# Odd Lots: Texas Is Still Texas — Only the People Issuing the Permits Changed

Source: Realpha Blog (blog.getrealpha.com)
Original article and charts: https://blog.getrealpha.com/en/blog/oddlots-2026-11-06-the-macro-implications-of-the-blue-tsunami/

> Notes on the Odd Lots episode published 6 November 2026 and recorded on 5 November, where Joe Weisenthal and Tracy Alloway talk with Skanda Amarnath of Employ America about the macro implications of Democrats taking both the House and the Senate: Bank of America's Michael Hartnett had said the S&P 500 could fall 10% on a Democratic win, the index rose about 1.5% the day after the count, and on the recording day Tesla was down 10% and SpaceX another 7.5%. Skanda's read is that six-year Senate terms tilt Fed appointments for four to six years and that inflation may sit near 3% for longer than the market expects. Every number is a reading from the recording date. Educational listening notes only — no investment advice, no price targets, and the judgements here may be wrong.

Published: 2026-11-06
Locale: en
Tags: midterms, podcast-notes, data centers, inflation, Fed appointments


![Early morning in a state capitol rotunda, a construction permit application open on a counter in the foreground beside an upturned rubber stamp with the ink pad still wet, the hall receding behind it into nothing but grey light falling from high windows](/covers/oddlots-2026-11-06-the-macro-implications-of-the-blue-tsunami-cover.png)

> Last night the east wind came again through the small tower; I cannot bear to look back at my old country under this bright moon. The carved railings and jade steps must still be standing — only the faces that were young have changed.
>
> —— Li Yu, "Yu Mei Ren" (Southern Tang, c. 978; my own translation)

The Odd Lots episode published on 6 November 2026 was recorded on 5 November, with Joe Weisenthal and Tracy Alloway hosting Skanda Amarnath, executive director of Employ America, on what changes now that Democrats hold both the House and the Senate. Before the vote, Bank of America's Michael Hartnett said the S&P 500 could fall 10% if Democrats won; the index was up about 1.5% the day after the count, while on the morning of the recording the Nasdaq was selling off, Tesla down 10% and SpaceX another 7.5%. Skanda offered two judgements: six-year Senate terms tilt the balance on Fed appointments for the next four to six years, and inflation may sit near 3% for longer than the market currently expects. These are all readings from the recording date.

## Democrats took both chambers — why didn't stocks follow the script?

Because the pre-election script was a bet on policy direction, and what got repriced after the count was political exposure at the single-stock level.

At the index level, the policy difference was cancelled out by gridlock: Democrats don't have the 60 votes it takes to break a filibuster, the president posted a photo of himself in a veto hat on Truth Social, and tariffs still move by executive order. So Hartnett's 10% did not happen, and the S&P rose the next day.

Single stocks are another matter. The damage concentrated in the companies whose assets sit on specific political conditions — SpaceX and Tesla, both of which put serious money into Texas, plus a few newer cloud companies with plans to expand their footprint. Tracy brought up Sterling Infrastructure, which builds data centres in Texas: a $1,000 stock in June, $255 on the day of the recording, and Joe said the chart looks like a dot-com chart. Skanda's reading is that the 2024 election trade is unwinding, and the footnote they added was John Arnold's line: political giving is the original wealth tax.

![From a single election-night node, several paths fan out to the right: one nearly flat line edging slightly up is the index, a grey dashed line is the pre-election script that never happened, and three clearly downward-diverging lines below are individual stocks, forming a fork in which the index stays flat while the single stocks fall hard.](/figures/index-vs-political-exposure-en.svg)

## The Senate changed hands — what's the most concrete channel into the economy?

Fed board nominations go through the Senate, and that channel jams the moment it changes hands.

The structural Republican advantage in the Senate comes down to one sentence: the Senate counts states, not people, so Kansas and Iowa get two votes each no matter how flat their population growth is. This time Democrats won South Carolina and Mississippi went down to the wire, so even if the wave turns out to be a one-off, the path has been shown to exist.

