Chiunque abbia un’auto o una moto conosce perfettamente la scena: il Brent scende del 5% sui mercati internazionali, ma il prezzo della benzina rimane pressoché immobile.

La convinzione diffusa è che il prezzo dei carburanti sia lo specchio fedele del greggio, ma in realtà, il costo del barile di Brent è solo il primo anello di una catena industriale e fiscale molto più complessa.

Ma quindi perché il prezzo alla pompa non segue (mai) davvero il Brent?

Subscribe now

La scomposizione del prezzo alla pompa e il “Waterfal Chart” dei costi

Il prezzo che paghiamo al distributore per un litro di benzina o gasolio è dato da tre macro-componenti:

• Prezzo Industriale (Materia Prima): la somma del costo del greggio acquistato dalle raffinerie, del costo di raffinazione, trasporti e logistica. Segue l’andamento dei mercati internazionali, ad esempio le quotazioni Platts del Mediterraneo per i prodotti finiti, non direttamente il Brent/WTI grezzo.

• Accise (Imposta Fissa): l’imposta statale calcolata per litro, non in percentuale sul prezzo. Quindi, indipendentemente da quanto sale o scende il prezzo del petrolio, il valore dell’accisa erariale rimane invariato (salvo interventi normativi del governo, promessi ma mai realmente eseguiti).

• IVA (Imposta Valore Aggiunto): Si applica in percentuale (attualmente al 22%) sul valore totale, cioè si paga l’IVA anche sulle accise stesse (un meccanismo perverso di tassazione sulle tasse stesse, su cui, al solito, nessun governo è mai intervenuto strutturalmente).

In Italia, la componente fiscale (Accise + IVA) pesa storicamente tra il 50% e il 60% del prezzo finale alla pompa, a seconda delle oscillazioni dei prezzi di mercato dei carburanti.

Share

La relazione con il Brent: Perché non si muovono all’unisono?

Generalmente si nota che quando il prezzo del greggio sale, i carburanti aumentano rapidamente, mentre quando scende, la discesa è più lenta (il fenomeno detto rockets and feathers). Questo disallineamento temporale non è pura speculazione come molti pensano, ma strutturale e dovuto a diversi fattori:

I prezzi dei prodotti raffinati seguono logiche proprie di domanda e offerta (ad esempio, l’aumento della domanda di gasolio in inverno per il riscaldamento o di benzina in estate per i viaggi).

C’è poi uno scostamento (lag) temporale legato alle scorte di prodotto : le stazioni di servizio e i depositi acquistano il carburante con contratti a termine o basati sulle giacenze dei giorni precedenti. Il carburante che compriamo oggi è stato acquistato a prezzi medi di qualche settimana fa (ma questo non impedisce al prezzo di schizzare in alto immediatamente a seguito di uno shock dei prezzi del greggio, più avanti vediamo perchè)

C’è poi un problema di rigidità fiscale: le accise sono un’imposta fissa e costituiscono una parte considerevole del prezzo, una diminuzione del 10% del prezzo del greggio si traduce in una riduzione percentuale molto più contenuta sul prezzo finale alla pompa.

La componente industriale: Raffinazione e Crack Spreads

Quando facciamo carburante non stiamo comprando greggio ma prodotti finiti per cui il punto di partenza fondamentale è la distinzione tra il prezzo del greggio (Brent o WTI) e le quotazioni dei prodotti raffinati (Platts CIF Med, RBOB Gasoline, Heating Oil).

• I margini di raffinazione: Perché il valore aggiunto delle raffinerie oscilla in base ai colli di bottiglia globali, alla capacità produttiva e ai costi energetici (es. il prezzo del gas).

• Logistica e Trasporto: I costi per muovere il combustibile dai siti di stoccaggio/raffinerie ai depositi costieri e locali (tramite navi, treni o autocisterne).

• Distribuzione e Margine del Gestore: Il margine lordo della compagnia petrolifera e del gestore dell’impianto (la pompa bianca o il marchio brandizzato), che deve coprire i costi di gestione, le utenze e il guadagno finale.

• Il caso del Gasolio (Diesel): Perché il diesel segue logiche di domanda diverse rispetto alla benzina (distillati medi legati al riscaldamento e alla logistica industriale pesante)

Le Dinamiche Dietro i Margini di Raffinazione

Il prezzo industriale dipende dal costo della materia prima, ma incorpora il margine di raffinazione (spesso misurato dal parametro Ural Med Cracking o dai crack spreads). Questo margine rappresenta il “valore aggiunto” della raffineria ed è guidato da dinamiche economiche piuttosto rigide:

• Domanda e Offerta di Prodotti Finiti: Le raffinerie non producono solo benzina, ma una gamma ampia di distillati (gasolio, Jet Fuel, oli combustibili, GPL). Se c’è una carenza globale di gasolio (legata ad esempio a inverni rigidi o a blocchi logistici), il margine di raffinazione del diesel impenna, anche se il prezzo del Brent resta stabile.

• Capacità Produttiva Globale e Colli di Bottiglia Strutturali: La transizione ecologica ed i bassi margini del decennio 2010-2020 hanno portato alla chiusura di molte raffinerie occidentali obsolete, riducendo la capacità di lavorazione globale ed aumentando la dipendenza da aree geografiche con una stabilità geopolitica ridotta. Questo significa che il sistema opera spesso vicino al 100% della capacità e che qualsiasi shock geopolitico o manutenzione straordinaria fa schizzare in alto i margini.

• Costi Energetici delle Raffinerie: Raffinare il petrolio richiede enormi quantità di energia. I processi di raffinazione richiedono generalmente elevate temperature e pressioni operative, con consumi energetici considerevoli. Quando i costi energetici (es. il gas in Europa) salgono, le raffinerie scaricano inevitabilmente questi costi sul prezzo industriale dei prodotti finiti per difendere i margini operativi.

• L’effetto “Rockets and Feathers” (Razzi e Piume): Quando il Brent sale improvvisamente, le raffinerie e i distributori adeguano subito i listini dei prodotti raffinati per proteggere il valore delle scorte e riacquistare materia prima a costi più alti. Quando il Brent scende, i prezzi alla pompa scendono molto lentamente (la piuma) perché si tende a smaltire le giacenze acquistate a un prezzo superiore nei giorni precedenti. (ok, ok, la speculazione ha un certo peso in questo meccanismo…)

Nel grafico abbiamo messo a confronto il prezzo del Brent con RBOB (Reformulated Blendstock for Oxygenate Blending), lo standard comunemente utilizzato nei mercati (in particolare al NYMEX) per quotare e contrattare i derivati sulla benzina, e l’indice Heating Oil, il contratto futures di riferimento utilizzato come benchmark globale per i distillati medi. I prezzi sono normalizzati a 100 per renderli graficamente confrontabili.

Il grafico mostra come, sebbene i trend abbiano lo stesso andamento, il valore assoluto di entrambi i derivati tende a separarsi sempre più marcatamente dal prezzo del Brent. è un pò quello che troviamo alla pompa di benzina oggi: il prezzo attuale è decisamente superiore a quello che avremmo trovato 2 o tre anni fa a parità di prezzo del barile di petrolio.

La radiografia della fiscalità italiana

Dopo aver visto come funziona la componente Industriale andiamo finalmente ad analizzare il peso dello stato.

Come abbiamo detto in precedenza, i carburanti sono soggetti ad una tassa fissa al litro (le accise) ed una tassa variabile in funzione del prezzo finale di vendita (l’IVA, attualmente al 22%).

Le accise

Lo stato italiano applica ai carburanti un’imposta fissa al litro la cui aliquota complessiva è stata innalzata più volte nel tempo, spesso per finanziare emergenze, calamità o esigenze di bilancio, fino a essere unificata in un’unica aliquota strutturale con la riforma del 1995 e resa definitivamente parte del gettito generale dello Stato dal 2013.

Fino al 2025 benzina e diesel avevano accise diverse (0,7284 €/l per la benzina, 0,6174 €/l per il diesel), con un vantaggio fiscale storico per il gasolio. Dal 1° gennaio 2026 è scattato il riallineamento previsto dalla Legge di Bilancio, nell’ambito della progressiva eliminazione dei Sussidi Ambientalmente Dannosi richiesti dall’UE, per cui l’accisa sulla benzina si è ridotta di 4,05 centesimi, quella sul diesel è aumentata della stessa cifra.

Oggi quindi abbiamo equiparato le due accise a 0,6729 € al litro. Un importo fisso, che rende quasi la metà del prezzo (senza IVA) immune alle fluttuazioni di mercato.

L’IVA e l’effetto moltiplicatore:

Attualmente l’IVA sui carburanti è al 22%, calcolato sull’imponibile complessivo, prezzo industriale + accise. Esattamente: l’IVA si applica anche sulle tasse, diventando una vera e propria tassa sulla tassa in una dinamica che nessuno, nonostante anni di inutili proclami elettorali, ha ancora avuto il coraggio di correggere, facendo in modo che il prezzo alla pompa in Italia sia ad oggi uno dei più alti di Europa, come mostrato nel diagramma sottostante

Con i prezzi medi reali comunicati dal ministero a fine agosto (benzina self a 2,016 €/l, gasolio self a 2,130 €/l) il prezzo dei carburanti è così composto:

Per la Benzina:

• Prezzo Industriale (Materia prima + Raffinazione + Distribuzione/Margini): ~48% del totale (circa 0,98 €/l).

• Accisa (Imposta Fissa dello Stato): Fissata a 0,6729 € al litro dal 1° gennaio 2026 (allineata a quella del diesel). Trattandosi di una cifra fissa in centesimi, il suo peso nominale in euro non cambia se il petrolio sale o scende.

• IVA (22%): Si calcola applicando l’aliquota sul prezzo industriale + accisa (circa 0,36 €/l).

• Peso Fiscale Totale (Accise + IVA): Si attesta intorno al 51% del prezzo pagato alla pompa.

Per il Gasolio (Diesel):

• Prezzo Industriale: ~50% del totale (circa 1,07 €/l) — spesso ha un costo industriale base leggermente superiore alla benzina, coerente con le dinamiche di domanda sui distillati medi viste nel paragrafo 5.

• Accisa (Imposta Fissa): Fissata a 0,6729 € al litro dal 1° gennaio 2026 — non più inferiore a quella della benzina, come storicamente accadeva, ma perfettamente allineata.

• IVA (22%): Applicata sempre sul valore imponibile complessivo (circa 0,38 €/l).

• Peso Fiscale Totale (Accise + IVA): Intorno al 50% del prezzo finale.

Complessivamente il peso fiscale sui carburanti si aggira attualmente attorno al 50% di quello che paghiamo alla pompa. Anche in questo caso, se ci confrontiamo con il resto della UE, siamo stabilmente nella parte alta della classifica…

La conseguenza pratica è che se il costo del greggio scende del 20%, questo si dovrebbe riflettere sul prezzo finale in una riduzione di circa il 10%, proprio perché la componente fiscale, che pesa per più della metà del prezzo alla pompa, rimane insensibile alle fluttuazioni di mercato.

Inoltre, come abbiamo visto sopra, il prezzo al dettaglio tende a seguire istantaneamente gli aumenti di prezzo (effetto razzo) ma a diminuire molto più lentamente in caso di cali del prezzo della materia prima (effetto piuma).

Conclusioni

Quando facciamo il pieno la domanda che dovremmo farci non è “quanto costa la benzina”, ma “cosa sto pagando davvero”. E, come abbiamo visto, la risposta è circa 50% costo industriale, il resto tasse. La parte industriale rappresenta una componente volatile, soggetta a geopolitica e logiche di mercato, mentre la componente fiscale è piuttosto rigida per via della struttura delle accise.

È questa combinazione, non solo la speculazione o il balzo del greggio, a spiegare perché il pieno resti la spesa che più difficilmente si smette di notare.

Guardare solo il prezzo esposto alla pompa, insomma, racconta solo una parte della storia. Il prezzo della benzina non replica mai fedelmente le oscillazioni del Brent ma è l’interazione tra il costo della materia prima, i margini di raffinazione (sempre più tesi dopo anni di sotto-investimento in Europa), e quella fiscale divisa in accise (fisse) e IVA (mobile).

L’Italia non è un’anomalia in Europa, ma si colloca in un gruppo ristretto di paesi (insieme a Malta, Grecia e Finlandia) dove il fisco assorbe più della metà del prezzo del carburante. E questa è la premessa indispensabile per valutare con lucidità le prossime proposte di intervento, se e quando arriveranno.

Marco

I also introduced my adaptive allocation model, which is, down to the bone, a quite simple rule-based algorithm designed to reduce exposure or step aside at all, when the market conditions become more hostile, instead of holding a static position through every market.

I also mentioned that the two most effective hedge instruments were Gold and SPY.

In this article I want to explain why a hedge is not optional for a strategy like this, walk through the actual backtest results and compare different hedge instruments.

And, more importantly, no matter how good the math looks, every strategy shall be mentally sustainable, so this cannot be the only thing in your portfolio.

Hedging fixes the strategy’s exposure to hostile regimes. Sizing fixes your exposure to the strategy. Neither one is optional, and neither one substitutes for the other.

Subscribe now

1. What “hedging” means in this context

If you hold a leveraged-ETF strategy, hedging usually means one of two things:

• Buying insurance — options that pay off when the position falls. This reduces the pain during drawdowns, but also caps the strategy’s returns.

• Dynamic reallocation — shifting capital away from the leveraged position and into a different, more stable asset when conditions turn dangerous, then shifting back when they improve.

The idea behind my model is the second one: keep the upside, avoid the worst of the drawdowns. And no — it’s not easy at all. My model is based on an active decision strategy, made systematically according to specific numerical conditions, to identify how much exposure to the convexity trap you should be carrying right now, given the current regime.

In another article, I introduced how the Kelly framework tells us that the growth-optimal leverage ratio depends on the ratio of expected return to variance. Static leverage assumes that ratio is constant — but it isn’t, because volatility and drawdowns cluster, and trends persist until they suddenly break. A hedge, in this sense, is just a mechanism for dynamically adjusting the effective risk (leverage) your portfolio is carrying, instead of locking in one number and hoping the market stays cooperative.

2. Hedging as a structural necessity, not an option

If you decide to use leveraged products, there’re three main reasons to hedge:

a. The daily reset punishes volatility itself, not just direction; in particular in choppy, directionless markets it makes capital bleed even when the underlying index is flat. Trying to reduce realized volatility during those regimes means mathematically reducing the σ² term (variance) that drives the decay.

b. A trend-following overlay is asymmetric. A model like this makes most of its money capturing sustained trends and is designed to do little, or step aside, during downturns. But since moving into cash is, most of the time, a wasted opportunity, rotating into an asset that tends to perform when tech-led momentum breaks down (Gold, historically, during flight-to-safety episodes) turns dead time into productive time.

c. Static diversification fails. We’ve seen that in 2020 and in 2022; regimes shift and traditionally uncorrelated assets went down at the same time. For example in 2022 persistent inflation pushed rates up forcing equities and bonds to fall together. A static hedge works until the correlation it relies on breaks. A dynamic one can, in principle, recognize the regime shift and adapt rather than ride it down.

3. The signals, at a high level

At a conceptual level, the model I developed is watching two things continuously:

• Trend strength in the underlying index (is momentum intact, decaying, or reversing?)

• The volatility regime (is realized/implied volatility calm, elevated, or spiking?)

When trend deteriorates and/or volatility regime shifts, the model reduces or rotates out of TQQQ exposure into the hedge asset. When conditions normalize, it rotates back. This risk-off/risk-on overlay is applied systematically, on a fixed weekly rebalancing schedule. No discretionary judgment, only systematic application of predefined rules.

4. Searching for the optimal hedge

I tested different hedges over different rolling windows: Gold, TLT, SPY, QQQ and BIL. Same adaptive allocation logic just swapping out only the hedge leg each time. Here’s what happened:

The first snapshot comes from a simulation starting in 2015

Backtest results - 2015 to date

In terms of risk-adjusted metrics, all the five hedged strategies beat the market and even the buy&hold TQQQ allocation. In terms of CAGR both the bond and cash hedged strategies provide a value lower than TQQQ, but higher than the non-leveraged market benchmarks. Within the 5 hedged allocations, Gold wins on every risk-adjusted metric — highest Sharpe, highest Sortino, highest Calmar, and the second shallowest max drawdown of the group, while also posting the highest raw CAGR. Only using cash (through BIL as a proxy) provides lower volatility, but at the cost of both overall and risk-adjusted returns.

TLT is the weakest hedge, and this isn’t a big surprise given what we saw in 2022 and partially in 2020: during the last two major crisis bonds stopped being decorrelated from equities, providing no drawdown protection.

QQQ and SPY as hedges essentially amount to de-leveraging rather than true diversification; it helps, but it doesn’t offer any decorrelation benefit and real drawdown protection.

Currently I settled on Gold as the hedge in the live model.

Let’s zoom in to give a closer look to the past 6 years. I run the same simulation but starting from 2020, 2022, 2024. Here the cumulative results:

Backtest results - 2020 to date

Backtest results - 2022 to date

Backtest results - 2024 to date

The pattern we can observe is quite similar to what we extracted from the 12 years simulation; the adaptive strategy beats the TQQQ buy&hold over almost every horizon. Nevertheless, in selected rolling windows (like 2024 to date), TQQQ beats the adaptive strategy only when bonds or cash are used as hedges, and this happens when using observation periods without any substantial and sustained drawdown (in the past couple of years we had mostly quick v-shaped market crashes).

5. The Gold-hedged strategy in detail

As I already mentioned, I’m actually using this strategy in my trading portfolio, with Gold as the hedge. For the records, I started live trading it in January 2025; the REAL application returned a staggering 85.8% CAGR with Sortino 2.416 and Calmar 3.417. To be honest those numbers are “inflated” by an almost uninterrupted bullrun and unprecedented performances from Gold. And to be totally honest, I don’t expect those numbers to repeat long term.

To analyze the strategy properly we need to go back to 2020 to include the two biggest recent bear markets.

Performance metrics, 2020–2026:

The result: roughly 2.7x the CAGR of QQQ and nearly 1.5x the CAGR of raw TQQQ buy-and-hold, while cutting TQQQ’s maximum drawdown by more than 30 percentage points, from -81.7% down to -49.7%.

That’s the good news. Now the part that actually matters more.

Subscribe now

6. The number that should scare you more than any of the above: -43.4%

The number you should take care of the most is not the total return. The painful part of this investing strategy is the drawdown.

Look again at 2022. Even with the hedge fully engaged, even with the best risk-adjusted metrics of any variant tested, the adaptive strategy still lost 43.4%, in a single year. Far better than the 79.7% of TQQQ alone but still a painful one.

That’s not a typo, and it’s not a flaw in the model; it’s the honest cost of retaining real convexity. The whole point of this strategy isn’t to eliminate drawdowns, because we can’t. The real scope is to make them survivable instead of terminal.

TQQQ unhedged lost nearly 80% that year and took roughly three years to fully recover. A -43% year, hedged, recovers. A -80% year tests whether your brain is wired to hold on for three years watching your capital sit at a fraction of its former value. Most people, me included, don’t. Most people will be prone to sell at the bottom, lock in the loss, and never recapture the upside that would have made them recover.

So ask yourself honestly: could you sit through a 40%+ drawdown in a single year without touching the position?

If the answer is “probably not,” the model isn’t broken, you just haven’t sized it correctly yet.

7. Sizing is the real risk management layer

This is the real point. Looking at the strategy in isolation, returns are staggering, but you can’t look at it as if it were your entire net worth. Because it shouldn’t be. In my own case, this strategy represents a small slice of my active trading capital, around 10%, and a considerably smaller fraction still when measured against my total financial picture.

The hedge inside the model reduces the strategy’s volatility. The allocation size reduces your volatility, the one that actually determines whether you will panic-sell during the next market crash.

And this is a part that doesn’t show up in any backtest table, because it isn’t a property of the strategy, it’s a property of you, applied on top of the numerical model.

Think of it as two independent layers of defense, stacked:

1. Intra-strategy hedging — asset rotation reduces the model’s own drawdowns and volatility drag.

2. Portfolio-level sizing — keeping the strategy itself as a modest fraction of your active capital, and an even smaller fraction of your total net worth, means that even the model’s own worst-case years become a manageable dent rather than a catastrophe.

A strategy that is mathematically optimal but psychologically unsurvivable is not actually optimal, because the investor who abandons it at the bottom captures none of the long-run compounding the backtest promises. The math and the psychology are an equation couple, to be solved together, or the math alone will fail.

Marco

🔒 Bonus: This Week’s Allocation

[Content below is for paid subscribers only]

Read more

In late 2025 and early 2026, the dominant view was that Brent would spend most of 2026 near $55 per barrel, pressured by excess supply and rising inventories.

We all know what happened then.

By August, we still have a market under an acute supply-disruption regime, spot Brent in the high-80s/low-90s range and most forecasts for Q4 clustered in $70–80 territory.

World oil supply went from ~107.8 mb/d in December 2025 and fell to 94.5 mb/d by May (13.6 mb/d below pre-conflict levels). After an initial de-escalation between US and Iran, we had a partial recovery to 101.5 mb/d by July, still well below 2025 levels

For LNG the situation is even worse. The Strait of Hormuz normally carries ~20% of global LNG supply and global production fell 11.4% in February only, to reach its lowest in may at 29.72 Mt, -24.1% compared to december 2025. TTF and JKM both hit multi-year Q2 highs (TTF +32% YoY to ~$16/MMBtu; JKM +45% YoY to ~$17.5/MMBtu).

The gap between supply and price is what we all see at the pump every day. It shows up just as clearly in the oil&gas industry’s earnings reports.

In the next few pages I’m trying to understand, by the real numbers, if companies with real Gulf exposure (Aramco, and to a lesser degree Shell/TotalEnergies via Qatar) took real production hits, compared to peers upstream operators with limited to zero exposure to the area.

We’ll see how, outside of the Gulf, most of the majors group captured the price upside with none of the volume downside, but even the exposed names still ended up ahead, because the price effect dominated the volume effect.

Subscribe now

Realized price vs. production vs. profit — per company

What “realized price” means, and why it’s the benchmark here: spot Brent/WTI is not what a company actually gets paid for a barrel. Realized price is the average price a company actually captured across its own production mix, after crude-quality discounts, regional differentials, hedging, and the oil-versus-gas split in its output. Companies can post very different realized prices in the same quarter simply because of the quality of crude they produce or how they’ve hedged it. It’s the number answering for “did this company get paid more?” independent of how much it actually pumped.

Let’s try to break down, where available, Upstream (E&P) segment earnings separately, isolating the price-and-production mechanism directly. Where it isn’t cleanly available, total (or adjusted) net income is used and flagged as such.

Eight of nine core-group companies show sharp Q1→Q2 acceleration; TotalEnergies is the lone exception (direct Qatar/Iraq/UAE shut-ins offsetting price gains); Aramco is flat but for a different reason (Q1 was already elevated on volume, not price).

What numbers actually show? Isolated from downstream noise, ExxonMobil and Chevron’s real upstream profit growth (+40% and +110%) is substantial.

Shell’s split is the cleanest illustration: Upstream (oil, largely undisrupted) grew steadily on price; Integrated Gas (where the Qatar damage actually sits) still grew despite real volume loss, because price more than compensated.

