What Are You Actually Paying For? The Valuation Deep Dive.

The Optical Engine Powering the AI Era — Part 4 of 5

Part 1 — What Lumentum DoesPart 2 — The Financial StoryPart 3 — NVIDIA's $2B BetPart 4 — The ValuationPart 5 — The India Angle

Let me open with a number that should stop you cold.

On the same week in mid-2025, Lumentum's stock was trading at $80 per share. Today, it's at $945. That's a 1,076% move in roughly twelve months.

The business didn't change 10× in quality. The technology didn't transform. The same fab, the same InP process, the same engineers. What changed was the market's perception of what this business is worth — and that perception swung from deep pessimism (during the inventory bust) to extraordinary optimism (during the AI-driven demand surge).

Which brings us to the only question that matters in Part 4: Is $945 a reasonable price to pay for what you're actually getting?

I'm going to be honest with you here in a way that a lot of financial content isn't. This post will not tell you to buy, sell, or hold. What it will do is show you exactly what assumptions the market is embedding at current prices, what happens to your money in three different scenarios, and what every serious investor in this name needs to understand before committing capital.

First: A Quick Primer on Valuation — Why It's Not Just "Is the Business Good?"

Many retail investors confuse two separate questions:

Question 1: Is Lumentum a great business?

Question 2: Is Lumentum's stock a great investment at $945?

These are not the same question. And the gap between them is where most investment losses occur.

Here's a simple analogy. You know that HDFC Bank is one of the greatest banks in India — arguably the best managed private sector bank in the country. But if someone offered to sell you HDFC Bank stock at 10× its book value and 60× next year's earnings, would you still buy it? Probably not — because you'd be overpaying for an excellent business, and overpaying for great businesses is how you generate poor returns.

Valuation analysis is the discipline of figuring out the gap between business quality and market price. A world-class business at a reasonable price is a great investment. A world-class business at an unreasonable price can be a mediocre or terrible investment. The two things travel together in the short run — when sentiment is hot, prices rise regardless — but they diverge in the medium run, where fundamentals ultimately determine returns.

With that framing established, let's look at the numbers.

The Current Snapshot: What Is the Market Pricing?

Before we get into scenarios, let me anchor on what we're working with.

The vital statistics as of June 2026:

That last number deserves a pause. The company guided for nearly $1 billion in a single quarter. Annualised, that's a ~$4 billion run rate entering FY2027. Twelve months ago, annual revenue was $1.65 billion.

How Expensive Is It? Multiples in Context.

The left panel shows how dramatically the EV/Revenue multiple has expanded — from 2.6× at the FY2024 trough to approximately 21× today. The historical average (ex-current) was 4.6×. The current multiple is 4.6× the average. The right panel compares Lumentum's multiple to peers — and the gap is striking.

Let me decode the key multiples clearly.

EV/NTM Revenue: ~16x

This means the market is valuing Lumentum at 16 times next twelve months' expected revenue. For context, TSMC — the world's most advanced chip manufacturer, growing at 26% annually — trades at ~11x. NVIDIA itself trades at ~22x. So Lumentum is priced comparably to NVIDIA on a revenue basis, even though NVIDIA has structurally higher margins and more diversified exposure.

When a semiconductor component company trades at NVIDIA multiples, the market is embedding two beliefs simultaneously: (a) growth will be extraordinary, and (b) margins will expand dramatically. Both need to happen. If only one does, the multiple contracts.

Forward Non-GAAP P/E: ~124×

Based on analyst consensus non-GAAP EPS of ~$7.60 for FY2026, the stock trades at ~124× forward non-GAAP earnings. This is an extraordinary multiple. For comparison, NVIDIA at its most frothy traded around 70–80× forward earnings.

The important nuance: this multiple compresses rapidly as EPS scales. If FY2027 non-GAAP EPS reaches ~$15 (at 43% revenue growth and margin expansion to 28%), the forward P/E on today's purchase drops to ~63×. And if management's FY2028 EPS target of ~$30 materialises, the forward P/E on today's entry price would be just ~31× — which is genuinely reasonable for a business with Lumentum's moat characteristics.

The compression trade: At $945, you're not really betting on the FY2026 multiple. You're betting that FY2027 and FY2028 EPS will be large enough to justify the multiple in retrospect. That's a bet on execution, not on current fundamentals.

What Does $945 Actually Believe? The DCF Decoder.

