Wall Street got the American election broadly right. It got Brazil wrong. And somewhere along the way, one of the most successful trades of recent years began coming apart.
For much of this AI boom, the winning formula seemed obvious. Own the companies building the infrastructure, avoid the businesses threatened by the technology, and trust that extraordinary investment would eventually produce extraordinary profits.
Then investors began asking who was going to pay for it all.
That question followed us throughout October, from the US Treasury market to the electricity bills of American households, from the ballot boxes in Brazil to two overlooked corners of the software industry. It also led us to make several portfolio decisions that have become particularly relevant now that the selling has intensified.
Because what has happened since is remarkable.
The major stock indices remain relatively close to their highs, yet beneath the surface some of the market’s most celebrated winners have suffered violent reversals. Hedge funds have been reducing exposure, yesterday’s neglected businesses have begun attracting buyers, and sentiment has fallen to levels normally associated with considerably more painful markets.
Is this the beginning of something much worse? Or are investors creating the conditions for another attractive entry into a bull market whose underlying economic drivers remain powerful?
Those are the questions connecting October’s research with the work we have already begun publishing to subscribers in November.
And the answers may be considerably more interesting than the headline indices suggest.
The bond market sent the first warning
Long before the election results arrived, the cost of money was already changing the investment landscape.
By early October, the yield on the thirty-year US Treasury had climbed to levels not seen for almost two decades. The consequences were spreading well beyond bond portfolios. Mortgages were becoming more expensive, companies were confronting higher financing costs, and investors were demanding greater compensation for owning businesses whose profits might not arrive for many years.
This was the subject of The Market Sets the Terms, the first public excerpt from October’s VMF’s Strategic Asset Allocation.
The uncomfortable part was that some of this pressure was coming from the very transformation we remain so optimistic about.
Artificial intelligence requires an extraordinary amount of capital before it can deliver its promised productivity gains. Data centres, advanced processors, electricity generation, transmission networks and new infrastructure are all competing for financing at the same time governments are borrowing heavily to meet their own commitments.
Our Abundance Shock thesis has always recognised that tension. Building a more productive future can be expensive, even if using the technology eventually makes large parts of the economy cheaper to operate.
But the bond market was beginning to impose a more immediate discipline.
Higher borrowing costs were putting pressure on the assets most dependent on cheap money. And that raised a question we had been investigating since September: what if the market causing so much discomfort was approaching the point where it could begin providing protection?
We decided the opportunity was becoming sufficiently interesting to act.
On 9 October, Tier One introduced a small allocation to long-duration US government debt, funded from short-duration Treasury exposure. Two weeks later, Alpha Tier made a related adjustment, this time reducing part of its China equity exposure to create additional protection against a disinflationary growth shock.
We were not predicting an imminent recession. Nor were we assuming that government bonds would protect investors against every possible crisis.
We were recognising that, after years of painful losses, the potential rewards from owning duration had improved, while the possibility of renewed risk-off episodes was becoming harder to ignore.
The distinction matters. A portfolio designed to benefit from a productive future should also be capable of surviving the financial stress involved in building it.
The future was on the ballot
Meanwhile, a very different version of the same argument was playing out in politics.
In October’s Alpha Tier, we returned to Strauss and Howe’s Fourth Turning framework, which has shaped our interpretation of the changing economic and political regime since VMF Research began publishing.
Our attention was on a problem that repeatedly appears during periods of institutional strain. Governments promise prosperity, security and national renewal. Businesses need capital to deliver those ambitions. Households want the benefits without unbearable costs. Eventually, someone has to pay.
The AI revolution was providing an increasingly visible example.
To Washington, a new data centre represents technological leadership and economic progress. To the family living nearby, it may represent higher electricity bills, pressure on local infrastructure and another reason why the cost of living refuses to become more manageable.
We argued that this conflict was moving directly onto the American electoral agenda, and that political resistance could become an important constraint on the economics of the AI infrastructure buildout.
The election subsequently confirmed the relevance of those concerns.
Democrats swept both chambers of Congress on 3 November, delivering a result whose broad direction Wall Street had anticipated. Brazil, however, offered a reminder of the limits of political forecasting. Markets had increasingly favoured a change in government after Flávio Bolsonaro’s strong first-round showing, only for Lula to win the 25 October runoff.
We returned to those outcomes in The Market Voted First, our public follow-up to October’s Alpha Tier research.
What interested us most was not simply who won.
It was the changing political willingness to support expensive ambitions without asking harder questions about their costs.
Data-centre demand may continue growing rapidly, while new restrictions, delayed approvals, higher utility obligations and more expensive financing change which companies can profit from meeting that demand.
For investors, the distinction is crucial.
A technological revolution can continue even as the investment returns migrate away from its original winners.
And that is precisely what we were beginning to observe.
The great AI rotation
One of the most revealing charts in October’s Tier One issue had nothing to do with artificial intelligence.
It showed that, by 5 October, only about 28% of S&P 500 constituents were trading above their 50-day moving averages.
The index appeared healthy. Most of its underlying stocks did not.
We explored that contradiction in A Narrow Road Higher.
At the time, the largest technology companies were providing much of the market’s leadership, while large sections of the equity universe were already struggling. Yet earnings expectations remained relatively strong, professional investors were cautiously positioned, and the historical seasonal backdrop was becoming more favourable.
We had good reasons to question the foundations of the advance without concluding that the advance had to end.
Then the rotation accelerated.
The popular strategy of owning AI infrastructure beneficiaries while betting against businesses expected to lose from AI began producing losses on both sides. The winners fell sharply. The presumed victims recovered. As leveraged investors reduced exposure, those movements reinforced one another.
