Price $31.10 · Base $20.96 · Price is +48% above base
Price (10/14 intraday)
$31.10
Base intrinsic value
$20.96
Bear → Blue-sky range
$14 – $42
Key driver
Tungsten (APT) persistence — price needs ≈$2,161/mtu at 12% margin
The Take — Overvalued. The +271% EPS "growth" is a tungsten inventory gain, and it is already reversing.
Strip the commodity windfall and FY26 was a flat year. Reported FY26 operating income of $473M (20% margin) includes $316M of favorable raw-material timing — Kennametal raised prices on a tungsten market that went from ~$920/mtu APT in January to $3,000–3,200 today, while still consuming cheaper inventory. Remove it and the restructuring add-back and underlying EBIT was $168M (7.1%) — below FY23 and roughly the FY25 trough. Volumes grew low-single digits. The Barchart screen saw "accelerating EPS"; the 10-K shows a price/cost lag.
The windfall is mean-reverting inside the company's own guide. Q1 FY27 guides adjusted EPS of $2.50–2.80 of which ~$2.25 is timing, yet the full-year guide carries only $0.39 of timing — the remaining three quarters therefore absorb roughly −$1.86/share (~$190M pre-tax) as cost catches price, even with tungsten flat. FY27 sales of $3.33–3.45B are ~40–43% price/surcharge and only 1–4% volume.
Normalize over ten years and the stock is ~48% too high. Ten-year adjusted EBIT averages $194M (9.1% margin); ex-COVID 9.6%. Base gives credit for a lot: APT settles at $1,500/mtu (≈4x the 2024 level, half today — China export controls stay, new Western mines arrive 2028+), revenue holds at ~$2.5–2.7B, and margin recovers to 10%, above the decade average. That is worth $20.96. At that point after-tax ROIC is ~9% vs a 9.2% WACC: Kennametal earns its cost of capital, no more — which is exactly what it has done for a decade.
Reverse DCF: $31.10 requires either a sustained 13.4% EBIT margin with APT at $1,500 (achieved only in FY18–19 and the FY12 tungsten spike), or 12% margin with APT holding ≈$2,161/mtu forever, or 10% margin with APT at ≈$4,759 — above today's record. The market is capitalizing the spike. Since July JPMorgan (twice, to $30), Morgan Stanley (cut, then Underweight), UBS ($35) and DA Davidson ($30) have all cut; consensus "Sell", mean PT $31.43 — the stock already trades at the Street's number with no margin of safety.
Call: Avoid / trim. No new money above ~$24. Holders: sell into the Q1 print (Nov 4) — the headline will look spectacular (~$2.65 EPS on timing), and that is the exit liquidity. Accumulation zone $18–21 (base, with the bear at $14 as the floor if tungsten round-trips to ~$800). Options: covered calls / a bear call spread $35/$40 into or just after Nov 4 (52-wk high $43.81; IV elevated into the print) — defined risk, because a further Chinese export-control tightening can gap tungsten — and the stock — higher. Don't short naked: dividend 2.6%, beta 1.3, and the blue-sky case (APT stays $3,100, 13.5% margin) prints $42.
Risk Profile — HIGH
HIGH
Tungsten (APT) price — the master variable. Long APT: revenue ≈ $2,016M×(1+0.131×(APT−500)/500); a fall from $3,100 to $1,500 cuts ~$850M of sales and, by FY26's own symmetry, ~$230M of timing losses. China (>75% of supply) export controls are the switch: tightening = up, quota relaxation (as with urea in 2026) = down.
HIGH
Inventory / working capital. Inventory doubled to $1,108M (vs $538M); OCF −$4M, FOCF −$79M in FY26; FY27 FOCF guided at only ~20% of adjusted NI and positive only in H2. Supplier prepayments to secure tungsten. Short: a sharp APT drop forces write-downs; long: release of ~$300M cash if it normalizes.
MED
Leverage creeping up. Debt $718M (+$500M term loan drawn in Sep-qtr to repay revolver), cash $96M, leases $46M, pension deficit $102M. ~1.2x FY27E EBITDA today, but ~2x on base-case normalized EBITDA (~$418M).
MED
Industrial cycle. Long: aerospace & defense (+43%), energy (+101% in Q4), earthworks. Short: transportation/auto (+7%), general engineering, EMEA (~30% of sales), Middle East conflict hitting customers.
MED
Demand destruction & substitution. 40%+ price increases invite customer destocking and recycling/substitution (scrap drill bits +17% w/w in India) — volume risk once the scramble ends.
LOW
Franchise. #2 global carbide tooling brand behind Sandvik; vertically integrated powders + recycling give some supply security vs smaller peers.
