% for the year. Management guided next quarter's revenue and EPS above what analysts were already expecting.

Quarter at a glance:

By every conventional measure, this was a blowout quarter, driven — like almost everything in this sector right now — by AI data center and communications demand. Coherent remains a recognized NVIDIA Ecosystem Innovation Partner for co-packaged optics, putting it at the center of the industry's push to move data faster between chips without burning more power to do it.

And the stock fell anyway. Shares dropped roughly 4–7% in the sessions following the report — a classic "sell the news" reaction, but one worth sitting with rather than dismissing.

Zoom out and the context helps explain it. Coherent shares have already climbed over 220% in the past 52 weeks. A company trading at a Price-to-Sales ratio well above its historical median of roughly 2.2x doesn't get much credit for merely meeting sky-high expectations — it needs to blow the roof off, and even record numbers can read as "priced in" when the run-up has been this steep.

The reaction rippled through the rest of the sector too, echoing the pattern we flagged in Issue #001: one major name reports, and the market reprices the whole group in sympathy — this time on the way down as much as the way up.

The warning most newsletters won't publish

Coherent's post-earnings drop isn't a one-off curiosity. It's a preview of what happens across this entire sector once expectations catch up with reality. Here's what to actually watch.

⚠️ Rip #1 — The Beat-and-Drop Pattern When a stock falls on genuinely good numbers, it's the market telling you the bar has moved higher than the news itself. Coherent, up over 220% in a year, needed more than a record quarter — it needed a surprise. Watch for this same pattern in Lumentum, AAOI, and POET the next time they report. A "beat" is not automatically bullish once a stock has already priced in years of growth in months.

⚠️ Rip #2 — Valuation Is Stretched, Not Broken Coherent's Price-to-Sales ratio now sits well above its historical median. That's not a crash signal on its own — but it does mean the margin for disappointment has shrunk. Earnings-based valuation metrics like P/E become less useful here; you're paying for a growth story, and growth stories are unforgiving the moment growth merely meets expectations instead of shattering them.

⚠️ Rip #3 — Watch Insider Activity, Not Just Insider Words Alongside the earnings beat, Coherent insiders sold roughly $1.1 million in shares over the prior three months, with no offsetting insider buying in that window. That alone proves nothing — insiders sell for all kinds of personal reasons. But paired with a stretched valuation and a post-earnings sell-off, it's one more data point worth weighing, not ignoring.

⚠️ Rip #4 — The Science Is Still Years Ahead of the Stocks This week's two Nature papers are genuine progress. Neither is a product, a customer, or a revenue line. The CCNY magneto-optical work and the École Polytechnique time crystal are foundational research — the kind that historically takes five to ten years to reach a commercial device, if it gets there at all. Don't let a great science story talk you into a timeline the technology hasn't earned yet.

The Rip doesn't mean the wave is wrong — Coherent's business is genuinely strong. It means the easy money in "just buy the earnings beat" has already been made. Read the reaction, not just the headline number.

The one actionable takeaway from this issue

This week's Shore: separate the earnings clock from the science clock — they're telling you two different things, and both matter.

The earnings clock just told you that even a record quarter isn't automatically a buy signal once a stock has run this far, this fast. The science clock told you something quieter but arguably more important: three independent breakthroughs in under twelve months — the Michigan excitonic switch, the CCNY magneto-optical exciton work, and now a working photonic time crystal — are converging from completely different institutions, on completely different problems, toward the same destination: computing built on light and quasiparticles instead of pure electronics.

That convergence is the real story. Not one lab chasing a headline, but multiple independent teams — funded separately, publishing separately — all running into the same wall electronics is hitting, and all finding different doors through it.

→ Don't chase the post-earnings dip or the post-earnings pop. Coherent's reaction shows the sector is now sensitive enough to expectations that entry timing matters more than it did six months ago.

→ Track the CCNY and École Polytechnique labs, not just their headline results. Foundational science tends to cluster — a DARPA-funded lab that just published once is a lab worth watching for the follow-up.

→ Keep POET on your watchlist. It remains the clearest public bridge between today's photonics revenue and tomorrow's excitonic computing story, and it didn't have an earnings-driven correction this cycle.

→ Hold the line on the photonics/excitonics distinction. Everything in The Break this issue is photonics — real, revenue-generating, and now showing its first signs of valuation fatigue. Everything in The Swell is still fundamental science. Investing in one on the timeline of the other is the single most common mistake we see in this sector.

〰️ See you next time. — The Excitonic Wave Editorial Team

Excitonic Wave is published at wave.excitonic.com for informational and educational purposes only. Nothing in this newsletter constitutes financial, investment, or legal advice. All data is sourced from publicly available records and believed accurate at time of publication. Past performance does not guarantee future results. Always conduct your own research and consult a qualified financial advisor before making investment decisions. The editors may hold positions in securities mentioned.