The Vote Moved the 2yr 12bp and the 30yr 3bp. Only One of Them Is a Political Price.

The 2yr par yield fell 12.0 basis points to 4.46% on 5 November and the 30yr fell 3.0 to 5.25%. The 2s30s spread, the gap between them, steepened 9 basis points to 79 - the 4th largest steepening in 212 sessions this year - and the long end took 1/4 of the front-end move. On the same session the S&P fell 4.36%, its largest one-day fall in the 44-session file, and TLT rose 0.21%: a hedge beta of -0.048, against -1.545 on 2026-10-22. A front-end rally that the long end does not follow is a repricing of the policy path, not of the fiscal path. Gridlock removed the fiscal channel; what is left of the political trade sits at 2 years.

There is a rule of thumb that says a bond rally is a bond rally. It is wrong, and the 5 November session is the cleanest demonstration of why. The 2yr par yield fell 12 basis points that day. The 30yr fell 3. The bond market did not reprice 'bonds'; it repriced the first 2 years far harder than the last 30. That difference is not noise. The front end of the curve is where the market writes down its expectation of what the Federal Reserve will do at each meeting over the next 2 years. The back end is where it writes down what it will cost the government to borrow for 3 decades, which depends on how much paper the Treasury has to sell and how much term premium (the extra yield demanded for holding a long bond instead of rolling short bills) investors insist on. Those are 2 different questions with 2 different answers, and only one of them is still political after a divided-government result. This piece separates them, prices the separation, and names the level at which the reading dies.

The curve moved 12bp at the front and 3bp at the back

Start with the session, because the shape is the argument. Between the 2026-11-04 and 2026-11-05 par closes, the 2yr fell from 4.58% to 4.46%, -12.0 basis points. The 3yr fell 11.0, the 5yr 10.0, the 7yr 7.0, the 10yr 6.0, the 20yr 3.0, and the 30yr 3.0.¹ The decay is monotonic. There is no tenor anywhere on that curve that fell more than the one in front of it, which is what a repricing of the near-term policy path looks like when it is not accompanied by a repricing of long-run fiscal risk.

The spread arithmetic follows directly. The 2s30s - the gap between the 30yr and the 2yr, the standard measure of how steep the curve is at its extremes - went from 70 basis points on 2026-11-04 to 79 on 2026-11-05, a steepening of 9. That is the 4th largest one-session steepening in 212 sessions of 2026. Only 3 sessions this year steepened more: 2026-03-27 at +13, 2026-07-29 at +15 and 2026-10-28 at +10.²

Two qualifications keep this honest. First, a steepener is not automatically a growth story: the curve can steepen because the front end rallies (bull steepening) or because the long end sells off (bear steepening), and they mean opposite things. This one is unambiguously the first - every tenor fell. Second, the level is still low in absolute terms. At 79 basis points the 2s30s sits at the 18.8th percentile of its 213 closes this year, against a 101 basis point median. The curve steepened; it did not normalise.

That shape is a policy move, not a fiscal move

The mechanism, stated plainly. The price of a 2yr note is set almost entirely by what the market expects the policy rate to average over those 2 years, so it moves when the expected path of the Federal Reserve moves. The price of a 30yr bond is set by that same expected path plus a term premium that compensates the holder for 3 decades of interest-rate risk and for absorbing whatever quantity of long paper the Treasury issues. A session that moves the 2yr 4 times as much as the 30yr is a session where the expected policy path changed and the term premium did not.

That is exactly what a divided-government result should produce. Before the count, the fiscal channel was live: one party holding the presidency and both chambers could pass a budget resolution, which meant a larger deficit, which meant more coupon supply at the long end, which meant a higher term premium. Gridlock closes that channel. It does not close the other one. The Federal Reserve sets the policy rate regardless of who holds the House, and the political variable that survives a split Congress is who runs the Fed and how fast a decision can be reversed - which is a question about the next 2 years, not the next 30.

