DK Street Journal

Jackson Financial sold $6bn of retail annuities in the second quarter, a third more than a year earlier, and reported net income that had almost nothing to do with it. The Lansing, Michigan company — which sells variable, fixed, fixed-indexed and registered index-linked annuities to individual American investors through independent broker-dealers, banks, wirehouses and registered investment advisers — earned $655m under the accounting rules in the June quarter after losing $424m in the March one. Its reported revenue line was $168m in the June quarter and $2,902m in the March quarter, because derivative marks run straight through it.

For an annuity writer, deposits are not revenue but additions to policyholder account value, and reported earnings are dominated by the quarterly remeasurement of the guarantees written into contracts, which moves opposite the hedges placed against them. What remains that a shareholder can bank is statutory capital, holding-company cash and the share count. On those meters, Jackson and Lincoln Financial are running in opposite directions.

The hedge is the business

Jackson's own numbers make the point cleanly. Its net hedge result in the quarter was a $2m gain. The index-linked book showed a $200m accounting gain but was near breakeven economically; the variable-annuity book showed a $200m accounting loss but was flat measured through Brooke Re, the Michigan captive reinsurer Jackson capitalized with about $1.9bn in 2024 specifically so that hedging to economics rather than to a statutory accounting floor would not wreck its capital ratios. Brooke Re took no contributions and paid no distributions in the quarter.

Underneath, the spread engine is widening rather than compressing. "Of the money that we put to work, the new money yield was roughly about 100 basis points ahead of our overall portfolio yield," chief financial officer Don Cummings told investors on the August 4 call, cautioning that the quarter might not repeat. That is a function of where rates sit: the ten-year Treasury was about 4.77% and the thirty-year about 5.25% in early September, near three-year highs. F&G Annuities & Life, the smaller fixed-index writer, likewise reported its portfolio yield rising to 4.91%, up 14 basis points sequentially.

The product mix is moving where the demand is. Registered index-linked annuities, which cap the upside and floor the downside instead of guaranteeing income for life, set an industry record of $23.3bn in the quarter, inside a record $123.9bn of total US annuity sales. Jackson's share of that was a record $2.3bn, up 69%. "We remain an industry leader with more than $26 billion in RILA assets," chief executive Laura Prieskorn said on the same call. Industry leader is generous: Equitable leads the category through its Structured Capital Strategies family, with Allianz Life, Prudential and Brighthouse also ahead of or alongside Jackson. Jackson is taking share in a market it does not lead.

The capital arithmetic is the payoff: $656m of after-tax statutory capital generated in the quarter, a risk-based capital ratio of 538% against a 425% target, about $1.4bn of holding-company liquidity, and guidance for at least $1.2bn of free capital generation this year with $900m to $1.1bn returned. Diluted shares have fallen from 88.7m in 2021 to about 70m.

Lincoln is the mirror

Lincoln Financial — the Radnor, Pennsylvania insurer that runs annuities, retirement plans, life insurance and a group disability and dental business — earned $439m of adjusted operating income, or $2.24 a share, up 3% and an eighth straight quarter of growth. Its reported net income was $1,332m. Behind that, annuity net outflows ran about $2.9bn, driven by traditional variable annuities, even as account balances hit a record $182bn. The growth engine is elsewhere: group protection balances rose 13% to $51bn, and "Group Protection extended its track record of excellent operating performance," chief executive Ellen Cooper said with the results.

Lincoln is still moving legacy risk off its own balance sheet — it agreed to reinsure roughly $6bn of life reserves with Talcott, costing about ten points of risk-based capital against a 420% buffer — and holding-company liquidity is $903m after a $500m subordinated issue. Its diluted share count has risen from 174.0m in 2024 to 196.4m.

What the market is pricing

Both stocks are cheap against earnings that mean something: Jackson at roughly 5.0x this year's consensus of $27.08 a share, Lincoln at about 5.5x $7.94. Against book value, which is the anchor that matters for an insurer, they have separated — Jackson at 0.96x, Lincoln at 0.74x, F&G at 0.68x and Brighthouse at 0.46x. Consensus has Lincoln's earnings flat this year; Jackson's are seen up 23%.

The whole life complex fell together between August 17 and 21 — F&G down 16.3%, Brighthouse 8.1%, Lincoln 7.7%, Jackson 3.5%, against a 0.9% decline in the S&P 500 — with no company news at either protagonist. The discoverable mechanism is credit quality: hedge-fund shorts against US life insurers have more than doubled to over $5bn, and Fitch put the US private-credit default rate at a record 6.0% in April, while regulators began requiring asset-adequacy reports on offshore and captive reinsurance from April 1. Brighthouse's own decline is deal mechanics: it closed at $53.17 against Aquarian's $70.00 cash price after Delaware's insurance department disclosed on August 17 that its review is ongoing.

Jackson's year has been earned by the only things that are countable here — capital generated, capital returned, shares retired — and that part of the re-rating is not a puzzle. What is spent is the discount: at 0.96x book, buying back stock adds far less per dollar than it did at two-thirds of book, and the next leg has to come from the spread and the fee base rather than from the share count. Lincoln's discount is paying for the outflows, the dilution and the reinsurance overhang, and until net flows stop being negative there is no obvious event that closes it.

Don Cummings, who spent the August call explaining why a $200m accounting gain was worth nothing, becomes chief executive on October 1.

