Nuclear Energy: Worth Getting Paid to Buy It at a More Reasonable Price? (as of 2026-09-01)

Updated 12 October 2026: what happened next, including to the BWXT put below, is at the end of this article. The original analysis is unchanged.

As of 2026-09-01. Educational and analytical only. Not investment advice. Company figures come from the companies' own Q2 2026 filings and earnings releases; screen figures come from the levelbox screener's own 2026-09-01 run at 0.20 delta, 30-day target. Where a cash-secured put is mentioned, its max loss and breakeven are stated. Nothing here is an instruction to buy or sell anything.

This article was generated by AI — levelbox's deep-dive tooling — and lightly reviewed. The data is point-in-time and can be wrong. If you spot a mistake, please tell us.

A friend recommends a nuclear stock. The story behind it is one of the better ones on the market right now. If you sell puts, the next question is whether you can get paid to wait for a cheaper entry.

We added a nuclear theme to the screener and ran nineteen names through it on 2026-09-01. Seven of them never reached the ranking.

The demand is real. The reactors arrive in 2028 at the earliest.

Big tech has committed to more than 10 GW of new nuclear capacity. Google signed the first US corporate SMR fleet deal with Kairos Power, targeting 500 MW. AWS is putting $20bn into Pennsylvania alongside Talen's existing nuclear site. GE Vernova's BWRX-300 at Darlington became the first small modular reactor under construction in North America, and TerraPower's Natrium plant in Wyoming has its NRC construction permit.

Here is the timing. No SMR operates in the United States today. First-of-a-kind units are realistically 2028 to 2030, and most planned deployments land in the early-to-mid 2030s.

None of that argues against the sector. It argues about which instrument suits it. A cash-secured put is a two-to-six week commitment that can leave you holding shares. A thesis that pays off in 2032 is indifferent to what happens between now and October.

Seven of nineteen never got ranked

Before scoring anything, the screener applies a floor: peak annual revenue of at least $100m. Anything below that is excluded with its number attached, rather than ranked low.

On 2026-09-01 that removed seven nuclear names:

excluded	peak annual revenue
OKLO	$0m
NANO Nuclear (NNE)	$0m
Lightbridge (LTBR)	$0m
NexGen (NXE)	$0m
Denison (DNN)	$9m
NuScale (SMR)	$37m
Energy Fuels (UUUU)	$78m

That list is most of what people mean when they say "nuclear stocks". The floor is there because a cash-secured put makes you an owner at the strike, and there is no way to judge a fair price for a company with no revenue to value.

Those names still carry the fattest premiums in the sector. The premium is fat because the valuation question is open, which is a different transaction from being paid to wait at a price you have already decided you like. Both are real trades. They pay you for different things.

What was left, and where BWXT landed

Twelve names cleared the gate. The run below is at 0.20 delta, 30-day target:

rank	name	segment	score	quality	growth	value	health
1	GEV	reactor OEM	0.56	0.63	0.33	1.00	1.00
2	URA	ETF	0.41	0.00	0.00	1.00	0.00
3	NUKZ	ETF	0.40	0.00	0.00	1.00	0.00
4	UEC	fuel	0.38	0.10	0.50	0.50	0.33
5	CEG	IPP	0.37	0.10	0.00	0.67	0.67
6	J	services	0.36	0.27	0.33	0.67	0.67
7	NLR	ETF	0.34	0.00	0.00	1.00	0.00
8	CCJ	fuel	0.34	0.20	0.33	0.67	0.33
9	BWXT	reactor OEM	0.34	0.37	0.00	0.67	0.33
10	TLN	IPP	0.30	0.37	n/a	n/a	0.00
11	VST	IPP	0.30	0.10	0.33	1.00	0.00
12	LEU	fuel	0.28	0.10	0.00	0.33	0.33

The three ETFs get scored on premium, liquidity and entry alone, so their 0.00 growth and health readings are a scoring convention and say nothing about the funds.

BWXT sits ninth of twelve. Mid-table, not disqualified.

BWXT: a good business, an awkward price

The operating figures deserve the attention they get. Q2 2026 revenue was $901.6m, up 18% year-on-year. Backlog hit $8.40bn, up 40%, on a trailing twelve-month book-to-bill of 1.7x. Management raised full-year guidance to $662–672m of adjusted EBITDA and $345–360m of free cash flow. In July it agreed to sell the medical business to Nordic Capital for up to $800m while keeping a minority stake, and closed the Precision Components Group acquisition. It builds the reactors for the US naval propulsion programme, which is about as durable a franchise as this sector offers.

The screener's own re-rating check agrees, and labels BWXT's revenue trend accelerating on 18.0% year-on-year growth. It attaches two qualifiers to that label: valuation context rich, and margins compressing.

Those qualifiers are what the score is picking up.

The multiple has run ahead of the growth. P/E of 41.96 against 14.1% earnings growth, PEG 1.67. The GARP lens scores that combination 0.00, the only zero among the non-ETF names except CEG and LEU.