TotalEnergies is the one name that breaks the pattern, an exception that proves the rule: it’s also the company with the most direct, disclosed Gulf production loss (~15% of total output shut in across Qatar, Iraq, and the UAE), large enough that even a genuine price tailwind wasn’t enough to move the bottom line. Where exposure was real and large, the price effect didn’t dominate — it’s only across the rest of the group, largely insulated from the Gulf, that it did.

US vs. Europe

The US/Europe split looks clean at first glance, but the real divide is exposure.

Exxon and Chevron aren’t actually Middle East-light. Exxon has an estimated 15-20% of total output linked to the region: multi-billion-dollar LNG partnerships with QatarEnergy (including North Field expansions), a stake in Upper Zakum offshore UAE, and a 50% joint venture in Saudi Arabia’s 400,000 b/d SAMREF refinery. Chevron, considered “underweight” in the Middle East, holds crude assets in the Saudi-Kuwaiti Partitioned Neutral Zone and operates major Tamar and Leviathan gas fields off Israel, which saw temporary state-mandated production halts during the conflict.

Yet their Q1 production kept growing. Exxon posted a Guyana-driven record, Chevron added 388kb/d YoY through the incorporation of Hess. Apparently the real disruption suffered in Middle East assets wasn’t large enough; their reported Q1 weakness traces to disclosed derivative mark-to-market and inventory timing effects (both companies said explicitly these would reverse), and Q2 confirmed it.

On the EU side, Shell, BP, and TotalEnergies actually felt the production problem. TotalEnergies reported ~15% of total production (360,000 b/d) shut in across Qatar, Iraq, and UAE offshore. Shell’s Integrated Gas output was cut sharply on Qatar/Pearl GTL damage, down from 909k to 580-640k boe/d. BP’s Q2 upstream volumes were reduced by a mix of Middle East disruption and scheduled maintenance. Eni doesn’t fit either camp. Production actually grew 9% in Q1 (Libya/Nigeria-driven, unrelated to Hormuz), yet adjusted profit fell 11% due to tax rate and geographic profit mix, not price or Gulf exposure at all.

ConocoPhillips clarifies this pattern: a US company, but its Q1 weakness came from real Qatar downtime and a 6% YoY drop in realized price. COP has a significant exposure to Qatar production, via partnerships with QatarEnergy in the North Field and Ras Laffan LNG facilities.

The actual line isn’t US vs. Europe but Gulf production exposure. We’re talking about upstream giants who diversified their businesses across the planet, although in different percentages. Exxon and Chevron were largely unaffected; ConocoPhillips, Shell, BP, and TotalEnergies got hit on the production side.

The Chinese majors

China’s three listed majors tell a coherent story of their own.

• PetroChina: Q1 realized price fell 8.5% YoY (China’s crude mix and domestic pricing arrangements aren’t aligned with Brent/WTI); profit growth came entirely from downstream with refining +57.7% and gas sales +39.7%. There’s no evidence of upstream price capture in Q1, but Q2/H1 data are still pending (board meeting scheduled on Aug 28).

• Sinopec: Q1 net profit up 27-29% YoY (~$2.5-2.6bn), with refining EBIT surging on inventory gains and improved refining margins; upstream EBIT actually fell 4%. H1 net profit RMB 25.63bn (~$3.81bn), up 19.3% YoY — but that implies Q2 fell well below Q1. Sinopec explicitly disclosed sourcing half its crude from the Middle East, a real supply-chain disruption, offset by a shift to non-Middle East crude and a refining margin windfall (+381.5% YoY refining profit)

• CNOOC: the one Chinese major that partially fits the Western pattern with Q1 net income $5.73bn, up 7.1% YoY, production also up 8.6%. Having limited Middle East exposure (Canada is its largest overseas contributor provides an interesting parallel to Equinor and Oxy: real production growth and price gains moving together, with essentially no Hormuz direct risk.

None of these companies show a clean “stable production, rising price, exploding profit” mechanism that defines the majority of western bigs. They offer two different alternative stories instead, downstream/refining capture (PetroChina, Sinopec) and organic production growth compounding with price rather than substituting for it (CNOOC).

Aramco bookend

The fact that Aramco makes a lot of money isn’t a surprise. But it’s worth pausing at how much compared to peers.

Saudi Aramco has a market cap ranging around $1.6-1.8 trillion, roughly 3x Exxon’s and 5-6x Chevron’s. It’s Q2 net income (~$33.0bn) is more than twice ExxonMobil’s and nearly three times Chevron’s in the same quarter. And this is the same quarter in which Aramco’s own production fell by roughly 25%, during the largest single oil supply shock ever recorded, larger than the 1973 embargo or the First Gulf War.

The company that lost the most volume still made more money than almost anyone else in this piece.

Aramco’s realized price was flat at $76.9/bbl in Q1 2026 (barely above the year-earlier $76.3), then jumped 41% quarter-on-quarter and 62% year-on-year to $108.10/bbl in Q2 2026 as the Hormuz shock fully hit the market.

The forward-looking question

Almost every oil major sits somewhere on the same line: a production shock and a price spike that started at the Strait of Hormuz, and ended, directly or indirectly, on someone else’s bill, from gas price at the pump, cost for utilities or grocery prices.

The industry that pumps the oil and ships it overseas just posted some of its strongest quarters in years, largely without producing more of either.

The open question is what happens on the way back down. If crude and LNG settle toward pre-crisis levels through H2 (which, looking at futures market) doesn’t seem to be underway) does the relief travel back to consumers as quickly as the cost did? Or does margin, once captured, tend to stay captured a little longer than it took to arrive?

Marco

Data sources

• Global supply/LNG: IEA Oil Market Reports (Feb-Aug 2026), IEA Gas Market Report Q3-2026

• Company results: Q1/Q2 2026 earnings releases, 8-K/10-Q/6-K filings, and earnings call transcripts for ExxonMobil, Chevron, ConocoPhillips, Occidental, Shell, BP, TotalEnergies, Eni, Equinor, Saudi Aramco, PetroChina

• PetroChina Q2/H1 2026: pending, board meeting scheduled August 28, 2026

And my conviction is exactly why I’m not going all in.

Subscribe now

A coin flip tale

Take a coin that lands your way 60% of the time. Win, and you double your stake. Lose, and it’s gone. You have a genuine edge, and over time you’re supposed to make a lot of money from it.

Now imagine going all-in on every flip. You win the first and double your money. You win the second and double again. But eventually the coin lands against you, and when it does, you’re broke.

That happens even though your edge was real. A real edge doesn’t protect you from ruin if your bet sizing is wrong.

This is also known as the Kelly criterion (at its easiest version), introduced by John Kelly, a WWII pilot and PhD in physics, in 1956, to answer a specific question: what fraction of your capital should you allocate, given a real edge, to maximize returns?

Kelly’s formula for this kind of bet is simple:

f* = 2p − 1

where

p is your win probability

f* is the fraction of your capital to bet each round.

At 60%, the optimal bet is 20% of your capital per flip. Bet less, and you leave money on the table. Bet more, and volatility slowly destroys your capital even though your average outcome is positive. In finance this is called volatility decay.

The formula is also sensitive to your edge. Push the win rate higher, and the optimal bet size grows fast:

Small changes in how good your edge really is moves the optimal size dramatically. That matters more than it looks, because in real markets you never actually know your edge with this kind of precision, you estimate it, and estimates are wrong more often than models admit.

The lesson is: Betting too much on a winning bet can lose you money. And this is the part most investors never internalize.

The answer to the question “why aren’t you going all-in on your best idea?” isn’t really about caution. It’s just math.

Investing isn’t a coin flip

Unlike traditional portfolio approaches that focus on maximizing expected return or minimizing volatility, Kelly maximizes the geometric growth rate of capital over many repeated investment decisions.

Let’s make a more credible example applied to the investing world. Suppose you have an opportunity where:

• Probability of success (p) = 50%

• Probability of loss (q) = 50%

• Gain if successful (g) = +20%

• Loss if unsuccessful (l) = -10%

f* = (pg − ql) / gl

Substitute and you’ve f*=2.5. The growth-optimal Kelly allocation is 2.5x your entire capital. This might look aggressive but the payoff asymmetry is highly favorable.

The expected return for a single bet is:

ER= 0.5 x 0.2 - 0.5 x 0.1 = 5%

Let’s see what happens to expected portfolio outcomes using Kelly fractions different from the optimal one.

Notice how quickly drawdowns explode once leverage rises.This is why most professional investors use fractions (½ , ⅓, ¼) rather than “Full Kelly”.

Now let’s take a look at how capital grows as a function of the fraction of allocated capital for this exact example, using the expected log-growth function:

G(f) = 0.5 ln(1 + 0.2f) + 0.5 ln(1-0.1f)

At 5× leverage, despite having a positive-expectancy strategy, your long-term compound growth drops to zero. Beyond that point, more leverage actually destroys wealth.

This curve has the typical inverted-U shape. Compound growth increases as leverage rises up to the Kelly optimum. Beyond that point, expected arithmetic returns continue to increase, but the volatility drag increases even faster, causing long-term compound growth to decline. Push leverage far enough and the expected growth rate eventually becomes negative despite the strategy possessing a positive edge.

A Kelly investor isn’t asking whether an asset will go up. He’s asking:

“Given my conviction and assets’ volatility, what allocation maximizes long-term compound growth?”

From coin tosses to leverage

The Kelly criterion introduced a new concept about risk:not just losses, but overbetting: a strategy can have a positive expected return and still produce terrible long-term results if position size is too large.

Leverage does mathematically what over-betting does in the coin-flip game. A 2x leveraged position means a bet size twice what an unleveraged position represents, applied not once but on every single trading day. If your true Kelly-optimal exposure to an asset is, say, 1.5x, running 3x leverage means every day you’re betting past the peak of the curve — the same mechanism that broke the all-in coin-flipper, just quieter and slower.

Real markets don’t hand you a fixed win probability, though — returns are continuous, not a discrete coin flip. For an asset with continuously compounding returns, Kelly’s formula takes a different form:

f* = (μ − r) / σ²

where

μ is the asset’s expected return,

r is the risk-free rate

σ² is the variance of its returns.

This works exactly like the coin-flip case, maximize long-run growth rate rather than expected return, but here the “edge” is the risk premium (μ − r), and the penalty for over-betting comes directly from variance instead of a win/loss probability.

Even in this case a higher edge drives f* up; more volatility pushes it back down, since higher variance means bigger swings around your bet/investment size.

The natural question now isn’t about leverage use in the abstract, but what does the market’s own history say about the optimal leverage.

What does 30 years of data say?

I did what every engineer or mathematician would: run the numbers.

I used QQQ and SPY historical daily returns as market proxy, calculating the Kelly-optimal leverage and resulting CAGR across five different lookback windows (10, 15, 20, 25, and 30 years), applying a realistic cost-of-leverage drag.

QQQ (Nasdaq 100): the 10- and 15-year windows both put the optimal leverage around 3.5x, with peak CAGRs above 37-43%. That’s the “AI bull market” era, an unusually strong, historically low-volatility sample. But stretch the window to 20 years, and the optimal leverage falls to 2.7x. Push further and it keeps reducing. At 30 years, which captures the dot-com crash and 2008 it collapses to just 1.3x.

SPY (S&P 500): the same shape holds, at slightly lower magnitudes: 3.5x optimal at 10 years, down to 2.1x at 20 years, 1.7xat 25 and 1.9x at 30 years.

The curves themselves are self explaining: past the optimal point, every single line turns over and starts falling, sharply. At 30 years, QQQ’s curve goes negative by around 3.3-3.5x leverage. This is the volatility-decay cliff, made visible.

Why volatility decay is a real mechanism

Ok, nobody would keep a 3x leveraged bet for 30 years. This isn’t even the point.

Volatility decay is the arithmetic consequence of gains and losses not matching the geometric reality of compounding. A 50% loss requires a 100% gain just to break even, and leverage doesn’t just scale your returns, it scales this asymmetry too. Every daily rebalance multiplies both winning days and losing days by the applied leverage factor, and because losses hurt compounding more than equivalent gains help it, high daily volatility grinds down a leveraged position’s terminal value even when the underlying asset’s average daily return doesn’t. That’s why every curve above turns over rather than climbing forever: past the Kelly-optimal point, extra leverage adds more decay than it adds return.

Why practitioners bet less than “optimal”

In real world, almost every practitioner deliberately bet less than Kelly’s optimal.

First of all, risk management. The growth curve is asymmetric around the peak. Near f*, the curve is nearly flat, but overshoot it and you give up far more than you gained on the way up. Betting half of f* gives up only a limited fraction of peak growth while keeping you away from that steep downside slope.

Second, f* is only as good as its inputs, and the expected return (μ) driving it is just estimated. Your true edge might be half or double what you calculated and sizing at half your estimated optimal keeps you closer to the true optimum even when your estimate runs higher.

Third, the Kelly formula gives you the destination but says nothing about the ride. In practice, it’s known to produce peak-to-trough drawdowns of 50% or more even when the edge is real and sustained. Mathematically optimal doesn’t mean psychologically survivable for most portfolios and Half-Kelly reduces the magnitude of drawdowns while giving up comparatively little long-run growth.

Finally, Kelly’s formula assumes one isolated and repeated bet. Real portfolios hold multiple correlated positions, face margin calls and redemptions, and add capital over time, frictions the formula doesn’t model.

“Half Kelly” is more rule of thumb, but the cost of betting too much is worse than the cost of betting too little, so a smaller error is still the safer (and less costly) to make.

Why this matters for AI and tech right now

All the discussion above brings us back to the AI and tech trade.

If you size your conviction off the last 5-10 years of market returns, you’re estimating your edge from an unusually strong, low-volatility sample, precisely the Kelly overestimation trap the math warns against. Using a more honest starting point (like the last 20-30 years) says a much more modest expected return than the recent past would suggest.

That’s exactly why I’m not 100% in AI and tech, and why I don’t run anywhere near the “optimal” size the last decade alone would suggest. More defensive, low-correlation positions help keep volatility under control and stop capital from decaying in choppy periods and make drawdowns “survivable”.

There’s also a practical risk management case.. In real portfolios assets don’t work in isolation so conviction itself is a risk factor. The more certain you feel about a trade, the more likely you are to be anchored in the recent past, and the more a single wrong catalyst can do to your capital.

Position limits, correlation caps, and rebalancing rules exist precisely to avoid the case where your own confidence is the less reliable parameter. That’s a different failure mode than volatility decay, but it points the same way: cap single-theme exposure regardless of how good the theory looks, keep enough uncorrelated capital that no one call, however well-reasoned, can impair the whole portfolio, and treat position sizing as a discipline that holds even when, maybe especially when, you’re sure you’re right.

Being right isn’t enough. You have to survive long enough to prove it.

Data: QQQ and SPY daily total returns, Yahoo Finance, through May 2026. Leveraged simulations assume daily rebalancing and a 4.5% annualized cost of leverage on the borrowed portion.

In the immediate aftermath of the initial strikes, IRGC initiated retaliatory maritime operations, deployed naval mines and launched drone and missile strikes against commercial vessels. Tanker and commercial transit through the Strait dropped near zero immediately establishing the de facto blockade and traffic collapse, despite the IRGC formally declared the waterway closed to vessels bound to/from the US, Israel, and allied ports only on March 27.

Normally, roughly a fifth of global oil transits and LNG trade passes through Hormuz and its closure created an immediate spike in both crude and LNG prices, increases that have become a sort of new normality in the energy market.

Subscribe now

Six months on, the primary export data — and now, second-quarter earnings let us measure the actual damage magnitude, who suffered the most and who profited from it regardless.

Saudi Arabia — partial bypass, still a record collapse

Let’s take a look on official Saudi’s crude export in the first months of 2026.

For context, Saudi Arabia’s monthly average was ~6,048 kb/d in 2024 and ~6,437 kb/d in 2025; May’s 3,434 kb/d is nearly half of that.

Saudi Arabia production found a safety valve in its East-West Pipeline (Petroline), which runs east-west moving crude up to the refinery sites and export terminal of Yanbu, on the Red Sea. Aramco pushed it close to its full ~7 mb/d capacity, avoiding a total collapse of its exports. To be totally fair, the May drop can be partly attributed to an increase of internal crude demand.

Kuwait — near-total shutdown

2024 monthly average: ~1,168 kb/d. 2025 monthly average: ~1,326 kb/d. Kuwait has no meaningful bypass option around Hormuz, so the consequences on export were harder: same shock, different outcome.

UAE — the sharpest hit, the fastest (and most fragile) recovery

UAE crude output collapsed to a record low of 1,908 kb/d in March (the lowest since 1984), down from 3,390 kb/d in February, a sharper production hit compared to Saudi Arabia. Emirates production partially recovered to 2,021 kb/d in April and 2,110 kb/d in May, still roughly 40% below the January/February baseline (~3,383-3,390 kb/d).

Even so, the full first-quarter picture undersells the swing: ship-tracking data (Banchero Costa/LSEG) put Jan-Mar 2026 UAE exports at 33.3 million tonnes (2.71 mb/d average), down just 2.9% y/y — barely visible in a quarterly average, because two calm pre-war months offset March’s collapse.

The export-side data shows a fast bounce back, and by June the recovery had reached production itself: output jumped to 3,820 kb/d in June and 3,780 kb/d in July — above the pre-war baseline. The IEA’s Oil Market Report had already flagged this trend building in April, when UAE’s Fujairah bypass flow hit 1.6 mb/d, up 450 kb/d month-on-month. By June, Reuters reported total UAE crude exports at a record 3.8 mb/d, matching the production surge and exceeding the pre-war baseline of roughly 3.1-3.3 mb/d. Reuters credited the recovery not just to the Fujairah pipeline but to direct US Navy support escorting tankers through the strait following the June US-Iran de-escalation agreement, a geopolitical unlock combined with the infrastructure one. This means also the current situation is brittle since the current geopolitical “stability” could reverse if hostilities resume (we’re going deeper on the July crude price spike below).

Qatar LNG — the gas-side companion

Ras Laffan, the world’s largest LNG facility, has been put offline since the March 2 strikes and QatarEnergy declared force majeure for long-term contracts on March 25. An estimated 17% of Qatar’s LNG export capacity is currently out due to drone attacks, and repairs are projected to take three to five years.

Wood Mackenzie’s own timeline puts the full restart of non-damaged trains not before the end of August at the earliest while the permanently damaged trains will stay offline for years regardless.

The IEA’s Q3-2026 Gas Market Report puts combined Qatar + UAE LNG loadings down 35 bcm year-on-year between March and June, with global LNG production down 4% (8 bcm) y/y over the same window — about three-quarters offset by non-Gulf supply growth in North America and Africa.

LNG export data show a progressively deteriorating situation from February to May.

In May only two LNG shipments (per NGI) left Qatar, but following the June 17 US-Iran de-escalation, Qatar LNG exports attempted a broad restart, with vessels massing at Ras Laffan premises in mid June.

A July 30 report notes the Al Areesh was the first QatarEnergy tanker to exit the strait in nearly three weeks, meaning transit was effectively blocked again from roughly mid-July to July 29-30. This also lines up exactly with the Brent/JKM price spike we saw in mid July.

Force majeure notice has been extended into September and, as of July 1, QatarEnergy had already cancelled 21 cargoes scheduled April-September to European buyer Edison alone. To protect Asian deliveries, QatarEnergy purchased 33 US cargoes in 2026so far versus just 4 in all of 2025.

Subscribe now

The broad (and brittle) recovery attempt

Vortexa’s shipment tracking shows total Gulf oil exports (Saudi Arabia, UAE, Kuwait, Iraq, Iran) rising from 7.67 mb/d in May to 9.57 mb/d in June, a real, broad-based recovery attempt across the region. These numbers, compared to June 2025’s baseline of 16.52 mb/d, still show an export 42% below normal conditions.

Saudi Arabia (6.49→4.32 mb/d) remained well short of their year-ago levels, and even Iran, Iraq and Kuwait, despite posting the largest percentage gains month-on-month, were still exporting a fraction of their pre-war volumes. Only the UAE (3.13→3.70 mb/d) managed to increase its numbers YoY, driving the recovery with shipping record volumes by relying heavily on its bypass infrastructure (the Habshan–Fujairah pipeline) and drawing down stranded floating storage. Additionally, spot tenders issued by ADNOC flooded the Asian market with discounted Emirati crude.

The recovery was real, in other words, but partial: June looked like a relief after March-May’s collapse, and the broad-based nature of the rebound across every single Gulf producer is the clearest sign that this was a geopolitical event.

This makes June-July numbers unreliable long-term, with a drop back to March-to-May numbers in case the conflict escalates again. The recent LNG and crude price spikes we saw in late July are unfortunately pointing toward a complicated situation rather than a resolution.

Who actually captured the windfall: Aramco vs. the US majors

The export data shows who lost volume. Q2 2026 earnings show who made money anyway — and the mechanism differs sharply between Saudi Arabia’s national champion and its US-listed counterparts.

Aramco’s profit surge came despite losing nearly a quarter of its production volume, entirely on the back of the price spike Hormuz itself created. Crude realized price jumped to $108.10/bbl vs $66.70/bbl a year earlier, while hydrocarbon output fell to 9.5 mmboe/d from 12.6 mmboe/d

ExxonMobil and Chevron, by contrast, had no comparable export exposure to the strait — their production sits mostly outside the Gulf — so they captured the same price windfall with none of the volume penalty, and added record US output growth on top. Chevron’s (which also benefited from the Hess acquisition and integration) ~400% earnings jump is maybe the clearest illustration

ConocoPhillips average realized price jumped 36% to $62.33/BOE (from $45.77), while production actually fell 143 MBOED y/y (down 4% adjusted for portfolio changes) to 2,248 MBOED; the same pattern we saw for Exxon/Chevron with price doing the work alone. Ironically the company announced a 42% stake in a Kirkuk (Iraq) JV and a re-entry into Syria, expanding into the same region whose disruption drove its own profit surge.

Single-quarter scale is also worth noting for context: Aramco’s ~$33bn alone is larger than Exxon and Chevron’s combined ~$26.6bn — even in a quarter where Aramco was selling meaningfully fewer barrels.

Bottom line

Every Gulf producer has the same constraint and none had a real alternative. Transport infrastructure played a role but geopolitics is what really matters.

On infrastructure alone, the region is split: Kuwait, with no real by-pass capacity, saw exports collapse to near-zero. Saudi Arabia, with roughly 3-5 mb/d of spare pipeline capacity to the Red Sea, absorbed a deep but partial hit. The UAE took the sharpest production collapse of any Gulf producer in March but had somewhere to route barrels once conditions allowed.

The June recovery that started that month was broad and simultaneous, all the Gulf countries including Iran posted gains. That timing lines up with the US-Iran de-escalation agreement and US Navy escort activity. Even so, the region-wide total (9.57 mb/d) remained 42% below June 2025 levels, so the return to normal is still a long way to go.

And the recovery didn’t hold, running into a fresh disruption by mid-July with strait transit effectively blocked again for roughly two weeks, exactly when Brent and LNG both spiked back up. The pattern that emerges across the whole dataset is the same one, repeated at every round: de-escalation unlocks flows; flows remain vulnerable to any reversal.

