ExpertsFMYou Ain’t Seen Nothing Yet with Daniel Clifton
The Compound and Friends · November 6, 2026
5 Selected Lessons:
case · 2 framework · 1 heuristic · 1 reasoning_pattern · 1
Hard-data-over-polls election forecasting
Source
The Compound and Friends — You Ain’t Seen Nothing Yet with Daniel Clifton (November 6, 2026) · at 2:12 ▸
When this applies
when forecasting any discrete, binary future event where public sentiment surveys are available but unreliable
The takeaway
Replace unreliable stated-preference signals with revealed-preference data from real, smaller-scale instances of the same behavior you are trying to predict.
Why it works
Polls are self-reported stated preference and are noisy/biased; actual voting behavior in real low-stakes elections reveals true revealed preference and trend direction before the big event.
One level up
Trust what people actually do over what they say they will do, especially when aggregated across many small real trials.
Where it stops holding
This stops working when no analogous smaller real-world data points exist, or when the underlying dynamics shift suddenly between the small trials and the big event.
Detail
procedure:
- Ignore public polling as unreliable ('garbage')
- Track special elections and off-cycle real votes as hard data
- Watch for shifts in voting margins among specific subgroups (party base, independents) across multiple small elections
- Use the accumulating trend to anticipate the larger election outcome
rationale: Real votes in lower-profile elections reveal genuine shifts in voter sentiment that polls miss or distort.
In their words
We used hard data. How are people voting in the special elections? And what you began to see were those Republican voters starting to go, I'm not sure I'm down with this.
Trace political-to-market causality through earnings-linkage weight, not headline intuition
Source
The Compound and Friends — You Ain’t Seen Nothing Yet with Daniel Clifton (November 6, 2026) · at 53:32 ▸
When this applies
when deciding how much a sector-specific political/regulatory event should move a broad market index
The takeaway
Judge the macro impact of a sector-specific shock by that sector's share of forward earnings/GDP growth, not by how 'important' the sector seems qualitatively.
Why it works
A sector-specific political shock normally stays contained to that sector's sub-industries, but if that sector carries an outsized share of expected GDP growth and earnings growth for the whole market, the shock propagates into broad index moves disproportionate to its sector-specific nature.
One level up
Concentration of forward growth expectations in a narrow segment turns sector-specific risk into macro risk.
Where it stops holding
This stops working when growth expectations are broadly diversified across many sectors, in which case a sector shock stays contained and doesn't move the broad index.
Detail
approach: Check what share of forward GDP/earnings growth is concentrated in the shocked sector before judging the likely breadth of market impact.
signals:
- Data centers represented roughly 75% of expected GDP growth next year
- Data centers represented roughly 60% of expected earnings growth
- Largest market-cap stocks tied directly to the sector
rationale: Midterm elections usually produce sector/sub-industry effects, not macro effects — but this one broke that pattern because of the sector's outsized growth-contribution weighting.
In their words
75% of the expected GDP growth next year and maybe 60% of the expected earnings or like it can't completely go away.
Midterm-year stock market base rate
Source
The Compound and Friends — You Ain’t Seen Nothing Yet with Daniel Clifton (November 6, 2026) · at 37:34 ▸ · at 45:56 ▸
When this applies
when any asset experiences a sharp uncertainty-driven decline tied to a recurring, historically-studied cyclical event
The takeaway
A sharp decline driven by resolvable political/cyclical uncertainty tends to be followed by an above-average recovery once the uncertainty clears, because the newly empowered actors have incentive to improve conditions before their own next evaluation.
Why it works
Markets price in worst-case political scenarios ahead of the event; once the election clears, incumbents (regardless of party) pivot toward growth-friendly policy ahead of the next election cycle, resolving the uncertainty discount.
One level up
Uncertainty discounts are typically overdone relative to the eventual outcome, because decision-makers are incentivized to self-correct once given power.