Cash conversion is thinner than the headline suggests. Free-cash-flow margin of 4.1% and debt-to-equity of 1.51. GE Vernova converts at 38.0% with debt-to-equity of 0.28. Both companies grow. They turn revenue into cash very differently.

The trend has not turned. BWXT trades 16.3% below its 200-day moving average and its support line slopes down. The health lens runs three checks and BWXT passed one, the re-rating check. Both price checks failed.

At $161.54 the stock sits 33% under its 52-week high of $241.82. The nearest support the model finds is $147.74, which is also the 52-week low, so nothing is established above it.

Get paid to wait, or just wait?

Selling a put means naming a price and collecting income while you wait there. It works once you can answer two questions. Would you be content owning this at the strike, and does the premium pay you enough for the wait?

On BWXT the first answer is arguable in a good way. A naval-reactor franchise raising backlog 40% is a reasonable thing to want to own at the right number.

The second is harder. On the 2026-09-01 close the 150-strike put expiring 2026-09-18 marked around $1.60 a share, about $160 for one contract at roughly 0.19 delta, with a modelled assignment probability of 18.6%. Max loss about $14,840 if the shares go to zero, breakeven near $148.40.

That breakeven sits just under the 52-week low, which sounds like room until you notice support is still sliding and the multiple is still rich. You collect about 1.1% of the capital you tie up for seventeen days.

BWXT has a mechanical problem too. It lists no weekly expiries and six expirations in total, and the quoted spread on that strike was wide. Turning positions over every few weeks needs a dense expiry ladder. Monthly-only names limit the cadence whatever you make of the business.

What the run suggests

The sector divides into what you can underwrite and what you cannot, and the screener draws that line for you at $100m of revenue. Fuel, enrichment, the reactor builders and the independent power producers have revenue and valuations you can argue with. The SMR developers have order books and optionality, and they did not make the list.

Among the twelve that did, the purest expression of the theme and the best-scoring stock were different names. GE Vernova topped the run at 0.56 while selling gas and wind equipment alongside its reactors, so it waters the theme down. BWXT gives you the concentrated version at 0.34. Choosing between those is a judgement about what you are buying, and no screen makes it for you.

A strong business can also be an awkward entry, which is why valuation and trend get scored beside quality. "Get paid to wait for your price" needs you to pick the price first. On this data, $150 is not obviously it.

The nuclear theme is live in the levelbox screener if you want to run it against your own delta and DTE settings. These numbers move. This is a snapshot of 2026-09-01.

Update, 12 October 2026: the electrons are winning

Six weeks on, the sector has split, and the split follows the timing argument above.

This section is an AI-assisted view, not financial advice. Tickers are examples used to describe the sector, not picks.

What the stocks did

From the 1 September close to the 9 October close:

Group	Names	Move
Power producers with existing nuclear fleets	Talen +28.9%, Vistra +16.9%, Constellation +6.3%	Up
Reactor OEM with gas and grid	GE Vernova +11.8%	Up
Engineering services	Jacobs −4.9%	Flat to down
Reactor OEM, nuclear only	BWXT −11.0%	Down
Fuel and enrichment	Cameco −8.5%, Centrus −15.4%, Uranium Energy −20.7%	Down
Small-reactor developers	Oklo −9.9%, NANO Nuclear −12.2%, NuScale −21.8%	Down
Sector ETFs	URA −11.5%, NLR −10.7%, NUKZ −3.6%	Down

The S&P 500 (SPY) rose 2.2% over the same period. The power producers took most of their gain in the first week of October: Talen rose 18.0%, Constellation 15.8% and Vistra 15.3% in the week to 9 October.

Why it split

Deals that add megawatts at existing plants.

• On 6 October Google signed a 20-year agreement with Constellation for 890 MW of new capacity from uprates, upgrades that raise the output of 11 existing reactors in Illinois, Pennsylvania and New Jersey. First power is due in 2028. A separate 2,700 MW supply agreement brings the total to about 3.6 GW (Reuters).
• On 5 October the Department of Energy made a conditional commitment to lend Vistra up to $4.2 billion for 433 MW of uprates at three of its nuclear plants, linked to its power agreement with Meta. The loan still depends on conditions being met (Utility Dive).

Both put new nuclear power on the grid in two to three years from plants that already run. A new reactor takes until the 2030s.

A permit for the first commercial small reactor. On 29 September the Nuclear Regulatory Commission issued a construction permit for TVA's Clinch River BWRX-300, four months ahead of schedule. The reactor design is GE Vernova Hitachi's. It is a real milestone for small reactors, and the company that gains most directly from it also sells gas turbines and grid equipment.