Nothing observed here looks structurally fixed and damaged infrastructures will take years to be fully operative again. Every export “recovery” is contingent on a ceasefire holding.

Nevertheless none of this volatility stopped anyone from profiting. Saudi Aramco’s Q2 earnings report tells us a story of price appreciation created by the Hormuz blockade, with a net income increase despite losing nearly a quarter of its own volume. Outside the gulf area, US giants like ExxonMobil, Chevron, and ConocoPhillips captured the same price windfall with none of the volume penalty..

Data sources

• Saudi Arabia & Kuwait export data: JODI-Oil World Database (official monthly submissions)

• UAE export data: Kpler, Vortexa, Bloomberg ship-tracking estimates (via Reuters/Gulf News/The National reporting)

• Qatar LNG & global gas context: IEA Gas Market Report Q2/Q3 2026

• Earnings: Aramco Q2/H1 2026 interim report; ExxonMobil, Chevron, and ConocoPhillips Q2 2026 earnings releases and filings

• Gulf oil export recovery: Vortexa (via chart, June 2026)

The picture so far

SpaceX’s public debut in June, 2026 was, under any metric, historic. The IPO was reported as 4x oversubscribed, with total demand exceeding 250 billion vs 75 billions made available.

The final amount actually raised, which includes the underwriters overallotment option, was $85.7 billion; the second IPO in history, Saudi Aramco in 2019, raised $25.6 billion (up to 29.4B with the greenshoe exercised). Alibaba raised $21,8B in 2014 once starting quoting at the NYSE.

Four days later, on June 16, the stock hit an all-time high of $225.64, pushing the company to a valuation near $3 trillion.

Subscribe now

Then, a near-uninterrupted slide. On July 31st, SPCX closed at $108.37, more than 50% below that June peak, with more than $1.2 trillion in market capitalization erased in six weeks. Of course the widespread tech sell-off we saw in July didn’t help, but still we’re in front of one of the sharpest post-IPO derating events in recent memory, on a company whose long-term bull case (AI infrastructure, Starlink connectivity, reusable launch, the “lunar economy”) hasn’t basically changed.

So the natural question isn’t just “why is it down,” but why now, before the company has even reported a single quarter as a public company.

Two catalysts, one date

On August 4, SpaceX will finally report Q2 2026 results — its first earnings release ever as a listed company. That alone would be enough to create volatility. But there’s a less-discussed mechanic on top of it: the earnings report release will trigger the company’s first insider lock-up release.

Up to 20% of eligible insider and employee shares — an estimated 911.5 million shares — become tradeable two trading days after the Q2 report. At current prices, this means $109 billion in stock becoming available, about 1.45 times the size of the original IPO itself. And a second wave of lock-up expirations will follow later in the year.

Two events, same trigger. The market knows, and this is one of the main reasons why the selling pressure has been building for weeks.

Front-running dilution or fundamentals deterioration?

A large share of the current drawdown looks less like a repricing of SpaceX’s business and more like the anticipation of future increase in available float.

This is a well-documented pattern around lock-up expirations generally, but it’s unusually acute here for a few structural reasons:

1. The float was purposely kept low at listing, with less than 5% of total shares available at the IPO. That scarcity has been a big driver for the initial price spike, given an initial demand 4x larger— and it’s exactly what makes the coming unlock so consequential.

2. The unlock tranche is unusually large; a supply event 1.45x the size of the original IPO offering at once, is unusual this early.

• Timing removes ambiguity: the unlock is contractually tied to the earnings date and there’s no scenario where good Q2 numbers avoid the oversupply two days later. This certainty gives short-term traders and institutional holders a very clean reason to reduce exposure ahead of time to avoid being caught on the wrong side.

The decline could be attributed to a combination of profit-taking by investors who allocated shares following the IPO, a revaluation of a market capitalization that implied over $1.8 trillion, and the liquidation of overly concentrated bullish positions. There is no specific operational failure at the company.

What the numbers actually say

Let’s put the narrative aside; the earnings consensus tells an interesting story.

Revenue estimates are tightly clustered: consensus sits around $6.7–6.9 billion for Q2, driven primarily by growth in the Connectivity (Starlink) segment, with Visible Alpha modeling a 35.9% operating margin in that segment offsetting losses in Space and AI.

EPS estimates are anything but tight. The consensus loss figure is around -$0.26 to -$0.28 per share — but the underlying analyst range spans from -$1.26 to +$0.33. That’s an extraordinarily wide dispersion for a “consensus” number, and it reflects a simple fact: with no reporting history as a public company, analysts are still guessing at SpaceX’s real cost structure across three very different businesses (launch, connectivity, and AI infrastructure).

Analysts remain generally bullish. SPCX has a strong buy consensus (23 buys, 6 holds, 1 sell) with an average price target suggesting a potential upside of 100% from current prices.

Morgan Stanley’s Adam Jonas has an Overweight rating and a $300 price target; while earnings could provide clarity on the growth trajectory of AI and Starlink, he points to two specific risks that could pressure the stock: an increase in investment guidance above the approximately $50 billion expected for 2026, or the announcement of a financing round by the end of the year. Goldman Sachs’ Eric Sheridan is equally optimistic, citing SpaceX’s position in launches, connectivity, and AI infrastructure.

HSBC is an exception, maintaining a Hold rating, suggesting that the current price already reflects much of the options stemming from future projects such as space data centers and the lunar economy, leaving very limited upside.

That split is significant. This is not a market turned bearish on SpaceX’s business, but uncertainty about near-term and reporting risk, added to a valuation that was already priced for a lot of stuff to go in the right direction.

Why conventional valuation struggles here

Looking at the range of fair value estimates here makes little to no sense at this point, considering how thin the underlying data actually is; the wide dispersion in target prices is more lack of reliable, certified data than just analysts disagreeing.

There are no audited public financials yet. August 4 is SpaceX’s first earnings report as a listed company. Every number in circulation today is a modeled approximation from private fundraising data and management commentary. No reported balance sheet, no verified segment breakout. So far, every “consensus” figure is mostly guesswork.

Standard valuation tools assume a maturity SpaceX doesn’t have. A price/sales multiple, a DCF, or a sum-of-the-parts model all assume you can reasonably forecast multi-year cash flows per business line. That holds for Starlink, which has subscribers, pricing, and a comprehensible model. It holds far less for the rest: launch, the AI and digital segments mostly not existing revenue lines, but some models are betting on infrastructure still being defined, and the a lunar/Mars “economy” with no timeline and no precedent to model against.

Let’s be honest: a meaningful share of the valuation is really a bet on execution and vision. The bull case is a bet that Elon Musk’s stated ambitions will get built, somehow, on a reasonable timeline.

SpaceX has a real track record of executing on things that looked implausible a decade earlier — but any valuation models that produce $300+ price targets are, to a significant degree, discounting a narrative, not a financial plan.

Conventional tools aren’t built to price a visionary founder’s multi-decade roadmap; applying them anyway creates the appearance of rigor around numbers that are closer to scenario-weighted belief.

This doesn’t mean the estimates below are worthless, but it’s worth uaing the resulting numbers with a grain of salt.

The incoming Q2 report will be the first time any of these models can be checked against an actual, audited number.

Is there a “fair value” here at all?

Given the scale of the drawdown, the obvious question is whether SPCX is now cheap, expensive, or simply unknowable, and the data available doesn’t support a confident answer.

Mapping the range of the available independent evaluation doesn’t help sp much.

Morningstar puts fair value around $62 a share, and at least one independent equity note has argued for a target as low as $21–28, built around the idea that the IPO scarcity premium is now unwinding alongside multiple contraction as the true float expands.

At the bullish end, Raymond James carries an (unrealistic) $800 target, and the broader sell-side consensus averages out to about $237, implying more than 100% upside.

Independent sum-of-the-parts models land all over this range too. One model built ahead of the IPO estimated a median fair value near $1.25 trillion in total equity value — roughly 30% below the IPO price and broadly consistent with where the stock trades today. Another sum-of-the-parts DCF (valuing the company across business lines) , produces a base case near $166 a share with upside scenarios well above $300, built on the assumption that SpaceX’s vertically integrated businesses compound in ways no single-segment peer can replicate.

Nobody disputes that Starlink is highly profitable and growing, or that launch remains dominant. This dispersion is more a disagreement about time horizon and multiple and whether SpaceX should be priced like an infrastructure company with real, current cash flows, or like an early-stage growth company.

The data isn’t sufficient to assign a fair value with any confidence, but it is sufficient to say the market has already moved a long way toward pricing the more conservative scenario.

What I’m watching on August 4

A few things will matter more than headlines and EPS for the stock’s mid-term direction:

• Capex guidance. anything above the $50 billion mark for 2026 is an alarm bell and could spook the market regardless of the revenue print, given projections that capex could rise toward $118 billion by FY2028.

• Starlink margins and guidance currently the profitability engine offsetting losses in Space and AI.

• Any commentary on financing plans. A capital raise before year-end would compound the dilution narrative.

• Management tone on the AI and lunar economy narrative, where much of the long-term valuation premium sits.

• Whether the stock stabilizes once the lock-up tranche actually clears — a classic “sell the rumor, buy the news” setup, if the anticipatory selling has already done most of the work.

Where this leaves investors

The thesis that the market is discounting the lock-up and earnings risk in advance is well supported by the data: the unlock is mechanically tied to the earnings date, the tranche size is unusually large relative to the float, and the sell-off has been gradual and broad-based rather than triggered by a single piece of bad news. At the same time, it would be a mistake to treat this as the only factor.

Starship’s operational setbacks, a still-demanding valuation multiple even after a 50% decline, and genuinely wide uncertainty around the company’s near-term cost structure are all real, independent contributors.

Marco

This article is for informational purposes only and does not constitute investment advice. Do your own research before making investment decisions.

Quarantuno giorni.

Tanto è passato dal 19 giugno, quando Donald Trump raccontò a Daniele Compatangelo - e al mondo intero - che Giorgia Meloni lo aveva implorato per una foto al G7 di Évian e che lui gliel’aveva concessa perché gli faceva pena, allo scoppio degli arrivi di migranti a Ceuta.

Tra 27 capi di governo europei, sapete chi ha preso per primo le sue difese? Esatto. Pedro Sánchez.

Sono passati 41 giorni. Meno di un mese e mezzo.

Quattro ore.

Tanto ci abbiamo messo per trasformare la crisi di Ceuta in un’occasione di propaganda. Migliaia di persone disperate che cercano di raggiungere una riva a nuoto (e decine di morti), e tu cosa ci vedi? Una possibilità di accaparrarti qualche voto facendo leva sulla paura dello straniero.

Alla prima occasione, in una giornata di caos - e di morti - sulle coste spagnole in Africa, la prima a saltargli addosso è stata proprio lei.

Sospendere Schengen con la Spagna, subito!

Una montagna di bugie, strafalcioni e tante, tante porcherie. Dalla prima all’ultima riga.

Subscribe now

Cosa è successo in cifre

In pochi giorni Ceuta è stata travolta da un’ondata migratoria senza precedenti: circa 60.000 persone provenienti dal Marocco hanno raggiunto, per lo più a nuoto, il territorio europeo, mettendo sotto pressione le strutture di accoglienza e riaprendo il dibattito sulle frontiere dell’UE.

A emergenza quasi conclusa, oltre 48.000 migranti risultano rientrati in Marocco, ma resta un costo umano intollerabile: il bilancio delle vittime è salito rapidamente nelle ultime ore a oltre sessanta morti accertati durante la traversata, e purtroppo il numero non è ancora definitivo. E restano anche molti interrogativi destinati a durare a lungo: sulla capacità dell’Unione Europea di gestire flussi improvvisi, sul ruolo dei Paesi di transito come il Marocco e, soprattutto, sulle cause profonde che continuano a spingere decine di migliaia di persone a rischiare la vita per attraversare il Mediterraneo.

Lo scontro politico sulla gestione dei flussi migratori e sulla sicurezza delle frontiere è ovviamente esploso, in Spagna e in Europa; noi non potevamo certo perdere l’occasione.

Ceuta e Melilla

Ceuta e Melilla sono due enclavi spagnole in territorio marocchino, vale a dire piccole aree sotto la sovranità spagnola ma interamente circondate dal Marocco, senza alcun collegamento terrestre con la madrepatria.

La loro origine risale all’espansione iberica nel Nord Africa tra il XV e il XVI secolo: Ceuta passò sotto controllo portoghese nel 1415 e poi spagnolo nel 1668, mentre Melilla fu occupata dalla Corona di Castiglia nel 1497. Dopo l’indipendenza del Marocco nel 1956, Rabat ne ha più volte rivendicato la sovranità, ma entrambe restano parte integrante della Spagna e, quindi, dell’Unione Europea. Per la loro posizione geografica rappresentano quindi uno dei pochi confini terrestri tra Africa ed Europa e da decenni sono uno dei principali punti di pressione dei flussi migratori verso il continente.

Entrambe hanno un regime speciale, proprio per via della loro posizione, che prevede doppi controlli al confine esterno: uno all’ingresso dal Marocco e un altro ancora in uscita verso il continente europeo.

Chi parte da lì verso la Spagna peninsulare o verso qualunque altro Paese dell’area viene fermato e identificato, documenti alla mano, come se uscisse da fuori Schengen.

Blocchiamo Shengen!

Partiamo da un presupposto: in questa vicenda, che si decide sulla frontiera EUROPEA in Africa, la Spagna è la vittima, bersaglio di un ricatto.

Ceuta, dunque. Molti, politici inclusi, non l’hanno mai sentita nominare o non hanno idea di dove si trovi. Io, che da bambino avevo, tra le altre, una passione per la Geografia, ricordo ancora quando andavo a cercare queste “anomalie” geografiche su quello splendido atlante che mio padre mi regalò quando avevo ancora 8 anni (sono abbastanza vecchio e no, Wikipedia non esisteva ancora, usavamo le enciclopedie di carta...)

Come abbiamo visto, Ceuta è fisicamente in Africa. Un piccolo fazzoletto di 18 chilometri quadrati, territorio spagnolo, dalla parte opposta dello Stretto di Gibilterra. Per uscire ed arrivare in Europa (continentale) serve una barca o un aereo. E, in accordo al trattato di Shenghen, bisogna passare i controlli di frontiera come se si provenisse da un paese extra-UE.

Per arrivare in Italia, le persone che hanno attraversato, clandestinamente, quella frontiera, dovrebbero prima passare i controlli d’identità per arrivare in Spagna, attraversare tutta la penisola iberica, valicare i Pirenei, attraversare tutta la Francia e presentarsi a Ventimiglia.

Sapete quali sarebbero gli effetti di una eventuale sospensione del trattato di Schengen con la Spagna su questa vicenda? Esatto: zero.

Ma l’elettore medio, che fino a due giorni fa forse non aveva mai sentito nominare Ceuta, ora è convinto che stia a due passi dal confine e che quella gente sarà sotto casa sua entro mercoledì prossimo.

No. Ceuta non è esattamente dietro Bordighera, e nemmeno a ridosso dei Pirenei. La porta che volevano sbarrare è sbarrata da 35 anni. La frontiera che il ministro degli Esteri della Repubblica Italiana ha chiesto di chiudere era già chiusa.

E infatti l’hanno chiusa lo stesso.

L’Italia ha effettivamente sospeso Schengen con la Spagna, parlando di “sicurezza nazionale tutelata”, “una scelta necessaria, prevista dai trattati”. Caro Ministro, il fatto che sia prevista dai trattati non la rende meno inutile.

Un atto simbolico, si, che colpisce chi viaggia legalmente tra Italia e Spagna (turisti, lavoratori, merci) e lascia del tutto indifferente il fenomeno migratorio, perché quel fenomeno si gioca su una frontiera esterna a migliaia di chilometri di distanza, già presidiata da controlli doppi, che nulla ha a che vedere con la libera circolazione interna a Schengen.

È la fotografia perfetta della differenza tra fare qualcosa e fare finta di fare qualcosa: un atto concreto, con conseguenze concrete sulla vita di chi attraversa regolarmente quella frontiera, per rispondere a un problema che sta da un’altra parte e che quell’atto non tocca minimamente.

Demagogia allo stato puro, ma certificata in Gazzetta Ufficiale.

Ma al nostro politico destropiteco medio, il quale non sarebbe in grado di trovare Ceuta su una cartina nemmeno se ci fosse scritto sopra a caratteri cubitali e cerchiato di rosso, questo non importa.

Piatto ricco, mi ci ficco. Una montagna di bugie, strafalcioni e tante, tante porcherie. Dalla prima all’ultima riga.

La colpa è del governo spagnolo che ha dato la cittadinanza a 500.000 irregolari

La Spagna ha approvato per decreto la regolarizzazione, tramite permessi di soggiorno e di lavoro di durata annuale, di persone che, clandestinamente, vivono e lavorano in Spagna da anni, senza precedenti penali, da prima del 31 dicembre 2025. Permettendogli di uscire dall’illegalità ed ottenere regolari contratti di lavoro e diritti. E anche pagare le tasse.

E questa norma riguarda persone che stanno in Spagna da anni, provenienti in gran parte dall’America Latina.

Esatto. Il capo della nostra diplomazia ha scambiato un permesso di soggiorno per una cittadinanza e un passaporto europeo.

Collegare la situazione di Ceuta alla regolarizzazione voluta dal governo Sanchez significa non avere ben chiaro il concetto di cittadinanza. O raccontare consapevolmente balle.

Però su qualcosa devo dare ragione a Tajani: questa sanatoria voluta dalla Spagna è una porcata. 500 mila irregolari, clandestini.

Scherziamo? Noi abbiamo fatto di molto meglio: 2002, legge Bossi-Fini, oltre 647mila persone regolarizzate.

Toma castaña.

Ma la Spagna rappresenta comunque un problema per gli ingressi in Europa

Certo. Ingressi irregolari registrati nel 2025 (dati Frontex):

Mediterraneo centrale (Italia, Malta): 66.328

Mediterraneo orientale (Grecia, Cipro): 51.399

Africa occidentale (Canarie): 17.280

Balcani occidentali: 12.525

Mediterraneo occidentale (Spagna): in aumento del 14% sul 2024, circa 30.000

Secondo i principi richiamati dai nostri, il primo Paese da buttare fuori da Schengen sarebbe l’Italia.

Ma il principio di solidarietà?

Da quattro anni sentiamo sempre la stessa storia: “Non possiamo essere lasciati soli”, “l’italia può affrontare il problema dell’immigrazione da sola”.

Cooperazione. Solidarietà europea. Redistribuzione. Una crisi globale da affrontare insieme.

Tutto svanito ora che la pressione si è spostata 1500 km ad ovest di Lampedusa

Il ruolo del Marocco

Il Marocco ha una unica arma diplomatica: i corpi dei propri giovani. E apre le frontiere per fare pressioni o per rappresaglia verso decisioni che non condivide (vedi gli accordi tra Spagna e l’Algeria della scorsa settimana). Ottenuto ciò che vuole richiude il rubinetto umano.

Non c’entra nulla la politica migratoria della Spagna. è un ricatto costruito sulle persone, e il nostro governo italiano punta il dito contro chi quel ricatto lo subisce.

Soldi e legittimazione in cambio di frontiere presidiate. Perchè chi ha appaltato i propri confini alla guardia costiera libica si guarda bene dallo spiegare al mondo come funziona il ricatto migratorio.

A volte mi chiedo come sia possibile che politici, perdonatemi il termine, “di spicco” possano uscirsene con dichiarazioni piene di strafalcioni, inesattezze, errori. È possibile che un leader politico non abbia uno staff in grado di verificare le sue affermazioni prima di fare dichiarazioni pubbliche? Basta una breve ricerchina su Google o magari chiedere a ChatGPT (nemmeno serve una licenza premium).

No, il punto non è fornire dati esatti, non è minimamente importante. Sapevano (almeno spero) dove si trova Ceuta e che quella frontiera è controllata e soggetta a vincoli specifici. Conoscevano benissimo che il numero di migranti irregolari entrati in Italia è ben superiore a quello spagnolo. E sapevano anche che gran parte della responsabilità di questa situazione ricade sul governo del Marocco.

Ma non importa la verità. Perché una bugia raccontata a caldo, magari condita con una bella dose di paura ed allarmismo, arriva subito a milioni di persone; una smentita o un’analisi approfondita e verificata arriva dopo giorni a poche migliaia di persone. E soprattutto perché una crisi umanitaria può valere mezzo punto nei sondaggi. Anche a costo di trasformarla, come è successo, in un atto concreto e inutile - basta che sia ben pubblicizzata sui giornali.

Che miseria senza fondo.

Marco

Subscribe now

1. The bold idea: every parking lot is a power plant

When we talk about the energy transition, our minds jump to hydroelectric dams, large solar plants on open land, wind turbines, battery farms.

France has taken a different starting point. Instead of asking where to build new solar capacity, the country is trying to use surfaces already existing, doing nothing but absorbing heat. Apparently obvious — not so easy to apply.

Under the loi APER (Loi relative à l’Accélération de la Production d’Énergies Renouvelables), France now requires existing outdoor parking lots larger than 1,500 m² to be covered with solar canopies over at least 50% of their surface, with compliance deadlines staggered between 2026 and 2028, and financial penalties for operators who don’t comply.

It’s a small regulatory tweak with an outsized logic: no new land use, no permitting fights over greenfield sites, no landscape controversy — just putting panels above cars that are already parked there.

2. Why this matters: the scale of the numbers

The French government’s own estimate is that the mandate could unlock roughly 9–11 GW of installed photovoltaic capacity. What does it mean for France’s energy landscape:

• Vs. nuclear: the nominal capacity of modern EPR-class reactors is roughly 1.4–1.6 GW. APER application could bring 7–8 “new” reactors.

• Vs. utility-scale solar farms: a large-scale solar farm typically ranges from 5 MW to 50 MW, while utility scale can reach 100 MW to 500 MW, spread across hundreds of football fields of space.

• Vs. households supplied: this range of capacity could cover the electricity needs of several million households — though this figure deserves a caveat, addressed below.

Figure 1 — Parking-lot potential capacity set against France’s existing nuclear and solar fleets.

Capacity isn’t production. For a solar installation, the nominal capacity is the maximum output under ideal conditions. What actually reaches the grid is measured in GWh/year, and depends on multiple parameters like solar radiation fluctuations over the year, orientation and tilt, inverter losses, ageing…

France’s existing solar fleet gives us a real-world benchmark: roughly 21 GW of installed capacity produced about 23 TWh in 2024 — an implied capacity factor of around 12–13%. Canopy installations sit in a similar range, maybe in the lower tier, since parking-lot geometry (elevation, orientation, partial shading from surrounding structures) is rarely as favorable as a purpose-built ground-mount farm.

Applying a 12–13% capacity factor to the 9–11 GW parking-lot estimate gives a plausible output of roughly 9.5–12.5 TWh per year. Two comparisons put that in context:

• French nuclear output: France’s nuclear fleet (61.4 GW installed) produced 373 TWh in 2025. The parking-lot program, even at the high end, would add the equivalent of about 2.5–3.4% of current nuclear output — real, but not remotely a substitute for the nuclear baseload.

