The latest odds (10 October)

Prices read from each market on Saturday 10 October; Kalshi shown as the range between buyers’ and sellers’ prices. A price of 62% means bettors put the chance at about 62%. It is not a forecast.

The House: close to a done deal

• The Cook Political Report, an independent US election rater (9 Oct): 213 seats at least lean Democratic.
• The line: a majority is 218, so the Democrats need only 5 of the 21 races rated too close to call.
• Betting markets: have favoured a Democratic House since the spring.

The Senate: the swing

• Today: Republicans 53 seats; Democrats 47, counting two independents who vote with them.
• Needed: a net gain of four seats.
• The seven close races (Cook, 6 Oct): Alaska, Iowa, Kansas, Maine, Michigan, Ohio and Texas are rated too close to call. Six of these seats are held by Republicans today, so each Democratic win there is a gain. Michigan is held by Democrats, but its senator is retiring, so Democrats must defend it too.
• They tend to fall the same way: when the national mood shifts, these races usually swing together, like coins that tend to land on the same side. The Senate result is less seven separate contests than one bet on the national mood.

One house or two: what each outcome changes

• About 92% of the odds point to stalemate. The two most likely outcomes:
  • Democrats win both houses (about 62%): they can pass bills, but the President can veto them, and they lack the votes to overrule him.
  • Split Congress, a Democratic House and a Republican Senate (about 30%): the House blocks the President’s party’s bills.
• Either way, new laws largely stop from January, and policy comes mainly from the President’s own orders.
• New stimulus only in the least likely case: a second one-party budget law needs Republicans to keep both houses (about 9%).
• Democrats win both, or a split: the difference is the Senate’s power over appointments, not laws.

The calendar that matters more than 3 November

• The outgoing Congress’s last two months (4 November to 2 January). Americans call this the “lame duck”. Republicans keep their 53 Senate seats until 3 January, so the President’s picks can still be approved whatever the result.
• 11 December. Government funding, extended in September, runs out. If Congress misses it, parts of the US government shut down.
• A possible early fix. Some Republicans have reportedly discussed raising the borrowing cap before January. If that happened, the new Congress would lose its main 2027 bargaining chip.
• Year-end deadlines. Chip factories must start construction by 31 December 2026 to qualify for a US investment tax credit. Wind and solar projects must be running by end-2027 to keep theirs.
• The borrowing cap: US$41.1 trillion; see the box below.

Regulation: the Senate decides who runs the regulators

These agencies set the rules for the markets most global portfolios are invested in. Laws already passed stay passed and the agencies keep working; what a change in the Senate alters is who leads them.

• The Fed (the US central bank): sets US interest rates, which feed into US dollar deposit rates, bond prices and borrowing costs worldwide. All seven board seats are filled and no term ends in 2027 (the next is January 2028), so the Senate matters here only if a seat opens.
• SEC (the US stock-market regulator): sets the rules for every share listed in New York, including Chinese companies listed there, and decides which new funds (ETFs) can launch. Two of its five seats filled since 2 October, both Republicans.
• CFTC (the US futures regulator): oversees trading in oil, gold and interest-rate futures, crypto derivatives and betting markets such as Kalshi, whose odds we quote above. One of five seats is filled, the chairman; filling the other four needs the Senate.
• Crypto: a bill to set US rules for crypto trading failed a Senate vote 49–50 on 15 September. A Democratic Senate would likely push it to 2027 or later. What already works is unaffected: the 2025 stablecoin law, trust-bank licences for crypto firms and the SEC’s own rules.
• Banks: capital rules (how big a safety buffer banks must hold) decide how much banks can lend and pay out to shareholders. Most of the easing has come from today’s regulators; the risk is to the next round, not this one.
• AI and data centres: there is no US national AI law; the limits are being set state by state. A change in Congress moves this little.

Sectors: where the outcome matters, and where it is overstated

The trap in the table: “a Democratic win means a clean-energy rally” confuses what Congress wants with what it can pass. Reversing the wind and solar cuts needs a President’s signature. Some investment firms disagree (see below).

Interest rates, the US dollar and the economy

The usual pre-election pattern has not shown up

• The usual path: since 1953 the 10-year yield has fallen by an average of 0.33 points in the six months before a November midterm (Morgan Stanley).
• This year: 4.45% on 4 May, 5.24% on 9 October; most of the rise came in September.
• What is driving it: government deficits, rising interest bills, inflation and a Fed that raised rates in September; not the vote.

