Newsroom
Policy
Regulation, taxation, and rules that reshape industries.
The Senate's reported grant-rule pause still has no defensible dollar value
Senate appropriators said a funding agreement through Dec. 11 would keep OMB's proposed grant rule from taking effect, but the measure had not passed and Cumulant did not locate official Senate text before publication. Public award records can measure commitments that agencies made, not applications or existing awards that the proposed rule would have changed, so a protected-dollar total would be false precision.

The EU can now fine general-purpose AI providers. We put the one-year odds of a first decision at 35%
The European Commission's power to impose fines under Article 101 began applying on 2 August 2026, although the detailed procedural regulation takes effect on 10 August. After tracing the required legal steps and correcting the closest DSA precedent to 833 days, we estimate a 35% chance of a first fine decision by 2 August 2027. [AI Office FAQ](https://ai-act-service-desk.ec.europa.eu/en/faq?faq_category_id=69) [Implementing Regulation 2026/1755](https://eur-lex.europa.eu/eli/reg_impl/2026/1755/oj/eng)

Removing ulcer-coded claims still leaves about $4.11 more Medicare spending per hospice day
Medicare payments for services outside hospice increased by $1.536 billion from FY2020 to FY2024. After mechanically removing the approximately $696 million increase in pressure-ulcer-coded carrier claims, CMS data indicate that the remaining spending rose by about $4.11 per hospice day, although the exact figure is unavailable because CMS has not published an unrounded FY2020 denominator matched to the final-rule data.

A higher household inflation estimate made up 90% of a constructed Russian rate-gap fall
The Bank of Russia cut its key rate by 0.25 percentage point to 14%, effective 27 July. In a month-end subtraction of the key rate minus households' one-year inflation estimate, the estimate's 2.3-point rise mechanically produced about 90% of the 2.55-point decline, but the calculation is neither a loan rate nor evidence of causation. [Bank of Russia](https://www.cbr.ru/eng/press/pr/?file=24072026_133000key_e.htm) [InFOM](https://cbr.ru/Collection/Collection/File/62217/inFOM_26-07.pdf)

For continuously taxed imports, 80% to 100% of Friday's new duty matches the old surcharge
At 12:01 a.m. Eastern on July 24, the temporary 10% Section 122 surcharge reached its stated endpoint as new Section 301 duties of up to 12.5% took effect. For an otherwise identical import taxed under both policies, 80% to 100% of the new duty numerically matches the old charge, but the available public data cannot support that range for the entire import basket.

Transit Receipts Would Cover 21% of the House Bill's Five-Year Highway Funding Gap
Transportation Secretary Sean Duffy urged Congress to eliminate the Mass Transit Account and direct its fuel-tax revenue to highways. Applying that proposal to CBO's estimates for H.R. 8870 would cover $28.941 billion of a $136.078 billion highway operating gap over fiscal 2027-2031, leaving $107.137 billion unresolved.

Euro-area business-loan tightening showed up mainly in price, not access
Among bank-finance applicants, a net 42% reported higher loan rates, while among firms for which bank loans were relevant, only a net 1% reported worse availability and 5% encountered a financing obstacle. The contrast suggests that the second-quarter squeeze operated mainly through price, although smaller firms and long-term borrowers faced more uneven access. [ECB SAFE](https://www.ecb.europa.eu/stats/ecb_surveys/safe/html/ecb.safe202607.en.html) [ECB Bank Lending Survey](https://www.ecb.europa.eu/stats/ecb_surveys/bank_lending_survey/html/ecb.blssurvey2026q2~baa6b60429.en.html)

Past tariffs point to U.S. importers, but Canada's new 50% tariff burden is not yet observable
President Donald Trump signed three proclamations adding a 50% duty to specified Canadian goods from August 19, measures that USTR says cover nearly $20 billion in imports. Studies of earlier tariffs lean strongly toward U.S. importers bearing most of the border-price increase, but they cannot establish how Canadian suppliers will respond to this different tariff. [White House](https://www.whitehouse.gov/fact-sheets/2026/07/fact-sheet-president-donald-j-trump-imposes-additional-tariffs-on-canada/) [USTR](https://ustr.gov/about/policy-offices/press-office/press-releases/2026/july/ambassador-greer-issues-statement-president-trump-imposing-section-338-tariffs-canada)