The arithmetic of the session says the market agrees. If the long end had repriced fiscal risk on 5 November, it would have moved at least as much as the front end. It moved 3 basis points against 12. Whatever the trigger was - and this piece makes no claim about the trigger, only about the response - the response was concentrated where the political information was concentrated.

The long bond stopped paying for risk

A curve fact becomes a portfolio fact on the same tape. On 5 November the S&P 500 closed at 7,353.73, down 334.89 points or 4.36%, the largest one-session fall in the 44 sessions this file covers. TLT, the long-duration Treasury fund, closed at 81.93, up +0.17 or +0.21%. The hedge beta - TLT's percentage move divided by the S&P's, the number that says how much protection a long-bond position actually delivered - was -0.048.³ Nearly nothing.

Put it against the other 4 sessions in the file where the S&P fell 1% or more and the contrast is stark. On 2026-10-22 the S&P fell 1.21% and TLT rose 1.87%, a beta of -1.545: the hedge worked exactly as advertised. The 5 observations run from -1.545 down to -0.048, and 5 November is the worst of them. That is a small sample and this piece will not pretend otherwise - but the direction is consistent with the curve: a long end that will not rally on a policy shock is a long end that will not rally to protect an equity position either, because both require the same thing, which is a fall in the 30yr yield.

The cross-check is in the volatility, and it needs care. VIX closed at 28.18 against 16.61, +69.7%, while MOVE - the rates-market equivalent, the implied volatility of Treasury options - went from 93.88 to 97.92, about +4.3%. The temptation is to read a collapse in rates vol against a spike in equity vol. Resist it. On their own long histories both are mid-distribution, and a 44-session file will manufacture exactly this divergence from a single large equity move. The defensible statement is narrower: one session's equity repricing did not travel into rates volatility, which is what a pinned long end would look like.

Where the front end actually sits

None of this means the front end is cheap. After a 12 basis point rally the 2yr is at 4.46%, which is the 81.7th percentile of its 213 closes in 2026 - still high in the year's range, which ran from 3.38% to 4.92%. The 30yr at 5.25% is at the 78.4th percentile of a range from 4.64% to 5.67%. Both ends are rallying off a high base, which is the opposite of a market that has priced a benign political outcome.

The real yield is the piece that disciplines the fiscal story. The 10yr TIPS real yield printed 2.64% on 2026-11-04 against 2.65% the session before, and the 10yr breakeven - the market's implied inflation compensation over 10 years - printed 2.24% on 2026-11-05 against 2.26%.⁴ Both are lagged relative to the 5 November par close and both are quoted to their own dates for that reason. The point they make is that the inflation compensation embedded in the curve has not moved in a way that would justify a term premium repricing: it is drifting down, not up.

The term premium estimate is the laggard and should be treated as such. The ACM 10yr measure - the New York Fed's model estimate of what investors demand for holding duration, rather than a traded price - last printed 0.9089% on 2026-10-30, from 0.9320% the session before.⁵ That is 3 sessions behind the curve move this piece is about. It is quoted so the ordering can be checked later, not to support a claim about 5 November.

The instrument and the level

The instrument is the 2s30s spread, currently 79 basis points, and the second instrument is the TLT-to-S&P hedge beta, currently -0.048. Neither is a forecast. Both are readings of the same thing: how much of a policy repricing the long end is willing to transmit.

The call is that the political content of this result prices at the front end and stops there. Gridlock removes the deficit channel, so the term premium has nothing new to price, so the 30yr behaves like a duration instrument rather than a fiscal instrument, so it neither sells off on the result nor rallies to hedge the equity consequence of it. The 2yr does all the work because the Fed is the only actor the result did not constrain. On the numbers: -12 against -3, a +9 basis point steepening that ranks 4 of 212, and a hedge beta of -0.048 on the worst equity session in the file.

