Kennametal's Tungsten Mirage: A $3.48 Quarter That Burned $212 Million of Cash

The toolmaker beat on price-raw timing, but the market is already looking through the beat to a cash conversion trap and a low-double-digit infrastructure volume hole.

KMT · Earnings Call · 2026-11-04

The beat the market refused to pay for

Kennametal reported fiscal Q1 2027 on Nov 4 with a headline that looked almost absurd for an industrial toolmaker: EPS of $3.48 versus $0.34 adjusted a year ago, EBITDA margin of 51.9%, and sales up 54% organically. The stock, however, has fallen 25% over the last 90 trading days and sits 27% below its Aug 4 high of $36.06. That disconnect is the story. The quarter was not an operating breakout; it was a tungsten price timing event that inflated reported margins while consuming cash. Sanjay Chowbey was explicit about the source: “We delivered sales and earnings above the midpoint of our outlook. Primarily driven by favorable tungsten price and cost timing alongside continued volume growth in metal cutting.” — Sanjay K. Chowbey, President and CEO · 2026-11-04 The "cost timing" phrase is doing enormous work. Price-raw timing contributed roughly $230 million to margins and about $3.14 to EPS, according to CFO Patrick Watson. That is why gross margin reached 58.9%, up 30.7 percentage points year over year and operating margin hit 41.1%. For a business that has historically earned gross margins in the low-to-mid 30s, this was not a new earnings power; it was a price/cost lag. Watson drew the distinction cleanly: “The key distinction here is timing. Our P&L benefits price realization relative to the costs recognized in earnings, while cash is being invested in higher cost inventory and supplier advances to meet our customers' needs.” — Patrick S. Watson, Chief Financial Officer · 2026-11-04 The cash statement proves the point. Free cash flow (less SBC) was -$106 million, down 298% year over year, and free operating cash flow was -$212 million. The company returned $15 million through dividends but bought back no stock; the repurchase program is paused.

Additionally, the value of our inventory balance has grown 3x over the last 18 months not because we have seen a threefold increase in physical quantities, but mainly because of tungsten prices. It is a substantial use of capital and a central focus of our cash conversion work.
Patrick S. Watson, Chief Financial Officer · 2026-11-04

That is the cleanest explanation of why this quarter's cash conversion problem is not a working-capital rounding error. It is a balance-sheet valuation problem tied to a commodity that has moved from a tailwind to a headwind.

The guidance cadence gives the game away

Management maintained FY27 sales of $3.35-$3.45 billion and introduced adjusted EPS of $5-$6. But the quarterly shape is a warning: “In Q2, we expect a benefit of $1.00 and in Q3 a headwind of $0.55 and in Q4 a headwind of $2.62.” — Patrick S. Watson, Chief Financial Officer · 2026-11-04 In other words, nearly all of the year's earnings leverage arrives in the first half, and the second half reverses hard as the price-raw timing unwinds. This is exactly what a commodity lag looks like when the underlying price stops rising. The changed variable is the tungsten price itself. Management now assumes tungsten stays around $2.7 thousand, down from the roughly $3 thousand level discussed on the prior call. That sounds helpful for cash, but it is bad for reported margins because the P&L had been benefiting from selling high-priced inventory into a rising price curve. With prices lower, the lag flips. The company now expects Q4 exit EBITDA margin in the low double digits, versus the mid-teens exit it had sketched 90 days ago. The long-term 16%-19% EBITDA target is still there, but only after stripping out tungsten noise.

The 4 levers I just covered give us confidence that we can deliver long term EBITDA margins of 16% to 19% assuming a stable elevated tungsten price environment. This margin target is consistent with our 23 Investor Day baseline. We have only revised this target to reflect the impact of higher tungsten prices which have increased 8 fold since they were issued.
Sanjay K. Chowbey, President and CEO · 2026-11-04

That is a reasonable through-cycle aspiration. It is not a Q1 earnings story. The market is treating it accordingly.

Infrastructure is the operating soft spot

Beneath the tungsten noise, the operating miss was in Infrastructure. Volumes fell in the low double digits, driven by three factors Sanjay laid out: price sensitivity in applications where tungsten content is high, a China mining slowdown after a May safety accident, and limited product substitution. The first factor is self-inflicted. Kennametal priced according to its own raw-material cost and found itself above market in price-sensitive applications. “We were slightly ahead in the market compared to some other players. We have seen the dynamics of that. We have already started to work accordingly.” — Sanjay K. Chowbey, President and CEO · 2026-11-04 Management is now "selectively pricing to win" targeted applications, which is code for giving back some price to recover volume. This is not a new debate. On the February call, Sanjay was already talking about redesigning products to use tungsten more efficiently: “We are also looking for ... how do we make our product more efficient in that regard?” — Sanjay Chowbey, President and CEO · 2026-02-04 The current quarter adds a new keyword, volume decline, but the underlying issue — how much price can Kennametal push before customers substitute steel or another solution — has been building for several quarters. The offset is Metal Cutting, which posted its fourth consecutive quarter of volume growth and mid-single-digit volume improvement, with share gains in aerospace and defense and energy. Power Generation revenue doubled from fiscal 2025 to fiscal 2026, and management continues to cite data-center power generation as a mid-teens growth market. That is a real tailwind, but it is still small relative to the Infrastructure hole and the tungsten balance-sheet drag. The global tape has also turned against data-center-adjacent hardware and power names over the last 30 days, so Kennametal's AI power-generation story is unlikely to rescue the narrative by itself.

What actually changed

The first change is that the tungsten trade has matured. A year ago, rising tungsten was a margin tailwind and a share-gain opportunity because competitors could not secure material. Today, it is a cash trap: inventory is worth three times more, supplier advances are higher, and the share repurchase is paused. The company's own language has shifted from "free operating cash flow" to cash conversion and "inventory efficiency." The second change is that management has been forced to quantify the unwind. The Q2-to-Q4 EPS cadence — +$1.00, -$0.55, -$2.62 — tells investors that the first-half beat is not a run rate. The third change is that Infrastructure's volume problem is now explicit. The company admitted it was ahead of the market on price and is now repricing to win back applications. On the prior call, Watson said the company wanted “some stabilization and clarity about where tungsten is headed” — Patrick Watson, Chief Financial Officer · 2026-05-06 before resuming buybacks. Tungsten has now fallen, but the buyback remains paused and the cash conversion challenge is not assumed away this fiscal year. That is the tell: the condition for restarting capital returns has not been met. The balance sheet itself is not the issue. Effective net cash was -$589 million, or net debt, 29% worse year over year, and interest coverage was 30.2x. Kennametal can carry this inventory. The question is whether it can convert it back to cash without giving away margin. For now, the stock is voting that the answer is not yet. The 25% drawdown over 90 days is not a reaction to a bad quarter; it is a reaction to a low-quality beat. The next proof points are Q2 cash flow, Infrastructure volumes, and whether tungsten stabilization actually allows the repurchase program to restart. Until then, the $3.48 EPS print should be read as a commodity-timing artifact, not a new earnings baseline.