Trump's Economy Approval Just Fell to a Net -31 — Which Puts the Record Refining Spread in the Crosshairs Before Nov. 3

Donald Trump's economic approval has fallen to a net -31, the lowest reading of either of his terms. Thirty-four percent approve of his handling of the economy and 65% disapprove, down from 38% and 60% in the CNBC All-America Economic Survey fielded Oct. 14-18 among 1,000 registered voters with a margin of error of plus or minus 3.1 points. Overall job approval sits at 40% against 59% — poor, but less poor than the economy, and that gap is the whole story. Voters are not rejecting the president; they are rejecting the price level.

Two readings do the work. First, 80% of Americans now rate the economy as fair or poor, the highest share since December 2023, and 63% are pessimistic about the economy and its outlook. Second, the pessimism has crossed into the president's own coalition: 35% of Republicans are pessimistic, up six points since July and the highest of either term. The driver is not abstract. 57% name the cost of food, gas and housing as the most important issue facing the country, ahead of government corruption at 36%, and 80% say prices are rising with 55% saying they are rising faster than usual. That is a price report, and it is why the same poll shows Democrats ahead 52% to 45% on control of Congress.

One more cut of the same data matters for positioning. Among the eight states with competitive Senate races, respondents prefer Democratic congressional control by four points — the Republicans' vulnerability is not confined to the House. And on Oct. 23, the day the survey was published, the Dow fell 652.82 points (1.23%) to 52,215.16 with the S&P 500 down 0.52% at 7,623.68, a session the market attributed to climbing oil prices. Markets traded the crude price and voters traded the pump price; both were looking at the same number.

September's CPI report is the mechanism behind the -31. Headline inflation rose 0.6% on the month and 3.7% on the year, with core at 2.6% after touching a five-year low of 2.4% in August. Energy prices are up 20.4% year over year and gasoline is up 35.3%. The pump itself sits at $4.29 a gallon nationally as of Oct. 23 — the first October in the series above $4, and 158 days above $4 in 2026, a record. Brent settled at $116.72 a barrel on Oct. 23, up 6.2% on the week and the highest since April, on the Iran war.

That leaves the administration with three other levers — fiscal, regulatory and legal — and each of them moves money between sectors rather than creating it. Oil would have been the only one that lowers the number households see at the pump. Everything else is a transfer: refunds already paid, fees about to be waived, or a margin about to be capped. With the record-low rating, the market's question changes from what will oil do to who pays for the relief.

The spending data does not show a collapse. Nominal retail and food services sales rose 0.8% in September to $745.0 billion after a revised 1.3% in August, and third-quarter consumer spending is tracking near a 3.4% annualized pace. Strip the gasoline station out and September's gain falls to 0.2%. That 60-basis-point wedge is the entire consumer story: the same nominal dollars are buying more fuel and less of everything else, while the household reports that the economy is bad.

One more detail explains why the sentiment has not yet broken the spending: the gap is being financed. The average rate on credit card accounts with assessed interest reached 22.36% in August, the highest in a year, and revolving balances sit at record levels. The consumer's purchase power is now a function of the rate on revolving credit, which the Fed just raised — that is the transmission channel from a record-low approval rating to the next two quarters of discretionary earnings.

Rates are the binding constraint, and they moved the wrong way. On Sept. 16 the Federal Reserve raised its federal funds rate 25 basis points to a 3.75%-4.00% range on a 12-0 vote, its first increase since 2023, with the statement conceding that \"inflation remains elevated\" while calling activity solid. Rising oil has since strengthened rate-hike expectations rather than easing them. With the 30-year mortgage well above 7%, the NAHB housing market index fell to 29 in October, the lowest since June 2012, and 43% of builders cut prices — the highest share of the 2020-2026 period — for a seventh straight month at an average 6% cut.