What the "inventory" really is: a ~$1.1B tungsten position bought near the top of a policy-driven price spike, financed partly with new debt. Kennametal is a tool company whose FY26–27 earnings are a mark-to-market on that position. If APT holds, the equity is fair-to-cheap; if it halves, the gain unwinds through the P&L.
Scenario values per share vs price
Bear: APT → $800, 8.5% margin · Base: APT → $1,500, 10% · Bull: APT holds $2,800, 12.5% · Blue-sky: $3,100 forever, 13.5%, no timing reversal. FY27 = guide midpoint in all cases; fade over FY28–29 with symmetric timing reversal. Prob-weighted (20/50/25/5): $24.54. Dashed line = price.
Driver tornado — value/share from base $20.96
No single swing reaches the price (best single case $26.5). Margin and discount rate dominate; tungsten level works through both revenue and the size of the timing unwind.
Red = >15% below price; amber = within −15%/+5%; green = above. Current APT CIF Rotterdam ≈ $3,000–3,200/mtu (SMM, mid-Oct 2026); Jan-26 ≈ $920; early-2025 ≈ $650; 2024 ≈ $340. The price is only covered in the top-right — high tungsten and above-average margins, permanently.
History — revenue vs adjusted EBIT ($M, FY17–FY26)
Blue = revenue; green = adjusted EBIT (ex-restructuring; FY26 ex-$316M tungsten timing). Earlier cycle: FY12 EBIT $416M on $2,736M (15.2%) at the 2011–12 tungsten peak, then FY15–16 operating losses (−$358M, −$175M incl. impairments) when tungsten and oil collapsed. This has happened before.
Normalization worksheet — 10 years ($M)
Year
Revenue
GAAP EBIT
+ Restructuring
− Tungsten timing
Adj. EBIT
Adj. margin
FY2017
2,058
94.8
+65.6
+0
160.4
7.8%
FY2018
2,368
290.3
+16.4
+0
306.7
13.0%
FY2019
2,375
328.9
+19.5
+0
348.4
14.7%
FY2020
1,885
22.3
+69.2
+0
91.5
4.9%
FY2021
1,841
102.2
+29.6
+0
131.8
7.2%
FY2022
2,012
218.1
-1.2
+0
216.9
10.8%
FY2023
2,078
192.4
+0.8
+0
193.2
9.3%
FY2024
2,047
170.2
+1.1
+0
171.3
8.4%
FY2025
1,967
143.1
+13.5*
+0
156.6
8.0%
FY2026
2,357
472.5
+11.4
-316
167.9
7.1%
10-yr avg
2,099
203.5
194.5
9.1%
GAAP EBIT and revenue from 10-K XBRL (SEC; FY17/18 restated values). Restructuring from XBRL; *FY25 estimated from adjusted vs GAAP; FY26 from the Q4 release ($11.4M adj.). Timing = company-disclosed favorable raw-material price/cost timing. R&D (~2% of sales) not capitalized — immaterial; SBC expensed.
Why 10 years: the 5-yr window (FY22–26) mixes a soft industrial patch with a commodity windfall; it averages 8.7% adjusted — similar to the 10-yr 9.1% — confirming there is no hidden margin step-up. Base 10% margin is above both, crediting completed restructuring savings and surcharge pricing discipline. Management gives no published mid-cycle margin target; the FY27 guide implies ex-timing EBIT ≈ $465M (≈13.7%) at $3,100 APT — that is the bull case, not the norm.
FCF history
OCF
Capex
D&A
FCF
SBC
FCF − SBC
FY2022
181.4
96.9
131.7
84.5
21.0
63.5
FY2023
258.0
94.4
134.0
163.6
24.7
138.9
FY2024
277.1
107.6
134.7
169.5
24.3
145.2
FY2025
208.3
89.0
136.5
119.3
22.1
97.2
FY2026
-4.0
76.9
143.2
-80.9
34.9
-115.8
FY26: "record" earnings, negative free cash flow — the profit is sitting in the inventory line ($593M build). Capex runs below D&A (~$90M vs ~$137M).
FY27 = guide midpoint: sales $3,390M, EBIT ≈$505M (adj. EPS $4.65 × 77.9M sh, 24% tax, ~$38M interest, ~$20M other income; incl. ~$40M timing). FY28–29 fade to steady revenue with a $230M timing reversal (symmetric to FY26's gain per $ of APT move) split across both years; working capital (33% of sales) releases as revenue falls — that is why FCFF rises in FY28–29 while EBIT falls. Reinvestment = ΔRevenue×33% + growth capex at sales/capital 2.6. PV explicit 1,173 + PV terminal 1,248 = 2,422; terminal share 52% (<70% ✓). r > g ✓.