This is the most important chart in this post. It reverse-engineers what assumptions must be embedded in today's stock price for a rational investor to pay $945. The answer: ~45–50% revenue CAGR for two years, margin expansion to 28–32%, and an exit multiple of ~31× FY2028 earnings. None of these individually is impossible. All of them simultaneously, without stumble, is demanding.

The concept I'm applying here is sometimes called a reverse DCF (Discounted Cash Flow analysis). Instead of projecting cash flows and solving for a price, you start with the price and solve for what growth and margin assumptions must be embedded.

The DCF assumptions at $945:

The current price implies approximately:

Now let me stress-test each of those:

Revenue CAGR of 45–50%: Demanding but visible. The Q4 FY2026 guidance midpoint of ~$985M implies an annualised run rate of ~$3.9B entering FY2027. Growing that to $8B by FY2028 requires another doubling in 18 months. Possible — but it requires the OCS backlog to convert rapidly, CPO revenue to begin ramping, and 1.6T demand to sustain or accelerate. There's a $400M OCS backlog, a CPO order already booked for H1 2027, and 35 million 1.6T transceiver units being forecast for this year alone. The demand is real. The question is execution.

Operating margin of 28–32%: Plausible and already trending. Non-GAAP operating margin is already at 30%+ in Q3 FY2026. Management guided 30–31% for Q4. So reaching 30%+ by FY2028 isn't an ambitious projection — it's essentially continuity of the current trend. This is the least speculative of the four assumptions.

Exit multiple of ~31× FY2028 EPS: The pivotal variable. If the AI infrastructure narrative stays intact in 2028 — which is an assumption about three years of sustained hyperscaler capex — a 30–35× multiple on a deeply embedded photonics infrastructure company is defensible. But if the narrative cracks — if there's a DeepSeek-style architecture shift, if AI scaling laws disappoint, if a competitor closes the technology gap — multiples will collapse faster than fundamentals.

No significant execution stumble: The hidden assumption. Every projection assumes no fab yield issues, no supply chain disruptions, no key customer loss, no balance sheet stress from capex. History says hardware companies stumble. The question is when and by how much.

Three Scenarios. One Price. Which Do You Believe?

Three scenarios, clearly labelled with their probability weights, implied FY2027 revenues, margins, EPS, and stock price targets. The orange dashed line represents today's $945. Read this chart as: what's the expected range of outcomes for someone buying at this price?

Let me walk through each scenario with the actual logic, not just the numbers.

Bull Case ($1,200–$1,400 target · 30–35% probability)

This scenario assumes the AI buildout accelerates beyond current forecasts. OCS reaches a $1B+ annual revenue run rate by end of FY2027 instead of FY2028. CPO revenue begins ramping in Q1 FY2027, adding a meaningful third revenue layer. The Greensboro fab comes online on time and without yield issues. 1.6T demand continues to surprise analysts on the upside.

Non-GAAP EPS in FY2027 approaches $22. The market, seeing a clear path to $30 EPS in FY2028, assigns a 40–50× multiple on FY2027 earnings. Stock reaches $1,200–$1,400.

What needs to be true: OCS outperforms its own backlog conversion schedule. CPO ramps faster than guided. Hyperscaler capex does not slow meaningfully. Fab execution is clean.

Base Case ($960–$1,100 target · 40–45% probability)

The analyst consensus plays out. Revenue grows 40–45% annually, margins sustain at 27–30%, non-GAAP EPS reaches ~$15 in FY2027. The market assigns a 55–65× forward P/E — consistent with a high-growth hardware company mid-cycle. Stock trades broadly in line with or modestly above today's price.

This is essentially the "nothing unexpected happens" scenario. Given that the stock has already moved 1,076% in a year, "nothing unexpected" translates to modest returns from here.

What needs to be true: Execution in line with guidance. No major positive surprises and no major negative surprises.

Bear Case ($450–$622 target · 20–25% probability)

A hyperscaler capex pause — for any reason, including macroeconomic concerns, AI model efficiency improvements reducing compute requirements, or geopolitical disruption — causes order book deterioration. Revenue growth slows to 15–20% in FY2027, margins compress back to 15–18% as volume drops and fixed fab costs bite. Management cuts guidance.

At that point, the market no longer views Lumentum as an AI infrastructure platform — it reclassifies it as what it was in FY2024: a cyclical component manufacturer. The multiple collapses from 120× earnings to 35–40×. Stock falls to $450–$622.

This is not an extreme scenario. This is exactly what happened in FY2024 from a much lower valuation base. The 2024 collapse started with inventory corrections and ended with GAAP losses of $546 million for the year. It can happen again. The difference this time is that you'd be entering at 10× the price.