The fall in fuel and small-reactor stocks was not about uranium. The spot price averaged $89.68 a pound at the end of August and $89.63 at the end of September, according to Cameco's published averages of UxC and TradeTech data. The long-term contract price held at $96.50. Two other things happened instead:

• New shares. Centrus sold about $500 million of stock at $199.64 a share on 9 September, with warrants attached. It closed at $142.40 on 9 October. Oklo set up a $1 billion at-the-market equity programme on 11 September, and Denison filed a US$700 million shelf.
• Higher rates. The 10-year Treasury yield rose from 4.79% on 1 September to 5.24% on 9 October. A company whose earnings arrive in the 2030s is worth less when money costs more today. That is a general effect of higher rates. We found no commentary tying it to these stocks specifically.

What happened to the BWXT put

The article above looked at a BWXT 150 put expiring 18 September, at about $1.60, and concluded that $150 was "not obviously" the right price. BWXT closed at $147.52 on 18 September, so that put would have been assigned. With the premium, the shares would have cost about $148.40. They fell to $134.35 on 28 September and closed at $143.80 on 9 October, about 3% below that cost.

BWXT held an investor day on 29 September. Secondary reports put its 2030 targets at $5.5 to 6 billion of revenue, and 2026 guidance was reported unchanged. The business case is intact. The trend is weaker: the stock is now 23.9% below its 200-day average and 39.7% below its 52-week high, with the 50-day average still falling.

The screener, re-run on 9 October

Same theme, same settings (0.20 delta), shipped screener run on 9 October:

rank	name	score	quality	growth	value	health	put strike / expiry	credit
1	URA	0.70	0.00	0.00	1.00	0.50	36 / 6 Nov	$62.50
2	GEV	0.63	0.63	0.33	0.67	0.67	940 / 23 Oct	$1,020
3	J	0.46	0.27	0.33	0.67	0.67	130 / 20 Nov	$197.50
4	NUKZ	0.43	0.00	0.00	1.00	0.00	57 / 20 Nov	$92.50
5	UEC	0.41	0.10	0.50	1.00	0.67	8 / 6 Nov	$17
6	TLN	0.41	0.37	n/a	n/a	0.33	335 / 23 Oct	$646
7	VST	0.39	0.10	0.33	0.67	0.33	150 / 30 Oct	$253.50
8	CCJ	0.37	0.20	0.33	0.67	0.67	83 / 23 Oct	$76.50
9	BWXT	0.35	0.37	0.00	0.67	0.33	130 / 20 Nov	$330
10	CEG	0.33	0.10	0.00	0.33	1.00	270 / 6 Nov	$525
11	LEU	0.21	0.10	0.00	0.33	0.00	120 / 20 Nov	$465

The same seven names stayed below the $100 million revenue floor. NLR dropped out of the ranking because no put could be priced at the target delta on that day.

Three changes since September:

• BWXT is ninth again, now of eleven. The 130 put expiring 20 November paid about $330, roughly 2.5% of the capital tied up for six weeks. That is about the same rate per day as the September put, from further away: the strike sits 9.6% below the 9 October close, against 7.1% for the September example. Max loss about $12,670 if the shares went to zero, breakeven near $126.70. The stock is also still below a falling 200-day average, which is the case covered in our post on value traps.
• Constellation fell to tenth even as its shares rose. Its value score dropped from 0.67 to 0.33 as the price climbed, while its health score reached 1.00. The screen likes the trend and is less keen on the price.
• The ETFs top the list. URA ranks first on premium, liquidity and value. As in September, its 0.00 quality and growth readings are a scoring convention for funds, not a judgement on the holdings.

How to read it

The September conclusion holds. The nuclear thesis pays off over years, and a cash-secured put is a commitment of weeks. Over these six weeks the market paid for megawatts it can see by 2028, from plants that already run, and marked down companies whose payoff is in the 2030s and which are raising capital to get there.

That fits the power row in our AI buildout scenarios update: power holds up in two of three scenarios. Contracted output from existing plants is the part most likely to survive a financing squeeze, and new builds are the part most likely to slip.

Dates ahead: GE Vernova reports on 28 October and Cameco on 30 October. Vistra is expected around 6 November. Constellation, BWXT and Oklo have not confirmed dates yet. The Fed decides on 28 October.

Sources for this update: daily closes to 9 October 2026, our own calculation; levelbox screener run of 9 October 2026 (nuclear theme, 0.20 delta); Reuters via TimesLIVE (6 October 2026) on the Google and Constellation agreement; Utility Dive (6 October 2026) on the Vistra loan commitment; NRC news release 26-079 (29 September 2026); Cameco published uranium price averages; Centrus, Oklo and Denison SEC filings (9 to 11 September 2026); US Treasury daily par yield curve. The BWXT investor-day figures are from secondary reports and are unverified.

This update is analytical and educational and is not financial advice. Tickers are examples, not recommendations. Figures are a point-in-time snapshot and can contain errors. This update was generated with AI assistance. If you spot a mistake, tell us.

Related: the power complex sits inside a wider chain. What actually breaks in the AI value chain looks at how much of the load forecast survives if the AI buildout halves, and why merchant power carries the highest leverage in the group.