Appalachian producers who shut in wells this summer have started bringing them back. The winter forward price at the TETCO M-2 hub, the benchmark for gas leaving the northern Marcellus, has narrowed its discount to Henry Hub by about 45 cents over five months, and gas flowing to the nine large US liquefied natural gas export plants rose to 18.3 billion cubic feet a day in early September from 17.2 in August as Texas facilities came back from maintenance. The spot price did none of this work: Henry Hub settled at $2.872 per million British thermal units on September 8, and the Energy Information Administration cut its 2026 forecast to $3.44 from $3.67 while projecting a record 3,985 billion cubic feet in storage at the end of October.

So the question for the three largest listed Appalachian and Haynesville producers is not whether the export pull arrives. It is how much of each company's gas is still for sale when it does. A hedge book layered years in advance is the difference between owning that curve and having sold it.

The book that is already sold

Expand Energy, formed when Chesapeake renamed itself in 2024 and now North America's largest gas producer with roughly 5,000 wells across the Marcellus and the Haynesville, had hedged 66% of its 2026 volumes and 41% of 2027 as of late July, using swaps and collars. It realized $2.42 per thousand cubic feet of gas in the June quarter, down 17.4% year on year; revenue fell 19.7% to $2.96bn and operating margin compressed to 22.3% from 34.4%. Management put standalone breakeven near $2.70 per thousand cubic feet before dividends, told investors the market stays oversupplied through at least the first half of 2027, and let roughly ten Haynesville well completions slip — "happy to allow these to float into 2027," on the July 29 call. It also bought $1.25bn of gas marketer Twin Eagle. "We're already North America's largest natural gas producer, and now we'll be its leading gas marketer," interim chief executive Michael Wichterich said on July 27. Expand has had no permanent CEO since February.

The book that is open

EQT, the Pittsburgh producer with 25 trillion cubic feet of proved reserves and 1.7m Marcellus acres, is the opposite exposure. Its 2027 hedges amount to roughly 1.5 billion cubic feet a day of swaps at $3.16 plus collars — thin against a company that sold 634 billion cubic feet equivalent in the June quarter at a differential of minus $0.67 to the NYMEX benchmark, with unit operating costs of $1.03 and $330m of free cash flow. The reported profit is where the openness shows: revenue down 29.2%, diluted earnings of $0.44 against $1.30. EQT is selling demand instead, including 325,000 dekatherms a day for ten years to a two-gigawatt West Virginia power plant, priced off PJM electricity rather than in-basin gas. The agreement "provides EQT a substantial premium over in-basin pricing," chief executive Toby Rice said on July 21.

Range Resources, the Fort Worth-based Marcellus operator with about 794,000 net acres, was the only one of the three to grow: revenue up 19.1% to $834m, operating margin 39.1% against 26.8%. It realized $3.53 per thousand cubic feet equivalent including hedges, of which just $0.16 came from settled contracts. The rest was mix — natural gas liquids at $29.10 a barrel, and full-year guidance raised to a premium of $2.00 to $2.50 over the Mont Belvieu benchmark. Range also added roughly $1.1bn of firm midstream commitments and guides transport, gathering and processing at $1.55 to $1.60 per thousand cubic feet equivalent, a fixed cost that solves takeaway and does not go away in a weak quarter.

What the shares are paying for

All three regained an uptrend during August after a spring slide, and over the past month Expand rose 6.3%, EQT 7.3% and Range 10.3%; over six months Expand and EQT are still down 8.5% and 10.9%. Consensus models 2027 earnings per share lower at every one of them — roughly 5% down apiece — which encodes the curve rather than any operating stumble.

The verdict is that the rally is real and misallocated. The export pull is a mechanism, not sentiment, but it is dated to late 2027 and beyond by Expand's own market view, and the storage table says nothing is tight today. Range earns its move outright: it grew, and it grew on barrels priced off crude. Expand's move is earned by cash rather than the curve — an 11.2% trailing free-cash-flow yield and the cheapest cash-earnings multiple of the three, on a book whose next year is largely pre-sold and cannot rise with the strip. EQT's is the one nothing in the reported numbers explains: the most expensive of the three at 12.1 times trailing and 13.5 times forward earnings, on the steepest revenue decline. What the market is buying there is the open volume itself.

Storage will peak near a record in seven weeks, and every molecule of it is a reason the tightness stays a 2028 argument. Expand has sold most of that argument forward. EQT has kept it.

Two premium footwear makers raised their full-year forecasts this summer while the apparel companies around them were cutting. Deckers Outdoor, which sells UGG sheepskin boots and HOKA performance running shoes through wholesale doors and its own stores, lifted fiscal 2027 earnings guidance to $7.35-$7.50 a share on July 23. Three weeks later Birkenstock, the German sandal maker founded in 1774 that manufactures in its own plants, raised its revenue and profit targets for the year.

Since those reports Deckers has fallen 18.6% and Birkenstock 20.7% from its post-earnings close. Over twelve months the two are down 28.9% and 32.2%. The reason that is worth taking apart is that the two declines have almost nothing in common: one is a pure re-pricing of unchanged earnings, and the other traces to a single day when a private-equity owner sold a billion dollars of stock.

Deckers: better margins, cheaper stock

Deckers' June quarter took revenue above $1bn for the first time, up 5.7%, with HOKA at $704m and UGG at $278m. Direct-to-consumer revenue rose 13% while wholesale grew 2.2%, and inventories fell 5% to $807.6m — no build ahead of the holiday quarter that decides UGG's year. Gross margin expanded 64 basis points to 56.4% even though tariffs cost roughly 150 basis points and the company lifted its assumed duty rate to 12.5% from 10%. That rate is not hypothetical: Section 122 duties expired on July 24 and were replaced the same day by a two-tier Section 301 forced-labor tariff on top of a 12.5% base rate, and Vietnam alone shipped 274 million pairs to the United States in 2024.

