BorgWarner
This is the archived full research. Read the story
It makes the turbochargers, timing chains and gearbox parts inside petrol cars, and from 2027 plans to sell gas turbines that power data centres on site.
The scores
Growth2/4 →
Organic sales rose 0.5% in 2025 and are guided down 1.5% to 3.5% in 2026; management itself says it is not outgrowing production.
Business quality3/4 ↓
About 30% of a two-firm turbo market and engine segments earning 15% to 18%, but electric cars remove the product and Volkswagen builds drives.
Financial position3/4 ↑
Net debt about 0.7 times cash profit and $1B of yearly free cash flow; not net cash, and $1.1B falls due in 2027.
Execution record3/4 ↑
Five straight quarters ahead of estimates and margins up on flat sales, but $1.27B of electric write-downs and a charging exit in two years.
Per share capture3/4 →
Diluted shares down about 8% in eighteen months on cash buybacks, but cash acquisitions behind $1.27B of write-downs did not reach shareholders.
Potential2/4 ↑
The turbine's only commitment is an unfiled framework with one private channel, no volume or term; more capacity and hyperscaler orders are still expected.
Risk profile3/4 →
Volkswagen and Ford are a quarter of sales and China about a fifth; the turbine rests on one partner, survivable because it is small.
Valuation4/4 ↑
About 16.9 times operating profit, 67% of a three-stage value near $72.47 at an 8% return: cheap, past the 20% margin.
Evidence quality3/4 →
The core rests on filed segment figures; every turbine figure is management's, from calls and a furnished 8-K, never a 10-K or 10-Q.
If it works 1.0× · if it fails you keep about 86%.
The Verdict
We rate BorgWarner Attractive. It makes turbochargers, timing chains and other parts for petrol cars; that side throws off cash, much of it handed back to shareholders, but sales are flat, engines are slowly going away and its electric-car businesses lose money, so the business is mixed. The price is about two-thirds of our value and near our gloomy case: it pays for the car parts alone, and the new gas turbine for data centres comes free. This is not a growth case, and insiders have been selling, but that cushion is enough. As a bet it is defensive.
The long caseData centres short of grid power are building their own, and buyers will take a new turbine maker if it can deliver sooner than the queue.
How we weighed it
The price against our value
In its favour
Price at 67% of our value
The price is about two-thirds of our value and near the bear case; the turbine is priced at nothing.
The business
Cuts both ways
Mixed
Engine parts that throw off cash, but the electric businesses lose money and sales are flat.
Growth against what the price assumes
Cuts both ways
Sales are guided down slightly, which the value assumes; cost cuts and buybacks carry the profit.
What could go wrong
Against it
Engines are going away, the power entry is late and through one partner, and insiders sold on the open market.
Against competitors
Cuts both ways
On this year's expected profit it sits between two rivals, so the comparison neither helps nor hurts.
How its group is behaving
Cuts both ways
It sits in none of our ranked groups; car production and the data-centre power trade move it.
The latest news and report
Cuts both ways
The last report held the margin goal but did not answer the turbine question.
The evidence under this - 9 claims
- Hybrids carry both turbos and electric parts, and hybrid output is forecast to grow (C19).
- PowerDrive earned $28M of adjusted profit in Q4 2025, and the battery segment's first-half loss narrowed to $4M (S02, S05).
- The inverter extension awards in Q2 2026 show electric parts still winning business (S05).
- The earnings-per-share raise in August came entirely from share repurchases; sales, margin and cash guidance were only held (S05).
- The GAAP margin guide was lowered from 9.8%-10.0% in February to 9.6%-9.8% in August (S02, S05).
- The CFO said the midpoint implies a second-half adjusted margin near 10.6%, below the first half, on sales about $7.0B (S06).
- The guided battery decline widened from about $210M to about $250M between February and August (S02, S05).
- The sales range narrowed down and the GAAP margin guide was lowered again in October; the earnings-per-share raise came from buybacks and a lower tax rate (S29).
- 2026-10-29editedre-struck on the Q3 2026 release and call: sales, GAAP margin and earnings per share ranges moved, adjusted margin held
- 11 February also carried the full-year results, 2026 guidance and record car-part awards, so not all of that day's move was the turbine (S02).
- Against the end-2025 multiple of about 7.9 times, before a 20% rally in early 2026, the premium would be nearer $3.2B; the baseline is a choice (S15).
- The August bond tender used cash to retire debt, which leaves this enterprise value roughly unchanged but is not yet in a filed balance sheet (S08).
- 2026-11-09editedRe-struck at the 6 November close ($48.81) after a 20% fall on no news; basis moved to the 9.3-9.6% GAAP margin guide of 29 October
- Management declined to link the $300M to the 2 GW or to size revenue at capacity (S03).
- The figure was reiterated in an 8-K furnished 'in response to questions', not in a filed report (S07).
- The product has not been through UL certification yet; component certification was due to start in September 2026 (S06).
- A product not yet certified or launched carries risk its established rivals do not, and a peer multiple applied to it gives that risk away (C12, C13).
- The premium also covers battery storage and inverters, which have no disclosed figures, so the turbine alone may need less than this or the others may be worth nothing (C38).
- The first sales are in 2027, so a year of waiting sits between today's price and the first profit (C12).
- No hyperscaler, order, volume or term has been disclosed (S06, S07).
- Management has not said what revenue 2 GW of capacity supports (S03).
- On the Q3 call management gave no capacity decision, customer or volume, said political conditions may slow final agreements, and said it would give no more shape until early 2027 (S30).
- 2026-10-29editedthe second-half capacity decision was not announced on the Q3 call; claim re-struck as undecided
The company
BorgWarner makes the parts inside a car's engine and gearbox: turbochargers, timing chains, clutch plates and, for electric cars, inverters and motors.