The Pentagon's 2027 materials deadline lacks the data needed to name the bottleneck
President Donald Trump's July 20 order tightens conditions for critical-material waivers on January 1, 2027, when an existing sourcing restriction also expands upstream to mining, refining and separation. The public record documents weak supplier visibility and past production stoppages, but it does not separate the time spent tracing origin, finding a compliant source, qualifying that source and obtaining a government decision, so no stage can yet be identified as the dominant delay. [White House](https://www.whitehouse.gov/presidential-actions/2026/07/securing-americas-defense-supply-chains-and-ensuring-domestic-acquisition-of-critical-materials/) [Acquisition.gov](https://www.acquisition.gov/dfars/225.7018-2-restriction.) [GAO](https://files.gao.gov/reports/GAO-25-107283/index.html)

Canada widened its Ebola entry rule without quantifying the added risk reduction
Canada will prohibit foreign nationals who were in the Democratic Republic of the Congo during the previous 21 days from entering beginning at 11:59 p.m. EDT on July 20. The available model covers measures already operating before the prohibition, while public records do not disclose how many additional travelers the new rule excludes or how much importation risk it removes.

The EU's €230 Billion Banking Figure Measures Liquidity Rules, Not New Lending
The European Commission presented a banking-reform direction on 17 July, highlighting about €230 billion of liquid assets whose transfer within cross-border banking groups is constrained. The underlying ECB calculation is a regulatory-liquidity stock, not a forecast of additional business loans, so the reform's economic effect remains unquantified.

ASC 842 Study Finds More Reported Non-Lease Leverage, Not Proof That Total Debt Fell
A new Federal Reserve working paper estimates that companies strongly affected by lease recognition reported a 3.0 percentage-point increase in non-lease debt relative to assets, while indicators of borrowing between reporting dates declined. The pattern is consistent with debt becoming more visible, but the study does not observe the proposed short-lived loans or establish that companies reduced their total economic debt. [Federal Reserve paper](https://www.federalreserve.gov/econres/ifdp/files/ifdp1441.pdf)

EU opens 11 Android features to AI rivals. Reliability remains unsolved
On 16 July 2026, the European Commission ordered Google to give rival AI services equally effective access to 11 Android features, with most changes due in Android 18 by 1 August 2027. The order removes a system-access disadvantage, but a controlled mobile benchmark in which three frontier models completed 47.8% to 55.6% of GUI-only tasks shows why equal access cannot guarantee equal performance. https://digital-markets-act.ec.europa.eu/developer-portal/interoperability/alphabet-specification-proceedings-interoperability-ai-services_en; https://ec.europa.eu/competition/digital_markets_act/cases/202629/DMA_100220_2683.pdf; https://aclanthology.org/2026.acl-long.278.pdf

The IEA's 9% mining-investment decline does not measure rare-earth factory finance
On [16 July 2026](https://www.iea.org/news/supply-concentration-export-restrictions-and-declining-investment-put-critical-mineral-security-at-risk), the IEA reported that investment by a sample of major mining companies fell in 2025. Our review finds that the figure cannot show whether financing for rare-earth refineries and magnet factories outside China rose or fell because those projects, financing instruments and disbursements were not measured separately.