Our read on the election and bonds

• A split government: less new borrowing, so fewer new bonds to sell; mildly helpful for long-dated bond prices.
• Republicans keep both houses: the only outcome with a second one-party budget law. For bonds: more tax cuts or spending means more US government borrowing and more bonds to sell, which tends to push long-dated bond prices down and yields up.
• But history gives no rule. Three months after past midterms that produced a split government, the 10-year yield moved:

What the large investment firms say:

• Brown Brothers Harriman, a New York private bank (1 October): rising interest costs and more government bond sales are “poised to lift the term premium”, the extra return investors want for lending for ten years rather than rolling over short-term loans. If yields rise on budget worries rather than growth, a weaker US dollar “could become the relief valve”. In the year after the 2002, 2006 and 2018 midterms the dollar fell 9.3%, 7.6% and 0.8%.
• Morgan Stanley (28 August): a US budget deficit of about 6% of the economy through 2027, and more fights over funding and the borrowing cap under a split Congress; the strain would show first in Treasury bills due near the deadline.
• TIAA, a large US retirement-savings manager (14 September): if Republicans keep both houses, bond prices could swing more should the result be read as a mandate; if Democrats take a house, money could move from US to non-US shares, with some dollar weakness.
• J.P. Morgan Asset Management (8 October): “pressure from the bond market should ultimately act as a constraint” on any fiscal expansion, and a Democratic Senate would reduce the risk to the independence of the Fed and similar institutions.

What history says about US shares

• A weak midterm year is not always about the election: the S&P 500 fell about 6% in 2018 and 19% in 2022, mainly because the Fed was raising interest rates fast.
• This year sits in the weaker group: BlackRock finds shares rose about 10.4% in the six months after a midterm when one party lost its full control of the White House and Congress, against 16.1% otherwise. Both likely results this year end that control. Weaker, but still a rise.

What it means for investors outside the US

• The US dollar
  • The large firms lean towards a softer dollar.
  • Brown Brothers Harriman counted falls of 9.3%, 7.6% and 0.8% in the year after the 2002, 2006 and 2018 midterms.
  • TIAA expects some dollar weakness if Democrats take a house.
  • For you: if your home currency is not the dollar, a weaker dollar trims the value of US dollar deposits, bonds and shares when counted back home.
• US interest rates
  • US rates set the tone for borrowing and deposit rates across much of Asia.
  • A 10-year US yield above 5% keeps longer-dated US dollar bonds paying well.
  • But it also makes borrowing dearer and weighs on share prices.
• US dollar cash
  • Many investors hold US dollar cash in Treasury bills or money-market funds.
  • The risk sits in bills due around the 2027 borrowing-cap deadline, not in bills generally.
  • See the borrowing-cap box above for the history and the dates.
• US shares
  • They make up a large part of most global portfolios, wherever the investor lives.
  • History has been kinder to the year after a midterm than to the run-up (table above).
  • TIAA also sees money moving towards non-US shares if Democrats take a house.

Questions to carry forward, and what each would tell an investor

What would change our mind

• If Republicans keep both houses, a second one-party budget law returns: more stimulus, and more pressure on long-dated bonds.
• If the borrowing cap is raised before January, the main 2027 bargaining chip disappears and a split Congress mostly means hearings.
• If a close Senate result is contested, by recounts or court challenges, the weeks from 3 November to 3 January could bring swings the odds do not show.

What it means for portfolios

The useful question is not who wins, but how much a business depends on Washington. Four groups:

• 1. Waiting on a law that has not passed: the most exposed
  • Who: US health insurers selling individual cover, such as Centene and Oscar Health.
  • Why: their customers lost the premium subsidies on 1 January; only Congress can bring them back.
  • Democrats win both houses: a real chance the subsidies return, a lift for these insurers. Republicans keep both: the subsidies stay gone.
• 2. Ruled by regulators: affected slowly, through the Senate
  • Who: crypto and trading platforms such as Coinbase and Robinhood, and the big US banks.
  • Why: their rules come from the SEC, the CFTC and the bank regulators, not from Congress directly.
  • If the Senate changes hands: the President’s picks for those regulators are slowed or blocked, so new rule changes come more slowly through 2027. Today’s rules stay either way.
• 3. Protected by a law already passed: largely safe until 2029
  • Who: wind and solar developers such as NextEra Energy and First Solar, whose tax credits were cut in 2025; drug makers such as Pfizer and Eli Lilly, whose price deals are signed.
  • Why: changing these needs the President’s signature, and he would veto it.
  • Until when: at least January 2029, when President Trump’s term ends.
• 4. Sensitive to interest rates: the 10-year yield matters more than the vote
  • Who: homebuilders, property trusts (REITs), utilities, smaller companies that borrow heavily, and holders of long-dated bonds.
  • Why: their value moves with the cost of borrowing, set by the 10-year US yield (now above 5%) and the 2027 borrowing-cap fight.
  • The election matters here only if it changes those two.

For most portfolios this is background to keep in view, not a reason to trade: a calendar of dated tests, not a reason to change course on 3 November.