A turbocharger uses exhaust gas to push more air into the engine, so a smaller engine does the work of a bigger one. Car turbos are its largest product, about 21% of sales. From 2027 it adds a product outside cars: a turbine generator that makes electricity at data centres.
The evidence under this - 3 claims
How it makes money
A carmaker picks the supplier for a part years before a model goes into production, then buys it for the life of that model. Volkswagen took 13% of 2025 sales and Ford 12%.
The engine side pays for everything. Its two engine businesses earned $1.92B of adjusted operating profit in 2025, more than all four businesses together, because the electric ones lost $122M.
The evidence under this - 3 claims
Where the money comes from, stream by stream
| Stream | Share of sales (H1 2026) | How it earns it |
|---|---|---|
| Turbos & Thermal Technologies | 40.0% | Turbochargers, thermal and emissions parts sold to carmakers and truck makers, per part, under multi-year programme awards. |
| Drivetrain & Morse Systems | 40.1% | Timing chains, cam phasers, transmission clutches and transfer cases for combustion and hybrid vehicles, sold per part on programme awards. |
| PowerDrive Systems | 17.4% | Electric motors, inverters and electric drives for electric and hybrid vehicles; still loss-making on its segment measure. |
| Battery Energy Systems | 2.8% | Battery packs for commercial vehicles, shrinking after the charging exit; future energy-storage products would sit here. |
Who runs it
Joe Fadool, a BorgWarner lifer who ran its turbocharger business, became chief executive in February 2025; Craig Aaron, the former treasurer, has run finance since March 2024. Insiders own under 1% of the company. Their yearly bonus rides on profit margin and cash, and most of their shares pay on the stock against rivals, sales growth beyond the car market and earnings per share. Most spare cash has gone into buying back shares.
The evidence under this - 4 claims
- The bonus margin and adjusted earnings per share are the company's adjusted measures, which leave out the impairment charges (C06), so the write-downs cost the plan little.
- Relative revenue growth, the gap management itself calls unsatisfactory (C10), carries a quarter of the performance units and nothing in the bonus.
The electric businesses built partly by earlier cash acquisitions were written down by $1.27B in 2024 and 2025 (C06).
Why it wins, and who could beat it
Why customers come back
Turbos need heat-proof alloys, precise balancing and plants near every big carmaker. Two companies, BorgWarner and Garrett, sell most of a $10B market, and BorgWarner's roughly 30% share spreads the engineering and tooling cost over more units than a newcomer could.
It shows in the margins: in 2025 the two engine businesses kept 15.2 and 18.4 cents of adjusted operating profit on each dollar of sales.
The evidence under this - 4 claims
- Garrett's $10 billion includes commercial-vehicle turbos, which BorgWarner's 21% excludes, so the share is understated if anything; nobody publishes the split.
- Chinese makers and in-house carmaker units sell turbos too, and neither company names their share.
The threat
The biggest threat is the biggest customer. Volkswagen has built 1 million of its own electric drives and is designing its next inverter in-house, and BorgWarner's annual report says several carmakers plan to make its parts themselves.
In data-centre power it is the newcomer: Bloom Energy is doubling its fuel-cell factory to 2 GW, and Caterpillar already sells turbines there.
The evidence under this - 6 claims
- Management says the turbine is lower-emission, faster to respond and fuel-flexible, and that time to power matters more than cost (S03, S06).
- Bloom and Caterpillar both describe demand outrunning supply, which leaves room for a new entrant (S12, S13).
The rivals: BE, CAT, MGA, GTX, and 1 more that is private, in-house or listed abroad.
Who else does this, rival by rival
| Ticker | Rival | Where it competes | How far along | Priced against it |
|---|---|---|---|---|
| Volkswagen ID. electric cars on Volkswagen's own APP550 electric drivePriced against it: Listed abroad | Turbos & Thermal Technologies | On sale | Listed abroad | |
| Volkswagen APP550 single-speed electric drive, built in Kassel | Drivetrain & Morse Systems | On sale | ||
| Volkswagen Group Components' in-house pulse inverter | PowerDrive Systems | Announced | ||
| BE | Bloom Energy solid-oxide fuel cellsPriced against it: Different product | Turbine generator system (industrial products, not a reportable segment) | On sale | Different product |
| CAT | Caterpillar gas engines and Solar Turbines gas turbinesPriced against it: One segment only | Turbine generator system (industrial products, not a reportable segment) | On sale | One segment only |
| MGA | Magna InternationalPriced against it: Yes | - | On sale | Yes |
| GTX | Garrett MotionPriced against it: Yes | - | On sale | Yes |
How big it could get
How big this could get
Carmakers buy about $10B of turbochargers a year, on Garrett's count, and hybrids, which still need one, are forecast to grow from 16 million a year in 2025 to 26 million in 2029.
The new market is vaguer. US data centres used 176 terawatt-hours of electricity in 2023, and a government laboratory expects 325 to 580 by 2028. Nobody measures how much of that is made on site, the part a turbine serves.
The evidence under this - 3 claims
Who has signed
For the turbine, one partner has signed: TurboCell, owned by the private data-centre builder Endeavour, under a supply agreement with no stated volume or length. Management expects more than $300M of first-year sales.
The evidence under this - 2 claims
- Management declined to link the $300M to the 2 GW or to size revenue at capacity (S03).
- The figure was reiterated in an 8-K furnished 'in response to questions', not in a filed report (S07).
- The product has not been through UL certification yet; component certification was due to start in September 2026 (S06).
The results
The numbers
Sales grew 0.4% over the twelve months to Sep 2026, about the same pace as the 1.6% in the year to Dec 2025.
Operating margin is 5.4%, down from 8.2% in the year to Dec 2023.