At 4 p.m., Brazil's final tariff list was missing. The concentration share was unknowable.
USTR had identified July 15 as its statutory deadline for responsive action, and Reuters reported that a 25% tariff announcement was expected that day. At 4 p.m. Eastern Time, no final action or product list was visible on the official pages checked, leaving the requested supply-concentration share impossible to calculate honestly. [USTR](https://ustr.gov/about/policy-offices/press-office/press-releases/2026/june/ustr-section-301-determination-brazils-unreasonable-acts-policies-and-practices), [Reuters](https://www.reuters.com/world/americas/brazil-braces-new-us-tariffs-washington-broadens-trade-push-sources-say-2026-07-15/), [USTR case page](https://ustr.gov/trade-topics/enforcement/section-301-investigations/section-301-brazils-acts-policies-and-practices-related-digital-trade-and-electronic-payment), [Federal Register](https://www.federalregister.gov/public-inspection/2026/07/15)

India's Russian oil arrivals fell 19.8%. The tariff's role cannot be isolated
Russian seaborne crude arrivals averaged 1.466 million barrels a day in the five full calendar months of the U.S. tariff, 19.8% below the preceding five months. The drop is real, but supplier sanctions, an EU fuel rule and shipping timing overlap it, so the comparison does not identify how much the tariff caused.

The public record does not show how the FTC reached its $13 billion Caremark savings ceiling
The FTC says its proposed Caremark settlement could preserve or produce up to $13 billion in patient savings over ten years. The public can reproduce one multiplication, but not the baselines, enrollment, adoption or cost-offset assumptions needed to estimate how much benefit the order would actually cause.

Rotterdam's investor rule shifted buyers but did not show cheaper homes
With H.R. 6644 expected to become law on July 11 without President Donald Trump's signature if the July 10 veto deadline passed without action, its investor-purchase restriction had not yet taken effect. Rotterdam's first-year evidence shows a clear shift from investor buyers to owner-occupiers, but its adjusted price estimate ranged from a small decline to an increase, a warning that access and affordability are different outcomes. [AP](https://apnews.com/article/trump-housing-bill-77ec340dcdd676c46c458813b461b1af) [GovInfo status](https://www.govinfo.gov/app/details/BILLS-119hr6644enr) [Rotterdam working paper](https://www.eur.nl/en/media/2025-07-paper-m-korevaar-buy-live-buy-let-1107251)

Insurers expect a weaker ACA risk pool in 2027, but their filings do not isolate its premium effect
Insurers in an early multistate sample requested a median premium increase of 14%, while disclosed medical-cost assumptions had a median of 10%. Some insurers separately attributed roughly 4 percentage points to an expected deterioration in the risk pool, but the published data cover different, partly unspecified groups and cannot divide the typical request into those components.

The $425,000 Newcomer: IgA Nephropathy's Sixth Drug Priced Near the Top, and the Proof Isn't What Bought the Premium
On July 7, 2026 the FDA cleared Vera Therapeutics' atacicept as the sixth disease-modifying drug for IgA nephropathy, at a $425,000 list price second only to the class's most expensive drug. Ranking all six shows the ladder tracks how new and how novel-in-mechanism a drug is more than the strength of its proven kidney benefit: the two full-approval drugs are the two cheapest, and the newer biologics list far above an independent value benchmark, while the two full-approval drugs sit closest to it. The catch that keeps this honest: the single priciest drug also has confirmed benefit, so proof plainly is not what commands the premium either way.

USMCA's 'No' Started a Clock. It Doesn't Reach the Auto Line Until 2028, and It Already Binds the Battery Plants
On 1 July 2026 the United States declined to reconfirm USMCA's full 16-year term, turning a one-time review into a yearly one and leaving a worst-case 'guaranteed runway' of 10 years that shrinks by one at each future review toward a hard 2036 sunset. Our finding: that runway still clears a typical 8-year auto-investment payback for plants approved through about 2028, so this year's paused factories are a tariff story, not a calendar one, but for the longest-lived capital the same treaty protects, battery and chip plants with 12-to-20-year paybacks, the clock already bites now.