Where it stops holding
This stops working when the underlying shock is structural rather than cyclical/political (e.g., a genuine earnings collapse, a permanent change in law that is actually enacted) — Clifton explicitly notes this pattern could break but says historically it 'never' has since 1938.
Detail
rule: Buy into midterm-year equity sell-offs driven by political uncertainty.
threshold: S&P 500 has not declined in the 12 months following a midterm election since 1938; average return +14.8%.
rationale: The market assumes the worst in a midterm year, that worst doesn't materialize, and newly elected officials start positioning for the next election two years out.
In their words
This is my sixth midterm election markets, and I've always said the SP 500 has not declined in the 12 months following a midterm election since 1938.
Louisiana's data-center property-tax model avoids backlash
Source
The Compound and Friends — You Ain’t Seen Nothing Yet with Daniel Clifton (November 6, 2026) · at 27:57 ▸
When this applies
when trying to build local acceptance for an unpopular but economically valuable project
The takeaway
Concrete, individually-felt economic benefits overcome local opposition far more effectively than abstract macro benefits; the benefit must be communicated, not just exist.
Why it works
Making the direct, visible economic benefit (lower property taxes, higher home values) salient to residents changes their cost-benefit calculus, whereas diffuse or invisible benefits (national GDP growth) do not move local sentiment.
One level up
Value you don't actively communicate to the people bearing the cost does not count politically, no matter how real it is.
Where it stops holding
This stops working when the non-economic objections (health fears, ecological concerns, noise) dominate regardless of tax benefits, or when the benefit cannot be structurally segregated/localized.
Detail
situation: Louisiana's Meta/Oracle-adjacent data center project (Benier Project) proceeded without the backlash seen in Virginia and other states.
decision: Louisiana segregated data center electricity load from residential consumers, embraced job creation messaging, and ran a PR campaign (e.g., teacher salary increases) tied to the benefits.
outcome: Loudoun County's property tax rate has fallen 10 years in a row due to data center tax revenue, raising home values by roughly $20,000 per homeowner, yet most residents are unaware of this benefit, unlike Louisiana's messaging.
lesson: Structural segregation of costs plus proactive communication of concrete individual benefits can neutralize opposition that purely economic facts (uncommunicated) cannot.
In their words
I authored a study in April that showed that if you have a data center, they pay your property taxes. And what that does is it leads not only to lower property taxes, which dwarfs any electricity increase, but it creates demand for housing.
Municipal bonds after 2010 midterm — chaos is a ladder
Source
The Compound and Friends — You Ain’t Seen Nothing Yet with Daniel Clifton (November 6, 2026) · at 37:34 ▸
When this applies
when a whole asset class is sold off indiscriminately because of a feared future policy action that has not actually happened yet
The takeaway
When a market sells an entire asset class on a feared-but-unconfirmed policy threat, the resulting mispricing can become an opportunity if the feared policy doesn't fully materialize.
Why it works
Indiscriminate fear-driven selling based on an anticipated-but-unrealized policy threat creates mispricing, because the market prices in the worst case rather than waiting for the actual outcome.
One level up
Panic-driven repricing of a probabilistic future event, rather than its realized outcome, is where asymmetric opportunity often lives.
Where it stops holding
This stops working when the feared policy actually does materialize in full force, validating the sell-off.
Detail
situation: After Obama's 2010 midterm loss, investors feared Republicans would cut state aid and force municipal bankruptcies, and sold municipal bonds broadly.
decision: Clifton implies the correct response was to buy into that panic rather than follow it.
outcome: It turned out to be 'the greatest trade ever' — investors could buy California municipal bonds tax-free yielding 12-13%.
lesson: Clifton uses this as a direct analogy for the current data-center/AI sell-off: wait out the panic, as range of outcomes narrow once policymakers show their actual hand.
In their words
And then, you know what? It turned out to be the greatest trade ever where you could buy California municipal bonds tax-free for, you know, like 12, 13%.