Running costs take 14¢ of each sales dollar, up from 11¢ in the year to Dec 2021.
Each dollar put into the business earns 5.9% a year after tax, about where it was (7.5%) in the year to Dec 2021.
Yearly sales growth has ranged from −0.8% to +16% over the last 5 years, with sales falling in 1 of them.
The share count fell 15% in 4.8 years.
All the numbers
Filed figures from the company's 10-K and 10-Q, read 2026-11-10. Fiscal years end in December.
| Dec 2021 | Dec 2022 | Dec 2023 | Dec 2024 | Dec 2025 | 12 mo to Sep 2026 | |
|---|---|---|---|---|---|---|
| Size and growth | ||||||
| Sales | $11.8B | $12.6B | $14.2B | $14.1B | $14.3B | $14.2B |
| Sales growth | +16% | +7.0% | +12% | −0.8% | +1.6% | +0.4% |
| Profit | ||||||
| Gross margin | 18% | 19% | 18% | 19% | 19% | 20% |
| Operating margin | 7.7% | 8.0% | 8.2% | 3.9% | 3.7% | 5.4% |
| Net margin | 4.5% | 7.5% | 4.4% | 2.4% | 1.9% | 3.4% |
| What growth costs, per sales dollar | ||||||
| Research and development | 5.0% | 5.5% | 5.1% | 5.2% | 5.0% | 4.8% |
| Selling, general and admin | 9.2% | 10% | 9.3% | 9.6% | 9.1% | 9.1% |
| Pay in shares | 0.4% | 0.5% | 0.4% | 0.4% | 0.5% | - |
| Capital spending | 4.4% | 4.9% | 5.9% | 4.8% | 3.3% | 3.7% |
| Cash | ||||||
| Cash from operations | $1.31B | $1.57B | $1.40B | $1.38B | $1.65B | - |
| Free cash flow | $792M | $947M | $565M | $711M | $1.18B | - |
| Balance sheet | ||||||
| Cash and short-term investments | $1.84B | - | - | - | - | - |
| Debt | $4.26B | $4.14B | $3.71B | $4.10B | $3.90B | $3.00B |
| Net cash | −$2.42B | - | - | - | - | - |
| Returns and shares | ||||||
| Return on capital | 7.5% | 7.0% | 7.9% | 4.5% | 3.6% | 5.9% |
| Shares (diluted) | 240M | 237M | 234M | 225M | 216M | 205M |
What growth costs
Growing has made BorgWarner neither more nor less profitable. Research, overheads and stock pay take about the same share of each sales dollar as five years ago, and the gross margin barely moved. The one cost that fell is spending on factories, as the electric-car push was cut back; the turbine plant and extra industrial research push costs up again this year.
The evidence under this - 3 claims
- Selling, general and administrative costs other than R&D rose to 4.3% of sales in Q2 2026 from 3.7% a year earlier (S04).
- Management plans $10M to $15M of extra industrial R&D in the second half of 2026, and guides capital spending up for the turbine plant (C38, C31).
None of the three has a disclosed order, and the chief executive said these products take 18 to 24 months to develop (S06).
The last report (29 Oct)
Quarter ended September 2026, filed figures.
The third quarter of 2026 held up: sales fell 2.9% to $3.487B, but 10.9 cents of adjusted operating profit on each dollar was kept, and management raised its yearly earnings-per-share goal to $5.20 to $5.35 while keeping its margin goal.
The turbine business moved forward only in construction: the North Carolina plant is built and trials start by year-end. Management named no customer, order or new capacity and said it will give no more detail until early 2027.
What it did to the thesisIt supported the steady half of the case, since the margin goal held. It did not answer the turbine question, so the part of the case that could grow is still waiting for an order.
The quarter in detail
Sales growth from the previous quarter, filed figures.
Earnings per share $1.39 against $1.25 expected.
The evidence under this - 4 claims
- The earnings-per-share raise in August came entirely from share repurchases; sales, margin and cash guidance were only held (S05).
- The GAAP margin guide was lowered from 9.8%-10.0% in February to 9.6%-9.8% in August (S02, S05).
- The CFO said the midpoint implies a second-half adjusted margin near 10.6%, below the first half, on sales about $7.0B (S06).
- The guided battery decline widened from about $210M to about $250M between February and August (S02, S05).
- The sales range narrowed down and the GAAP margin guide was lowered again in October; the earnings-per-share raise came from buybacks and a lower tax rate (S29).
- 2026-10-29editedre-struck on the Q3 2026 release and call: sales, GAAP margin and earnings per share ranges moved, adjusted margin held
- Management declined to link the $300M to the 2 GW or to size revenue at capacity (S03).
- The figure was reiterated in an 8-K furnished 'in response to questions', not in a filed report (S07).
- The product has not been through UL certification yet; component certification was due to start in September 2026 (S06).
- No hyperscaler, order, volume or term has been disclosed (S06, S07).
- Management has not said what revenue 2 GW of capacity supports (S03).
- On the Q3 call management gave no capacity decision, customer or volume, said political conditions may slow final agreements, and said it would give no more shape until early 2027 (S30).
- 2026-10-29editedthe second-half capacity decision was not announced on the Q3 call; claim re-struck as undecided
No order, customer or volume was named, and management said political conditions may slow the finalising of program agreements (S30).
What management did with the money
It hands back much of what it makes. In 2025 it turned $1.21B into spare cash and returned about $630M, and the share count fell from 224.8 million in 2024 to 206.3 million by June 2026.
Insiders in 2026Q2: 14 purchases and 10 sales, from their Form 4 filings.
The evidence under this - 2 claims
The price, and where we disagree
In short. Cheap. The price is about 67% of what we think the business is worth, at an 8% yearly return.
