Powell Industries (POWL) -- Deep Philosophical Analysis

The Seduction of Cyclical Peak Earnings

April 2026

The Core Question: Is This a Transformed Business or a Cyclical Peak?

Here is the central tension of Powell Industries: the management has genuinely done something remarkable -- they have taken a cyclical oil and gas equipment company and diversified it into data centers, electric utilities, and industrial automation. Revenue has more than doubled. Margins have gone from barely positive to almost 20% operating margin. The backlog sits at a record $1.6 billion.

And yet.

When I look at Powell's history, I see something that should give any investor pause. This is a company that earned $0.06 per share in fiscal 2021. Negative $0.62 in fiscal 2018. Negative $0.80 in fiscal 2017. The revenue went from $662 million to $396 million. Management at that time was also talking about diversification and growth markets. They were not wrong -- they were simply early, and the cycle crushed them anyway.

Charlie Munger would say: "Show me the incentives and I'll show you the outcome." When I look at the incentive structure around Powell today, I see a CEO selling shares, a CFO selling shares, and the founding Powell family -- Thomas Powell, the 10%+ owner -- liquidating tens of millions of dollars of stock in February and March 2026 at $500-600 per share pre-split. These are not people who believe their stock is cheap.

Moat Meditation: Custom Engineering Is Not a Fortress

The bulls argue that Powell's custom-engineered switchgear creates meaningful switching costs. And within a given project, they are right -- once you have specified Powell equipment for a $500 million LNG facility, you are not switching mid-construction. That is a real advantage.

But Warren Buffett's moat framework asks a different question: can the company raise prices without losing customers over a 10-20 year horizon? Here the answer is more nuanced. At the project bidding stage, Powell competes against Eaton, Schneider Electric, ABB, and Siemens -- companies with revenue bases 20-40 times larger. These are not sleepy competitors. They are all investing aggressively in the same data center and utility markets that Powell is celebrating. Eaton in particular has been vocal about its electrical infrastructure strategy and has far greater resources to throw at the opportunity.

Powell's moat is real but narrow. It is the moat of a skilled craftsman in a world where large factories can produce similar products. The craftsman can charge a premium when demand exceeds supply -- as it does today -- but when supply catches up, pricing power evaporates. This has happened before to Powell: the 2016-2018 period saw gross margins compress from 19% to 13% as capacity came online and the oil cycle turned.

The Remsdaq acquisition adds an interesting wrinkle. Electrical automation (SCADA, monitoring, controls) is stickier than hardware because it involves software integration, data protocols, and ongoing service relationships. If Powell can successfully cross-sell Remsdaq's automation into its existing customer base, the moat could widen. But we are talking about a small U.K. acquisition that has been owned for less than a year. The optionality is real but unproven.

The Owner's Mindset: Would Buffett Own This for 20 Years?

I think Buffett would admire several things about Powell: the fortress balance sheet ($501 million cash, zero debt), the 50+ year dividend history, the founding family involvement, and the disciplined approach to capacity expansion. This is a company that invests $40 million across its Houston facilities rather than making leveraged, transformative acquisitions.

But I think Buffett would also recognize that this is fundamentally a project-based business with inherent earnings volatility. The "lumpy" nature of large project wins ($100M+ LNG contracts, $75M data center orders) means that individual quarters and even fiscal years can deviate wildly from trend. You cannot compound earnings at 15% per year when your operating income can swing from +$218 million to negative in a bad cycle.

More importantly, Buffett pays attention to what he calls "the gap between perception and reality." Right now, the market perceives Powell as a secular growth story -- a data center beneficiary, an LNG infrastructure play, an electrification winner. And these narratives contain genuine truth. But the reality is that Powell is still a cyclical manufacturer where 30% of backlog is oil and gas, margins are at all-time highs that no serious person believes are sustainable, and revenue growth has decelerated from +45% to +4% in just two years.

Buffett's most famous aphorism applies: "Be fearful when others are greedy." The 36x move from $21 to $756 (pre-split) in three and a half years is the market being greedy.

Risk Inversion: What Could Destroy This Business?

Inverting the question, as Munger teaches, the existential risks are manageable:

1. A severe recession would hurt but not kill. Powell survived FY2017-2018 with losses and maintained its dividend. The $501M cash pile provides ample runway.

2. Data center spending rationalization is the most likely near-term catalyst for a correction. If hyperscalers pull back on CapEx -- as they periodically do -- Powell's commercial and industrial orders would slow significantly.

3. Competition from Eaton/Schneider ramping capacity to serve data centers could compress margins. When the giants decide to prioritize a market, small niche players feel the pressure.

4. Tariff disruption to component supply chains could delay projects and margin.

None of these would destroy the business. Powell has survived worse. The question is not survival but returns: buying at 49x peak earnings and suffering a 50-60% drawdown to fair value is perfectly consistent with the business surviving and eventually thriving.

Valuation Philosophy: The Cyclical Trap

Here is the most important thing I can say about Powell's valuation: cyclical stocks are cheapest on P/E at the peak and most expensive on P/E at the trough. This is the classic cyclical trap that catches momentum investors.

At 49x trailing earnings, a naive screener would say "this is expensive." But many investors look at the $5.13 post-split TTM EPS, project $5.50 or $6.00 for next year, and conclude the forward P/E of "only" 42-46x is justified by the secular growth story. They are looking at the multiple and not asking whether the earnings are sustainable.

On normalized, mid-cycle earnings, even assuming a structurally higher floor from data centers and utilities, Powell probably earns $0.70-1.00 per share post-split. At $252, that is 250-360x mid-cycle earnings. The stock is priced for a world in which the cycle never turns, margins never compress, and Powell becomes something it has never been -- a consistent, predictable grower.

This is not a reasonable assumption for any company. It is an especially unreasonable assumption for a custom-engineering manufacturer with a violent cyclical history, serving industries (oil & gas, data center construction) that are themselves cyclical.

The fair value range of $107-160 post-split (using 20-30x peak earnings) implies the stock needs to fall 37-57% to reach reasonable entry territory. This is not a small gap. This is a chasm.

The Patient Investor's Path

The correct approach to Powell Industries is to:

1. Acknowledge the quality. This is a well-run company with real secular tailwinds, an exceptional balance sheet, and a management team that has genuinely improved the business. It deserves a place on the watchlist.

2. Refuse to pay the price. Quality alone does not justify any price. At 49x peak earnings, you are paying for a decade of flawless execution with no cyclical downturn. History says that is a bet you will lose.

3. Wait for the cycle. The entry opportunity will come. It always does with cyclical industrials. It might come from an oil price decline, a data center spending pause, a broader market correction, or simply the natural rhythm of mega-project ordering patterns becoming lumpy. When it comes, margins will compress toward 20-22% gross, revenue will decline or stagnate, and the stock will re-rate sharply lower.

4. Buy aggressively at $112-136 post-split. At those prices, you own a quality niche manufacturer with a fortress balance sheet, real secular tailwinds, and the potential for another up-cycle. The margin of safety will let you compound through the next expansion without worrying about overpaying.

5. Target 3-5% portfolio allocation at entry. Powell is too small and too cyclical for a larger position, but at the right price, it can be a meaningful contributor to portfolio returns.

The hardest part of value investing is watching a stock go up without you. Powell has gone from $21 to $252 post-split, and every day it goes higher, the regret of not buying gets louder. But the discipline to wait for value -- to buy quality at a discount rather than quality at any price -- is what separates investors from speculators over a full cycle.

The quality is undeniable. The price is not. Wait.