The Sanction Whose Formula Pointed the Ceiling Up After the Market Came Back Down
On its first live six-month recalculation, the EU's dynamic cap on Russian crude pointed to a ceiling near $64 a barrel, roughly 45 percent above the $44.10 frozen in place, even though spot Urals had already fallen back to the mid-$50s. Brussels froze the formula by hand. The freeze is not blocking a rise the market justifies; it is blocking one only the rule's 22-week rear-view mirror still sees, and that view clears once the spring war spike ages out of the window in early 2027.

The Fed's 3.8% Median Is a Coin Toss, Not a Dot
On 8 July the Fed released the minutes of its June meeting, the first on-record account of a committee whose June dot plot carried only 18 dots because new chair Kevin Warsh withheld his own projection. Recompute the median by hand from the primary table and 3.8% turns out to be no policymaker's view but the empty midpoint of a dead-even 9-to-9 hike question, so reading it as a committee that leans toward hiking mistakes a tie for a majority.

The Antitrust Division That Stopped Going to Court
A deal-friendly Justice Department was supposed to be swapping tough divestitures for soft promises. Coding every formal merger settlement from its own court filings, the opposite is true: five structural divestitures, one mixed case, zero behavioral-only decrees. What thinned out was not the divestiture. It was the courtroom.

LG Energy Solution Booked a Profit; a US Subsidy Booked It for Them
The world's number two battery maker reported a return to operating profit on July 6, powered by its pivot from electric-vehicle cells to storage batteries for AI data centers. Strip out one line of US taxpayer money and the quarter is a 128 billion won operating loss, from a business that was at breakeven a year earlier, which means the reroute is not yet paying its own way.

The 9-0 Loss That Came Back Deeper: CMS Now Proposes Paying 340B Drugs 37% Less, and the Survey Is Why
On 2 July 2026, CMS proposed paying 340B hospitals average sales price minus 33.4 percent for outpatient drugs, a deeper cut than the ASP minus 22.5 percent the Supreme Court struck 9-0 in 2022. The analysis finds CMS could cut deeper precisely because it ran the acquisition-cost survey the Court had demanded, and that budget neutrality recycles the entire cut back into other payments, so the real story is redistribution across hospitals and patients, not net savings for anyone.

The Fed Built a Bank-Killer That Can't Kill Banks: Why the 'Skinny' Payment Account's Zero-Interest Rule and $1B Cap Are Brakes, Not Bugs
On 20 May 2026 the Fed voted 6-1 to propose a stripped-down 'Payment Account' that would give stablecoin issuers direct access to its plumbing. Crypto voices called it the end of commercial banking; our arithmetic finds the same proposal's two other rules, no interest and a $1 billion cap, make large-scale deposit flight mathematically impossible. The real erosion is smaller, and runs through a different pipe.

The 100% Drug Tariff Where the 100% Hits Almost Nothing
On July 31 the US switches on a Section 232 pharmaceutical tariff with a 100% headline rate on patented medicines. Rebuild the $213.8B import base lane by lane using the proclamation's own carve-outs, and the full 100% reaches only a sliver of the value, while the average tariff actually collected lands around 8 percent. The headline rate is a lever to extract drug-pricing concessions, not a levy built to collect, though the tariff overall still bills importers billions, and a scheduled ratchet points upward.

The Treaty Didn't Expire. Its Ceiling Did.
When Washington declined to renew USMCA on July 1, not one tariff line changed and markets barely moved. But the pact's guaranteed lifespan stopped being something a single renewal could stretch toward 2042 and became a roughly 10-year countdown to July 1, 2036, a horizon that now shrinks each year unless all three governments affirmatively agree to extend it. Any cost would hide not in prices but in the long-lived investments that quietly don't get made.

The $650 Million Meter That Only Runs If the Deal Survives
UK Culture Secretary Lisa Nandy is days from triggering a media-plurality probe into Paramount's ~$110B takeover of Warner Bros. Discovery. Paramount pre-installed a ticking fee, $0.25 per WBD share for every quarter the deal stays open past 30 September 2026, about $650M a step, but the analysis finds signing the intervention notice does not commit that cash: the fee bites only on a slow-but-successful close, a blocked deal triggers a larger ~$5.8B reverse termination fee instead, and whether the meter ever arms turns on whether the UK, alone among the open regulators, drags the deal past 30 September.