"Both HOKA and UGG maintained solid momentum and continued to capture high level of full price consumer demand," chief executive Stefano Caroti said on the July 23 call.

The honest problem is HOKA: growth halved to 7.7% from 14.5% the prior quarter, and Bank of America cut its price target to $115 from $120 citing moderating HOKA sales and more discounting. But the nearest rival slowed in the same three months. On Holding, the Swiss running brand, missed on second-quarter sales and cut full-year growth guidance to the low-20% range from at least 23%, and its shares fell as much as 22% to a two-year low. Performance running decelerated; HOKA did not lose to On.

Against that, the arithmetic of Deckers' decline is stark. A year ago the shares were $117.17 against a next-year consensus of $6.89; today they are $83.37 against $7.53. Forward earnings estimates rose 9.3% and the multiple fell 35%, from 17.0 times to 11.1. Consensus still sits above the company's own guidance range.

Birkenstock: the seller, not the business

Birkenstock's June quarter grew 15% in constant currency to EUR 720m, direct-to-consumer sales outgrew wholesale for the first time in two years, and own retail rose 50% as store count reached 124. Growth is roughly one-third price and two-thirds pairs sold, in line with management's stated target mix, with closed-toe styles gaining 500 basis points of penetration. Adjusted gross margin of 59.2% was down 130 basis points, all of it FX and US tariffs. Capital spending of EUR 110-130m is going into production at Wittichenau, Arouca and Görlitz; net leverage is 1.8x and falling.

The shares jumped 15.2% to $41.78 on the August 13 guidance raise. "Given our currently undervalued shares, we will look for opportunities to continue our buybacks," chief executive Oliver Reichert told investors that day. The next morning, an affiliate of L Catterton priced 25,523,226 shares at $39.35, about 4% below the prior close, with an option on 3.8m more; Birkenstock received nothing and simultaneously repurchased 12.8m shares. Volume that day ran near five times normal. At $33.14 the stock is 15.8% below where the seller cleared.

What the apparel comparison shows

Lululemon, whose second-quarter revenue fell 4% to $2.4bn on comparable sales down 9%, cut full-year revenue guidance to a 5-7% decline and earnings to $9.48-$9.73 a share; the stock lost 17.4% in one session on September 4. Oxford Industries, owner of Tommy Bahama, also lowered its outlook. Footwear fell in the same week — Deckers 5.0%, Birkenstock 6.1% — but on forecasts moving the other way.

So: the market is treating raised guidance and cut guidance identically. Part of Deckers' de-rating is earned — a brand growing 8% against a low-double-digit full-year promise is worth less than one growing 15% — but a 35% cut to the multiple against a rising estimate is more than that gap explains. Birkenstock's is harder to defend on fundamentals at all; an accelerating, de-levering manufacturer priced at 9.3 times trailing enterprise value to earnings before interest, tax, depreciation and amortization and 1.82 times book looks like the residue of a supply event the tape has not worked through.

Deckers has told investors HOKA will grow low double digits this fiscal year, and delivered 8% in the first quarter. The holiday quarter settles both that promise and UGG's, and it is the only evidence that will change either argument.

No news came out of Booking Holdings on 8 September, and the owner of Booking.com, Agoda, Priceline, KAYAK and OpenTable closed 5.7% lower at $182.20 — with every other listed online travel agency falling alongside it. Expedia Group, which runs Brand Expedia, Hotels.com, Vrbo and the business-to-business unit Expedia Partner Solutions, fell 5.6%. Airbnb and Tripadvisor each lost about 3%, MakeMyTrip 4.1% and Trip.com 1.3%. Trade coverage described a broad selloff that swept across travel services companies with no company-specific catalyst; intraday, Booking was quoted down 7.0% at $179.76 before recovering into the close. The macro backdrop was Brent crude near $99 a barrel intraday on renewed Middle East fighting, roughly 60% market-implied odds of a Federal Reserve rate increase on 15-16 September, and consumer discretionary among the least-wanted sectors.

That matters because the two largest names in the group reported June quarters that moved in opposite directions, and were sold within a tenth of a percentage point of each other regardless. What the session priced was the channel both companies rent — paid search — not the profit and loss statements underneath.

Booking earned its de-rating

Booking's revenue of $7,352m grew 8.1% year on year, against 16.2% in the March quarter and 16.0% in December. Room nights grew 5% while gross bookings grew 9%, so the dollar growth came more from price and mix than from travellers. Marketing expense rose 11% to $2,371m, or 4.7% of gross bookings, the company citing traffic-mix changes and incremental paid marketing — traffic got more expensive as volume growth slowed. Operating margin still improved, to 34.0% from 33.1%, and diluted earnings per share of $2.53 was flattered by a share count down to 770m from 815m after a record $3.7bn of second-quarter repurchases. Chief financial officer Ewout Steenbergen framed the year as intact: full-year guidance is "at a high single-digit level for gross bookings and revenues and at mid-teens level for EPS at the high end," he told investors on the 4 August call, with seven of twelve months carrying Middle East war impact.