![On the left, one very large block and one very small block stand for two states with a huge population gap; on the right, each connects to two seat cells of exactly the same size, so the areas differ greatly while the votes are identical.](/figures/senate-counts-states-not-people-en.svg)

The appointments reasoning is the hardest part of the episode. Your own party scrutinises your nominees far less than the opposition does — in the years Republicans held the Senate, Trump's Fed picks Judy Shelton, Steve Moore and Marvin Goodfriend all failed to make it. Powell's term as governor runs to the beginning of 2028, and Skanda's guess is that the seat does not get filled inside the term, because it has to pass through a whole row of Democratic veto gates. Add six-year terms: whoever won this week stays seated, which tilts things for four to six years. He was careful on the other side too — the Fed board carries much less partisan character than the Supreme Court, and Chris Waller still votes in the same direction as Philip Jefferson — so the phrase he used was more sand in the gears.

![A timeline shows, on top, a solid bar for Powell's governor term running to 2028 followed by a long dashed stretch of vacancy, and below it a long bar for the senators elected this time spanning six years, with the overlap marked as a four-to-six-year tilt of the scales.](/figures/fed-seat-confirmation-timeline-en.svg)

Tariffs run the same arithmetic. Pulling trade policy authority back from the executive would need Republicans in the ag states to work with Democrats — he named John Thune as someone who doesn't like aggressive tariffs, since farmers carry the cost — and it would take a group of them opposing their own president at once, which he files as a tail scenario. Gridlock also carries a bill few people price: policy gets harder to pass in a crisis. Divided government after the 2018 midterms was followed 18 months later by COVID, and the CARES Act only got through on a lot of coordination. He won't promise that happens again.

## With the governors' seats flipped, does data centre construction stop?

Power-related data centre approvals sit mostly at the state and local level rather than the federal one, and the states leading the buildout all changed colour — Texas, Georgia, Pennsylvania, Ohio, Virginia, even Wisconsin. Skanda's read is that economics still sets the pace.

The political change is already filtering down to companies. One of the Bloomberg terminal headlines on the recording day was OpenAI changing how it handles data centre efforts: no more NDAs, and no more having local residents pay for the data centres' electricity. For years even blue states handed out tax breaks for data centres as economic activity worth attracting; over the past 12 months it turned toxic. Skanda named Josh Shapiro and Greg Abbott as the two most pro-data-centre governors he can point to, and both pivoted — Shapiro pulled the pivot off, Abbott didn't, and he thinks that looks more like the macro environment at work.

The economic signals showed up earlier, and had nothing to do with the election. Markets have stopped rewarding every dollar of capex, capital efficiency is starting to be worth more than raw capex, and the funding mix keeps moving: cash first, then borrowing in the bond market, and now Oracle issuing new equity, with the pressure filtering down the cap stack. The hyperscalers are locked into spending against each other in an arms race, and what they spend is a cost — invisible on EBITDA, hiding in the D and the A. Skanda's line I wrote down: what was once the best business model in the world, ads, now looks like an industrial. Demand doesn't look gone, though — H100 rental prices rose the day before — except Joe's reminder that the compute can get built in Malaysia.

![A three-layer capital stack runs cash, corporate bonds and new equity from top to bottom, with an arrowed line travelling down the right side to show the funding source moving lower over time.](/figures/funding-moves-down-capital-stack-en.svg)

The surprise he calls most underrated is the Texas business court. Texas set it up to compete with Delaware for corporate charters, and Musk's move to Texas is said to be connected to that; then Democrats won every statewide race. "Texas is the perfect business environment" always rested on political conditions holding, and those conditions turned out not to be that popular. California ran the other way: the wealth tax failed, while Proposition 42, put on the ballot by billionaires to bar new taxes on business holdings and financial assets, is winning by a wide margin.

## Does inflation just get stuck around 3%?

Skanda's read is that it doesn't reach 4%, but it may sit near 3% longer than the market thinks. Three reasons, none of them from the labour market.