• French solar output: in 2024 solar plants in France gave ~23 TWh of total generation; parking lots alone could add the equivalent of roughly 40–55% on top of France’s entire existing solar output — simply from previously unused surface.

Figure 2 — Capacity vs. actual annual production (TWh) across the same three references.

Parking-lot solar is a marginal addition next to nuclear, but a major structural addition to France’s solar base specifically — this is fundamentally a solar-expansion policy to facilitate the energy transition, not a nuclear-replacement one.

3. Europe is moving in the same direction

France is not working in isolation, though the mechanism varies sharply by country.

Germany has taken the regional route: five Länder (Baden-Württemberg, North Rhine-Westphalia, Rhineland-Palatinate, Lower Saxony, Schleswig-Holstein, later joined by Hesse) each passed their own parking-lot solar obligations since 2022–2023, with thresholds ranging from 35 to 100 spaces. The key difference from France is that every German rule applies only to new-build parking lots, leaving the existing stock untouched. One estimate puts Germany’s technical potential from existing lots alone at up to 24.6 GW, roughly France’s scale, but with no law currently reaching it.

Spain has no equivalent mandate at all, yet has built out solar aggressively through strong resource quality and self-consumption incentives — mostly via utility-scale ground-mount projects rather than retrofitting built infrastructure. The Netherlands and Austria are pushing in the same general direction through subsidies and planning frameworks rather than binding law.

Europe increasingly treats parking areas as wasted, already-sealed land — well suited to decentralized generation with no new land take; countries diverge on mechanism: compulsion versus incentive, and — as with Germany — future stock versus existing stock. That last distinction is exactly where Italy’s story begins.

Figure 3 — How five European approaches to parking-lot solar compare on scope and bindingness.

4. And then there’s Italy

Italy receives more annual solar irradiation than France — its south and islands rank among the sunniest parts of the continent. Yet no comparable nationwide parking-lot solar obligation exists today.

The contrast is striking, and it isn’t about resources. France is legislating deployment. Italy is largely relying on individual investment decisions.

5. The paradox of Italian solar

Italy’s paradox is not a lack of ambition on solar generally — the country has added significant utility-scale and rooftop PV capacity in recent years, helped by tax incentives and a genuinely favorable resource base. The paradox is specifically about this surface type.

The numbers make that clear, and they now line up year-for-year against France’s own data. Italy’s cumulative installed solar capacity reached 37 GW at the end of 2024 and 43.5 GW at the end of 2025 (source: Terna/IEA-PVPS), generating 36 TWh in 2024 and a record 44 TWh in 2025. France, over the same two years, went from 24.5 GW to 30.4 GW of installed capacity (source: RTE), with generation rising from about 24.8 TWh in 2024 to 32.9 TWh in 2025. In both years, on both metrics, Italy’s solar fleet is larger than France’s. National targets under the PNIEC call for 80 GW of capacity and 100 TWh of annual production by 2030.

Figure 4 — Italy’s solar fleet vs. France’s, matched year-for-year (2024 and 2025). Sources: Terna/IEA-PVPS (Italy), RTE (France).

This is what makes the missing parking-lot mandate so pointed. Italy isn’t behind on solar as a technology choice — it’s ahead. What it lacks is a structural policy pointed specifically at the country’s vast stock of unused parking surface, the same surface France decided, however imperfectly, to legislate against.

In a 2025 report, Legambiente explicitly proposed making solar installation mandatory on parking lots over 1,500 m² — modeled directly on what France has done. Separately, a bill was introduced in April 2024, offering tax incentives for solar canopies and a simplified permitting process focused specifically on open-air parking lots. Neither has become binding law. Italy still has no legislation comparable to France’s.

Meanwhile, the private sector has quietly demonstrated the model works at scale — just not through Italian regulation. Carrefour has an agreement to install solar canopies across 350 hypermarket and supermarket parking lots by 2027, covering 180,000 parking spaces for a total capacity of 350 MW, expected to produce around 450 GWh per year — basically proceeding under French law.

So the resource is there, the private-sector proof of concept is there, and even the civil-society and legislative proposals exist on paper. What’s missing is the binding mechanism that turns “could” into “will.”

6. Real investment or political theatre?

Before holding France up as the model Italy should copy, it’s worth being honest about how solid that model actually is.

In May 2025, France’s own National Assembly passed an amendment that relaxes the mandate for solar installations on parking lots larger than 1,500 square meters, allowing owners to comply with either solar canopies, shade-providing trees, or a combination of both rather than solar panels specifically. This implies that there is no longer any formal obligation to solarize these parking lots, since simply planting trees will be sufficient to comply with the law. A lawyer following the file separately noted that the legal framework for solarizing parking lots has shifted repeatedly since 2021 — through amendments, exemptions, and delays — creating ongoing uncertainty for the industry.

This matters for how the 9–11 GW figure should be read. It was a ceiling calculated under the original, stricter version of the law. Every parking lot that opts for trees instead of panels is a data point moving actual outcomes away from that ceiling. France set an ambitious structural target, then partially walked it back under pressure, but the framework is still there.

That doesn’t make the policy meaningless. Even a diluted mandate reaching into the existing parking lots is structurally different from Italy’s approach, which currently reaches into none of it.

7. The missed opportunity sitting in plain sight

Strip away the policy mechanics and the underlying fact doesn’t change: Europe is full of potential solar panel installation areas with no need to consume additional soil. The technology to turn them into decentralized power generation is neither new nor expensive relative to utility-scale alternatives. No new land take, no agricultural trade-off, no landscape objection that a solar farm on a hillside inevitably attracts.

France’s experience is itself a useful lesson for whoever legislates next, Italy included. A mandate written without accounting for the operators who will look for the cheapest compliant option is a mandate that will get watered down. The lesson is: legislate carefully, with enforcement and incentives that don’t leave a cheaper non-solar loophole sitting right next to the solar one.

For Italy specifically, the ingredients are already on the table: a favorable solar resource, a private sector that has already shown the economics work, and civil-society and legislative proposals that mirror the French model almost line for line. What’s missing isn’t proof of concept. It’s the decision to do so — and, if France’s experience is any guide, to legislate in a way that actually works.

Marco

L’Offerta Pubblica di Acquisto e Scambio (OPAS) promossa da Poste Italiane su Telecom Italia (TIM) ha ridisegnato i confini del settore delle telecomunicazioni e dei servizi digitali in Italia. A poche settimane dalla chiusura del periodo di adesione, il quadro per l’investitore si è semplificato attorno a due punti:

● Lo spread di arbitraggio tra TIM e il valore teorico dell’offerta è ormai azzerato: comprare TIM oggi per aderire all’OPAS non offre più un rendimento significativo rispetto al rischio assunto.

● Tutta la creazione di valore residua si è spostata su Poste Italiane, sostenuta dalle sinergie industriali dell’operazione: nelle ultime settimane BofA, Barclays, Intermonte, Banca Akros e Websim hanno alzato in sequenza il target price sul titolo, con upside stimati tra il 9% e il 30%.

In conclusione: TIM, come posizione a sé stante, ha esaurito il suo potenziale di rendimento da arbitraggio. La tesi di investimento sull’operazione oggi può reggersi solo su Poste Italiane e sul valore che l’incorporazione di TIM porta al suo business.

Subscribe now

2. Cosa prevede l’OPAS

Meccanismo dell’offerta

Per ogni azione TIM portata in adesione, Poste Italiane offre agli azionisti attuali:

● €1,67 in contanti (componente fissa).

● 0,218 azioni Poste Italiane di nuova emissione/concambio (componente variabile).

Ai prezzi attuali, il valore dell’offerta dipende per oltre il 78% dall’andamento di Poste in Borsa.

Condizioni principali e tempistiche

● Soglia minima: 66,67% delle azioni con diritto di voto di TIM.

● Integrazione e delisting: è prevista la fusione societaria post-operazione con conseguente delisting di TIM da Piazza Affari.

● Periodo di adesione: aperto fino all’11 settembre 2026.

● Data di regolamento: prevista per il 18 settembre 2026.

3. Valore reale dell’offerta (analisi di sensitività)

La formula fondamentale per determinare il valore teorico dell’OPAS è:

Valore OPAS = 1,67 + 0,218 × Prezzo Poste

Poste Italiane scambia attualmente attorno ai 27€, ai minimi degli ultimi 30 giorni rispetto al picco di 29,17€ toccato il 2 luglio 2026. Rispetto al primo trimestre 2026 (quando quotava tra 20€ e 23€), il titolo ha registrato un rally sostenuto dall’annuncio dell’operazione.

Nell’ultima settimana TIM ha scambiato tra €7,53 e €7,67, sostanzialmente in perfetta linea con il valore teorico dell’OPAS al prezzo attuale di Poste (€7,556). Questo conferma che il mercato ha già considerato l’operazione conclusa: l’arbitraggio sul prezzo TIM/Poste è di fatto bloccato.

4. Perché BofA (e non solo) è diventata più positiva su Poste?

Nell’arco di pochi giorni tra fine luglio 2026, diverse case d’affari hanno alzato in sequenza il target price su Poste Italiane, un movimento che segnala un consenso crescente sulla creazione di valore dell’operazione TIM. Il quadro che emerge dai report è molto più articolato del semplice “sconto convertito in premio”.

La valutazione BofA: due scenari pesati alla pari

BofA ha portato il target price su Poste a 32€ per azione (da 28,8€), confermando il giudizio Buy, con un potenziale di rialzo di circa il 15% rispetto ai prezzi di fine luglio. Il punto centrale del report è metodologico: BofA costruisce il target come media pesata al 50% di due scenari valutati con metodo Sum-of-the-Parts (SoTP):

● Poste standalone, senza l’operazione TIM: fair value stimato a 28,3€.

● Poste con il 100% di TIM incorporato: fair value stimato a 35,8€, sostenuto principalmente dalle sinergie in fase di implementazione, dai maggiori utili di TIM e dai multipli più elevati tipici del settore telecomunicazioni rispetto ai servizi postali/finanziari.

Il messaggio implicito è che anche nello scenario più prudente (l’operazione non si concludesse) Poste avrebbe comunque un fair value superiore ai prezzi attuali; lo scenario di integrazione piena aggiunge però del valore aggiuntivo rispetto alla semplice somma delle parti..

Le determinanti della revisione al rialzo

● Sinergie di costo ante imposte: circa 0,5 miliardi di euro l’anno a regime, con alcune stime di mercato che arrivano fino a 0,7 miliardi includendo anche il rinnovo dello spettro, la condivisione della RAN e l’ottimizzazione degli accordi sulle torri di trasmissione.

● Sinergie di ricavo: oltre 0,2 miliardi di euro, derivanti soprattutto da cross-selling e riduzione del churn sulla base clienti combinata.

● Beneficio fiscale (DTA): l’utilizzo delle imposte differite attive di TIM permette di realizzare gran parte di queste sinergie con un carico fiscale quasi nullo, per un beneficio netto stimato attorno a 0,7 miliardi di euro.

● Impatto sull’EPS: l’operazione ha il potenziale di incrementare l’utile per azione di Poste di circa il 25% entro il 2029, secondo le stime BofA.

● Consolidamento del mercato telco italiano: l’Italia resta uno dei mercati europei meno consolidati, con quattro operatori e un ARPU (ricavo medio per utente) tra i più bassi dell’Unione Europea — quasi la metà rispetto ai mercati più concentrati. Un passaggio da quattro a tre operatori (con Iliad spesso indicata come possibile protagonista di un ulteriore consolidamento) aggiungerebbe, secondo le stime BofA, circa 0,2 miliardi di euro di utile netto per ogni euro di incremento dell’ARPU.

● Infrastruttura digitale e data center: BofA attribuisce un peso strategico crescente alla capacità di TIM in ambito data center (125 MW installati, 16 strutture attive più una in realizzazione), che si combina con la rete fisica capillare di Poste per posizionare il gruppo come infrastruttura portante della digitalizzazione italiana, in un mercato — quello dei data center italiani — atteso più che raddoppiare gli investimenti entro il 2027.

Il coro di revisioni al rialzo delle altre case d’affari

BofA non è un caso isolato. Nell’arco di pochi giorni si è formato un consenso piuttosto ampio sul potenziale dell’operazione:

Alcuni dettagli utili a leggere questi numeri: Banca Akros stima un valore attuale netto delle sinergie post-closing (atteso nel 2027) pari a 3,6 miliardi di euro, a fronte di un premio pagato agli azionisti di minoranza TIM di circa 0,8 miliardi — un differenziale che si traduce direttamente in creazione di valore per gli azionisti Poste. Kepler Cheuvreux, pur mantenendo un giudizio Hold, quantifica in circa 2,10€ per azione l’extra-valore derivante dall’integrazione con TIM rispetto alla valutazione standalone di Poste (che stima a 27,40€). Intermonte, dal canto suo, valuta anche leve di creazione di valore a medio termine non ancora pienamente incorporate nel prezzo — tra cui il risanamento del mercato telco e la valorizzazione delle torri — per un potenziale aggiuntivo di 5,1€ per azione, di cui solo il 30% è già riflesso nel target price attuale.

Il quadro complessivo è quindi quello di un consenso convergente, costruito con metodologie diverse (SoTP, DCF, comparables), che colloca il fair value di Poste in un intervallo ampio ma coerente tra 29,50€ e 35,20€: un premio significativo rispetto ai prezzi di mercato di fine luglio 2026.

5. TIM come investimento standalone

Domanda: se l’OPAS non esistesse, compreresti TIM oggi?

Il cambio strutturale post-NetCo

A luglio 2024 TIM ha completato la cessione della rete fissa (NetCo) al fondo KKR per 22 miliardi di euro. Questo ha permesso di ridurre l’indebitamento finanziario ma ha trasformato TIM in una ServCo focalizzata unicamente sui servizi:

● Pro: bilancio più leggero (asset-light), minor carico d’interessi.

● Contro: riduzione della base patrimoniale, assenza di un asset difensivo come l’infrastruttura fissa e concentrazione in un mercato domestico retail iper-competitivo.

Multipli comparativi con i peer europei

Se andiamo a confrontare i multipli di TIM con i principali player europei nel campo delle telecomunicazioni (v. tabella) è chiaro che TIM non presenta uno sconto tale da giustificare una scommessa standalone, soprattutto alla luce dell’assenza di dividendo rispetto ai peer continentali.

6. Per chi ha TIM in portafoglio: l’arbitraggio non è più la tesi

Con TIM che scambia stabilmente tra 7,53€ e 7,67€, cioè in linea con il valore teorico dell’offerta al prezzo attuale di Poste (7,556€), lo spread residuo tra prezzo di mercato e valore dell’OPAS è ormai marginale e in alcuni casi negativo al netto dei costi di transazione. Non esiste più, in pratica, un rendimento da arbitraggio che giustifichi l’acquisto di TIM in vista del regolamento del 18 settembre.

Questo significa che chi possiede — o valuta di comprare — TIM oggi non può più basare la propria tesi di investimento sulla differenza tra prezzo e valore dell’offerta, ma deve necessariamente ragionare come se stesse comprando un’esposizione indiretta a Poste Italiane, diluita dal fatto che una parte del controvalore (1,67€) è fissa in contanti e non partecipa all’eventuale rivalutazione. Il ragionamento si riduce quindi a una domanda sola: le sinergie industriali che il mercato sta iniziando a prezzare su Poste (si veda la sezione 4) giustificano la posizione, oppure conviene semplicemente spostarsi sul titolo Poste, che partecipa per intero all’upside stimato dagli analisti e non sconta il collar implicito della componente fissa dell’offerta?

Resta un rischio da tenere presente, indipendentemente dalle sinergie: un ripiegamento di Poste sul mercato trascinerebbe con sé anche il valore implicito di TIM, dato che oltre il 78% del controvalore dell’offerta dipende dal prezzo di Poste. Il profilo di rischio/rendimento di TIM, in altre parole, è oggi essenzialmente quello di Poste, ma senza parteciparvi appieno.

7. Sviluppi di lungo periodo: cosa diventa Poste Italiane

Al di là del singolo target price, il punto più interessante per un investitore con orizzonte pluriennale è capire in cosa si sta trasformando Poste Italiane. L’operazione TIM si inserisce in una traiettoria più ampia, già delineata nel piano strategico 2024-2028 (”The Connecting Platform”), che punta a ricavi per 13,5 miliardi di euro e un EBIT di 3,2 miliardi entro il 2028, con un impegno a un payout ratio di almeno il 65% e un dividendo minimo di 1€ per azione a partire dal 2026.

Da operatore postale a piattaforma infrastrutturale

L’integrazione con TIM accelera un percorso che Poste aveva già avviato con la SuperApp e il wallet di pagamenti: diventare un punto di accesso unico — fisico e digitale — per decine di milioni di clienti. Con l’aggiunta della base clienti e dell’infrastruttura TIM, il gruppo combinato arriverebbe a servire circa 46 milioni di clienti tra servizi bancari, assicurativi, di telecomunicazione ed energia, sfruttando al contempo la rete di uffici postali — la più capillare del Paese — anche come canale di accesso ai servizi pubblici per i cittadini meno digitalizzati.

Infrastruttura per l’intelligenza artificiale e sovranità digitale

Il tema più strategico nel medio-lungo periodo è probabilmente quello dei data center. Il mercato italiano dei data center è in forte espansione: da 7,1 miliardi di euro investiti tra il 2023 e il 2025 a oltre 25 miliardi di euro di nuovi progetti annunciati per il triennio 2026-2028, secondo l’Osservatorio del Politecnico di Milano. In questo contesto, la combinazione tra i 16 data center di TIM (125 MW installati, con ulteriore capacità in sviluppo) e la rete fisica distribuita di Poste su tutto il territorio nazionale posiziona il gruppo per intercettare la tendenza verso data center di prossimità (”edge”), alternativi ai grandi hub centralizzati, per cui gli operatori con asset capillari sono considerati particolarmente ben posizionati.

L’operazione si inquadra inoltre in una spinta più ampia verso il cloud sovrano europeo: come già avvenuto in Germania e Francia, anche in Italia si starebbe delineando un operatore infrastrutturale sostenuto dallo Stato (il MEF resta azionista sia di Poste sia, indirettamente, di TIM), in grado di offrire capacità di calcolo distribuita per settori strategici come pubblica amministrazione, difesa e sanità. Resta un tema aperto, e non banale, il costo dell’energia in Italia, strutturalmente più alto rispetto a Francia e Spagna, che pesa sulla competitività di questi investimenti.

Consolidamento del mercato telco domestico

Un ultimo sviluppo di lungo periodo, meno dipendente dall’esecuzione di Poste e più dal contesto di mercato, è la possibilità che l’operazione faccia da catalizzatore per un ulteriore consolidamento tra 4 e 3 operatori nel mercato telefonico italiano, storicamente iper-competitivo e con l’ARPU più basso d’Europa. Un simile riassetto, se si realizzasse, aggiungerebbe valore ricorrente al business telecomunicazioni del gruppo combinato indipendentemente dalle sinergie dirette tra Poste e TIM già messe a budget.

Nel complesso, gli sviluppi di lungo periodo rafforzano l’argomento per cui il vero motore di rendimento dell’operazione non è il differenziale di prezzo a breve termine, ma la trasformazione strutturale di Poste Italiane in un gruppo diversificato tra servizi finanziari, postali, assicurativi, telecomunicazioni e infrastruttura digitale — una tesi che, per sua natura, si gioca su un orizzonte di anni e non sul completamento dell’OPAS a settembre 2026.

8. La domanda finale

Oggi TIM non è più una scommessa sulla trasformazione del business telecom italiano. È diventata, di fatto, un modo indiretto e parzialmente diluito di esporsi a Poste Italiane.

La vera domanda per l’investitore, quindi, non è più “quanto vale TIM rispetto all’offerta”, ma se abbia senso mantenere questa esposizione indiretta oppure spostarsi direttamente su Poste, che oggi raccoglie il consenso degli analisti come principale beneficiaria dell’operazione — sia nello scenario di integrazione piena, sia, secondo BofA, in quello standalone.

Comprare TIM oggi offre un’esposizione indiretta e diluita a Poste, senza più un premio da arbitraggio a compensare il rischio di storno. Comprare Poste offre accesso pieno all’upside potenziale stimato dagli analisti (tra il 9% e il 30% a seconda della banca) catturando appieno il valore generato dalle sinergie dell'operazione.

Let’s start with the uncomfortable truth that many crypto enthusiasts prefer to sidestep: Bitcoin’s price is, at this stage, almost entirely driven by speculation.

This isn’t an insult. It’s a structural observation.

The on-chain utility of BTC — as a medium of exchange, as a settlement layer, as a programmable financial instrument — remains embryonic relative to the capital sitting on top of it. Institutional adoption is real, but it is still largely limited to ETF exposure (BlackRock’s IBIT has accumulated over $50B in net inflows since launch — faster than any ETF in history — and currently holds approximately $47–49B in AUM as of mid-2026) and treasury reserve plays by a handful of companies. Saylor’s Strategy convertible debt structure is clever financial engineering, not a sign that Bitcoin has become embedded in the global financial plumbing.

What drives price, then?

Narrative cycles, liquidity conditions, leverage in the derivatives market, and the reflexive feedback between retail sentiment and institutional positioning. This doesn’t mean BTC has no value — scarcity is real, the network effect is real, and the idea of a non-sovereign store of value resonates in an era of persistent fiscal expansion. But potential value and current price are two very different things, and conflating them is how retail investors consistently get hurt.

The honest framing is this: BTC trades like an early-stage tech asset with a fixed supply. Its price reflects what people believe it might become, not what it currently does. Until institutional application matures beyond ETF wrappers and balance sheet hedges, that gap between narrative and utility will remain the dominant source of volatility.

The Analyst Divide: Even the Experts Don’t Agree

If you’re looking for consensus on Bitcoin’s fair value, you won’t find it — and that itself is informative.

But in mid-2026 the story has become richer than a simple bull-vs-bear split. The most revealing signal isn’t any single price target. It’s how dramatically those targets have moved within the same institutions over just a few months.

Start with Citigroup. In December 2025, a joint research note from analysts Alex Saunders, Dirk Willer, and Vinh Vo put the 12-month base case at $143,000 — with a bull scenario of $189,000 if ETF demand accelerated and U.S. digital-asset legislation passed on schedule. By March 2026, that base case had been cut to $112,000, with legislative delays and slowing network activity cited as the primary drivers. Then, in a July 1, 2026 note, Citi cut again — to $82,000 — and made a move that deserves more attention than the headline number: the bank simultaneously lowered its 12-month net spot ETF inflow assumption to zero, down from a prior estimate of $10 billion. Three cuts in seven months. Base case down 43%. The institutional demand tailwind that was supposed to drive the 2025–2026 cycle is no longer part of Citi’s base case at all.

JPMorgan’s position is structurally different and arguably more intellectually honest about the distinction between near-term price calls and long-run valuation. The bank has maintained a $266,000 long-run theoretical target derived from a gold-parity thought experiment — matching Bitcoin’s market cap to private-sector gold investment at approximately $8 trillion — while being explicit that this figure is “unrealistic” in the near term. The more operational near-term figure JPMorgan has cited is around $170,000, conditioned on market stability and Strategy’s continued resolve as the largest corporate holder.