Expedia did not

Expedia's revenue accelerated to 14.0% growth, its fourth straight quarterly improvement, and operating margin reached 23.9% from 14.0% — operating income nearly doubled. Gross bookings of $33.9bn rose 12%, with the business-to-business arm, which white-labels Expedia's hotel supply to airlines, banks and offline agents, up 21% and consumer bookings up 8%. "We exceeded the high end of both our top and bottom-line expectations for the fifth quarter in a row, growing bookings 12%, revenue 14%, and adjusted EBITDA 23%," chief executive Ariane Gorin said on 5 August. Chief financial officer Derek Andersen raised full-year margin guidance the same day. Expedia's take rate, 12.7% of bookings against Booking's 14.4%, is lower precisely because wholesale distribution is its fastest-growing channel — and it bought back only $200m of stock in the quarter, so the earnings growth came from operations.

Booking trades at 17.2x forward earnings against 19.9x trailing, 13.3x trailing enterprise value to EBITDA, on a 6.8% free-cash-flow yield; years of debt-funded buybacks have left book equity negative. Expedia trades at 13.2x forward against 16.5x trailing and 7.8x EBITDA, a roughly 24% forward-earnings discount to a slower-growing peer. In July, value investor Scott Black of Delphi argued Booking's then-17.5x multiple was not warranted by the size of its advantages and preferred Expedia; the price has come to him, but so have the estimates, which is why Booking's multiple has barely compressed.

The verdict

Booking's 15% slide from its 7 August close is mostly a business fact: growth halved, units grew half as fast as dollars, and acquisition cost rose. Expedia's 9.5% one-month decline has no such explanation — guidance went up, margin went up almost ten points, and share count did nothing. The residual on both is the same fear, and it is a repeat: Booking fell 25% in February 2026, its worst month since April 2010, and Expedia 32%, on the identical worry. Booking's own data does not corroborate it yet — Steenbergen said large-language-model referrals, paid and unpaid, are under 1% of total room nights and have not moved much. BTIG's Jake Fuller, on the launch: "We get the market's reaction but see the risk as overstated."

The detail worth holding is in the plumbing of Google's agentic hotel booking, live in AI Mode since 27 August with Booking.com, Expedia and Hotels.com as launch partners. The agencies remain merchant of record, so the commission survives the first cut. What they lose is the chat context ahead of it — the inspiration signals that tell them what a traveller wanted before they asked for a room.

Domino's Pizza sold more orders in its second quarter and collected less on each one. US same-store sales rose 0.1%, the softest quarter in over a year, and the company attributed the increase to higher customer transaction counts partly offset by a lower average ticket — the reverse of the pattern that has carried most quick-service chains through three years of menu inflation.

That matters because Domino's does not earn its money the way its revenue line reads. The chain runs three segments — US stores, international franchising and supply chain — across more than 22,500 locations, and the supply-chain arm, which mills dough and distributes cheese, meat and cardboard to franchisees at a cost-plus markup, generated $731.7m of the quarter's $1,194.4m in revenue, up 6.5%. The profit is elsewhere: royalties levied on what franchisees ring up. Income from operations rose 2.6% excluding currency, and Domino's credits higher US and international royalties plus supply-chain margin dollars.

The cheese line

Domino's defines its "food basket pricing change" as the cost of what an average US store buys from its distribution centers, and states plainly that as those prices move, supply-chain revenue, cost of sales and gross margin move with them. Last quarter the basket inflated 2.2%, supply-chain gross margin widened 0.2 percentage points and that segment's adjusted operating income rose $11.7m, or 18.1%. The direction is now reversing: CME 40-pound cheddar blocks were quoted at $1.4750 a pound on September 2, against $1.5400 in late May. Cheaper cheese subtracts from reported revenue while leaving franchisee food costs lower — a revenue decline that is good news at store level.

The meters that decide store count

Unit economics are improving, not eroding. Domino's added 209 net stores in the quarter, 26 in the US, and reaffirmed guidance of 175-plus net US openings and roughly 800 international. Average US franchisee store profit reached about $166,000 in 2025, up $4,000. Where closures happened — the master franchisee shutting low-volume restaurants in Japan and France — Japan profits rose 19% on lower revenue. Orders arriving through DoorDash and Uber Eats are priced at a premium on those menus so the economics land roughly neutral for the operator.

"In a quarter where the broader U.S. QSR industry continued to face pressure on consumer demand, Domino's generated order count growth across both our delivery and carryout businesses," chief executive Russell Weiner said in the July 20 release. It was his last quarter in the job; Weiner retires October 1 and chief operating officer Joe Jordan succeeds him.

The category is worse than the operator

Papa John's, which owns roughly 600 of its 5,650 restaurants and therefore carries labor and food cost on its own books, saw North American comparable sales fall 8.3% and revenue drop 8.8% to $482.4m. It suspended the dividend to fund franchisee incentives, is closing about 300 North American restaurants by the end of 2027, and carries $733.5m of debt at 3.3 times leverage against negative book equity of -$432.7m. Its board rejected Irth Capital's $47-a-share approach in March; the stock is $22.82. At about seven times trailing earnings before interest, taxes, depreciation and amortization, it is cheap against a base consensus expects to shrink 9.2% this year. Pizza Hut fared worse still — US system sales down 5% — and Yum agreed to sell the brand for $2.7bn.

So the category is genuinely shrinking, and Domino's is the one taking orders inside it. Its de-rating — to about 19 times trailing and 17.7 times forward earnings, from 23.7 times at the end of 2025 and 28.1 times at the end of 2023 — is earned to the extent that a chain compounding on royalties needs franchisee sales to grow, and they have stopped. What it does not explain is the rest: earnings still rose, units still opened, and consensus models profits up again next year and the year after.