The market has priced the turbine in and out twice. The shares jumped 22.5% on the day it was announced in February, gave it all back within a month, and have since fallen about 20% from late September with no company news.
The next results, in February, settle the full-year margin. Management says it will decide on more turbine capacity and give more detail in early 2027.
The evidence under this - 2 claims
- 2026-11-09edited6 November close added
- No hyperscaler, order, volume or term has been disclosed (S06, S07).
- Management has not said what revenue 2 GW of capacity supports (S03).
- On the Q3 call management gave no capacity decision, customer or volume, said political conditions may slow final agreements, and said it would give no more shape until early 2027 (S30).
- 2026-10-29editedthe second-half capacity decision was not announced on the Q3 call; claim re-struck as undecided
How we value it
- Method
- Cash the business will produce, in three stages
- Required return
- 8%
- Long-run growth
- 4%
- Moat
- An average moat
| Bear | Base | Bull | |
|---|---|---|---|
| Sales growth, years 1 to 3 | 0.9%, 2.9%, 2.6% | 1.9%, 5.9%, 5.2% | 2.4%, 7.4%, 6.5% |
| Normal operating margin | 7.7% | 9.6% | 10.6% |
| Return on new capital after year ten | 8% | 12% | 15% |
| Value a share | $38.32 | $72.47 | $93.59 |
- Price
- $48.81
- Priced on
- 6 Nov
- Price as a share of the base value
- 67%
- Starting growth the price pays for
- -8.6%
- Yearly return the price offers on the base case
- 9.4%
- Risk check: valued at 10%, the rate for a cyclical or indebted business
- $42.61
Where each figure comes from
| Figure | Source | Where exactly |
|---|---|---|
| Sales, the last twelve months | Filed report | Quarterly report (10-Q) for the period to 30 Sep 2026, filed 29 Oct 2026; the four quarters to that date. |
| Operating margin today | Filed report | Quarterly report (10-Q) for the period to 30 Sep 2026, filed 29 Oct 2026; the four quarters to that date. |
| Tax rate | Filed report | Quarterly report (10-Q) for the period to 30 Sep 2026, filed 29 Oct 2026; the four quarters to that date. |
| Sales per dollar of new capital | Filed report | Annual report (10-K) for the period to 31 Dec 2025, filed 11 Feb 2026: the year's sales over its operating capital. |
| Shares | Filed report | Quarterly report (10-Q) for the period to 30 Sep 2026, filed 29 Oct 2026; the diluted share count. |
| Cash and investments | Filed report | Quarterly report (10-Q) for the period to 30 Sep 2026, filed 29 Oct 2026; the balance sheet on that date. |
| Debt | Filed report | Quarterly report (10-Q) for the period to 30 Sep 2026, filed 29 Oct 2026; the balance sheet on that date. |
| Sales growth, years 1 to 3 | Analysts' forecasts | FMP's consensus for the years to 31 Dec 2026, 31 Dec 2027, 31 Dec 2028 and 31 Dec 2029: the average of 4 to 11 analysts' revenue estimates, 1.9%, 5.9%, 5.2%. FMP publishes how many analysts, not who they are. |
| Normal operating margin | Filed report | The best operating margin it has filed before amortisation of past deals, over five years and the latest quarter: a cyclical business is held to its record. |
| Return on new capital after year ten | Our method's rule | Set by our reading of the moat: the 8% rate with none, 12% with an average one, 15% with a strong one; the bear case earns the rate. |
| The bear and bull cases | Our method's rule | Our method's spread: the bear case takes half the growth and a fifth off the margin, the bull a quarter more growth and a tenth more margin. |
| Price | Market price | FMP's closing price on 6 Nov 2026. |
What each case rests on
Bear caseGrowth runs at half the analysts' path, 1%, 3% and 3%, the margin settles a fifth lower at 7.7%, and new capital earns only the 8% rate: the downturn test.
Base caseRevenue grows 2%, 6% and 5% over three years, the analysts' path (C62), then fades toward 4%, closing 30% of the gap each year; the margin moves to 9.6%, the analysts' year-three figure after stock pay and on the filed basis (2.1% less).
Bull caseGrowth runs a quarter above the analysts' path, 2%, 7% and 6%, the margin reaches 10.6%, and new capital earns 15%.
What moves the base value
| Lower | Higher | |
|---|---|---|
| Growth in years 1 to 3, a quarter lower or higher | $69.53 | $75.55 |
| The normal margin, 15% lower or higher | $58.97 | $85.98 |
| Capital per dollar of growth, a quarter more or less | $69.77 | $74.10 |
| Return on new capital after year ten, at the rate or one step up | $55.15 | $79.40 |
| The rate, a point higher or lower | $54.55 | $102 |
| Long-run growth a point lower | $64.15 | - |
Value a share as growth and margin change
| Normal margin ↓ · growth → | 50% | 75% | 100% | 125% | 150% |
|---|---|---|---|---|---|
| 7.7% | $50.46 | $52.42 | $54.46 | $56.60 | $58.83 |
| 8.6% | $58.59 | $60.97 | $63.47 | $66.08 | $68.80 |
| 9.6% | $66.71 | $69.53 | $72.47 | $75.55 | $78.77 |
| 10.6% | $74.84 | $78.08 | $81.48 | $85.03 | $88.74 |
| 11.5% | $82.96 | $86.64 | $90.49 | $94.51 | $98.71 |
Growth is shown as a share of the base case's. Green: the price is at or under 80% of that value. Red: at or over 120%.
How we read the multiple
A write-down cut last year's profit. On this year's expected operating profit the whole company, debt included, costs about 8.7 times, against 9.5 before the turbine was announced, about $0.7B less than the car parts alone are worth:
- Its first year should earn about $45M of profit.
- At the 25 September price, $1.4B needed about $0.9B of turbine sales a year.