Standard Chartered offers perhaps the most dramatic revision arc of all. Geoff Kendrick, the bank’s global head of digital assets research, had a $300,000 year-end 2026 target as recently as mid-2025. That was built on two legs: continued corporate treasury accumulation following Strategy’s playbook, and accelerating institutional ETF inflows. By December 2025, Kendrick cut the target to $150,000, noting that the corporate treasury leg had stalled as the companies involved no longer had the valuation incentives to keep accumulating aggressively. “Future Bitcoin price increases would effectively be driven by one leg only,” he wrote — meaning ETF buying alone. One leg supporting a $300,000 target is a $150,000 target. By February 2026, with ETF outflows mounting and Fed rate cut expectations fading, the target came down again to $100,000, accompanied by a warning that a dip to $50,000 was possible first. As of mid-July 2026, Kendrick has reaffirmed the $100,000 year-end call — with BTC trading around $64,000, that implies roughly 56% upside from current levels in about five months.

On the long-duration bullish end, ARK Invest’s Cathie Wood still holds a $1 million+ target over a multi-year horizon, and Bernstein has maintained a $150,000 year-end 2026 call through the drawdown. Peter Schiff, structurally bearish as ever, argues the asset has no intrinsic cash flow and therefore no rational floor.

On the other side of the spectrum, voices like Peter Schiff remain structurally bearish, arguing that BTC is a greater fool asset with no intrinsic cash flow and therefore no rational floor.

More nuanced skeptics — including several macro strategists at major banks — acknowledge the speculative demand but question whether BTC can survive a prolonged global liquidity contraction without reverting to four-figure prices.

The spread between the most bearish near-term scenario (Citi’s $53,000 recession floor) and the most bullish long-run target (ARK’s $1 million+) is close to a 20x range. When institutions with access to the same on-chain data, the same ETF flow metrics, and the same macro backdrop produce conclusions this divergent — and when the same institution revises its own base case by 43% in seven months — the honest conclusion is not that some analysts are smarter than others. It’s that nobody has a reliable model for pricing this asset. BTC is not mispriced the way a DCF analysis can reveal a mispriced stock. It is unpriced in any fundamental sense, and the institutional consensus is still being written in real time.

Subscribe now

The Halving Theory: Elegant, Useful, Probably Dead

For years, the four-year halving cycle was Bitcoin’s most compelling structural narrative. The logic was clean: every ~210,000 blocks, the block reward is cut in half, supply issuance drops, and — assuming demand holds — price rises. The 2012, 2016, and 2020 cycles appeared to validate the theory with almost suspicious regularity.

Here’s the problem: the halving’s impact on supply is now structurally marginal.

With over 20 million BTC already mined out of a 21 million cap — a milestone crossed on March 9, 2026, meaning more than 95% of all Bitcoin that will ever exist is now in circulation — the incremental supply reduction from each halving is smaller in absolute terms than the previous one. The April 2024 halving dropped the daily issuance from ~900 BTC to ~450 BTC per day, a rate that will halve again at the next halving expected in April 2028. At current prices, that’s a relatively modest amount of daily sell pressure removed — meaningful at the margin, but trivially small compared to the billions in daily spot and derivatives volume.

More importantly, Bitcoin’s price has become increasingly correlated with global liquidity conditions — specifically with M2 money supply trends and the Fed’s balance sheet trajectory. The 2022 bear market wasn’t a halving anomaly; it was a liquidity event. The 2023–2024 recovery tracked the Fed’s pivot narrative and the anticipation of ETF approval almost perfectly. The halving happened in April 2024, but the price had already moved months before.

This doesn’t mean the halving is irrelevant — narrative power has real market impact, and traders position around it. But treating it as a mechanical price catalyst with predictable 12–18 month lag effects is increasingly a relic of a smaller, less liquid market. BTC now moves with risk assets, responds to macro surprises, and breathes with global dollar liquidity. The halving is a background condition, not a driver.

Bitcoin price (log scale) across four halving cycles. Blue = ~12 months pre-halving. Green = ~12 months post-halving. Source: approximate monthly close prices.

Floors, Ceilings, and the Heatmap: Useful Tools, Wrong Timeframe

Does talking about a BTC price floor make sense? In the long run, probably not in any precise way. In the short run, there are more useful frameworks.

Liquidation heatmaps — tools that visualize the concentration of leveraged positions across price levels — have become one of the more reliable short-term indicators in crypto. The logic is straightforward: large clusters of long liquidations below the current price act as magnets during downside moves, as market makers and algorithmic traders have structural incentives to push price into those zones. The inverse is true for short liquidation clusters above the market.

Platforms like CoinGlass publish real-time liquidation data, and watching how price interacts with these levels provides genuine edge over a 24–72 hour horizon. A dense liquidation cluster at $85,000 doesn’t predict whether BTC goes to $150,000 or $60,000 over the next year — but it does tell you something about where short-term volatility is likely to concentrate.

BTC liquidation heatmap, Jul 21–22 2026. Brighter colours indicate higher liquidation density at that price level. Source: CoinGlass.

The same logic applies to open interest-weighted support and resistance levels, funding rates, and options market positioning (particularly the gamma exposure of major dealers around monthly expiry dates). These are legitimate tactical tools.

What they are not is a basis for long-term valuation. Using a liquidation heatmap to justify a multi-year investment thesis is like navigating a cross-ocean voyage with a weather radar. Useful for the next few hours. Useless for the destination.

So, Does Investing in Bitcoin Actually Make Sense?

This is the question the article has been building toward, and the honest answer is: it depends on what you’re trying to do, and whether you’re being honest with yourself about it.

If you’re seeking a non-correlated store of value that protects purchasing power across decades — the “digital gold” thesis — the evidence is mixed at best. BTC’s correlation with the Nasdaq during stress periods has historically been uncomfortably high. It behaves like a risk asset when risk assets sell off, which is precisely when you’d want your store of value to hold.

If you’re seeking asymmetric upside exposure to a technology that might become foundational financial infrastructure — and you can size the position such that a complete loss doesn’t derail your financial plan — there’s a rational case. The optionality is real, even if the probability of the most bullish outcomes is uncertain.

If you’re buying BTC because price has been going up and you don’t want to miss out — that’s speculation, and there’s nothing wrong with speculating, as long as you call it what it is.

The deeper question isn’t whether Bitcoin has value. It’s whether you have a clear thesis, a defined position size, and an exit framework. Most retail investors who have been hurt by crypto didn’t lack information. They lacked process.

What’s your process?

Marco

Subscribe now

Recent debate over retail access to leveraged products raises a broader question: has market volatility actually increased, or has the market become structurally more sensitive to shocks?

Key findings

- Traditional volatility measures do not show a permanent, structural increase — volatility remains episodic, not chronically elevated.

- Margin leverage is at a record high in nominal terms and relative to GDP.

- Options activity has grown for six consecutive record years.

- 0DTE (zero-days-to-expiration) options now make up a majority of SPX option volume.

- Retail investors dominate the fastest-growing, most leveraged corners of the ETF market.

- Modern market structure appears to amplify volatility *bursts* even while average volatility stays moderate.

1. A Niche Regulatory Change Could Hint at a Structural Change

Regulators are once again debating who should be allowed to access leveraged products and under what conditions. The Financial Industry Regulatory Authority (FINRA) scrapped the rigid $25,000 account minimum and trade-counting limits, allowing traders to execute unlimited day trades so long as they meet standard risk requirements.

A rule change around access to leveraged products may look like a niche regulatory story but it is not. It sits at the intersection of three powerful market trends: retail participation, embedded leverage, and the transformation of volatility from a risk metric into a tradable product.

And it’s worth sitting with the implication: if regulators feel compelled to revisit access rules, it may be because the market itself has grown more dependent on leverage and speculation than it was a decade ago.

That’s the real question we’re trying to answer — not “is volatility high right now,” but “has the *machinery* that produces volatility changed shape.”

2. How Do We Measure Volatility?

Most investors use the word “volatility” loosely, which is part of why the debate is so muddled. There are really three distinct concepts.

1. Realized volatility — what actually happened. Typically measured as 20-day or 60-day annualized standard deviation of returns.

2. Implied volatility — what the options market *expects*. The standard proxy is the VIX, which measures the market’s expectation of S&P 500 volatility over the next 30 days.

3. Shock frequency — arguably the most interesting and least discussed metric. Rather than asking “how volatile is the market on average,” this asks “how often does the market move violently.” Practical measures include the count of ±1% and ±2% daily moves, and the number of trading days per year the VIX closes above 25 or above 30.

3. Is Volatility Actually Higher?

Let’s test the hypothesis: is volatility now permanently higher than in previous market eras?

The data pushes back on this. As of mid-2026 the VIX has mostly traded in the 15–19 range, well inside its long-run “normal” band, with an average in the high teens over the trailing year — not far from where it has sat during many calm stretches since the 1990s. Spikes still happen — the VIX touched the low 30s in March 2026 around a geopolitical flare-up — but they resolve quickly rather than persisting.

That pattern is consistent with the historical record going back to 1995: volatility does not drift steadily upward over decades. Instead it sits in a low, boring range for long stretches and then jumps sharply around specific catalysts — the dot-com crash, the Global Financial Crisis, the Covid shock, the 2025 tariff correction — before mean-reverting.

The market is not necessarily more volatile “all the time”. If something structural has changed, it isn’t the average level of volatility — it’s what happens around the edges and the frequency of violent spikes.

That distinction matters because it’s the edges, not the average, that retail investors actually feel, and the frequency of those spikes tends to rattle them the most.

4. The Leverage Explosion

Question: has leverage actually increased?

FINRA-reported investor margin debt hit a record $1.53 trillion in June 2026, up roughly 51% year-over-year and up nearly 8% in that month alone — the third consecutive monthly increase. Margin debt relative to GDP is now sitting near record territory as well, well above its multi-decade historical average.

Leverage behaves like fuel, not like a spark. It doesn’t create a shock on its own, but it changes the amplitude of whatever shock does occur:

- When prices rise, leveraged gains compound faster than unleveraged ones, which pulls in more leverage.

- When prices fall, losses compound the same way — except now the investor doesn’t get to choose the exit. A drop in account equity below maintenance requirements triggers a margin call, and margin calls create forced sellers at exactly the moment the market can least absorb them.

The key takeaway here is that more leverage does not, by itself, create volatility. It amplifies whatever volatility shows up.

5. Volatility Has Become a Product

Historically, the retail default was to buy stocks and hold them. Today, a large and growing share of trading activity is not really “investing” in the traditional sense — it’s trading volatility itself, through instruments explicitly built for short-term, leveraged directional bets: standard options, 0DTE options, leveraged ETFs, and inverse ETFs.

U.S. options markets recorded roughly 15.2 billion contracts traded in 2025, up about 26% from 2024 — the sixth consecutive record year for options volume.

The trend line slope increases dramatically in the back half of the chart — that’s the point.

The market now contains far more instruments specifically designed to express short-term views, and their growth rate has outpaced the growth of the underlying equity market itself.

6. The 0DTE Revolution

Most investors — even sophisticated ones — underestimate how large this corner of the market has become.

Zero-days-to-expiration SPX options accounted for roughly 59% of total SPX options volume during 2025 — meaning a majority of all S&P 500 index options activity now expires the same day it’s traded.

0DTE options aren’t a long-term volatility story — they’re an intraday one. Their significance comes from what they force market-makers to do: as retail and institutional flow pushes dealers short or long gamma over the course of a single session, dealers hedge continuously, and that hedging activity itself becomes a source of intraday price pressure, independent of any new information entering the market.

7. Retail Investors: Cause or Catalyst?

It’s easy to slide into “blame retail traders” territory. The data supports a more careful framing.

Reuters has recently reported that retail investors account for nearly 90% of trading activity in leveraged single-stock ETFs — arguably the most aggressively amplified retail-accessible product on the market today.

But there’s an important distinction: retail participation on its own is not inherently destabilizing — retail investors have always been part of the market. The relevant question isn’t “are retail traders active,” it’s “what happens when retail participation concentrates in the *most leveraged* corner of the product shelf.”

Retail traders are increasingly concentrated in products explicitly engineered to amplify short-term market moves — which changes the *composition* of volatility risk even if it doesn’t change participation levels themselves.

8. The Role of Algorithms

A popular misconception is that algorithmic and high-frequency trading is the root cause of modern volatility.

Both the SEC and FINRA treat algorithmic trading as a core, largely beneficial part of modern market structure — it’s a major source of liquidity — while also acknowledging that it carries its own set of risks, particularly around how quickly it can transmit stress through the system.

Algorithms are not the root cause of volatility. They’re the transmission mechanism — the wiring that lets a shock originating anywhere in the leverage-options-retail chain propagate through the entire market almost instantly.

`Leverage = Fuel` · `Options = Accelerant` · `Algorithms = Transmission` · `Volatility = Output`

---

9. A New Framework: Increased Structural Volatility Sensitivity

None of the individual pieces above — leverage, options activity, retail concentration, market concentration — fully explains modern volatility on its own. Put together, they suggest a framework where the market sensitivity to short term volatility seems structurally increased.

This isn’t meant as a rigorous, back-tested quantitative metric — it’s a conceptual lens. The point isn’t precision; it’s that these four inputs move together, and when they all expand simultaneously, the market’s *sensitivity* to a shock rises even if its *average* volatility does not.

Conclusion

The evidence does not support the idea that markets are permanently more volatile than in the past. What it does support is a more specific and more interesting claim: markets have become structurally more sensitive to shocks, because leverage, options activity, retail participation, and rapid execution mechanisms have all expanded at the same time.

Volatility has not disappeared, nor has it steadily increased. It has changed form. Modern markets behave less like a rough sea and more like a compressed spring: calm for long stretches, but capable of releasing enormous energy when the wrong trigger is pulled.

Marco

I recently came across a market commentator making a point that, on the surface, sounded almost too obvious to bother stating: “a truly guaranteed, risk-free 8% annual return.” He was talking about the long-run average of the US stock market — and from a certain angle, sure, that’s a fair statement. But historical averages can’t be confounded with promises.

No, the S&P 500 has never signed a contract with anyone.

But it stuck with me, because it’s really a small piece of a much bigger idea — one that I think is worth unpacking properly, because it separates good investors from people who risk losing their savings to someone with a nice suit and a confident voice.

I’ve a rule I want to plant in your head today: whenever the words “guaranteed” and “investing” show up in the same sentence, it’s a big warning sign, not a selling point.

Subscribe now

Why “guaranteed” and “market return” can’t coexist

Markets go up over time — but in lumps, with drawdowns, with years that test your stomach lining and your conviction at the exact moment you need both the most. That’s not a bug in the system. It’s the actual mechanism that generates the return in the first place.

This is a core idea in finance that gets lost in casual conversation: risk and return are not two separate things you can unbundle. The reason equities have historically outperformed cash and bonds over long horizons is because they carry the risk that cash and bonds don’t. Volatility isn’t a design flaw you can remove, at least not entirely — consider it the toll you pay for participating in the compounding machine at all. Remove the toll, and you’ve removed the machine. Entirely.

So when someone says “guaranteed 8%,” what they’re actually saying — whether they realize it or not — is “I’ve found a way to give you equity-like returns with none of the equity-like risk.”

I’ll tell you a secret… nobody has found that! Not you, not me, not Warren Buffett, not even a quant fund with a hundred PhDs and a supercomputer. If it existed at scale, every pension fund, sovereign wealth fund, and central bank on the planet would already be doing it exclusively, and every other asset class would cease to exist because all the capital in the world would flood into the free-lunch machine.

There is no free lunch in the investing world. There’s just a lunch bill someone hasn’t shown you yet.

So, next time you find someone promising you a risk-free 6, 8 or 10% return, well, they’re just lying.

The 30-year picture

Zoom out, and the drawdowns look different — not because they weren’t real (they were, and living through them was genuinely hard), but because they sit inside a much longer upward path:

$10,000 invested in the S&P 500 at the start of 1996, with dividends reinvested and nothing else touched, would be worth roughly $237,000 today — a compound annual growth rate of about 10.75% over thirty years. The shaded bands mark the dot-com crash, the Global Financial Crisis, the COVID crash, and the 2022 bear market. Every one of them looks, from this distance, like a dip in an otherwise upward-sloping line. None of them looked like that while they were happening.

That’s really the tension at the heart of this whole article. The long-run number is genuinely, historically strong. It is also not a number anyone could have promised you in 1996 — nobody knew in advance that this particular thirty-year window would deliver this particular result, and there’s no guarantee the next thirty years replay it. What you can say with real confidence is that the number came from staying exposed to risk the entire time, drawdowns included, not from someone removing the risk while preserving the return.

What a real drawdown actually looks like

It’s easy to nod along with “markets are volatile” as an abstract idea. It’s a different thing to see the number next to your own account balance while it’s happening. So let’s look at what the last few real bear markets actually did to the S&P 500, on a pure price basis:

A few things jump out of that chart:

• The dot-com crash (2000–2002) took the index down roughly 49% from its March 2000 peak, and it took until October 2007 — about seven and a half years — for the index to close at a new high.

• The Global Financial Crisis (2007–2009) was even deeper: a peak-to-trough decline of nearly 57%, with the index not reclaiming its old high until 2013 — again, over five years.

• The 2020 COVID crash was the sharpest and fastest ever recorded: a 34% decline in just 33 days — but also the fastest recovery on record, a new high inside about five months, thanks to an extraordinary policy response.

• The 2022 bear market took the index down about 25% over roughly nine months, and it took about two years to reclaim the prior high.

Two lessons sit inside that data. First, depth and recovery time are only loosely related — a smaller decline can take longer to recover from than a bigger one, because recovery speed depends more on the underlying cause (a liquidity panic vs. a valuation reset vs. a structural credit crisis) than on the size of the drop itself. Second, and more importantly for this article: every single one of these episodes eventually recovered. That’s the actual case for staying invested through volatility — not that the volatility isn’t real, but that time, plus a diversified, low-cost approach, has historically gotten investors through it.

Notice what’s completely absent from that story: a guarantee. Nobody selling S&P 500 exposure in 2007 promised you it would take “only” five years to get your money back. The recovery is a historical pattern, observed after the fact — not a contractual outcome anyone could have promised you in advance. That distinction is the whole point of this article.

This isn’t an argument against passive investing — it’s the opposite

I want to be careful here, because I don’t want this to read as market pessimism or an argument for hiding your money under a mattress. Quite the opposite. I’m one of the loudest advocates you’ll find for low-cost, broadly diversified ETF investing as the backbone of a long-term portfolio. The evidence on buy-and-hold, market-cap-weighted exposure held over decades is about as robust as anything in empirical finance.

But “historically resilient over long horizons” and “guaranteed” are not the same statement, and conflating the two is exactly where people get hurt. A 30-year historical average is a description of what happened to a specific set of markets over a specific set of decades. It is not a forward-looking promise, and it’s especially not a promise about your specific 5-year or 10-year window, which might land squarely inside a drawdown.

Good investing accepts this uncertainty and manages it — through diversification, time horizon, position sizing, rebalancing. Bad actors sell you the removal of uncertainty instead, because uncertainty is uncomfortable and certainty is an easy thing to buy.

A case study, unfortunately close to home

We just watched this exact dynamic play out in real time here in Italy. A well-known former politician and media personality was arrested over an alleged scheme built around “collective bets” — pooled money from supporters — with, you guessed it, guaranteed returns attached. The scheme is now unraveling into what looks like a textbook Ponzi structure: new money coming in was used to cover the promoter’s own personal bills and to pay out earlier participants, creating the illusion of a working strategy for as long as new money kept flowing in.

Nothing was actually being invested in any productive, value-generating sense. There was no underlying engine producing the returns. There was only ever the next depositor’s check.

This is, structurally, the same mechanism behind every Ponzi scheme going back to Charles Ponzi himself in 1920, through Bernie Madoff’s multi-decade fraud, through dozens of smaller regional schemes that never make international headlines. The details change — postal reply coupons, split-strike conversion strategies, collective sports betting pools — but the skeleton is always identical: no real return-generating activity, payouts funded by new contributions, and a charismatic front person whose main asset is trust rather than expertise.

The part that matters most isn’t the fraudster — it’s the trust shortcut

Here’s what I think is the more important lesson, and it’s not really about the one man in this specific case. It’s about all of us.

People handed over their savings largely because he was famous. Not because the strategy made economic sense. Not because the math behind “guaranteed returns on collective sports bets” could survive five minutes of scrutiny. Not because anyone independently verified where the money was actually going. They just trusted him.

Fame and familiarity filled in for due diligence.

And notice how absurd the underlying premise was once you say it plainly: guaranteed returns on sports betting. Yes. Sports betting. By construction, a negative-expected-value activity for the bettor once you account for the bookmaker’s margin. There’s no version of “guaranteed” that logically attaches to that activity. The fact that thousands of people didn’t stop at that sentence and ask “wait, guaranteed how, exactly?” tells you everything about how trust substitutes for analysis when a familiar face is doing the talking.

This is a well-documented pattern in behavioral finance, and it has a name worth knowing: affinity fraud — fraud that specifically exploits trust within a community, whether that community is built around religion, ethnicity, a shared hobby, or simply shared exposure to the same TV personality. Madoff’s victims were disproportionately drawn from Jewish community and country-club networks where he had built social credibility over decades. The mechanism is always the same: social trust is used as a substitute for financial verification, because verification is effortful and trust feels free.

A checklist worth keeping

Whether you’re evaluating a fund, a strategy, a professional portfolio manager, or a guy your cousin met at a dinner party, here are the questions that actually protect you:

• Can the return be explained mechanically? Not “trust me,” but an actual description of what’s being bought, sold, held, or hedged, and why that activity would plausibly generate the stated return.

• Is risk mentioned at all? Every legitimate return stream comes with a description of what can go wrong and roughly how much you could lose. If risk isn’t part of the pitch, the pitch is incomplete by design.

• Does the return stream correlate with anything explainable? Real strategies tend to have some logical relationship to market conditions — momentum, credit spreads, volatility regimes, macro cycles. Fabricated return streams tend to be suspiciously smooth, because someone is quietly managing the reported number rather than the actual portfolio.

• Is the person’s credibility based on fame or on verifiable expertise? These are not the same currency, and one cannot be exchanged for the other, no matter how comfortable it feels to pretend otherwise.

• Would the same pitch survive being said in plain language? “Guaranteed returns on sports bets” sounds absurd stated directly. Scams often rely on jargon, complexity, or social pressure specifically to avoid ever being stated directly.

• Is there independent custody and reporting? Legitimate investment structures separate the person managing the money from the person holding it, with independent statements you can verify. If one person controls the narrative, the funds, and the reporting, you are trusting a story with no external check.

The takeaway

Passive investing works because it accepts market risk, spreads it as broadly as possible, and lets time and compounding do the heavy lifting — not because it eliminates risk. Active and quantitative strategies, momentum-based or otherwise, work the same way at their core: skill and process can tilt the odds in your favor, but nobody tilts them to zero volatility, and certainly nobody honestly guarantees them.

So the next time you hear “guaranteed” attached to a return number — whether it’s coming from a slick fund pitch, a friend’s cousin, or a famous face on television — treat it exactly like a smoke alarm. Don’t analyze the smoke. Just get out.