The 10% Wall That Sunsets Itself
A universal 10% import surcharge dies by operation of law at 12:01 a.m. on July 24, with no vote and no signature. Yale's tariff data shows the average rate Americans actually pay steps down only about 2 points, from 11.8% to 9.7%, because permanent statutes were quietly rebuilt underneath it. Whether July 24 delivers a real, temporary dip or almost nothing turns on a sequencing detail Yale has not pinned down.

The grid added a record 58 gigawatts. The risk NERC just raised to its top alert tier can't be measured in gigawatts at all.
A record year for building power plants and a rare top-tier reliability alert landed two weeks apart, and the instinct is to net them out. Our analysis finds the two answer different questions, the new supply settles whether there is enough power, while the alert is about what happens in the split seconds after a fault, so they are orthogonal (at right angles: they do not trade off), and reading one as a cushion against the other turns a headline into a blind spot.

A 'Forced-Labor' Tariff Whose Discounts Went to the Countries With More Forced Labor
On June 2, 2026, USTR proposed a forced-labor tariff on 60 economies, splitting them into a 10% and a 12.5% tier. Rank the major trading partners by how much forced labor they actually have and the rate does not follow: Pakistan, the highest-prevalence economy on the list, got the 10% discount, while India, the second-highest, pays 12.5%, and China is taxed at the same 12.5% as low-prevalence Norway. What the rate tracks is USTR's own stated test, whether a country has a forced-labor import-ban law or signed a reciprocal-trade deal, not the amount of forced labor. It reads like the struck-down IEEPA tariff wall being rebuilt on court-tested authority under a human-rights label.

The $278 Default Option
On July 1, servicers began sending 90-day exit notices to roughly 7.5 million borrowers still parked in the defunct SAVE plan, and the option they get by doing nothing is the income-blind 10-year Standard plan. For a representative low-income borrower, plan-formula arithmetic puts that default at about $278 a month, roughly 5.5 times the $50 bill a single application for the new RAP plan would set.

The stablecoin yield ban didn't move the interest. It locked in who already keeps it.
Congress barred stablecoin issuers from paying holders any of the roughly $3.80 that every $100 of reserves now earns each year. This week's 140-company OUSD launch and the stalled CLARITY Act show the money flowing where it already flowed, up the chain, to the distributors, with the one law that could have changed that now forbidding it. The holder's legal share from the issuer is fixed at zero.

The $6 dividend: run Alaska's own payout rules on Altman's 5% pitch and the check nearly vanishes
Sam Altman has discussed handing Washington a 5% stake in OpenAI, and in every leading US AI lab, inside a fund modeled on the Alaska Permanent Fund. Run Alaska's actual payout rules and the OpenAI stake pays about $6 per American per year, and the fund would struggle to cash even that: nearly a quarter of it would be hard-to-sell private stakes, including a company projecting $115 billion of cash burn through 2029, and in four prior government-stake deals, markets treated the equity as a win for shareholders rather than the public.

June's Jobless Rate Fell to 4.2% While Employment Dropped 507,000: The Improvement Is an Exit, Not a Hire
On 2 July the U.S. unemployment rate fell to a one-year low of 4.2%, yet the same household survey showed employment down about 507,000 and the labor force down about 720,000. The arithmetic is airtight: the rate fell because the count shrank, not because the jobless found work. Who left and why is far less settled, and that gap matters because the Federal Reserve and markets mostly read the headline.

Trump's Housing Bill Bans the Wall Street Landlords Who Were Already Leaving
The 21st Century ROAD to Housing Act becomes law in early July, and its marquee 'Wall Street landlord' ban restricts only for-profit investors controlling 350 or more single-family homes, roughly 1 percent of purchases, and a cohort that has been a net seller for six straight quarters. The provision aims at a target already walking out the door. The one clause that independent analysts said would genuinely cut new housing supply, a seven-year forced-sale rule on build-to-rent homes, was stripped out by the House before passage.