- At today's price the turbine needs nothing to earn its keep.
- The full-year margin and turbine orders decide the next move.
Against its own history. Dearer than its own normal, but that normal includes years of write-downs, and the rise since last year came from the price paid, not from sales.
The evidence under this - 6 claims
- 11 February also carried the full-year results, 2026 guidance and record car-part awards, so not all of that day's move was the turbine (S02).
- Against the end-2025 multiple of about 7.9 times, before a 20% rally in early 2026, the premium would be nearer $3.2B; the baseline is a choice (S15).
- The August bond tender used cash to retire debt, which leaves this enterprise value roughly unchanged but is not yet in a filed balance sheet (S08).
- 2026-11-09editedRe-struck at the 6 November close ($48.81) after a 20% fall on no news; basis moved to the 9.3-9.6% GAAP margin guide of 29 October
- Management says 'more than' $300M, and the first year is a partial ramp, so a full year could be larger (S03).
- A turbine business could deserve a higher multiple than the car-parts core, which would lower the sales needed (S12, S13).
- Battery storage and microgrid inverters are also in the premium and have no disclosed figures at all (S06).
- 2026-11-09editedPremium re-struck: the price now holds none for the turbine
- All five companies' prices have been lifted by the same data-centre demand, so their multiples may be a theme's rather than a durable level (S12, S18, S19).
- Data-centre power is a minority of each company's sales, so their whole-company multiple prices mostly trucks, mining and construction machines, or home generators (S18, S19, S12).
- A product not yet certified or launched carries risk its established rivals do not, and a peer multiple applied to it gives that risk away (C12, C13).
- The premium also covers battery storage and inverters, which have no disclosed figures, so the turbine alone may need less than this or the others may be worth nothing (C38).
- The first sales are in 2027, so a year of waiting sits between today's price and the first profit (C12).
Priced against its rivals
The competitors its own annual report names, each confirmed against the rival's filing. Measured on company value for each $1 of operating profit, cheapest first:
| Ticker | Company | company value for each $1 of operating profit |
|---|---|---|
| MGA | Magna International | 8.3× |
| GTX | Garrett Motion | 15.2× |
| BWA | BorgWarner | 19.5× |
The middle of the other 2 is 11.8. BorgWarner Inc. is dearest of 3.
Rivals' figures come from a data provider, so read the spread of the group rather than any single pair.
The trailing profit the table ranks includes last year's write-down. On this year's expected profit it sits between Magna and Garrett, so its higher price is not clearly its own.
Private or foreign rivals that cannot be priced: Robert Bosch GmbH, Denso Corporation, Hitachi, Ltd., Valeo, Schaeffler Group, Contemporary Amperex Technology Co., Limited.
The bet: what it pays if it works
This is a bet on one new business inside BorgWarner: a turbine generator for data centres, with no sales yet and production from 2027. The car-parts business is counted separately, at the price per dollar of profit it carried before the turbine was announced, and it covers most of today's price. Even if turbine sales grow past management's first-year figure, the company is worth about what it costs, so the bet is expensive.
If the turbine never sells, shareholders keep the car-parts business, worth about a tenth less than today.
| Turbine generator system (industrial products, not a reportable segment) operating profit in two to three years | Stated first years of turbine sales needed | At the stated multiple |
|---|---|---|
| $45.00M | about 1 at 15% kept | 0.9× |
| $70.00M (the big win) | about 2 at 15% kept | 1.0× |
| $135.00M (market forecast) | about 3 at 15% kept | 1.1× |
| $136.00M | about 3 at 15% kept | 1.1× |
Forecasts put US on-site gas for data centres near 20 GW a year by 2029; 1 GW, half BorgWarner's installed capacity and one rival-sized order, at $900 per kW earns about $135M. On that forecast, $135.00M pays 1.1× from today's value: an Expensive bet on the market case. Turbine generator system (industrial products, not a reportable segment) would be worth about $2.70B, 1.6× what the price pays for it.
To pay 2× (a Fair bet) it would need about $781.47M of Turbine generator system (industrial products, not a reportable segment) operating profit (about 17 stated first years of turbine sales at 15% kept); to pay 5× (a Cheap bet), about $2.87B of Turbine generator system (industrial products, not a reportable segment) operating profit (about 64 stated first years of turbine sales at 15% kept), at the same multiple.
Each row is Turbine generator system (industrial products, not a reportable segment) alone, plus the rest of the company (Turbos & Thermal, Drivetrain & Morse and PowerDrive: the car-parts core) counted at about $12.26B.
The steps it has to take
- Done Master supply agreement with TurboCell, Endeavour's turbine subsidiary
- In progress Decision on turbine capacity beyond the initial 2 GW due 2026-12
- Not started A turbine order with a stated volume and term, or a named second customer due 2027-03
- Not started Production starts in Hendersonville, North Carolina due 2027-12
- Not started First-year turbine sales above $300M at about a 15% margin due 2028-12
The detail behind the label
The evidence under this - 4 claims
- Management declined to link the $300M to the 2 GW or to size revenue at capacity (S03).
- The figure was reiterated in an 8-K furnished 'in response to questions', not in a filed report (S07).
- The product has not been through UL certification yet; component certification was due to start in September 2026 (S06).
- No hyperscaler, order, volume or term has been disclosed (S06, S07).
- Management has not said what revenue 2 GW of capacity supports (S03).
- On the Q3 call management gave no capacity decision, customer or volume, said political conditions may slow final agreements, and said it would give no more shape until early 2027 (S30).
- 2026-10-29editedthe second-half capacity decision was not announced on the Q3 call; claim re-struck as undecided
- 11 February also carried the full-year results, 2026 guidance and record car-part awards, so not all of that day's move was the turbine (S02).