Marco

Markets are down, and everyone’s got a one-line explanation ready. Iran. The Fed. AI is a bubble. Pick your poison.

The honest answer is messier: three separate stress points are hitting at the same time, and only one of them is actually new news. The other two have been building for months and just needed a trigger.

Subscribe now

The trigger: geopolitics or fundamentals?

The proximate cause is unambiguous. President Trump declared the fragile Iran ceasefire “over” at the NATO summit in Ankara on July 8, telling reporters “we’ll probably hit them hard again tonight” after U.S. strikes the previous night. He did it.

As a consequence oil prices jumped and yields on government bonds across the globe rose as investors priced in higher inflation following renewed U.S. strikes. Brent crossed $76, WTI above $72, both up more than 5% on the day. Oil prices extended gains after a U.S. official said Treasury is revoking the license permitting Iranian oil sales, with the threat level to shipping through the Strait of Hormuz raised to “severe.”

This is pure macro/geopolitical, not structural. Equity markets are doing exactly what they should do when a war premium re-enters energy prices: sell risk, buy oil-linked names. The energy sector ETF jumped over 2% while tech dropped around 2% as investors booked profits. Asian markets took it hardest — South Korea’s Kospi fell 5.35% with trading briefly halted via circuit breaker, and Japan’s Nikkei dropped 2.11% — which tells you this is also compounding with Point Two below.

The IMF’s own numbers show why this matters beyond a headline: the fund expects oil prices to rise nearly 32% in 2026 and global consumer prices to climb 4.7%, up from 4.1% in 2025 — and that forecast assumes the Strait of Hormuz reopens this month, which is now in doubt. If Hormuz stays disrupted, this stops being a one-day scare and becomes an actual stagflation input.

The slow burn: a Fed that won’t just cut

Here’s what most headlines are missing. This isn’t a market that sold off because the Fed did something. It sold off partly because the Fed won’t rule out doing something ugly. New Chair Kevin Warsh’s first FOMC minutes, released July 8, showed the committee split nine-to-nine on whether to hike rates before year-end. Six of those nine projected two separate 25-basis-point hikes. That’s not the Fed pivot the market spent 2025 pricing in.

Warsh has been busy dismantling the Powell-era communication playbook — ending forward guidance, shrinking the post-meeting statement, and declining to submit his own dot — while the Fed’s own inflation projection for 2026 jumped to 3.6% headline from 2.7% in March, alongside a rate hold at 3.5%–3.75%. Bank of America has already flipped its call: three quarter-point hikes this year, taking the funds rate to 4.25%–4.50%, up from a prior base case of no change. Even Warsh himself has admitted the tension out loud — he said he’d “have a hard time” calling policy restrictive given what’s happening in financial markets, calling the picture “uneven.”

This is structural in the sense that it’s not going away with one good CPI print. It’s the reintroduction of two-way rate risk into a market that had priced in only one direction.

The real structural question: is the AI trade cracking?

This is the part with legs. The semiconductor selloff has wiped out more than a trillion dollars in market value this week, with Intel down over 20%. The Philadelphia Semiconductor Index crashed 6.3% on July 1 and another 5.5% on July 2 — before Iran even reignited. That timing matters: the chip unwind started as an internal valuation story, and the Middle East just piled on.

The bear case (personified by Michael Burry himself.) Burry has gone from one short to a full-blown thesis. He’s disclosed short positions against Nvidia, Micron, Tesla, Caterpillar, and the SOXX semiconductor ETF, framing it as a bet against “fear of missing out, the greater fool theory, and public commitment bias.” He points out Micron’s stock is more extended above its 200-day moving average than at any point since 1984 — “not even during the dot-com peak” — despite the company being, in his words, a serial “destroyer of capital” one quarter in three. His broader thesis: the AI trade is circular — chipmakers rally because hyperscalers spend on AI infrastructure, equipment makers rally because chipmakers expand capacity, and every new capex announcement gets read as demand validation rather than questioned. When Samsung and SK Hynix announced a new Korean chip hub, Burry called it “the beginning of the end.” In May he said the setup felt like “the last months of the 1999–2000 bubble.”

Worth noting for credibility, not just headline value: Burry isn’t purely bearish right now. He’s also buying — adding to JD.com, and initiating new positions in Flutter and DraftKings on a bet that prediction markets’ regulatory threat fades — which suggests this is a rotation call as much as a crash call.

The bull case is not nothing. Wedbush’s Dan Ives calls this “3rd inning, 1 out in a 9-inning game,” and Goldman notes Nvidia’s forward P/E of 21.7 is actually cheap against its own five-year average of 72. Morgan Stanley’s Michael Wilson argues the S&P’s forward P/E has already fallen 18% from its six-month peak — a drawdown rarely seen outside recession or aggressive tightening, neither of which is his base case — and that earnings, crucially, never actually collapsed. The counter-argument to “1999” is structural too: today’s AI leaders carry net margins above 25% against the S&P average of 13%, real cash flow, and measurable revenue — a fundamentally different starting point than the profitless dot-com names of 2000.

The honest tension, though: annual AI infrastructure spending near $400 billion still dwarfs the roughly $100 billion in actual enterprise AI revenue generated so far, a gap that can’t persist indefinitely. BofA’s own Bubble Risk Indicator has semiconductors at 0.91 — objectively stretched, even if not proof of an imminent crash.

Put a number on that gap and it gets sharper. Sequoia’s David Cahn calculated roughly a $600 billion annual revenue shortfall between what hyperscalers spend on AI infrastructure and what the AI ecosystem actually generates in sales — a gap that’s widening, not closing, as 2026 capex accelerates faster than revenue. The bull rebuttal isn’t naive, though: Bain’s own framework says the industry needs roughly $500 billion in annual capex to eventually justify $2 trillion in revenue — a 4x multiple that assumes the payoff arrives over years, not quarters. Both things can be true. The chart above isn’t proof of a bubble popping; it’s proof the market is now old enough in this cycle to start asking the question out loud.

Forecast: V-shaped bounce or the start of something worse?

This is where it gets genuinely interesting, because the market has already run this exact experiment once this year. Back in March, U.S. stocks pulled back more than 4% as energy costs spiked with the first leg of the Iran conflict — the S&P fell to a low around then rallied nearly 19% off it, and international developed markets fully recovered their 13% conflict-related decline. That’s the textbook V: geopolitical shock, oil spike, sharp but shallow equity drawdown, full recovery within weeks once the news stopped getting worse. If Trump’s “ceasefire is over” escalation resolves the same way — a bad week or two, then de-escalation — this second leg likely rhymes with the first.

The technical backdrop doesn’t scream crash — yet. VIX closed around 16 on July 8, elevated but nowhere near panic territory (30+ for real fear, 40+ for capitulation). Morgan Stanley’s Mike Wilson — who correctly called both the 2025 “rolling recession” and the April 2025 tariff-shock trough — argues the current chip/tech correction is “mature in time and price,” noting half of the Russell 3000 is already down 20%+ from highs, which is usually a sign a correction is ending, not beginning. His own caveat is the one to watch: if crude breaks and holds above $100 for a sustained period, the read shifts from “correction in a bull market” to something more serious — and Hormuz disruption is already blocking roughly 20 million barrels a day of tanker flow that strategic reserves can only partly offset.

The bull case for a fast recovery, laid out by Fundstrat’s Tom Lee, isn’t dismissing the pain — it’s timing it differently. Lee’s roadmap assigns 60% odds to a 10–20% drawdown sometime between August and October, driven by sticky core inflation, elevated margin debt, and oil supply risk — but explicitly not now, and explicitly followed by a resumption of the rally toward 8,000–8,800 on the S&P by year-end. In other words: his framework says this week’s action is premature for the “real” correction, which suggests the current bounce-back window may still hold.

The bear case for something structurally worse rests less on this week’s headlines and more on positioning. One widely-circulated technical read from early summer flagged the combination of a low, complacent VIX, collapsing market breadth, deteriorating technicals, and the historically weak May–October seasonal window — plus an active geopolitical conflict — as the highest correction risk since early 2022, while stressing this describes elevated risk, not certainty. That note’s own counterargument is worth carrying forward too: AI capex is still the single largest driver of GDP growth this cycle, and a “catch-up” of lagging stocks is just as plausible a resolution as a “catch-down” of the leaders — though it also flagged that oil at these levels acts as a direct tax on the consumer names that would need to do the catching up.

My read: the ingredients for a repeat V are mostly there — a single geopolitical trigger, contained (so far) technical damage, and a market that’s already proven this year it can absorb an Iran shock and recover in weeks. The thing that would flip this from “buy the dip” to “this is different” isn’t the equity tape itself, it’s oil. A Hormuz disruption that holds through Q3, pushing Brent durably above $100, changes the transmission mechanism from “one bad week for risk assets” to “actual stagflationary drag on consumer spending and corporate margins” — at which point the Fed’s already-split hawkish wing gets the ammunition to hike into a slowing economy, which is the scenario nobody wants. Watch oil, not the S&P, for the real signal.

So: structural, macro, or fear?

All three, stacked, and that’s the actual story:

● Fear/technical, short-term: the Iran escalation is a shock that can reverse on a single diplomatic headline. Don’t overweight it.

● Macro, medium-term: a Fed that’s genuinely split on hiking, not just “on hold,” is a real regime change from the 2024–25 cutting cycle narrative. This doesn’t resolve quickly.

● Structural, longer-term: the semiconductor/AI capex trade was already digesting a valuation reset before Iran reignited. Whether that’s a healthy “mid-cycle” correction (bulls) or the first crack in a self-referential bubble (Burry, Hartnett) is the real question for the second half of 2026 — and it won’t be answered by this week’s headlines. It gets answered by TSMC and Intel earnings, and by whether enterprise AI revenue actually starts closing the gap with the capex.

On balance, base case is V-shaped, not the start of a bear market — but it’s a conditional base case. The condition is oil staying off $100, and that condition is entirely in the hands of two governments neither of which reports to any of us.

Marco

Who I am:
I'm a chemical engineer with 20+ years experience in the O&G sector. A finance enthusiast, trying to break down complex ideas in simple insights. Active Quantitative Investor, developing proprietary algorithms for index sampling and quantitative trading portfolios. Verified 45% CAGR over the past 3 years on eToro (marcogarattoni). Currently managing a public portfolio with 4000+ copiers and $10+ M under copy (relentlessalpha).

Il 1° luglio 2026 il default cambia verso. Da oggi, un neoassunto nel settore privato senza una posizione previdenziale attiva che non esprima esplicitamente una scelta avrà il TFR automaticamente conferito al fondo pensione negoziale previsto dal proprio contratto collettivo.

Entro i 60 giorni dalla nuova assunzione, può comunicare la propria opposizione all’adesione automatica, mantenendo il TFR in azienda (o versandolo al Fondo di Tesoreria INPS a seconda delle dimensioni dell’impresa).

Una volta scaduti i 60 giorni (o dopo aver firmato l’adesione esplicita), il conferimento al fondo pensione diventa definitivo e non revocabile. Non sarà più possibile riportare il TFR in azienda per gli anni successivi.

È la prima volta in vent’anni che il legislatore inverte l’inerzia invece di limitarsi a regolarla. Vale la pena fermarsi un secondo a chiedersi perché — e soprattutto se, per chi in passato ha optato per lasciare il TFR in azienda (o INPS), conviene fare lo stesso percorso volontariamente.

Subscribe now

Cosa cambia davvero dal 1° luglio

La Legge di Bilancio 2026 (L. 199/2025), con le linee guida operative COVIP pubblicate il 22 giugno, ridisegna il meccanismo del silenzio-assenso:

• Adesione automatica per i neoassunti del settore privato (esclusi i domestici), senza bisogno di manifestare alcuna volontà iniziale.

• 60 giorni di finestra per rinunciare e mantenere il TFR in azienda (o destinarlo a una forma diversa).

• Irreversibilità asimmetrica: una volta scaduti i 60 giorni, l’adesione al fondo è definitiva — non si torna più al TFR in azienda. Al contrario, chi rinuncia oggi può sempre scegliere di aderire al fondo in un secondo momento.

• Il meccanismo si applica anche a chi cambia lavoro dopo il 30 giugno 2026 e ha già un’adesione attiva con versamento di TFR presso un’altra forma. Resta escluso solo chi ha riscattato integralmente la posizione precedente, o chi è iscritto a un fondo ma vi versa solo contributi (non TFR).

• Confluisce non solo il TFR, ma — salvo eccezioni per le retribuzioni più basse — anche il contributo a carico del datore di lavoro previsto dal contratto collettivo. È la parte che, restando in azienda, semplicemente non esiste: nessuno te la versa se non aderisci.

• Il fondo di destinazione è quello previsto dal CCNL; in mancanza di accordi specifici, confluisce nella forma residuale.

Il messaggio implicito della norma è chiaro: vent’anni di evidenza COVIP su rendimenti e tassazione hanno convinto il legislatore che il “non scegliere” non è neutro. È un costo, e da luglio quel costo lo paga chi rinuncia attivamente, non più chi non fa nulla.

Per chi lavora già da anni, però, la legge non cambia nulla in automatico: il TFR accumulato finora resta dove è. La scelta — se farla — resta del lavoratore, e i numeri che la giustificano sono gli stessi di sempre.

I numeri, senza sconti

Rendimento medio annuo composto, decennio 2015-2024, al netto di costi di gestione e fiscalità — fonte COVIP:

Tre cose saltano all’occhio.

1. il TFR in azienda non è un disastro, ma nemmeno un investimento. La rivalutazione per legge (1,5% fisso + 75% dell’inflazione) è costruita per agganciarsi al costo della vita, non per batterlo. In termini reali, sostanzialmente piatto: hai conservato, non hai costruito.

2. la linea garantita — quella “per chi non vuole rischiare” — in diversi fondi ha reso meno del TFR in azienda nello stesso periodo. Garantire il capitale nominale non significa proteggerlo dall’inflazione. Sono due garanzie diverse, e la seconda — quella che conta sul lungo periodo — il comparto garantito non la dà.

3. solo bilanciata e azionaria hanno prodotto rendimento reale positivo significativo. La bilanciata, circa mezzo punto sopra il TFR in termini reali. L’azionaria, su 10 anni, più del doppio.

E i fondi di categoria, nello specifico?

Le medie di sistema sono un punto di partenza. Ma ogni fondo negoziale ha la sua storia, i suoi gestori, i suoi costi — e i dati pubblici, fondo per fondo, raccontano qualcosa che le medie nascondono.

Cometa (metalmeccanici, oltre 524.000 iscritti)

Cometa è il più grande fondo negoziale d’Italia e dal 1° gennaio 2026 offre quattro comparti: Monetario Plus (uscita/RITA), Sicurezza (garantito, anche per i “silenti”), Reddito (bilanciato prudente, default per le adesioni esplicite), Crescita (bilanciato dinamico, fino al 65% in azioni). Il confronto anno per anno con la rivalutazione del TFR mostra bene cosa significa “rischio” in pratica:

Il 2022 è il caso di scuola: inflazione alle stelle, il TFR si rivaluta dell’8,28% in un solo anno, mentre azioni e obbligazioni crollano insieme e tutti i comparti Cometa chiudono in doppia cifra negativa. Chi guardava solo quell’anno poteva pensare di aver fatto la scelta sbagliata. Ma nel 2023-2024-2025 Reddito e Crescita hanno recuperato e superato pienamente i livelli pre-crisi, mentre la rivalutazione del TFR — finita l’inflazione alta — è tornata sotto il 2%. Sull’orizzonte di una carriera, il 2022 pesa come un incidente di percorso, non come un verdetto.

Sul fronte costi, Cometa è tra i più efficienti del sistema: l’ISC (Indicatore Sintetico dei Costi) a 10 anni è di circa 0,16% per Reddito e 0,18% per Crescita, contro una media di sistema dello 0,38% per i fondi negoziali bilanciati, dell’1,44% per i fondi aperti e del 2,14% per i PIP. Su un montante di 100.000 €, un punto percentuale di ISC in più può erodere fino al 18% del capitale accumulato in 35 anni — circa 18.000 €, non di mercato: di costi pagati.

C’è poi il contributo datoriale: chi versa almeno l’1,2% della retribuzione minima contrattuale riceve dal datore il 2% (2,2% per gli under 35) — soldi che il TFR in azienda semplicemente non prevede.

Fonchim (chimici e farmaceutici, oltre 170.000 iscritti, oltre 7 mld in gestione)

Tre comparti: Garantito (obiettivo: rendimenti almeno pari al TFR, con garanzia di restituzione del capitale a determinate condizioni — destinazione automatica dei “silenti”), Stabilità (bilanciato a prevalenza obbligazionaria, 70/30, default per i nuovi iscritti), Crescita (bilanciato a prevalenza azionaria, 60% azioni / 40% obbligazioni).

Il dato più interessante arriva dalla comunicazione ufficiale del fondo sul rendimento dall’avvio del comparto Crescita (2003): un rendimento medio annuo composto del 4,79%, contro un’inflazione media nello stesso periodo dell’1,78% e una rivalutazione media del TFR del 2,45%. Un differenziale reale di oltre 3 punti percentuali all’anno, su più di vent’anni — esattamente l’orizzonte per cui questi strumenti sono pensati. Il 2025, da solo, ha visto Crescita chiudere a +8,08%.

Fondenergia (settore energia e petrolio, oltre 2,2 mld nel solo comparto Bilanciato)

Tre comparti — Garantito (orizzonte breve, fino a 5 anni dal pensionamento, garanzia di rendimento minimo del 3,5% a scadenza sui conferimenti), Bilanciato (orizzonte 10-15 anni, benchmark misto 61% obbligazionario / 25% azionario / resto in alternativi — private equity, infrastrutture, real estate), Dinamico — più un profilo Life Cycle introdotto dal luglio 2025. Il comparto Bilanciato ha chiuso il 2024 con un rendimento netto di +4,84%, su un patrimonio di oltre 2,2 miliardi di euro.

La contribuzione è interessante anche sul fronte del contributo aziendale: chi versa il 2% della RAL riceve dal datore un ulteriore 2,55% — più del contributo del lavoratore stesso. I “silenti” finiscono d’ufficio nel comparto Garantito, senza il contributo aziendale: un doppio costo dell’inerzia, in rendimento e in soldi che restano sul tavolo.

Fopen (settore energia elettrica)

Tre comparti — Obbligazionario Garantito, Bilanciato Obbligazionario, Bilanciato Azionario. La nota informativa del fondo confronta esplicitamente, nei suoi materiali, il rendimento del comparto garantito con il coefficiente di rivalutazione del TFR: è il fondo stesso a riconoscere quel benchmark come il vero punto di paragone per chi sceglie la linea più prudente — un’ammissione implicita che “garantito” non significa automaticamente “miglior risultato a lungo termine”.

Il filo comune

In tutti questi fondi vale la stessa regola: la dispersione dei rendimenti fra comparti dello stesso fondo è enormemente più ampia della differenza fra fondi diversi nello stesso comparto. Scegliere Cometa o Fonchim conta meno di scegliere, dentro quel fondo, la linea garantita o quella bilanciata/dinamica. La decisione che pesa di più non è “quale fondo”, è “quale comparto dentro il fondo” — ed è esattamente la decisione che il meccanismo del silenzio-assenso, di default, non fa nel modo più favorevole: i “silenti” finiscono quasi sempre nel comparto più prudente, qualunque sia il fondo, e spesso senza il contributo del datore di lavoro.

(I dati sui singoli fondi sono ricostruiti da documentazione ufficiale dei fondi stessi — bilanci, note informative, comunicazioni agli iscritti — aggiornata a fine 2025/inizio 2026. Cambiano nel tempo: se sei iscritto a uno di questi fondi, il dato puntuale e aggiornato è sempre sul sito del fondo, sezione “rendimenti” o “valori quota”.)

Perché la differenza non è “qualche euro”

Capitalizzare un differenziale di 2 punti percentuali reali all’anno per 30-35 anni non è un dettaglio contabile: è interesse composto su un orizzonte lungo, e il cervello umano fatica a intuirlo perché non si comporta linearmente.

Prendiamo un caso concreto: RAL 50.000 €, 35 anni di lavoro, 7% di TFR accantonato ogni anno (3.500 € annui), capitalizzato ai rendimenti medi nominali discussi sopra. Il montante finale, a parità di contribute:

Tre numeri da guardare con attenzione. La linea bilanciata supera il TFR in azienda di circa 40.000 €. La linea azionaria di oltre 100.000 €. E la linea garantita — quella percepita come “la scelta sicura” — finisce 35.000 € sotto il TFR in azienda: un comparto che protegge il capitale nominale può comunque produrre un risultato finale peggiore della non-scelta di default, semplicemente perché rende troppo poco per troppo tempo.

Questi numeri sono una proiezione a rendimento costante, non una previsione: servono a dare l’ordine di grandezza, non la cifra esatta che incasserai nel 2061. Ma l’ordine di grandezza è il punto. Non stiamo parlando di qualche centinaio di euro di differenza. Stiamo parlando di un differenziale che, da solo, vale più di uno stipendio annuo lordo.

E quella differenza non l’hai persa per una scelta di investimento sbagliata. L’hai persa per assenza di scelta — fino al 30 giugno 2026, il comportamento di default. Da luglio, il default si inverte; ma per chi lavora già da anni, l’inerzia resta un costo che si paga ogni mese in cui non si controlla dove va quel 7%.

Il rischio non è uguale al rendimento

Sarebbe disonesto fermarsi al confronto sui rendimenti medi senza dire la cosa più importante: sono medie su un decennio specifico, non un destino garantito.

La linea azionaria che ha reso il 4,5% reale composto negli ultimi dieci anni è la stessa linea che nel 2022 — inflazione al 4,2%, mercati in caduta — ha prodotto rendimenti negativi su quasi tutti i comparti, garantiti compresi. Chi guarda solo la media decennale rischia di sottovalutare cosa significa “stare dentro” un comparto azionario per 30 anni: non una linea retta che sale, ma anni di +12% alternati ad anni di -10%, con il rischio reale di vendere nel momento peggiore per panico.

È anche il motivo per cui COVIP stessa, nella Relazione 2024, ha sollecitato un cambio di paradigma sul default: oggi chi non sceglie nulla finisce spesso nel comparto garantito, quando l’evidenza di lungo periodo suggerirebbe l’opposto per chi ha decenni di orizzonte davanti — un’allocazione life-cycle, più azionaria da giovane, più prudente vicino alla pensione. Le nuove regole di luglio risolvono solo metà del problema: spostano il default da “TFR in azienda” a “fondo pensione”, ma non toccano il default dentro il fondo, che resta spesso il comparto più prudente.