The test arrives with cheese. Third-quarter supply-chain revenue will likely fall on a cheaper basket while franchisee food costs ease — a quarter where the reported line and the underlying business point opposite ways, handed to a chief executive in his first weeks.

Sherwin-Williams put a new price list into its roughly 5,000 North American stores on September 1 — an 8% architectural paint increase — and told investors to expect its historical realization rate of 60% to 70%, with the remainder phasing into 2027 through contract customers.

That lag is the business. A paint company sells a spread between an input basket it does not control and a price it sets by letter rather than by quote, on gallons that move with existing-home turnover. Both halves of the basket are now inflating on supply disruption. Brent crude stood at $97.41 a barrel on September 8, up 11% in a month and 47% on the year after attacks halted operations at Saudi energy facilities including the 400,000-barrel-a-day Jazan refinery; resins, solvents and additives are crude derivatives. Titanium dioxide, the pigment that whitens the can, went through four rounds of increases in the first half totaling roughly $450 to $500 a tonne, driven by sulphur prices up more than 90% this year — the sulphuric-acid route consumes three to four tonnes of acid per tonne of pigment. Over the past month Sherwin-Williams fell 11.6%, PPG 8.1% and Axalta 5.9%.

The squeeze arrives at the gross line first

Sherwin-Williams, which develops paint and sells it mostly through its own stores to professional contractors, is not obviously a company in trouble. Revenue growth has accelerated four quarters running, from 3.2% year-on-year to 7.5% in the June quarter, and operating margin widened to 18.13% from 17.41%. But gross margin went the other way, slipping to 49.17% from 49.38%. Part of the top-line acceleration is bought rather than earned: Consumer Brands grew mid-teens with the contribution from Suvinil, the Brazilian decorative paints business acquired from BASF for $1.15bn and closed last October, and only mid-single digits without it.

The gallons are real but modest. Paint Stores Group same-store sales rose 4.2%, split between low-single-digit price increases and low-single-digit volume growth, with residential repaint up a mid-single-digit percentage. The demand meter underneath has not turned: existing-home sales edged down 1.7% in July and have never recovered from the 30-year low of 2023. Management's guidance assumes no broad demand recovery. "We know growth will need to come from what we do, not from what the market gives us," chief executive Heidi Petz told investors on the July 28 call. Her chief financial officer, Ben Meisenzahl, was blunter about the cost side: "We are seeing the impact of higher oil and related cost pressures... We expect inflation in our raw material basket to be up in the high single digit range in the second half."

PPG, which sold its US and Canadian architectural business — the Glidden and Pittsburgh Paints brands — to American Industrial Partners for $550m in December 2024 and is now an industrial, aerospace and auto-OEM mix, put a number on the industry's recovery rate. "We covered about 90% of our cost of goods sold inflation with pricing, and we expect to reach 100% coverage by the fourth quarter," chief executive Tim Knavish said on July 29. Ninety percent coverage is under-recovery, described as progress.

The one uptrend belongs to a merger

Axalta is the only one of these names with a positive twelve-month return and the only one whose 50-day average has sat above its 200-day since early July. Its June quarter was genuinely good: Refinish sales up 6% to $545m, record adjusted profit before interest, tax, depreciation and amortization of $305m, and net leverage of 2.2x, the lowest in company history. "We far exceeded our normal run rate with more than 1,900 new net Body shops secured in the first half," chief executive Chrishan Villavarayan said on July 28. Reported operating income still fell 17.3%, on merger costs.

Those merger costs are the point. Shareholders of both companies approved an all-stock merger of equals with AkzoNobel on August 5 at a fixed 0.6539 ratio, tethering Axalta's price to AkzoNobel's and to antitrust review — a US Federal Trade Commission second request, with reporting that an in-depth European probe over powder coatings and refinish overlap may follow. Underneath the ratio, the unit meter shrinks: repairable claim volume fell 9.7% in 2025, and management expects collision claims down mid-single digits, with back-half volume resting on body-shop wins.

What the businesses earn

The part of this the businesses earn is the margin: cost programs, share gains, an acquisition, and price letters that stick two-thirds of the time. What none of them earns is a demand inflection, because there isn't one — the gallon count is crawling and the repair count is falling. Sherwin-Williams still trades at 29.8 times trailing and 27.0 times forward earnings, against 31.6x at the end of last year and against PPG at 15.7x trailing and 14.0x forward; barely two turns have come out while the business improved, which leaves the multiple, rather than the operations, carrying the risk. Axalta's 11.1 times trailing enterprise value to that same profit measure is supported by the numbers, but its share price now answers to a merger ratio and a regulator.

The next paint price increase will be decided in the Strait of Hormuz rather than in the housing market, and it will reach contract customers sometime in 2027.

Salesforce shares had their second-best session on record on 27 August, gaining 22.6%. The quarter behind the move showed revenue of $11.345bn, up 10.8% from a year earlier — of which $456m came from Informatica, the data-cataloguing and integration business Salesforce bought for about $8bn and consolidated from 18 November 2025. Remove revenue that was not in the year-ago base and the underlying company grew roughly 6%, against 13.3% in the April quarter.