- Against the end-2025 multiple of about 7.9 times, before a 20% rally in early 2026, the premium would be nearer $3.2B; the baseline is a choice (S15).
- The August bond tender used cash to retire debt, which leaves this enterprise value roughly unchanged but is not yet in a filed balance sheet (S08).
- 2026-11-09editedRe-struck at the 6 November close ($48.81) after a 20% fall on no news; basis moved to the 9.3-9.6% GAAP margin guide of 29 October
- 2026-11-09edited6 November close added
Where we disagree with the market
BorgWarner is becoming a data-centre power company, not an ageing engine-parts supplier.
The car parts are worth about what buyers paid before the turbine; at today's price the turbine is free, and growth beyond it is unproven.
The enthusiasm is easy to see: nine of ten analysts on the last call asked about data centres, and buy ratings rose from 50.0% to 68.8% since January. But the turbine appears in no annual or quarterly report, and every figure is management's.
We would be wrong if more turbine capacity or a large cloud order with a stated volume arrives, or if car-part margins rise while sales fall.
What analysts say, and the evidence
The evidence under this - 7 claims
- 2026-11-09edited6 November close added
- No hyperscaler, order, volume or term has been disclosed (S06, S07).
- Management has not said what revenue 2 GW of capacity supports (S03).
- On the Q3 call management gave no capacity decision, customer or volume, said political conditions may slow final agreements, and said it would give no more shape until early 2027 (S30).
- 2026-10-29editedthe second-half capacity decision was not announced on the Q3 call; claim re-struck as undecided
- Management declined to link the $300M to the 2 GW or to size revenue at capacity (S03).
- The figure was reiterated in an 8-K furnished 'in response to questions', not in a filed report (S07).
- The product has not been through UL certification yet; component certification was due to start in September 2026 (S06).
- A product not yet certified or launched carries risk its established rivals do not, and a peer multiple applied to it gives that risk away (C12, C13).
- The premium also covers battery storage and inverters, which have no disclosed figures, so the turbine alone may need less than this or the others may be worth nothing (C38).
- The first sales are in 2027, so a year of waiting sits between today's price and the first profit (C12).
What could go wrong, and how we would know
The risks
- Engines are going away. Its profit comes from parts that exist only where there is an engine, while its electric parts have lost money and cost $1.27B of write-downs.
- A late entrant in power. It reaches data centres through one private partner, against Bloom and Caterpillar, which already ship at the scale it is building.
- Shrinking sales. Sales without currency moves are guided to fall 1.5% to 3.5% this year; cost cuts and buybacks, not sales, are growing earnings.
- Insiders are selling. Insiders made 30 open-market sales in a year, none under a pre-arranged plan, including the chief executive in May.
The evidence under this - 9 claims
- Hybrids carry both turbos and electric parts, and hybrid output is forecast to grow (C19).
- PowerDrive earned $28M of adjusted profit in Q4 2025, and the battery segment's first-half loss narrowed to $4M (S02, S05).
- The inverter extension awards in Q2 2026 show electric parts still winning business (S05).
- Management says the turbine is lower-emission, faster to respond and fuel-flexible, and that time to power matters more than cost (S03, S06).
- Bloom and Caterpillar both describe demand outrunning supply, which leaves room for a new entrant (S12, S13).
The chief executive said the capacity decision may serve 'a different market, which we're also active with our partner, Endeavour' (S06), so the channel may widen, but still through the same partner.
- The earnings-per-share raise in August came entirely from share repurchases; sales, margin and cash guidance were only held (S05).
- The GAAP margin guide was lowered from 9.8%-10.0% in February to 9.6%-9.8% in August (S02, S05).
- The CFO said the midpoint implies a second-half adjusted margin near 10.6%, below the first half, on sales about $7.0B (S06).
- The guided battery decline widened from about $210M to about $250M between February and August (S02, S05).
- The sales range narrowed down and the GAAP margin guide was lowered again in October; the earnings-per-share raise came from buybacks and a lower tax rate (S29).
- 2026-10-29editedre-struck on the Q3 2026 release and call: sales, GAAP margin and earnings per share ranges moved, adjusted margin held
What we still do not know
- How much the 2 GW turbine plant can sell in a full year, and at what price.
- What margin the turbine really earns. The only figure is an analyst's paraphrase of the finance chief.
The evidence under this - 2 claims
- Management declined to link the $300M to the 2 GW or to size revenue at capacity (S03).
- The figure was reiterated in an 8-K furnished 'in response to questions', not in a filed report (S07).
- The product has not been through UL certification yet; component certification was due to start in September 2026 (S06).
- The figure reaches the record only through an analyst's paraphrase, not a management sentence (S03).
- A first year of production usually carries launch costs that a steady-state margin does not (S06).
What would change our view
- Attractive→NeutralThe price rises within 20% of fair value
It needs:
- ○The share price rises to about $57.98, 80% of what we think the business is worth, so it is no longer cheap enough.about ≈$55.74 today, -7% above the last close
- Attractive→UnattractiveThe price runs 20% above fair value
It needs:
- ○The share price rises to about $86.96, 20% above what we think the business is worth.about ≈$78.27 today, 30% above the last close
- Attractive→UnattractiveThe launch slips or the core margin breaks
It needs any of:
- ○The 2027 turbine start or the $300M first-year sales figure is pushed back or withdrawn.watched as it happens
- ○Management cuts this year's adjusted operating profit guidance below 10.5 cents per dollar of sales.checked at the next results, due by 28 Feb
We would also be wrong if
- The electric-drive business still loses money in 2027.
- Turbine sales in the first year come in below $300M.
What is coming
- December 2026 Turbine capacity decision: Whether demand beyond Endeavour is real enough to add capacity, ideally with a volume or a second named customer; silence would leave potential where it is.