Nota metodologica

Rendimenti medi annui composti 2015-2024 per linee garantita/bilanciata/azionaria dei fondi pensione negoziali e per la rivalutazione del TFR: Relazione Annuale COVIP 2024 e documentazione statistica COVIP. Inflazione media 2014-2024 stimata sull’indice FOI. La simulazione con RAL 50.000 €, 35 anni, 7% di TFR annuo e rendimenti nominali costanti (2,4% TFR; 1,3% garantito; 3,4% bilanciato; 4,6% azionario) è un esercizio di calcolo a tassi costanti, non una previsione: i rendimenti reali variano di anno in anno, anche in modo marcato, come mostrano i dati di Cometa 2021-2025. I dati sui singoli fondi di categoria (Cometa, Fonchim, Fondenergia, Fopen) sono ricostruiti da documentazione ufficiale dei fondi — bilanci, note informative, comunicazioni agli iscritti, dati FIM-CISL — aggiornata a fine 2025/inizio 2026. Le novità sull’adesione automatica dal 1° luglio 2026 sono basate sulla Legge di Bilancio 2026 (L. 199/2025) e sulle linee guida operative COVIP del 22 giugno 2026.

Questo articolo non è consulenza finanziaria o previdenziale personalizzata.

Subscribe now

Bitcoin is sitting at a 52-week low. Not “near” one — at one. The current 52-week range runs from roughly $58,100 to the $126,198 all-time high printed on October 6, 2025, and price is now parked at the bottom of that range, around $59,000-60,000.

To find the last time BTC traded consistently at these levels, you don’t look back weeks. You look back to September 2024 — in the early phases of the spot ETF rally that took it through $69k, then $73k, then on to six figures. Every dollar of gains booked across roughly twenty months of bull market has been given back. This looks like a full round trip than a simple pullback.

The weekly chart confirms it: lower highs, lower lows, every week since the October peak. Eight months of grinding decline, down roughly 53% from the top, with last week’s close below $60,000 removing the final psychological floor on the way down.

The bull market didn’t end with a bang. It ended with a slow bleed — and the next stop on that bleed has a price tag at $50,000.

The Liquidity Magnet

Forget narrative for a second. Markets move toward liquidity, not toward fairness.

Between $59,000 and roughly $49,800 sits a thick, uncollected pool of resting orders — stop-losses under recent lows, liquidation triggers on leveraged longs that piled in during the $100k-plus euphoria, and limit buys that never got filled on the way down. That pool is a magnet. Price doesn’t need a reason to visit it beyond the fact that it’s there.

This is how markets have always worked, in crypto, but also out of it. Liquidity pools below price get swept before any “real” reversal, because market makers and large players profit from triggering the stops, not from avoiding them. A clean break below $60k — which just happened — removes the last psychological speed bump between current price and that $50k zone. The path of least resistance is down, not because Bitcoin is broken, but because that’s where the unfilled orders are.

That doesn’t mean $50k is a floor. It means $50k is a target.

The Silent Accumulation

Here’s the part the headlines skip: somebody is buying this dip aggressively, and it isn’t the people you’d expect.

Glassnode data (via a recent Nexo study) shows that in the ten days following BTC’s break below $60,000, retail wallets and mid-sized whales — addresses holding up to 1,000 BTC — accumulated more than 260,000 BTC. That’s happening while institutional flows go the other way: spot ETFs have seen north of $1.5 billion in net outflows over a recent three-day stretch alone.

Two very different investor bases are making two very different bets. Institutions are de-risking into a macro environment that’s giving them every excuse to (rate uncertainty, AI capex rotation pulling money out of crypto, the Fed playing it cautious). Mid-sized holders are doing the opposite: treating sub-$60k as value and buying in size.

That’s the textbook setup for a structural floor, not proof of one. Accumulation at these levels builds a wall of cost-basis support — but walls get breached when the selling pressure above them is heavier than the buying pressure below. Right now, that’s an open contest. 260,000 BTC bought in ten days is real demand. $1.5 billion in three-day ETF outflows is real supply. Neither side has won yet.

The Clock Is Ticking: Why October-November Matters

Macro bear-market bottoms aren’t gentle. They’re violent. 2018, March 2020, even pieces of 2022 — they all share the same shape: a sharp, fast capitulation flush that liquidates the last of the leveraged longs in days, not months, followed by a slow base-building process afterward. Nobody rings a bell at the bottom. The bottom rings itself, loudly, all at once.

The current macro-bottom thesis circulating among technicians points to October-November 2026 as the likely window for that flush. That’s not numerology — it lines up with a few real catalysts:

• MiCA enforcement tightening across European exchanges, which mechanically reduces liquidity and forces compliance-driven selling or fund migration.

• Sustained ETF outflows, which remove the institutional bid that was propping up price discovery through 2025.

• Idiosyncratic stress points in crypto-adjacent structured products (leveraged treasury vehicles, corporate BTC-holding entities under balance-sheet pressure) that can act as forced-seller catalysts if conditions tighten further.

None of these alone breaks Bitcoin. Together, in the wrong sequence, they’re the kind of black-swan combination that turns an orderly liquidity sweep into a disorderly capitulation.

Two Scenarios for H2 2026

Scenario A — The Liquidity Sweep & Autumn Bottom (base case). BTC grinds lower through Q3, drawn toward $49,800-50,000. The move down is sharp enough to flush remaining leverage but doesn’t break market structure entirely. It coincides with the October-November macro-bottom window. The 260k-BTC accumulation cohort gets reinforcements, the wall holds, and BTC reverses into a new uptrend heading into early 2027.

Scenario B — Black Swan & Deep Capitulation (tail risk). ETF outflows accelerate, MiCA enforcement locks up more European liquidity than expected, and a structural stress event (a forced unwind somewhere in the leveraged-BTC-vehicle space) breaks the $50k floor cleanly rather than bouncing off it. Panic selling overtakes the liquidity-magnet mechanics, and BTC searches for a lower historical floor — pushing any real recovery past November and into 2027.

The Bottom Line

$50,000 isn’t a doomsday number. It’s where the order book says price is going next, mechanically, regardless of how anyone feels about it. Whether it’s a bounce or a breakdown depends on a fight that’s already underway: 260,000 BTC of fresh accumulation versus $1.5 billion-plus in ETF outflows, with regulatory tightening as the wildcard that decides which side wins.

Trade the level, not the narrative. The magnet is real. What happens when price gets there is the only question that actually matters.

Marco

This is market analysis, not investment advice. Crypto markets are volatile and these scenarios are probabilistic frameworks, not forecasts.

34%. 20%. 21%. 17%.

These aren’t the multi-year liquidation metrics of a distressed micro-cap company or a highly leveraged speculative vehicle. These are the four separate, major peak-to-trough drawdowns experienced by Micron Technology Inc. ($MU) from its all-time highs in less than 6 months within the span of this single calendar year.

Yet, following a historic fiscal Q3 blowout—where revenue surged 345% year-over-year to $41.46 billion and non-GAAP gross margins expanded to an insane 84.9%—the stock immediately captured an explosive 15% upward gap in after-hours trading.

These violent structural swings look like chaos, fear, or a market completely detached from reality. From a quantitative perspective, however, these metrics map perfectly to a clean mathematical and behavioral framework. They demonstrate a foundational truth about equity-risk premiums: volatility is not structural failure; it is the natural variance of a high-beta asset experiencing complex path dependency.

Subscribe now

1. Volatility Is the “Rent” You Pay for Structural Alpha

In institutional asset pricing, investors frequently seek asymmetric upside while maintaining the psychological expectation of a smooth, linear equity curve. This is an operational paradox. In high-beta, structurally transforming technology sectors like semiconductors, deep drawdowns are not a structural bug; they are an explicit feature of the asset class.

If an asset exhibits a systematic Beta significantly greater than 1.5 relative to the broad market index, a minor macro-driven sector rotation or a systemic liquidity squeeze will amplify price corrections exponentially.

The institutional reality is simple: you cannot mathematically or structurally capture the vertical 300%+ multi-year extensions of a supply-chain chokepoint unless you are willing to underwrite the periodic volatility tax. When standard deviation expands during aggressive growth phases, a 20% to 30% localized retracement falls cleanly within standard 1-sigma and 2-sigma normal boundary distributions. It is simply the cost of carrying an asset that possesses extreme, non-linear earnings power.

2. Path Dependency vs. Operational Reality

A structural failure occurs when an investor confuses path dependency (the temporary trajectory of price discovery on a chart) with fundamental reality (the structural earning capacity of the enterprise).

While institutional algorithmic desks and macro funds were aggressively liquidating long semiconductor positions earlier this week—driving a sharp 17% correction into the earnings print—the operational baseline of Micron’s business did not degrade by a single dollar.

To the contrary, the operational landscape remained remarkably rigid. High-Bandwidth Memory (HBM) nodes—the ultimate hardware bottleneck preventing next-generation AI clusters from experiencing data starvation—remained completely sold out under non-cancelable, fixed-price contracts through the entirety of the next calendar year. The market was experiencing a tactical liquidity squeeze and a positioning reset, entirely divorced from the firm’s secular micro-fundamentals:

3. Behavioral Traps: The Disposition Effect & Sizing Symmetries

From a behavioral finance lens, these sharp drawdowns trigger severe execution errors among market participants. The most pervasive of these is the Disposition Effect—the psychological bias where investors rapidly harvest small, marginal gains to lock in a sense of safety, while stubbornly holding onto deeply underwater, structural losers.

When a core structural winner like Micron drops 20% or 21% in a week, retail participants routinely capitulate, selling at the localized statistical trough right before an asymmetric fundamental catalyst re-rates the entire security.

To survive and exploit this systemic behavior, position sizing must match the asset’s underlying volatility profile. If an allocation is sized too large, the emotional pain of a 34% drawdown forces irrational liquidation. If it is sized appropriately within a multi-factor risk framework, that identical drawdown transitions from an emotional crisis into a highly systematic, structural rebalancing opportunity.

Conclusion for the Portfolio Manager

The next time financial media outlets run sensationalized headlines decrying the “end of the technology cycle” during a localized 15% semiconductor pullback, look past the noise. Review the contract lock-ins, evaluate the gross margin expansion, and compute the structural variance.

The drawdowns aren’t a signal to abandon the thesis; they are the exact price of admission for capturing structural alpha.

Marco

Eleven days. That’s how long it took for SpaceX’s public market debut to go from “fourth most valuable company on Earth” to a case study in IPO mechanics gone sideways.

If you only looked at the headlines, you’d think this was a disaster. “SpaceX stock crashes.” “SpaceX tumbles 16%.” “SpaceX shares sink.” All true, all dramatic, and all missing the more interesting story underneath: this isn’t really about SpaceX the business. It’s about float, leverage, and the gap between a $1.75 trillion valuation and what the underlying numbers can actually support in year one.

Let’s walk through what happened, why it happened, and whether the pullback is creating an entry point or just delaying a more honest repricing.

Subscribe now

The Anatomy of the Slide

This wasn’t a single bad day — it was four in a row. After pricing its IPO at $135 and opening at $150 on June 12, SPCX rocketed to an all-time high of $225.64 on June 16, briefly pushing the company’s valuation past Amazon and Microsoft. Then the unwind started: down 5% on Wednesday, 3.6% on Thursday (markets were closed Friday for Juneteenth), a brutal 16.4% drop on Monday — the stock’s worst single day since listing — and another roughly 3.6% slide on Tuesday, landing around $154.

Strip out the noise and the math is almost anticlimactic: shares are now up only about 14% from the $135 IPO price. Most of the speculative hype from launch week has already evaporated. That’s not a company falling apart. That’s a market correcting a structure that was never built to hold its initial price.

Float Mechanics: The Squeeze That Had to Unwind

Here’s the part most retail commentary skipped. SpaceX’s IPO floated only around 4.2% of total shares outstanding — an unusually tight float for a listing of this size. When retail demand met unhedged call-option activity shortly after the debut, it created a textbook mechanical squeeze: a small number of tradable shares chasing enormous buy-side interest, pushing the price well past anything justified by fundamentals.

Squeezes like this don’t resolve gently. Once the momentum-chasing capital stops rotating in, there’s nothing left to support the price at those levels, and the air comes out fast. That’s effectively what we watched happen between June 16 and June 22.

The Lockup Calendar: The Real Overhang Through Q3

Long-term holders actually need to underwrite this, because the timeline is already set.

Right now, roughly 4.2% of shares are floating. But that number is set to expand dramatically over the next ten weeks. According to IPO filing data, there’s a 20% insider unlock tied to SpaceX’s first earnings report (expected early-to-mid August), an additional 10% unlock that triggers if the stock trades 30% above the IPO price, and two further 7% unlocks scheduled around August 21 and September 10.

Stack those together and insiders could be in a position to sell up to 44% of total shares by early September — an increase in tradable float of roughly 900% from today’s level.

That’s the structural ceiling on this stock for the next quarter. It doesn’t matter how good the next earnings print looks if a meaningful chunk of the float is simultaneously becoming sellable for the first time.

Anyone treating this correction as “the bottom” needs to have a clear view on how the market absorbs that supply — because it’s a known, dated, almost mechanical event, not a surprise risk.

The Bond: Refinancing, Not Distress

The other headline driving Tuesday’s weakness was confirmation that SpaceX is seeking at least $20 billion in senior unsecured notes — its first-ever bond offering as a public company. On the surface, raising debt one week after the largest equity offering in history looks alarming. It isn’t, once you understand what the money is actually for.

The bond is reportedly being used largely to repay bridge loans SpaceX took out earlier this year to fund its acquisition of Musk’s xAI. The banking syndicate behind that bridge financing — Bank of America, Citigroup, JPMorgan, Goldman Sachs, and Morgan Stanley — is expected to run the new notes offering as well. In other words, this is balance-sheet cleanup: swapping expensive short-term bridge debt for longer-term, properly priced bonds. That’s standard corporate finance, not a liquidity scramble.

The market’s reaction actually confirms this read. Moody’s, Fitch, and S&P all assigned investment-grade ratings to the offering. Rating agencies don’t hand out investment-grade marks to companies they think are in trouble. If anything, the credit market is sending a calmer signal than the equity market is: solvency isn’t the concern here. Valuation is.

The Cash Cushion Nobody’s Talking About

If the bond news worried you, this should put it in context: SpaceX disclosed $100.8 billion in cash and equivalents as of last Friday. Against a post-offering debt load of roughly $29 billion, that’s an extremely conservative balance sheet for a company of this size and capital intensity.

This matters because it directly undercuts the “SpaceX needs cash and is scrambling” narrative that tends to spread alongside debt headlines. A company sitting on over $100 billion in cash isn’t issuing bonds because it’s desperate — it’s issuing bonds because debt is cheaper than equity dilution, and because locking in investment-grade financing now, while ratings are fresh and favorable, is simply good treasury management. The optics of “more debt right after the IPO” are worse than the substance.

What the Fundamentals Actually Say

Strip away the structural noise and look at the business itself. Early analyst coverage has initiated at “Sector Weight” — essentially a Hold — and the numbers explain the caution: a reported 2025 net loss of $4.9 billion on $18.7 billion in revenue, followed by a further $4.28 billion loss in Q1 2026. This is a company still burning significant cash while scaling two extremely capital-intensive businesses simultaneously: launch/Starlink and AI infrastructure.

None of that makes SpaceX a bad business. It makes it a business whose current multiple is pricing in a lot of future execution that hasn’t happened yet. Institutional money initiating coverage at Hold rather than Buy, even after a 16%+ pullback, tells you Wall Street isn’t convinced the valuation has fully caught up with reality.

Separating the Hype From the Business

For anyone evaluating a position, it helps to split SpaceX into its actual operating segments rather than treating it as one undifferentiated story stock:

The Core Moat — SpaceX retains effective dominance over global commercial launch capacity, and Starlink has built a genuinely sticky, recurring revenue base, now reportedly serving over 12 million subscribers. This is the part of the business with the clearest, most defensible economics.

The AI Infrastructure Play — Through deals like the up-to-$6.3 billion compute leasing agreement with Reflection AI, SpaceX is aggressively scaling its Colossus data center infrastructure. This is the higher-growth, higher-uncertainty segment, and it’s also the part most directly tied to the AI-infrastructure narrative that helped inflate the IPO valuation in the first place.

The bull case rests on these two segments compounding together. The bear case is that the market priced both segments as if they were already at AI-hyperscaler maturity, when in practice one is still being built out and the other faces an unusually loaded supply calendar in the months ahead.

The Bottom Line

Chasing SPCX near $200 was a poor risk/reward trade from day one — that part of the move was always going to mean-revert. As the price grinds back toward the $135 IPO baseline, the margin of safety genuinely improves for patient, long-term investors. The credit market’s investment-grade vote of confidence and the $100 billion cash cushion both support the idea that this is a structural unwind, not a fundamental breakdown.

But “improved margin of safety” isn’t the same as “floor is in.” With up to 44% of shares potentially becoming sellable by early September, the next two and a half months carry a known, dated supply event that the current price may not yet reflect. This looks less like a clean entry point and more like a market that’s still searching for where real, sustainable demand meets a much larger float.

For a structured, diversified portfolio, small initial exposure with room to add through the lockup window is a more defensible approach than trying to call a single bottom now.

Marco

Lo strumento finanziario più usato dagli italiani per proteggere i risparmi dall’inflazione potrebbe in realtà essere strutturalmente progettato per erodere il potere d’acquisto reale. Perché? Perché lo Stato tassa proprio quella “rivalutazione” pensata per coprire l’inflazione.

Il prodotto si chiama BTP Italia Sì (durata 5 anni, scadenza 23 giugno 2031), emesso dal 15 al 19 giugno 2026. Offre una Cedola Reale Minima Garantita dell’1,6% annuo (pagata semestralmente come 0,8%) ed è indicizzato all’Indice FOI (l’indice Istat dei prezzi al consumo per le famiglie di operai e impiegati, esclusi i tabacchi).

Secondo la normativa italiana, la tassazione è al 12,5% (contro il 26% standard sugli strumenti azionari) ed è completamente escluso dal calcolo ISEE fino a 50.000 €. Zero commissioni bancarie se acquistato in fase di collocamento, e il 100% del capitale è garantito a scadenza. Prevede inoltre un Premio Fedeltà dello 0,6% extra a scadenza per i risparmiatori retail che acquistano in collocamento e mantengono il titolo fino al 2031.

Il BTP Italia Sì funziona diversamente dai vecchi BTP Italia classici. Con il Vecchio Sistema, il capitale stesso veniva rivalutato ogni sei mesi, quindi la cedola successiva veniva calcolata su un capitale rivalutato più alto (capitalizzazione composta). Con il Nuovo Sistema, la cedola viene calcolata con una semplice somma aritmetica applicata direttamente al capitale nominale iniziale:

Cedola Semestrale = Capitale Nominale × (Tasso Reale Garantito/2 + Tasso di Inflazione Semestrale)

Se l’inflazione diventa negativa (deflazione), la componente inflattiva si azzera, ma il tasso reale minimo garantito dell’1,6% resta protetto.

Il freno occulto sui tuoi rendimenti

La morte della capitalizzazione composta (Vecchio vs. Nuovo Sistema):

Passando dalla rivalutazione composta del capitale a una semplice somma aritmetica, il MEF (Ministero dell’Economia e delle Finanze) ha di fatto messo un tetto al potenziale di crescita del capitale durante i cicli di alta inflazione. Nel vecchio sistema, se l’inflazione saliva, il capitale cresceva e la cedola successiva veniva calcolata su quella base più ampia. Nel nuovo sistema, il capitale resta fermo a 100. In 5 anni, questa modifica matematica apparentemente piccola erode sensibilmente il rendimento di lungo periodo, perché non guadagni mai “interessi sui tuoi aggiustamenti per l’inflazione.”

La trappola fiscale dell’inflazione:

Il paradosso definitivo delle obbligazioni indicizzate all’inflazione è che lo Stato tratta la tua perdita di potere d’acquisto come una plusvalenza. Se l’inflazione è al 3% e lo Stato ti dà un adeguamento del 3% per non farti perdere terreno, quell’adeguamento viene tassato al 12,5%. Non ti stai arricchendo, stai solo cercando di non impoverirti — eppure vieni tassato comunque. Di conseguenza, il tuo rendimento reale netto sarà quasi sempre inferiore al reale aumento della tua spesa per spesa e bollette.

Un esempio numerico concreto

Vediamo come funziona la matematica nella realtà. Immagina di investire 10.000 € nel BTP Italia Sì, e ipotizziamo un tasso di inflazione semestrale costante dell’1,25% (che equivale a un’inflazione annua del 2,5% — in linea con l’obiettivo BCE del 2% e con la recente traiettoria dell’indice FOI italiano).

Applicazione della formula semestrale:

Cedola Semestrale = 10.000 € × (0,8% + 1,25%) = 205 €

Tasse e freno patrimoniale:

• Cedola Lorda Semestrale: 205,00 €

• Meno Tasse 12,5%: -25,63 €

• Meno Imposta di Bollo semestrale (0,1%): -10,00 €

• Cedola Netta Semestrale: 169,37 €

I totali a 5 anni: Nei 10 semestri totali, incasserai 1.693,75 € di cedole nette. A scadenza, ricevi il tuo capitale di 10.000 € più un Premio Fedeltà netto di 52,50 € (0,6% lordo, meno tasse). Il tuo profitto monetario netto totale è di 1.746,25 €, per un Tasso di Rendimento Interno (TIR) netto nominale del 3,51%. Sembra ragionevole — finché non sottrai il 2,5% di inflazione annua ipotizzata. Il tuo rendimento reale netto è di appena lo ~0,97% annuo. Lo Stato ha tassato la tua protezione dall’inflazione e ti ha restituito una crescita del potere d’acquisto vicina allo zero.

A chi conviene davvero?

A chi è adatto:

Questo bond è ideale per risparmiatori prudenti che hanno una spesa certa e prevedibile in arrivo esattamente entro 5 anni (es. ristrutturazione casa, spese universitarie, acquisto auto). Se hai liquidità inattiva che non puoi permetterti di rischiare sui mercati azionari, il BTP Italia Sì funziona come un eccellente “posteggio”. Non ti farà arricchire, ma assorbirà in modo sicuro buona parte del morso dell’inflazione, dandoti la tranquillità di una garanzia statale al 100% sul capitale.

A chi non è adatto:

Se hai 30, 40 o anche 50 anni e il tuo obiettivo è l’accumulo di ricchezza nel lungo periodo o costruire una pensione complementare, questo prodotto è completamente sbagliato per te. Bloccare il capitale per 5 anni a un rendimento reale netto vicino allo zero è un incubo in termini di costo opportunità. Su orizzonti lunghi, rinunciare a un portafoglio azionario diversificato in favore di obbligazioni governative indicizzate all’inflazione significa lasciare volontariamente che il tuo potere d’acquisto resti fermo.

In conclusione

Non esiste un investimento intrinsecamente buono o cattivo (beh, a meno che non sia fraudolento o garantito in perdita). Esistono solo investimenti mal allineati alla tua vita. Per vincere la partita finanziaria, devi guardare oltre il marketing istituzionale e capire se un prodotto è davvero coerente con il tuo orizzonte temporale, la tua tolleranza al rischio, la tua età e le tue esigenze di capitale.

Paul Krugman called Elon Musk a “human Ponzi scheme.” That’s a strong claim from a (Trump’s words) a “Deranged BUM” who’s. accidentally, also a prominent economist and Nobel laureate — and it’s mostly right, but not entirely.

Krugman published his piece on the eve of SpaceX’s Nasdaq debut, the largest IPO in history. He wasn’t wrong to raise the alarm. He might have swung a little wide, but before you decide what to do with SPCX — buy it, ignore it, or wait — you need to understand where Krugman’s argument holds, where it doesn’t, and what most of retail investors are totally missing.