The gap did not stop at Salesforce. All five names in marketing and customer-relationship software turned their trends upward between 16 July and 1 September, and all five rose over the following thirty trading days — Salesforce 29.1%, HubSpot 15.9%, Braze 13.4%, Klaviyo 8.3% and Sprout Social 1.8%, an equal-weight gain of 13.7% against a twelve-month equal-weight loss of 27.4%. What reversed was a story rather than a set of growth rates: months of positioning for AI to destroy per-seat software unwound, and Adobe and Autodesk rose alongside Salesforce in the same session. "This SaaSpocalypse narrative has been such nonsense," chief executive Marc Benioff told CNBC on 26 August.

What Salesforce actually reported

Gross margin fell to 76.65% from 78.10%, and GAAP operating income was flat year over year at $2.33bn. The genuinely good number was contracted revenue due within twelve months, which reached $33.5bn and grew 14% in constant currency — three points faster than revenue, and what management called the fastest bookings growth in four years. Free cash flow was $1.1bn, up 81%.

The per-share arithmetic is the loudest part of the story. Diluted shares fell from 962m to 821m over the year, a 14.7% reduction, driven by a $25bn accelerated repurchase that delivered 103m shares upfront. It was funded with $25bn of new debt, which lifted quarterly interest expense from $67m to $473m, and Salesforce told investors in May that full-year operating and free-cash-flow growth would be roughly 4–5%, half the earlier guidance, specifically to absorb that cost. The share count shrinks; the cash flow it concentrates shrinks with it.

Agentforce annual recurring revenue passed $1.5bn, up more than 240% — on a definition widened this quarter to include Slackbot and other AI products, and against a subscription run-rate near $43bn. Two-thirds of the $10.82bn subscription book is still billed per human login, and that seat-billed core grew 8% in constant currency.

The cheapest name grew fastest

HubSpot, which sells a bundled marketing, sales and service platform to mid-market businesses, reported June-quarter revenue of $911.7m, up 19.8%, and swung its GAAP operating margin from minus 3.2% to plus 4.8%. Its shares are down 51% over twelve months. The reason sits in its own release: 7,000 net new customers against guidance of 9,000–10,000, a forward pace cut to 5,000–6,000, net revenue retention of 102%, and third-quarter revenue guided up 14%. Customers reached 306,446, up 14%, while average subscription revenue per customer rose 4% — growth carried almost entirely by logos, and logos are what management cut.

Against a trailing year of gross profit — the comparison that works when HubSpot's GAAP earnings are near zero and its trailing price-to-earnings ratio reads 86x — HubSpot trades at 4.35x, down from roughly 10.9x twelve months ago while its gross profit grew 19%. Salesforce trades at 6.0x, up from 4.47x on 29 July, though a year ago, before the de-rating, it was near 8x. Klaviyo, whose e-commerce messaging business grew 26.4% with retention at 109%, sits at 5.26x; Braze, growing 30.2% but losing money at a minus 13.0% operating margin, at 6.46x. Sprout Social, the slowest at 10.8% growth, is cutting about a fifth of its headcount: "we expect to reduce our overall non-GAAP cost structure by at least $50 million on an annualized go-forward run rate," chief executive Ryan Barretto said on the 6 August call.

What the move earns

Salesforce's re-rating is earned by two things — bookings growth of 14% and a share count 14.7% smaller — and by neither organic revenue nor operating income, which did not grow at all. Its 15.0x forward earnings multiple rests on consensus that embeds $2.53 per share of unrealized gain on the Anthropic stake, with consensus earnings then falling 2.9% the following year. HubSpot's discount, meanwhile, prices a deceleration management has guided to but has not yet printed. Within one category, the name that re-rated hardest is the one whose organic growth halved, and the name left cheapest is the one still compounding near 20%.

Salesforce's sales chief Miguel Milano committed on the August call to organic re-acceleration in the second half. The third-quarter print is where that either arrives unaccompanied by another acquisition, or where a company that bought $456m of growth has to explain why it needed to.

The market shrank; Rush did not

Rush Enterprises, which sells and services heavy trucks through a network of North American dealerships, retailed 3,172 new Class 8 trucks — the heaviest tractor category — in the United States in the June quarter, down 0.2% from a year earlier. The US Class 8 retail market fell 6.7% to 54,718 units over the same stretch on ACT Research's count, which lifted Rush's share of it to 5.8%. Management guided the second half of 2026 to be "considerably stronger" than the first as freight rates improve.

That matters beyond one quarter because heavy trucks are ordered months before they are delivered, so the order book is the forward read on an industrial cycle that has been shrinking for two years. Two listed companies sell into it, and both are describing the same inflection.

"Our backlog is as big as it has been in a couple years, to be honest with you, where we sit right now. And I will tell you, we are basically sold out," W. Marvin Rush, the company's president, chief executive and chairman, told analysts on the July 29 call.

A stock dividend, not a sell-off

Rush's price history says the opposite. The Class A shares closed at $76.55 on August 31 and opened the next morning at $50.99 — precisely the three-for-two ratio of the stock dividend paid that day — before closing at $48.83 on 173,997 shares, ordinary volume. The board declared the split alongside a post-split quarterly dividend of $0.14 a share, taking Class A shares outstanding from roughly 61.1m to roughly 91.7m. What looks like a collapse is arithmetic; the roughly 4% the stock lost beyond the split ratio that session is the whole of the real move.