What would confirm it. The 2s30s holding above 70 basis points through the next front-end rally, with the 2yr making the larger move on each of them. The hedge beta staying below 1.54 on any further session where the S&P falls 1% or more. A 30yr auction tailing while the 2yr rallies - the cleanest single signal that duration supply and policy expectations have come apart.

What would break it. The 2s30s back below 70 basis points with the 30yr making the session's largest move, which would say the long end is still the fiscal instrument and the 5 November shape was a single-session artefact. A hedge beta back at or below -1.545 on a 1% equity decline, which would restore the long bond's hedging property and contradict the pinning read. And on the political mechanism specifically: any credible fiscal expansion passing a divided Congress, which would reopen the channel this piece says is closed.

Where this is wrong

First, on the sample. The hedge evidence rests on 5 sessions, and 5 observations cannot establish a beta. If the long bond hedges the next 3 equity declines at -1.545 or better, the portfolio claim in this piece is dead regardless of what the curve does. The curve evidence is stronger - 212 sessions - and should be weighted accordingly.

Second, on attribution. This piece deliberately makes no claim about what triggered the 5 November equity decline. It claims only that the fixed-income response was concentrated at the front end. A single-session curve move can be driven by positioning, by a data surprise, or by a funding squeeze rather than by any political read, and none of those would support the gridlock mechanism. The mechanism is a hypothesis consistent with the shape, not a proven cause of it.

Third, on the level. If the 2s30s steepens past 101 basis points - the 2026 median - on front-end strength alone, the curve is normalising rather than repricing policy, and the distinction this piece draws between a policy move and a fiscal move stops doing work. The falsifier is the level, not the direction.

Notes

1. Par yields: data/treasury_par.csv, Treasury daily par yield curve, 2026-11-04 and 2026-11-05 closes, 213 sessions from 2026-01-02. Yields quoted are par closes, not intraday marks. Session changes are par close to par close.

2. The 2s30s is the 30yr par yield minus the 2yr par yield, in basis points, computed on every close in the ledger. The 212 one-session changes are differences of that series. The 4 sessions at or above 9 basis points are 2026-03-27 (+13), 2026-07-29 (+15), 2026-10-28 (+10) and 2026-11-05 (+9).

3. Equity, TLT, VIX and MOVE levels are from the daily snapshot files, which carry the prior session's close: the 6 November snapshot carries 5 November marks for these series. Hedge beta is the TLT percentage change divided by the S&P 500 percentage change on the same session. The 5 sessions where the S&P fell 1% or more in the 44-session file are 2026-11-05 (-4.36% / +0.21%), 2026-10-30 (-1.82% / +0.43%), 2026-10-22 (-1.21% / +1.87%), 2026-10-14 (-1.11% / +0.66%), 2026-10-16 (-1.09% / +0.10%).

4. Real yield and breakeven: FRED DFII10 and T10YIE, daily. The newest prints are 2026-11-04 and 2026-11-05 respectively and are quoted to those dates; they lag the par close used for the curve arithmetic.

5. Term premium: FRED THREEFYTP10, the ACM 10yr estimate maintained by the New York Fed, in percent. It is a model estimate rather than a traded price and its newest print is 2026-10-30, 3 sessions behind the 2026-11-05 par close. No claim in this piece depends on it.

6. Volatility context: VIX and MOVE are quoted for the 5 November session only. Both series sit mid-distribution over their full histories and the 44-session snapshot file overstates any divergence between them; the piece states only that one session's equity repricing did not travel into rates volatility.

7. All 4 charts are generated by script from the series named above. Chart 1 plots the par yield change by tenor; chart 2 the 2s30s path with the 4 steepening sessions marked; chart 3 the S&P and TLT returns on the 5 down sessions; chart 4 the 2yr against 30yr scatter for every session the 2yr fell.

8. No claim is made about the cause of the 5 November equity decline, and no forecast is made. Levels are descriptive as of 2026-11-05.