The damage is already visible in the numbers. D.R. Horton generated $9.23 billion of revenue and $3.19 of diluted earnings in its fiscal third quarter ended June 30, with inventory up from $26.0 billion three months earlier — builders are building speculatively into a market where nobody is taking delivery. Price cuts of 6% on a 43% of the market are the release valve, and they come out of builder margin rather than out of the mortgage rate.

Every other concession on the table is administrative. The one that would actually move pump prices is crude, and the White House has ruled it out until after the vote. Brent's climb to $116.72 on Oct. 23 is what turns $4.29 at the pump into a political liability, and it is happening under a war the administration chose. That is the trap: the same spread that fills the pump price is the spread that funds the profit Washington now needs to argue about.

The money is real and it is enormous. Valero Energy reported $8.8 billion of refining operating income in the third quarter against $1.6 billion a year earlier, on a $39.24-a-barrel refining margin, and earned $6.80 billion, or $23.90 adjusted a share, with a $1.20 dividend and a 52% payout of operating cash flow. That is the profit pool that converts $116 crude into $4.29 at the pump. The market has already made a judgment on it: Valero trades at roughly seven times trailing EBITDA on a record quarter, and its market value has climbed from about $77 billion at the end of June to roughly $121 billion. The equity is priced for the spread to be taken away rather than for it to persist.

The same squeeze runs through Exxon Mobil, which told the market that stronger refining margins would add about $4.5 billion to third-quarter earnings against roughly $1.3 billion of upstream earnings lost to lower crude. Major producers are paid on the spread between what they sell and what they refine, and that is the same spread Washington now wants to compress. Political pressure on pump prices lands first on the refining margin, because that is where the retail price is manufactured — while the only lever the president has actually named, cheaper crude, lands on the barrel and therefore on upstream earnings.

The second relief channel is the one already spent. On Feb. 20, 2026 the Supreme Court held that the emergency-powers statute does not authorize the president to impose tariffs; the Court of International Trade ordered Customs and Border Protection to refund the duties collected under it, and CBP began accepting refund applications in the summer. The average effective tariff rate fell to 7.1% by August 2026. Roughly $166 billion of duties is slated for refund, and replacement Section 301 and Section 338 levies of 10% to 12.5% have recovered less than 60% of the revenue the ruling surrendered.

For the equity read, the important fact is that this relief is banked rather than pending. It is already inside reported gross margins, and it does not repeat unless Congress makes it repeat. The vision of the election cycle is that a Democratic majority would pursue legislation constraining executive tariff power and seeking tariff refunds for consumers and small businesses. Any further tariff relief after Nov. 3 is therefore a 2027 story, and the pre-election relief has already been spent — which narrows the November price-relief menu to energy and housing, the two channels that do not reach a household's pump price.

The remaining tariff file is autos, and it is the one place where an election can make things worse rather than better. The 25% Section 232 duties on imported vehicles and parts survived the court ruling because they are not IEEPA measures and are not refundable, and US automakers had absorbed about $35.4 billion of tariff costs by Oct. 6. The unresolved front is Canada: the threat to raise tariffs on Canadian autos, parts and steel to 50% on Jan. 1, 2027 remains formally in effect, and Reuters reports Toyota and Honda would be most exposed because the two account for the large majority of what Canada builds.

This is the cleanest listed expression of the trade-conflict file, and it is a reminder of what an approval-driven policy shift looks like from outside the United States. A 50% Canadian auto tariff would land on the one region where Toyota's US profit has already been written down, into a US consumer that is rating the economy fair or poor — and none of that is fixed by a mortgage-insurance pilot.

Put the chain together: a record-low economic approval rating is the arithmetic of gasoline at $4.29, headline CPI at 3.7% and a Federal Reserve that just hiked into it. The president has ruled out the one lever that lowers the pump price before the vote, so November's relief must be paid for by someone else — and the two candidates are the refining spread and the tariff refund, one of which has already been spent. The trade is therefore not a bet on the consumer, it is a bet on which balance sheet absorbs the cost of the election.