Every dated event ahead
- December 2026 Turbine capacity decision: Whether demand beyond Endeavour is real enough to add capacity, ideally with a volume or a second named customer; silence would leave potential where it is.
- February 2027 Q4 2026 results and 2027 guidance: Whether two years of awards turn into growth above vehicle production, as the chief executive promised, and what the turbine adds to the 2027 plan.
- July 2027 2.65% notes due: Whether the remaining 2027 notes are repaid from cash or refinanced, after part was tendered in August 2026.
- December 2027 Turbine production starts in Hendersonville: Whether 2 GW of capacity is installed and shipping on time; first shipments would move the turbine from a management figure to a filed one.
Since we wrote this
- 9 NovPrice moveShares fell 20% on no news; no turbine premium leftThe 9.2 times line was crossed at 8.7 times; the line is redrawn at 8.0. The full-year margin leg waits for February, so the view holds.
- 9 NovThesis reviewAttractive: priced at 66% of a value near $73.53Re-valued in three stages at an 8% return, the method now drawn on every thesis: the price sits past the 20% margin of safety.
- 29 OctEarningsQ3 2026: margin guide held, no turbine order on the callQ3 2026 results and the earnings call
- 5 OctThesis reviewPilot: the bet now names what it ridesThe owner asked that each bet name the macro trend behind it, after BWXT's bet was missing its chief executive's case for nuclear over gas.
- 4 OctThesis reviewOwner's checklist: revenue streams, cost of growth, management, catalystsThe owner's checklist found the method silent on who runs the company, what growth costs, where the money comes from and what is dated ahead; this pass fills those four blocks from the proxy and filings.
- 26 SepThesis writtenNeutral, medium convictionA fairly priced, cash-rich car-parts business, with about $1.4B in the price for a data-centre turbine that rests only on management's figures.
- 26 SepThesis reviewNeutral, medium conviction, heldValued like other generator makers rather than like car parts, the $1.4B in the price pays for about one full year of turbine sales, not three. Priced at the plan, not beyond it.
Also checked 6 times with nothing new to act on.
Sources 32
Sources - 32
Every claim above cites one of these, or states the reasoning that derived it. The quote is frozen at the date it was read: a link can rot, and an agent fetch returns a reading of a page rather than the page. Backs counts the whole claim ledger, so it can name a claim this page does not surface.
- S01primarycompany filing2026-02-11Item 1 business and segment descriptions, the turbocharger share of sales, customer concentration, the named competitors, the insourcing risk factor, eProduct and Foundational revenue, R&D, share count on the coverbacksC01C02C03C05C06C13C21C22C56C60
- S02primarycompany filing2026-02-11segment adjusted operating income, the $624M impairment, the non-GAAP reconciliation, the TurboCell award, original 2026 guidance, full-year cash flow and balance sheetbacksC02C04C06C07C11C28C31C45
- S03primarytranscript2026-02-11Q4 2025 earnings call transcript (FMP)data/cache/transcripts/BWA-2025Q4.jsonmanagement's account of the turbine generator: 65% of content controlled, the capacity-to-revenue question left unanswered; analysts' questions are a separate origin from the answers
So first of all, the $300 million does not relate directly to the two gigawatts.
backsC12C14C31C53 - S04primarycompany filing2026-08-05debt schedule and revolver, liquidity, the tariff-refund treatment, the buyback authorisation, 203,666,354 shares at 31 July 2026, no mention of the turbine generatorbacksC13C28C29C37C55
- S05primarycompany filing2026-08-05Q2 2026 segment sales and adjusted operating income, updated 2026 guidance, the $1B buyback increase, first-half cash flow and balance sheetbacksC07C08C09C28C29C40C45
- S06primarytranscript2026-08-05Q2 2026 earnings call transcript (FMP)data/cache/transcripts/BWA-2026Q2.jsonmanagement's answers on the turbine, battery storage and inverters; ten analysts' questions, nine of them touching the data-centre products
what we've clearly stated is Craig and I are not pleased with the current outgrowth that the company is witnessing
backsC10C12C15C16C34C38C39 - S07primarycompany filing2026-09-04the company reiterating, 'in response to questions', production in Hendersonville in 2027, 2 GW of initial capacity and more than $300M of first-year salesbacksC11C12C13
- S08primarycompany filing2026-08-10the note series outstanding, the any-and-all offer for the 7.125% 2029 notes and the $720M waterfall cap including up to $250M of the 2027 notesbacksC30
- S09primarycompany filing2026-02-19the rival turbo maker's size of the OEM turbocharger industry, and S&P Global Mobility's forecasts of turbo adoption, hybrid output and battery-electric share that Garrett quotesbacksC17C19
- S10primarycompany filing2026-03-27Magna names BorgWarner and BYD/FinDreams among the key competitors of its Power & Vision segmentbacksC42
- S11primarygovernment data2025-01-15Lawrence Berkeley National Laboratory, 2024 United States Data Center Energy Usage Report (news release) Lawrence Berkeley National Laboratorythe size and growth of US data-centre electricity use, measured by a national laboratory for the Department of Energy
The total data center electricity usage climbed from 58 TWh in 2014 to 176 TWh in 2023 and is estimated to increase between 325 to 580 TWh by 2028.
backsC25 - S12primarycompany filing2026-02-13the incumbent engine and turbine maker describing prime-power orders from data-centre customersbacksC24C46
- S13primarycompany filing2026-02-09a shipping on-site power rival doubling factory capacity from 1 GW to 2 GW by the end of 2026, and a utility procurement of up to 1 GWbacksC23
- S14primarycompany release2026-02-10Volkswagen Group press release: 5 million electric drives produced; APP550 sets the benchmark Volkswagen AGBorgWarner's largest customer describing its own electric-drive and inverter production
1 million units produced at the Kassel plant. Next step for the technology: Pulse inverter developed in-house will be a key component in strengthening future electric drives for the Electric Urban Car Family by allowing for faster product development, cost reductions and more independent manufacturing.