Because the most interesting part of this story isn’t the valuation. It’s the exit queue.

Subscribe now

What Krugman Actually Argued — and Where He’s Right

Krugman’s core thesis isn’t that SpaceX is a fraud. It’s that Musk’s wealth follows a self-fulfilling faith loop: investors believing in his genius pile into his companies, rising valuations enhance his reputation for genius, which attracts more investors — who push valuations higher still.

“We have a term for enterprises that look successful because they keep drawing in new investors and keep drawing in new investors because they look successful,” Krugman wrote. “They’re called Ponzi schemes.”

Musk has long history of promises that haven’t materialised. The Hyperloop. Commercial Boring Company tunnels. Fully self-driving Teslas. A Mars colony. All of them by 2025. Neuralink has been implanted in a handful of patients and gone no further. Tesla robotaxis are operational in Austin — and nowhere else. Waymo, not Tesla, is winning the autonomous vehicle race in the real world.

That pattern matters, because SpaceX’s valuation isn’t being priced on its current business. It’s being priced on the assumption that Musk will keep delivering miracles — and on the further assumption that retail investors will buy in simply because they believe they’re buying a stake in his genius.

There’s also something more structurally troubling: the index inclusion game. Nasdaq and FTSE Russell both changed their rules to allow SpaceX fast-track entry into major indexes after just 15 trading sessions — a process that normally takes a minimum of a year. Luckily the S&P500 didn’t change its rules. Roughly 52% of mutual fund assets now sit in index-tracking funds, so millions of ordinary Americans will end up holding SPCX not because they chose to, but because the passive allocation machinery decided for them.

The bet gets placed on your behalf.

That’s the Ponzi-like dynamic Krugman is pointing at. And he’s not wrong.

Where He Overstates the Case

But Krugman is also making a category error — and it’s worth separating the legitimate critique from the political theatre.

SpaceX, of course, is not a Ponzi scheme. SpaceX does have an underlying business.

The company generated $18.7 billion in revenue last year. Starlink, its satellite internet service, has 9 million paying subscribers and is the only genuinely profitable division of the business. The Falcon 9 rocket has become the global workhorse of commercial launch — with a track record of reliability that no competitor, state-sponsored or private, can match. Reusable rockets aren’t a promise: they’ve flown hundreds of missions and landed. The Starship programme, though still in development, has made more tangible progress in the last two years than any other heavy-lift vehicle in history.

The underlying business is genuinely solid — and that distinction matters enormously when you’re deciding whether the criticism is about the company or about the price.

Mine is about the price.

The Valuation Is the Problem

SpaceX priced its IPO at $135 per share, raising $75 billion and valuing the company at $1.75 trillion before a single share changed hands. On its first day of trading, the stock closed at $160.95 — briefly pushing the market cap above $2 trillion.

To put that multiple in context:

SpaceX - IPO day (June 2026): ~107x

Tesla - Q4 2021 peak mania: ~26x

Nvidia - Early 2024 AI peak: ~42x

Amazon - Mature growth phase: ~5x

SpaceX is being asked to justify a price-to-sales ratio roughly four times Tesla’s most euphoric valuation — on its first day as a public company, before it has spent a single session reporting earnings to markets, with a Q1 2026 $4.28 billion quarterly loss on its books.

That loss is real. The company spent aggressively on Starship development, satellite constellation expansion, and launch infrastructure. That capital allocation isn’t crazy — it’s the cost of building infrastructure at planetary scale. But it does mean that the justification for a $2 trillion valuation rests entirely on what SpaceX might become, not on what it currently earns.

For the valuation to hold, the narrative needs to stay perfect for a very long time. SpaceX needs to win major Department of Defense contracts. Starlink needs to keep growing and eventually dominate emerging-market internet access. Starship needs to achieve operational status and unlock a new economy of scale in launch costs.

The AI compute angle — Starship potentially serving as a launch platform for satellite-based AI infrastructure — needs to become real.

Every one of those is plausible. None of them is guaranteed. And you are paying for all of them, today, at full price.

The Exit Queue Nobody Is Talking About

Here is the part of this story the financial press is underreporting — and it’s the part that matters most to anyone holding or considering SPCX in the next six months.

The early investors in SpaceX are sitting on gains that are almost impossible to comprehend.

Founders Fund, Peter Thiel’s firm, put in approximately $600 million across multiple rounds over nearly two decades. At the IPO price, that stake was worth more than $50 billion — roughly an 80-fold return, and one of the largest single venture wins in history. Sequoia Capital invested about $2 billion and is looking at returns exceeding $20 billion. Andreessen Horowitz holds a position now worth over $10 billion — the largest return in that firm’s history. These firms got in when SpaceX was worth less than $1 billion, more than 15 years ago.

Not one of them sold a single share in the IPO itself. The offering was entirely new shares issued by SpaceX — zero secondary sales.

They couldn’t sell. And the staggered lock-up schedule tells you exactly when they can.

The structure, drawn from SpaceX’s S-1 filing, works like this:

• After Q2 2026 earnings: up to 20% of eligible insider shares unlock immediately

• Bonus tranche: an additional 10% unlocks if SPCX trades 30%+ above the IPO price for five of ten consecutive trading days

• Days 70, 90, 105, 120, and 135: 7% tranches free up at each milestone

• After Q3 2026 earnings: a further 28% of locked shares become eligible for sale

• Day 180 (mid-December 2026): the remaining shares subject to standard lock-up become fully tradeable

Elon Musk himself sits outside this schedule entirely — he and certain major backers committed to a full 366-day lock-up. That’s a meaningful positive signal, and it’s worth noting.

But Musk is one person. The aggregate supply picture from everyone else is significant.

Every one of these venture funds — Founders Fund, Sequoia, Andreessen Horowitz, DFJ, Valor Equity Partners — has a legal obligation to return capital to their limited partners. Their LPs are pension funds, endowments, and institutional allocators who have waited years for liquidity. The fiduciary clock is ticking. These are some of the most sophisticated early-stage investors on earth, sitting on 15-year-old positions, 100x gains, and the first real exit window of their lives.

The staggered structure blunts the supply impact by spreading it out. It does not make that supply disappear.

Every unlock gate is a date on which rational, fiduciary-constrained investors have both the legal right and the institutional obligation to sell. Not all of them will. Not immediately. But the pressure is systematic — and the market will begin pricing it in well before each gate opens.

The clock started on June 12th.

What a Rational Investor Should Do With This

The company is exceptional. The valuation is stretched. The exit overhang is real and underappreciated. Those three facts can all be true simultaneously — and for SPCX, they are.

What that implies in practice:

If you’re an index fund investor, you don’t have a choice. SPCX will be in your Nasdaq-100 tracker within 15 sessions. That’s not a reason to panic — it’s just a reason to know what you own.

If you’re an active investor considering a position, the question isn’t whether SpaceX is a great company. It is. The question is whether this is the right price. At 107x sales with a $4.28 billion quarterly loss and a wall of insider supply coming through the back half of 2026, the answer for most investors should be: wait.

The IPO premium will not last forever. The lock-up gates will create selling pressure. There will likely be better entry points in Q3 and Q4 — particularly around the Q3 earnings report, when the largest single post-IPO unlock (28%) becomes eligible.

If you already hold it from the IPO, that depends entirely on your time horizon and risk tolerance. If you believe in the Starship thesis and the AI infrastructure angle over a 5-to-10-year horizon, the entry price matters less. If you’re a short-to-medium-term investor, the risk/reward looks unfavourable from current levels.

Krugman’s “human Ponzi scheme” framing is too blunt and too politically charged to be fully useful. But the structural observation underneath it — that SPCX’s valuation is being sustained by narrative and passive-flow mechanics rather than current earnings power — is correct.

The company is real. The business is real. The valuation is the bet. And the most informed sellers on the planet are sitting in a queue waiting to take the other side of that trade.

Know what you’re buying.

Marco

Enjoyed this piece? If you found it useful, feel free to share it. And if you’re thinking about the broader IPO mechanics — passive flow, index inclusion, and what it means for your portfolio — that’s worth a deeper look too.

Wars, as a rule, end. Markets price them as temporary. The playbook is familiar: a spike in volatility, a flight to safety, a gradual normalization once the guns fall quiet. The Iran crisis, now in its fourth month, is following a different script — and markets are only beginning to understand why.

What began on Feb 28 as a surgical US-Israel military operation has become the largest oil supply disruption in recorded history. The de-facto blockade of the Strait of Hormuz has wiped from the market approximately 10 billion cubic feet per day of LNG and an estimated 12-18 mbpd of crude. Commercial shipping traffic dropped more than 90% after the outbreak of hostilities. The economic aftershocks are radiating outward in ways that a ceasefire alone cannot reverse.

The paradox at the heart of this crisis: a conflict (that looks temporary) is likely producing permanent economic rewiring. The disruption is not just removing barrels from the market — it is simultaneously destroying demand and reshuffling every winner and loser across the global energy complex. Understanding which of those forces dominates is the defining investment question of 2026.

Subscribe now

I. The Ceasefire-With-No-Oil Purgatory

Diplomatic language has a way of making complex problems sound tractable. Negotiations, ceasefires, frameworks for withdrawal — the political vocabulary of resolution is fluent and fast. The physics of oil supply isn’t.

Even if a diplomatic settlement were reached tomorrow, the Strait of Hormuz would not immediately reopen for business. The waterway has been the scene of active military operations, including attacks on tankers, the placement of naval mines, and systematic strikes on port infrastructure throughout the Gulf. The best-case scenario — what energy analysts have labelled the “Tehran Tollbooth” — requires Iran to entirely cede operational control of the strait and grant free passage to all vessels. That is a political outcome that would require extensive verification, demining operations, and a rebuild of insurer confidence before commercial traffic normalizes.

Meanwhile, the market has been sedated. Crude oil hovering around $100 per barrel might look like a crisis premium, but it is actually an underpricing of the structural damage. The mechanism behind this illusion is the US Strategic Petroleum Reserve. The release of roughly 400 million barrels from the SPR has papered over a supply loss estimated at 12 to 18 million barrels per day — a number that, unmasked, would put prices at levels not seen in the modern era.

The World Bank’s April Commodity Markets Outlook confirmed the scale: global oil supply crashed by 10.1 million barrels per day in March alone, the largest single-month disruption in history. Brent crude recorded its highest monthly rise ever — roughly 65% — before partially retracing on ceasefire hopes. Those hopes have since stalled, and the SPR buffer has a finite life.

The inventory clock is ticking. As summer peak-demand season progresses and reserves approach what traders call “tank bottoms” — the point at which strategic stockpiles are effectively depleted — the SPR fiction ends. At that point, the market will have to price reality.

The LNG situation is even worse. The strait carried approximately 10 billion cubic feet per day of LNG, with about 93% of Qatar’s and 96% of the UAE’s LNG exports transit through the strait. Qatar — the world’s second-largest LNG exporter — suspended production entirely. Unlike oil, there is no alternative route for LNG: no bypass pipeline exists, so the impact on global gas markets is structurally total for Gulf-origin volumes.

II. The Demand Destruction Trap

There is a more dangerous story embedded in the price charts — one that threatens oil-producing nations and energy investors far more than the supply disruption itself.

High energy prices kill demand. This is not a theory; it is the mechanism that has triggered every major transition away from fossil fuels in modern history. The 1970s oil shocks did not just hurt consumers — they gave birth to the nuclear industry, the first serious wave of fuel efficiency standards, and the seeds of what would eventually become the modern renewables sector. The 2026 crisis is operating on the same logic, but at a fundamentally faster pace — what took a decade in the 1970s is playing out in months.

Current estimates suggest that elevated energy costs are wiping out between 3 and 8 million barrels per day of global demand. Some of this is cyclical — people driving less because petrol is too expensive, airlines grounding routes where yields no longer cover fuel costs, factories running at reduced capacity. This demand will return when prices normalize.

But a substantial portion is turning structural, and that is where the danger lies. Structural demand destruction does not reverse. A factory that converts to electric arc furnaces does not switch back to oil-fired kilns when prices fall. A household that installs rooftop solar and heat pumps does not go back to the grid utility. A government that fast-tracks an EV fleet programme does not cancel it when geopolitical risk premiums subside.

This distinction — cyclical versus structural — is the focus of the most uncomfortable conversation happening inside energy boardrooms right now.

The Brent futures curve is already pricing in that world: front-month contracts at $94, December 2027 at $78, converging toward the $80 floor that structural demand loss implies. The market’s own forward view is skeptical that today’s elevated prices are durable.

The uncomfortable arithmetic before every energy executive is this: endure $140 oil for one volatile year, or suffer $80 oil for five years as permanently destroyed demand resets the long-term price floor. The $80-for-five scenario is not a fringe view — it is what structural demand loss at scale looks like in a discounted cash flow model.

III. Why the Supply Switch Won’t Flip Back On

Markets are pricing in a V-shaped recovery. The consensus narrative runs roughly as follows: conflict ends, Hormuz reopens, Persian Gulf barrels flood back to market, prices normalize. This narrative is wrong on almost every dimension.

The physical bottlenecks alone make a rapid recovery impossible. Naval mines deployed in the strait require methodical clearance — a process measured in months, not days, even under optimal conditions. Insurance markets, which effectively collapsed for vessels transiting the region at the outbreak of hostilities, will not re-engage until sustained safe passage is demonstrated. The world’s largest tanker operators — Frontline, Euronav, DHT — suspended Hormuz transits entirely, and rebuilding that commercial confidence requires a track record, not just a ceasefire declaration.

The infrastructure damage compounds the problem. Beyond the strait itself, the conflict has involved strikes on energy infrastructure across the Gulf Cooperation Council — Saudi Arabia, Kuwait, and the UAE have all absorbed attacks on oil facilities and coastal installations. Saudi Arabia’s Ras Tanura refinery — 550,000 barrels per day, the kingdom’s largest — shut down entirely after coming under attack, a single facility loss larger than many OPEC members’ total output. Production systems that have been shut down for extended periods do not simply restart at the flick of a switch. Reservoir pressure management, wellbore integrity checks, pipeline recommissioning — these are complex industrial processes that take time even without battle damage to contend with.

Saudi Arabia’s East-West Pipeline and the UAE’s Fujairah bypass route offer partial alternatives to Hormuz, but their combined theoretical capacity of roughly 8.8 mb/d covers less than half of the strait’s pre-conflict throughput of 20 mb/d — and that assumes both pipelines are running at full capacity with no infrastructure damage, which they are not.

And looming over all of this is the inventory reality. Global observed oil inventories drew by 129 million barrels in March and a further 117 million in April, according to IEA data. The cumulative supply losses from Gulf producers have now exceeded one billion barrels. The SPR releases that have masked the crisis cannot continue indefinitely — and once they run out, there is no second buffer.

IV. Oil & Energy Stocks: Who Weathers the Storm — or Profits From It

Not all energy stocks are equal in a crisis of this configuration. Geography, operational profile, and balance sheet quality are separating this cycle’s winners from its casualties in ways that require a more granular lens than simply “energy is up.”

The Obvious Winners: US Shale and Gulf of Mexico Operators

The clearest beneficiaries are producers geographically insulated from the disruption. US shale companies — EOG Resources, Devon Energy, Occidental Petroleum, Diamondback Energy, ConocoPhillips — are cashing windfall margins at a safe remove from Hormuz. The S&P 500 Energy Index has outperformed every other major sector in 2026, up approximately 29% year-to-date, and shale pure-plays have led that charge with double-digit returns since the conflict began. The Permian Basin, producing around 6.5 million barrels per day, is structurally insulated and operationally mature.

The math at current prices is compelling. Enverus estimates the marginal breakeven for US shale at around $70 per barrel WTI today. At $100 Brent, the margin per barrel for efficient operators is extraordinary. ExxonMobil’s cube-development approach in the Permian — drilling multiple stacked horizons from a single surface pad — allows it to target doubling output to 2.5 million barrels of oil equivalent per day by 2030 without proportional capex increases. Scale and technology are compounding the windfall.

The Nuanced Winners: LNG and Midstream Infrastructure

If shale is the obvious trade, LNG exporters are the elegant one. Cheniere Energy has surged approximately 45% year-to-date, trading above $280 as of mid-March, as European and Asian buyers scramble for alternatives to Persian Gulf supply. With over 95% of capacity contracted for the next decade and new CCL Stage 3 trains completing in 2026, Cheniere offers both crisis-driven upside and long-term revenue visibility that is rare in this sector. Venture Global, with large-scale Louisiana Gulf Coast facilities, has seen similar demand dynamics.

Midstream infrastructure — Enbridge, Kinder Morgan, Williams Companies — operates on a different logic entirely. Fee-based revenue structures mean these companies earn whether volumes are high or low, as long as pipelines are moving product. They are volume-agnostic in a crisis that is volume-disrupted. Enbridge’s 30% share of North American oil transport gives it structural relevance that geopolitical shocks cannot easily displace.

The Split Verdict: Integrated Majors

ExxonMobil and Chevron hit all-time-high market capitalizations in the first weeks of the conflict, up roughly 30% year-to-date as of mid-March. But the integrated picture is more nuanced than the headline numbers suggest. Upstream profits are being partially offset by refining margin compression — refineries configured for Middle Eastern heavy-sour crude grades are facing feedstock disruptions that their downstream margins cannot fully absorb.

Chevron’s portfolio diversification through its Hess acquisition provides some insulation that pure-plays lack. But “integrated” in this crisis means exposure in both directions, not just one.

The Exposed Names

On the other side of the ledger: oilfield services companies face a paradox. Higher prices theoretically drive drilling activity, but geographic disruption and operational uncertainty have caused SLB, Baker Hughes, and Weatherford International to decline sharply since the conflict began — down 12%, 12%, and 18% respectively from their pre-war peaks. Activity in the Middle East, historically a core revenue geography for services firms, has effectively paused.

Asian import-dependent economies — Japan, South Korea, India — face the most acute structural exposure. India has pivoted sharply toward Russian crude to partially compensate, while China is leveraging its position as both a major importer and the world’s dominant clean energy manufacturer to play both sides of the transition.

The Dual Stress Test

The right analytical lens for this environment is not simply “who wins at $140” — it is who passes the dual stress test: operational profile to exploit $140 today AND balance sheet durability to survive $80 for five years if demand destruction sets the long-term floor. That combination — short-term harvest capacity plus long-term resilience — is what separates the compounders from the commodity plays.

V. The Clean Energy Accelerant: How the Iran Crisis Is Reshaping the Green Transition

The oil shock generating record windfall profits for fossil fuel producers is simultaneously funding and accelerating the transition that will eventually obsolete their core business.

The Investment Pivot

Global energy investment is forecast to reach $3.4 trillion in 2026 — of which $2.2 trillion is flowing into clean energy categories. Solar PV alone is attracting $365 billion, more than $1 billion every single day, according to IEA projections. The Iran crisis did not create this trajectory, but it has injected urgency into it. As of May 1, 2026, twenty-three countries across five continents had made public clean energy and electrification announcements specifically citing energy security concerns triggered by the Hormuz disruption.

The IEA’s executive director, Fatih Birol, said it directly: “I expect one of the responses to this crisis will be an acceleration of renewables. Not only because they are helping to reduce emissions, but also because they are a homegrown domestic energy source.” Countries that diversified their energy mix before the crisis are absorbing the shock with measurably less economic damage — a real-time validation of the transition investment thesis that is difficult to argue with.

Policy Acceleration

The crisis is giving governments political cover to fast-track measures that had previously stalled against fossil fuel industry lobbying and public cost concerns. The UK government introduced rules requiring heat pumps and solar panels in all new homes. The pattern is the same everywhere: energy security has become the argument that overrides the objections.

In the United States, March 2026 marked a milestone: for the first time on record, the country generated more electricity from renewables than from natural gas across a full calendar month. The ripple effect is global. Indonesia’s President Prabowo Subianto announced a commitment to deploy 100 gigawatts of solar within three years. The Philippines fast-tracked 1.47 gigawatts of renewable energy and storage. In the UK, over 27,000 households installed rooftop solar in March alone — a single-month record — as families sought energy cost certainty that fossil fuels can no longer provide.

The 1970s oil shocks, as the Carlyle Group’s Jeff Currie and former NATO commander Admiral James Stavridis have observed, gave birth to solar, wind, and nuclear as industries — driven by security imperatives, not environmentalism. Today’s crisis is operating on the same logic, but with technologies that have already crossed the cost threshold. IRENA data now shows that 24/7 renewables — solar, wind, and battery storage — can beat new coal and gas on cost in many parts of the world. The security argument and the economic argument now point in the same direction.

The Investment Paradox

Clean energy ETFs attracted over $3 billion in April 2026, the largest monthly net inflow on record. The source of that capital is, in part, the windfall profits being generated by the oil sector itself — institutional investors using elevated energy stock dividends and buyback proceeds to rotate into long-duration clean energy positions. The fossil fuel crisis is, with grim efficiency, funding its own succession.

Chinese exports of solar technology, batteries, and electric vehicles all reached record highs in March 2026. As the world’s dominant clean energy manufacturer, China is positioned to be the primary industrial beneficiary of a disruption it had no hand in creating.

The Caveat: The Transition Is Accelerating, Not Frictionless

The clean energy narrative in this crisis risks overreach if left unqualified. Supply chains for the transition are not immune to geopolitical pressure. Rare earth minerals, polysilicon, and copper — the physical inputs for solar panels, wind turbines, batteries, and EV motors — face their own concentration risks and trade tensions. China’s decision to scrap export incentives on solar panels in March 2026 is already expected to raise infrastructure costs abroad, a reminder that decarbonization has its own geography of vulnerability.

The transition is real, it is accelerating, and the Iran crisis has supercharged it. But it is not a clean escape from geopolitical risk — it is a reallocation of that risk onto a different set of supply chains.

The Fork in the Road

The energy industry’s existential question has been reframed by this crisis. The debate is no longer only about the price of a barrel of oil in the near term. It is about which world is being priced in — the world where the supply disruption resolves and oil resumes its role as the dominant global energy commodity, or the world where structural demand destruction has already reset the long-term equilibrium.

The $140-versus-$80 arithmetic is real and urgent. But it hides a deeper question. Even the $140 scenario is, in structural terms, self-defeating — every month of $140 oil accelerates the demand destruction that produces the $80 world. The crisis is eating its own tail.

History’s most significant oil shocks were not necessarily the ones that sent prices highest. The 1973 embargo mattered less for its peak price than for the permanent behavioural and policy changes it triggered. If 2026 is the year when demand destruction shifted from cyclical to structural — the year when energy security stopped being an argument for fossil fuels and became an argument against them — then the Iran conflict may be remembered not as the last crisis of the oil era, but as the first crisis of its unravelling.

The last oil shock may not be the one that makes prices highest. It may be the one that makes high prices irrelevant.

Sources: World Bank Commodity Markets Outlook (April 2026); Dallas Fed Research (March 2026); IEA World Energy Investment 2026; Morningstar energy sector analysis; Kpler market research; Britannica / Congress.gov on Hormuz disruption; Zero Carbon Analytics; Fortune, CNBC, Axios energy coverage.