FERC Gave the Grid 60 Days to Stop the Data-Center Power Subsidy. Three Years of It Already Cleared.
On June 18, 2026, federal regulators ordered six grid operators to rewrite how data centers and other large new users connect to the grid and how their costs are shared, so those costs stop landing on everyone else's bills. Tracing each dollar back to the auction that set it shows the order is forward-only: it cannot touch the data-center costs already locked into the 2025/26, 2026/27, and 2027/28 bills, and the earliest delivery year a new rule could plausibly reshape is 2029/30.

The ACA Premium Spike Is Mostly Sticky Care Cost. Only a Thin Policy Slice Can Round-Trip.
Insurers requested their largest average increase in more than five years for 2026 coverage, an estimated 26%, and early 2027 filings point to a second straight double-digit year. A decomposition of the typical request shows only about 4 percentage points are the kind of reversible, subsidy-driven markup that a credit extension could refile away before the 12 August window closes. The rest is locked-in medical cost growth, and the healthy enrollees the markup was hedging against have largely already left.

The 17-Day Tariff Cliff: Why the Replacement May Not Be in Force When the Old Tariff Dies
When the 10% Section 122 global tariff automatically expires at 12:01 a.m. on 24 July 2026, the forced-labor Section 301 tariffs meant to replace it would have to clear their entire post-hearing stage in about 17 days, roughly one-third of the 50 days that stage took in the 2018 China action. Our finding: a legal-instrument gap is likely, but what importers actually pay barely moves, because older tariff layers hold the effective rate near 8 percent, far above the 2.3 percent paid before this escalation.

The Credit System Is Racing to See Buy-Now-Pay-Later. It Can Only See the Healthiest Slice.
Across 2025 and 2026, FICO, HUD and the US Senate have all moved to wire buy-now-pay-later into credit scores and mortgage math, each assuming the credit file can already see it. Our analysis finds the file sees only a minority of US pay-in-4, and the slice it sees is the most heavily underwritten and lowest-loss part (Affirm's), while the grocery-financing, loan-stacking, increasingly-late slice that is growing fastest stays dark. So the file is built to register household stress late and faintly for the people least able to absorb it.

The SAVE borrowers who open the mail get the lowest bill. The ones who don't get the cliff.
Starting around July 1, 2026, the roughly 7.5 million borrowers parked in the SAVE plan's forbearance begin getting notices to pick a new student-loan plan within at least 90 days, or be swept onto a higher-payment plan automatically. Built from the statute, the math inverts the popular framing: among the plans this group can actually choose, the new Repayment Assistance Plan carries the lowest monthly payment, and the costly outcome is the one that requires no decision at all. The catch is that RAP's low payment buys cash-flow relief at the price of the longest payoff.

Who really pays a digital services tax, and who would pay the tariff against it
On 26 June 2026 President Trump threatened a 100% tariff on any country that taxes US tech firms' digital revenue. The best available incidence evidence, drawn from millions of Amazon marketplace prices, shows local sellers and shoppers bear most of that tax rather than US shareholders. If the tariff is ever collected instead of merely threatened, it would fall on American importers, meaning the policy could tax Americans to answer a tax foreigners largely levy on themselves.

The May trade blowout: one number, two stories, and only part of it round-trips
The U.S. goods deficit hit a 14-month high of $105.8 billion in May 2026. The half everyone reported, surging imports, is a datable tariff front-run that should reverse by autumn. The half nobody led with, falling exports, is murkier: much of it is the same oil-price illusion that inflated the import side and will also reverse, but consumer-goods exports falling 9.2 percent is a cleaner demand signal that may not, and that residual is what the Fed has to weigh.