Penske confirms it, and is priced on something else

Penske Automotive, which runs 320 retail automotive franchises, 23 CarShop used-vehicle centers and 37 commercial-truck dealerships selling Freightliner and Western Star, reported Class 8 orders up 170% year over year in the quarter and up 231% in June alone, against an industry backlog of 186,000 units — about 8.5 months of production. It guided commercial-truck deliveries from 6,000 in the first half to 10,000 in the second. Its stake in Penske Transportation Solutions, the truck-leasing venture, contributed $57m of equity income, up 7%, with fleet utilization above 80% after the fleet was cut from 414,000 units to 380,000.

The truck side is not large enough to carry the company. Quarterly revenue rose 6.0% to $8.51bn while operating income fell 7.6% to $337.6m, and 2025 operating income of $1.28bn was 6.5% below 2024.

In any case the shareholder outcome no longer turns on operations. On July 22, Penske Corporation and Mitsui & Co., which already beneficially own 72.6% of the company, offered $210 a share in cash for the rest, about $13.8bn of equity value; the special committee of independent directors retained Moelis & Company and Paul, Weiss on August 10 and warned that no transaction is certain. The shares have closed between $216 and $220 every session since August 17 and stand at $218.42, 4.0% above the bid. Trailing earnings of 16.09x and forward earnings of 16.06x are effectively the same multiple, which is what a price set by an offer looks like.

Where the gross profit sits

Rush's parts, service and collision revenue of $605m, up 1.5%, produced 64% of total gross profit at an absorption rate of 130.8% — service gross alone covering fixed overhead one and a third times over. On the car side the same annuity is doing heavier lifting against a weaker vehicle line: Lithia Motors' aftersales gross rose 3.1% last quarter while its same-store gross profit fell 2.7%. Nor is inventory financing about to get cheaper for anyone. The Federal Reserve has held its policy rate at 3.50–3.75%, and after the Jackson Hole speech by Chair Kevin Warsh, futures priced roughly a coin flip on a September hike.

So the evidence for a commercial-truck upturn is unusually concrete — share gained in a falling market, orders up at a second dealer, a backlog measured in months of production — while the evidence for a car-retail recovery remains a service department offsetting a deflating vehicle gross. What the truck cycle does not do is set either share price. Penske's is fixed by a bid its own controlling owners can revise or withdraw, and Rush's, once the split is stripped out, has barely moved on the best order commentary it has given in two years.

A sold-out backlog is a promise about the next two quarters, and Rush has now put its name to it. The trucks still have to be built.

York Space Systems told investors on August 13 that it now expects $375-405m of revenue in 2026, against a prior range of $545-595m. The company did not lose a competition. Its chief executive, Dirk Wallinger, attributed the cut to a change in how the Pentagon issues work: awards are increasingly made as indefinite-delivery, indefinite-quantity vehicles — vendor pools that authorize spending but pay nothing until individual task orders are funded — instead of a run of discrete requests for proposals. Roughly 30% of York's planned new business for the year came out, with supply-chain delays pushing a similar amount into 2027.

"Through 2026 and increasingly in Q2, we've observed a meaningful shift in how the U.S. government is acquiring spacecraft systems, moving away from a rapid succession of larger RFPs to an IDIQ approach that is slow to start, but often faster to accelerate task orders later," Wallinger said on the August 13 call. That distinction matters well beyond York, because all three listed spacecraft builders sold off together this summer and only one of them has a forecast that broke.

The one that broke

York, a Colorado prime that builds S-CLASS, LX-CLASS and M-CLASS satellite buses for national-security customers, delivered an excellent quarter operationally: 21 satellites launched to finish its Space Development Agency Tranche 1 Transport Layer commitment at 42-for-42, gross margin up 13 percentage points to 24%, backlog up 9% to $592m. None of it stopped the revenue line from contracting. Adjusted losses before interest, tax, depreciation and amortization widened to $9.5m as selling and research costs rose 52%. The share damage arrived before the guidance cut, around the August 5 announcement that finance chief Kevin Messerle was leaving for a job outside the space industry: the stock fell 12.7% and then 18.5% on consecutive sessions. JPMorgan cut York to Neutral days later and dropped its target to $17 from $38.

The two that did not

Intuitive Machines, the Houston builder of lunar landers and, since acquiring the former Maxar satellite business now called Lanteris, of communications spacecraft, reported June-quarter revenue of $206.2m — more than four times the year-ago figure — with gross profit of $35.9m against a gross loss a year earlier. Contracted backlog reached a record $1.8bn, roughly double the $943m on the books in early May, and full-year guidance of $900m-$1bn was reaffirmed. Its largest booking of the quarter was a $600m order for three geostationary communications satellites built on the Lanteris bus. "This quarter demonstrates that Intuitive Machines is no longer dependent on a single market, customer or mission cadence for growth," finance chief Peter McGrath said on August 13.

The honest caveat is on the margin line and the share count. A $14.7m estimate-at-completion adjustment on the IM-4 lander helped drag gross margin from 39.0% in the March quarter to 17.4% in June, the company burned $60m of operating cash, and diluted shares rose 51% year over year. The at-the-market equity sold to date was priced at a volume-weighted average of $26.81 — well above where the shares now trade.

Redwire, which sells solar arrays, star trackers and deployable structures into space programs and, through its Edge Autonomy unit, uncrewed aircraft and surveillance payloads to defense ministries, posted record revenue of $117.1m, a record 27.8% gross margin and a record $542.1m backlog on a book-to-bill of 1.42. Estimate-at-completion adjustments were net neutral in the quarter, after years in which program charges held full-year 2025 gross margin to 5.2%. "We are reaffirming our full year 2026 revenue forecast in the range of $450 million to $500 million," finance chief Chris Edmunds said on August 6. Defense Tech revenue of $61.9m exceeded Space's $55.2m: the satellite-component company is now, by revenue, a drone company.