backsC20 - S15secondaryvendor data2026-09-26BWA research pack (FMP, FINRA, SEC XBRL), frozenprice and return series, the price-move split, the multiple against its own history, analyst rating history, price-target drift, earnings surprises, short interest, the filed cohort's multiplesbacksC32C33C35C41C45
- S16primarycompany filing2026-09-26each open-market trade in the year to September 2026 with its Rule 10b5-1 checkbox read from the filingbacksC36
- S17primarycompany release2026-09-26that the turbine generator reaches buyers only through Endeavour
TurboCell is available exclusively through Endeavour’s Edged Infrastructure solutions division.
backsC15 - S18primarycompany filing2026-02-10Cummins sells standby and prime power generators up to 3.5 MW to data-centre customers, with data-centre demand extending six to eight quartersbacksC46
- S19primarycompany filing2026-02-18Generac's large-megawatt diesel generators sold into the supply-constrained data-centre marketbacksC46
- S20secondaryvendor data2026-09-26FMP trailing ratios for Cummins, Caterpillar, Generac, GE Vernova and Bloom Energyenterprise value and trailing operating margin, from which enterprise value to trailing operating profit is struck for eachread, but no claim rests on it
- S21primaryindustry data2026-06-25SemiAnalysis - US Grid Constraints: Towards 40GW+ of Behind-The-Meter Datacenter by 2028? SemiAnalysisan independent research firm's forecast of US data-centre additions (+21GW in 2026 to +84GW by 2030), the behind-the-meter share from 2028 and a BTM equipment market above 50GW a year by 2029; the free preview of a paid report, read through ir_fetch
Our research suggests that BTM will power well over half of new US datacenters in 2028+ , and the Total Addressable Market (TAM) for DC BTM equipment to cross 50GW/ year by 2029.
backsC49 - S22secondaryanalyst2026-05-14RBC Capital Markets insight - Natural gas powers the data center boom RBC Capital Marketsa bank's tally, from Bloomberg and company reports, of announced, ordered and under-construction behind-the-meter gas capacity for US data centres, split by technology (gas turbine nearly 70GW, engine about 18GW, unknown 13GW, mixed 3GW); read through ir_fetch
As of 2026, data center developers have announced approximately 101 gigawatts of behind-the-meter natural gas generation capacity in the United States. While some of these projects may be speculative, over 57 gigawatts have already publicly disclosed equipment orders, and approximately 7 gigawatts are already under construction.
backsC50 - S23primarytranscript2026-04-30Caterpillar Q1 2026 earnings call transcript (FMP) CATdata/cache/transcripts/CAT-2026Q1.jsonthe incumbent genset maker's sixth prime-power agreement of a gigawatt or more (up to 2.1GW for ProPower) and its large-engine capacity plan, 50GW a year plus about 15GW more; read through transcript_fetch (FMP)
This represents the sixth agreement with at least 1 gigawatt of Caterpillar equipment for prime power applications. ... From a gigawatt standpoint, we gave the 50 gigawatts. ... we estimate this will give us another 15 gigawatts capacity annually when we're done with this installation.
backsC51 - S24primarytranscript2026-08-04Cummins Q2 2026 earnings call transcript (FMP) CMIdata/cache/transcripts/CMI-2026Q2.jsona genset rival adding 20GW of capacity, the larger step in 2028, and a multiyear hyperscaler agreement for several gigawatts of backup gensets; read through transcript_fetch (FMP)
the additional investment in capacity, 20 gigawatt incremental capacity across basically all of our plants and our supply chain. And we'll see some of that coming online next year. So we'd expect some step-up and then the bigger step-up happening in '28.
backsC51 - S25primarycompany filing2026-06-25a listed small-turbine generator maker's average revenue per megawatt shipped, about $1.5 million in fiscal 2026 and $1.4 million in fiscal 2025, on $56.9 million of microturbine product revenue; read through filing_fetchbacksC52
- S26primarytranscript2026-05-06Q1 2026 earnings call transcript (FMP)data/cache/transcripts/BWA-2026Q1.jsonthe chief executive agreeing that 2GW of turbine capacity is 'multiples' of the $300M first year, without giving a price; read through transcript_fetch (FMP)
but just ballpark like 2 gigawatts is multiples of $300 million of revenue. Is that fair to say? ... Yes. We haven't provided pricing. So yes, multiples is a fair way to think about it.
backsC53 - S27primarycompany filing2026-03-19beneficial ownership as of March 3, 2026 and the 5% holders, the executive biographies, the 2025 Management Incentive Plan metrics and payout, the 2025-2027 long-term incentive designbacksC57C58C59
- S28primarycompany filing2024-02-082021 to 2023 income statements recast for the PHINIA spin-off, and gross and net R&D for those yearsbacksC56
- S29primarycompany release2026-10-29the October guidance ranges for sales, margins and earnings per share
BorgWarner Reports Strong Third Quarter 2026 Results, Increases 2026 Adjusted EPS Guidance
backsC08 - S30primarytranscript2026-10-29Q3 2026 earnings call transcript (FMP)data/cache/transcripts/BWA-2026Q3.jsonmanagement's account of turbine launch progress and the absence of a capacity or customer announcement; analysts' questions are a separate origin from the answers
We will not be giving you any kind of color until we get into early 2027.
backsC16C61 - S31primarycompany filing2026-11-09trailing revenue, operating profit, stock pay, tax, debt, cash and the diluted share countbacksC63
- S32secondaryvendor data2026-11-09the analysts' average revenue and operating profit for the next three fiscal years, and the closebacksC62C63