MiCA's deadline was sold as a doorway to all of Europe. The arithmetic says it is a turnstile.
On 1 July 2026 the EU's MiCA transitional period ends with no extension, and Binance has confirmed it will wind down most EU services after failing to secure a license in time. Cumulant finds that only about one in six already-registered crypto firms converted to the new EU-wide license, a survival rate that looks as harsh as the UK's tough FCA regime. But the resemblance is largely an arithmetic accident: French data suggest the thinning came mostly from firms that never applied, and among firms that did apply, MiCA's pass rate looks higher, not lower, than the FCA's.

The Tariff Refund Where Finality, Not the Merits, Decides Who Gets Paid
The Supreme Court voided about $166 billion of IEEPA tariffs for every importer, and U.S. Customs is paying most of it back through a streamlined claims system that does not require a lawsuit. The analysis finds the real fight is over a smaller, finally-liquidated slice, and that whether it is recoverable turns on liquidation timing and who filed suit, not on the law, which is now the same for everyone.

The Loudest Stablecoin Deadline Does Not Touch USDT
Six US agencies are racing to a July 18, 2026 rulemaking deadline under the GENIUS Act, but a close read of the statute shows that date governs new American issuers, not whether USDT can keep trading. The legal cliff for the roughly $186B offshore giant is July 18, 2028; the one ruling that could spare it runs on a 210-day clock that cannot start until Treasury builds a filing process it has not finished; and the deadline that bites first may be set by exchanges, not regulators.

Data Centers Drove PJM's Record Power Price. The Cost of Their Growth Is Spread Across Everyone's Peak, Not Billed to the Newcomers.
On 1 June 2026 PJM's record $329.17/MW-day capacity price began landing on bills in 13 states, and on 18 June the federal regulator ordered all six US grid operators to justify or rewrite their large-load rules by 17 August. Reading the rate mechanics shows the catch is not that data centers dodge the bill, running flat around the clock, they pay a large absolute share, but that the extra cost their growth created is socialized across all customers' peak demand rather than charged to the loads that triggered it, and a federal fix would still have to clear both a wholesale rule change and 13 state retail dockets to change who pays.

Bayer's Roundup win clearly erases about $0.8 billion in verdicts. The market added $7.5 billion.
On 25 June 2026 the Supreme Court ruled that federal pesticide law blocks one type of Roundup cancer suit, and Bayer had its best stock day in 23 years. The ruling clearly removes only about $0.8 billion in verdict exposure, yet the market added roughly $7.5 billion, nearly ten times as much, while Bayer kept pushing its live $7.25 billion settlement and its litigation reserve actually rose. The gap is less a verdict than a bet on questions the Court did not decide.

RAP is sold as the 'affordable' student-loan plan. The arithmetic says it is only cheaper in the middle, and never cleanly.
On July 1 about 7.5 million borrowers begin a 90-day clock off the court-voided SAVE plan and toward the new Repayment Assistance Plan, marketed as the simple, affordable option. We rebuilt RAP's monthly payment and set it against the New IBR plan it sits beside: for a single borrower with no dependents RAP is cheaper across a broad band from roughly $29,300 to $80,000 of income, but because RAP re-rates your whole income at every $10,000 step, the advantage arrives as a sawtooth, not a smooth discount. Which side of those edges a household lands on decides whether 'affordable' is true for them, and even inside the band the savings flicker.

The Tenth: Why France Can Hold the Digital Tax Canada Gave Up
On 26 June 2026 President Trump threatened a 100% tariff on goods from any country that keeps a digital services tax, a threat strikingly like the one that pushed Canada to abandon its own tax within 48 hours a year earlier. Our analysis finds France can hold firm for three reinforcing reasons, not one: its goods exports to the US are about a tenth of Canada's as a share of its economy, it negotiates behind the 27-member EU rather than alone, and a February 2026 Supreme Court ruling stripped most of the legal force from the tariff weapon itself. Because all three differ at once, the Canada-France contrast is suggestive, not a controlled experiment, but it points to a hard pattern in trade brinkmanship: leverage flows to whoever can most afford to walk away.