What the prices are actually saying

No company-specific bad news is discoverable at either name inside the last month; both peaked on August 17 and drifted. The likelier reading is that a scarcity premium drained out of the group after SpaceX's June 12 listing, the largest initial public offering on record, on whose debut day Redwire and Intuitive Machines each fell at least 10%. Redwire's own catalyst is datable: it established a $500m at-the-market equity program on June 9 and fell 15.2% that day, having already grown its diluted share count 146% in a year while cutting debt 75% to $48.9m.

The three now sit in different places. Intuitive Machines trades at 2.63x forward sales and 9.6x forward gross profit, against 18.7x sales in early May with no forecast cut in between — the clearest gap between what the order book did and what the price did. Redwire is the most expensive of the three on forward sales at 5.53x, and its trailing 6.17x is above the 5.36x it carried in May, so its slide is a retracement from a June peak rather than a de-rating. York, at 0.64x book value, is priced below the accounting value of its net assets by a market that has decided slow task-order conversion is permanent.

The grouping itself is the thing to discard. One of these companies had its revenue shrink; one quadrupled it; one now earns most of its money from drones. Whether York's IDIQ backlog converts in 2027 is a question about federal contracting speed, and it will be answered on a schedule no share price can pull forward.

On April 10, 2026, the last condition under Hawaiian Electric Industries' Maui wildfire settlement fell away — a judgment on insurers' subrogation claims became final after every carrier stipulated to dismiss its appeal — and the Honolulu utility authorized the first of four annual payments of $479m. The money did not come from the bank it sold. It came from stock it had sold eighteen months earlier and parked in a special-purpose vehicle built to hold exactly that payment.

That distinction is the whole story of what Hawaiian Electric's shares now represent. The company generates and distributes power across Oahu, Maui, Hawaii Island, Lanai and Molokai on a grid connected to nothing else, earning a regulated return on a small island rate base. Sitting in front of that return are three more $479m installments, due each April through 2029.

The funding stack, in order

Hawaiian Electric priced 54,054,054 shares at $9.25 in September 2024, raising roughly $575m of new common stock expressly to fund its settlement contribution. Diluted shares have gone from 109.8m in 2022 to 173.2m in the June 2026 quarter, a 58% increase. Consensus has earnings per share at $0.84 this year and $1.09 next — against the $2.20 to $2.25 the company earned in 2021 and 2022, because the same rate-base earnings are now divided among far more shares.

The sale of 90.1% of American Savings Bank brought $405m of cash at the end of 2024 — less than one installment — and went to reducing holding-company debt. It also removed the non-utility earnings leg that used to smooth the utility's results. As of June 30 the discounted settlement liability stood at $1.30bn, while unrestricted cash was $52m at the holding company and $186m at the utility, $238m in all. The $1.3bn of consolidated liquidity management cites is almost entirely undrawn at-the-market equity capacity and revolving credit — the next installment gets funded by issuing something.

The engine that has to refill it

June-quarter core net income was $22.5m, or $0.13 a share, down from $0.20 a year earlier; utility core income fell to $32.6m from $42.5m. Reported profit of $123.2m is not comparable — it includes a $153.9m pre-tax non-cash gain from remeasuring the settlement liability, which is why a trailing price/earnings ratio of 8.5x means nothing here. The utility's trailing-twelve-month earned return on equity is 6.1% against an authorized 9.5%, and 2026 carries $28m of wildfire insurance premiums that lost deferral treatment, flood response, and the August 16-17 damage from Tropical Storm Lala, which cut power to more than 197,000 customers.

The repair is a rate case. "Our total $170 million proposed base rate increase is phased in over 2 years, with $125 million of the increase proposed to take effect beginning in 2027," chief executive Scott Seu told investors on August 7. An interim decision is expected by December 18, with a final order in April 2027 — after which, regulatory affairs head Joe Viola said on the same call, "we're going to rebase rates right now for the next five-year multi-year rate plan."

Bondholders won; the equity waited

Every favorable development of the past year accrued mostly to creditors. Moody's upgraded in April; "in July, S&P upgraded HEI and Hawaiian Electric one notch to double B-minus," chief financial officer Paul Ito said, after the commission approved roughly $350m of wildfire-mitigation cost recovery in June. Hawaii's Act 258 permits securitizing those costs, with a financing order expected late this year. In February the state Supreme Court held that subrogating insurers "do not have a protectable interest that allows them to intervene" in the settlement. The company still owes $1.44bn of undiscounted installments, pays no dividend, and cannot chase the load growth lifting mainland utilities — data-center demand does not reach an island with no interstate transmission.

The shares are down 10.3% over twelve months, the worst in a regulated-electric group in which most names rose, and they barely moved on August 31 when California's wildfire legislation knocked 23% off Edison International. This is company-specific. At 1.09x book value against Pinnacle West's 1.68x, the discount is real but modest; PG&E trades at 0.91x. The market is not pricing insolvency. It is pricing a business whose allowed return is fixed, whose share count already rose, and whose cash goes out every April.

The rate plan being set this winter binds through roughly 2032. The last settlement installment is due in 2029 — meaning the terms decided in December will govern what shareholders earn across every remaining payment, and for three years after the last one clears.