First, energy. The closure of the Strait of Hormuz, plus refining capacity problems tied to the Russia-Ukraine conflict, blew out jet fuel and pushed airfares up considerably — so when you hear that services inflation is running hot, the source is energy pushing airfares, and the labour market needn't be the first suspect. Diesel is expensive and the effect on consumers hasn't shown up yet: food retailers absorb the transport and energy costs first, and whether they pass them into food and food-services prices is the part that reaches core. Second, the AI boom itself: laptops have moved, memory-linked goods have moved, more could still move, so this piece lasts longer than people expect. Third, tariffs: plenty of categories have already moved, cars have not, that sector runs under a different tariff regime, and the on-off Canadian tariffs carry costs that haven't reached consumers either.

![On a scale running from 2% to 4%, a band near 3% is boxed off as the stuck zone, and three arrows above push into it from the left, solid ones for costs already passed through and dashed ones for costs that have not.](/figures/inflation-stuck-near-three-en.svg)

On the cycle itself, his verdict on the non-AI economy is: fine. Not gangbusters, not broken; housing is soggy, and homebuilders rising the day before was a rates move. Then the line I wrote down: all booms tend to be imbalanced. The AI-sensitive slices of spending — tech hardware, tech software, industrial equipment — add up to something close to 8% of GDP. The reasoning turns here: if AI capex cools, this kind of thing tends to be binary, on or off, and rarely slides down gently; once it's off, counting on housing to pick up the slack quickly is hard.

![On the left, a large GDP block has a very thin slice cut from its base to stand for roughly 8% of AI-sensitive spending; on the right are two blocks for on and off, with the sloping dashed ramp between them crossed out to show it rarely glides down gradually.](/figures/ai-capex-share-and-binary-switch-en.svg)

## So what can I actually do with this episode?

The company I hold picked the right state and the right policy direction — does that count as a moat? After this episode my answer is that it's a condition with an expiry date.

California's two measures pointed in opposite directions, so "blue states are unfriendly to capital" doesn't work as a general rule either; the test goes back to the specific wording of that state's specific ballot. When I now hear that some state is a given industry's natural home, my first move is to find out who issues the permit and how often that person faces an election.

The second layer is keeping cause and correlation apart. Skanda said the people who spent heavily in 2024 took the win as proof their own contribution worked, so they ran the same playbook again this year and it didn't land; he doesn't think Elon was the cause of Trump's win, though plenty of Democrats believed he was. Joe's bit left me laughing and then cold: a group of people who pride themselves on updating their priors walked straight into the biggest blind spot, and we all have Dunning-Kruger — the difference is that this crowd believes it doesn't.

The third layer is how fast things move. Joe said the comparison being made is the post-Watergate sweep, where the common ingredients were high inflation and a deeply unpopular president, and Reagan was in office eight years later. Tracy's example is closer to home: her dad is Texan, from a small town outside Dallas, and her aunt was a closeted Democrat right up into the early 2000s, unwilling to tell friends and family.

## Worth a look

- The Odd Lots episode of 6 November 2026, the source for this piece (the episode mentions it was recorded on 5 November)
- The previous Odd Lots Election Day special, with Nate Silver's pre-count read on the Senate
- Employ America's research and Skanda Amarnath's public writing
- The text of California's Proposition 42, and the wealth tax measure on the same ballot
- The statute creating the Texas business court, and the debate on competition with Delaware corporate law

## The one thing to take with you

**When you judge whether something is going well, treat "who gets to decide right now" as a variable that changes hands, not as background.**

Every thread in this episode points to the same place. The Texas business court is still standing and the people issuing permits changed; the Fed board seat is still there and the hundred people who confirm it changed composition; data centre power approvals still sit with states and localities, and those hands changed colour. None of these institutions broke. What broke was the premise about who decides that held when they were built — a premise that never made it onto anyone's list of assumptions.

![Two panels sit side by side: on the left a platform rests firmly on a single pillar labelled as someone's nod or some rule, while on the right that pillar's place is only a dashed empty outline, so the platform tilts.](/figures/who-decides-is-a-variable-en.svg)

Here is something I have tried that works outside investing. Take one thing in your life that is running smoothly — your position at work, your kid's after-school arrangement, that fixed weekly slot for exercise, how care for a parent is divided — and write down the one person whose nod it depends on, or the one rule that makes it possible. Then write a second sentence: on the day that person transfers out or that rule changes, what is the first thing you do. Write it on a piece of paper today; two sentences is enough.