What causes this: Hyperscaler capex moderation, China-US diplomatic normalisation (reducing geopolitical moat), significant delay at Greensboro fab, or a technology shift (CPO maturity making pluggable optics obsolete faster than Lumentum can pivot).

Historical Drawdowns: The Stock Has Done This Before

The chart shows two historical drawdown cycles for $LITE — the 2020 peak-to-trough at -52% and the devastating 2022–2024 bust at -72%. The projected bear case drawdown of -34% to -52% from today is not anomalous for this stock — it's a repeat of patterns that have already occurred twice in the last five years.

This matters for position sizing more than anything else in this post. The quality of Lumentum's business did not prevent a -72% decline between 2022 and 2024. The technology remained differentiated. The customer relationships remained intact. The InP process knowledge didn't evaporate. And yet the stock fell from ~$135 to ~$38.

Why? Because the market re-priced the business from "scarce, high-growth component supplier" to "cyclical hardware company with falling revenues." The fundamental business didn't change much. The sentiment changed completely.

At $945, you are buying at 10× the 2024 trough price. If sentiment shifts again — if hyperscalers blink, if AI scaling shows cracks, if a competitor catches up — the magnitude of the re-rating would be severe. Not because Lumentum becomes a bad business, but because it becomes a less-loved business. And at these multiples, that is enough.

The Probability-Weighted Math

Let me pull this together into a single expected value calculation.

Expected return from $945 entry, probability-weighted across scenarios:

That's a 7.3% expected return from a stock that has delivered 1,076% in twelve months. The risk-reward at current levels is not asymmetric in your favour. The upside is real but limited relative to the downside magnitude. An S&P 500 index fund historically returns 10–12% annually with far lower volatility.

This is the honest math of where the stock sits. It doesn't mean the stock can't go to $1,400 — it absolutely can, especially if the bull case plays out. But the expected return, probability-adjusted, is not the home run that the past year's performance suggests.

What Would Make It More Compelling?

Two things could make the entry point more attractive than it is today:

1. A 20–30% correction back to $650–$750

At ~$700, the bull case generates ~$700 in additional upside (+100%), the base case generates ~$330 (+47%), and the bear case generates ~$150 of additional loss (-25%). The probability-weighted math flips significantly in favour of the buyer. This is why the $622 technical support level has meaning — at that price, the risk-reward looks genuinely attractive.

2. A clear Q4 FY2026 beat + FY2027 guidance above consensus

If Q4 revenue comes in at $1.05B+ (above the $1.01B guidance ceiling) and management guides FY2027 revenue at $5B+ with explicit OCS quarterly targets being raised, the current price would look more justified. This is the "growth outpaces valuation" scenario where paying a high multiple makes sense because the earnings catchup is faster than expected.

Neither of these requires predicting a crash or waiting forever. They require patience — and a clear entry framework rather than emotion-driven buying.

Where Does This Leave You?

Lumentum at $945 is a stock that could go to $1,400 and make you look brilliant, and a stock that could go to $500 and make you regret the entry, both in the same 18-month window.

The business quality is genuine. The demand tailwinds are real. The NVIDIA partnership is not theatre. OCS is not a fantasy — it's a $400M+ backlog with named hyperscaler customers.

But the price today offers thin margin of safety. You are not buying a misunderstood, undervalued business. You are buying a fully understood, richly valued business and betting that it executes its way out of the valuation. That's a bet many intelligent investors have made profitably — but it requires position discipline, time horizon clarity, and the emotional resilience to hold through what history says will be a violent drawdown somewhere in this cycle.

The most important sentence in this post: At $945, the bear case drawdown (-34% to -52%) is larger in absolute dollar terms than the bull case upside (+27% to +48%). The stock needs to go up more to make the bull case than it needs to go down to reach the bear case. That asymmetry tells you everything about entry discipline.

In Part 5 — the final post — we bring all of this home specifically for the Indian investor: the LRS mechanics, the currency math, the tax structure, whether systematic accumulation makes sense, and the one question that matters more than any other: is there an Indian-listed company that gives you this exposure without the friction?

Manu | Curious Investing Insights

Applying institutional analytical depth to markets that matter for Indian investors

This is not investment advice. Do your own research. The author may hold positions in securities discussed. All scenario analysis involves assumptions that may prove incorrect.

Tags: Lumentum LITE · Valuation · DCF Analysis · Bull Bear Base Case · AI Infrastructure · US Stocks for Indian Investors · Investment Framework · Risk Management