This is the mechanism behind degrossing. A manager who needs to reduce risk may sell a position they still fundamentally like, while buying back a short position in a business they still fundamentally distrust.
The price moves can become extraordinary without an equally extraordinary change in the underlying companies.
And that brings us to one of the most important investment themes we have been developing for months.
The Application Layer.
We have argued for some time that the next phase of the AI opportunity could increasingly reward businesses capable of using cheaper intelligence, rather than only those supplying the infrastructure required to produce it.
That distinction is now attracting considerably more attention. But identifying a promising part of the technology stack is only the beginning. We still need to find companies with durable customer relationships, attractive economics and valuations that offer a reasonable return on the capital invested.
October presented two particularly interesting opportunities.
Both businesses operate in specialised software markets where customers depend on systems built around years of accumulated knowledge, regulatory requirements and operational experience. Both have experienced dramatic share-price declines as investors questioned whether AI would make their products easier to replace.
And both continued to generate meaningful cash.
In The AI Trade’s Biggest Blind Spot, we explained why writing software and replacing a dependable business system are not necessarily the same thing.
A hospital cannot afford to discover that a new application fails to preserve critical records. A public transport operator cannot casually replace a system responsible for coordinating thousands of daily journeys. The customer pays for much more than the code on the screen.
That does not make these businesses immune to disruption.
It does mean that the market may be overlooking some of what customers actually value.
October’s paid VMF’s Security Selection followed that argument into two new company recommendations, examining their acquisition economics, cash generation, competitive advantages and valuations.
The public article explains the opportunity. The complete investment work, including the two companies and our Model Portfolio decisions, remains with subscribers.
November is making the question more interesting
By the time the elections were over, October’s rotation had developed into something more violent.
The major indices still offered a relatively reassuring picture, but investors concentrated in several fashionable AI exposures had suffered a very different experience.
Our subsequent The Index Tells Half the Story provided another window into this dispersion, including how some previous winners were retreating while several neglected positions were finally beginning to attract capital.
The selling also changed the mood.
During the first week of November, CNN’s Fear & Greed Index fell into Extreme Fear. By 13 November, the reading had recovered to 39, still in Fear territory. Investor surveys and other positioning indicators were showing similarly uncomfortable levels of confidence.
That deterioration has occurred without the sort of broad-index collapse investors usually associate with such anxiety.
Now consider the calendar.
November begins what has historically been the stronger six-month period for US equities. The period following American midterm elections has also tended to produce favourable twelve-month returns, although those historical patterns offer no guarantee about the next several weeks.
The combination deserves attention.
Investors have reduced exposure. Confidence has deteriorated. Valuations in selected industries have fallen considerably. Meanwhile, many of the long-term investment themes we have been following remain intact.
There are still important obstacles. Real yields and the US dollar have recently strengthened, financial conditions remain demanding, and a genuine deterioration in earnings could make today’s lower valuations less attractive than they appear.
But the possibility of a meaningful recovery is becoming increasingly interesting.
The question is whether the market has already discounted more disappointment than the businesses themselves are likely to deliver.
Which brings us to where our research is going next.
Where we went shopping
Last Friday, 13 November, we introduced a new investment idea to subscribers in a very specific niche of the AI economy.
It is an area where expectations have retreated meaningfully, even as we believe the longer-term commercial opportunity remains substantial. The recent repricing has changed the relationship between what investors are being asked to pay and what the industry might ultimately become.
We are not treating a fallen price as sufficient evidence of value. The investment still has to justify itself through real adoption, attractive economics and the ability to turn technological progress into shareholder returns.
But the setup has become interesting enough for us to act.
And it brings the argument back to the question running through October’s research.
What happens when investors become so concerned about the cost of building the future that they begin overlooking the businesses capable of profiting from it?
We will explore that question in this week’s articles on VMF’s Market View, making part of the underlying research available to free readers.
Paid subscribers received the complete recommendation and its portfolio implementation on Friday.
That is how we intend these publications to work.
October was not a month in which we abandoned our convictions. We reconsidered their risks, introduced protection where we believed it was warranted, and looked for opportunities where prices had become more attractive.
November is already providing a new test of those decisions.
The future has not become less interesting. The market has simply become less willing to pay for it.
And for investors prepared to distinguish between the two, that may be where the next opportunity begins.
Good investing!
Vasco Marques de Freitas, CFA, CMT
Read the full analysis and access our complete portfolio recommendations.
Promotional communication for VMF Research’s independent financial research. Not personalised investment advice.
Important Disclosure
This article was produced by VMF Research, Lda., under the responsibility of Vasco Marques de Freitas, CFA, CMT, Founder and Chief Executive Officer. It contains general investment research and market commentary, including references to investment recommendations. VMF Research’s activity as a producer of investment recommendations is subject to the applicable European market abuse framework, including Regulation (EU) No 596/2014 and Commission Delegated Regulation (EU) 2016/958, and to supervision by the Comissão do Mercado de Valores Mobiliários (CMVM) in Portugal. Such supervision does not constitute regulatory approval or endorsement of the research.
The article summarises research published during October 2026 across VMF’s Strategic Asset Allocation, VMF’s Security Selection and Alpha Tier, together with public excerpts and subsequent market developments through November 2026. Original research, investment decisions and Model Portfolio changes retain their respective publication dates. Subsequent events, including election results and market movements, should not be interpreted as information available when the original recommendations were made. The underlying publications contain the complete analyses, valuation assumptions, methodologies, principal risks and dated disclosures.
This content is not personalised investment advice and does not take account of any reader’s financial circumstances, investment objectives, knowledge, experience or risk tolerance. It does not constitute portfolio management, order execution or an offer to transact in any financial instrument. The article also promotes VMF Research’s paid research publications.
The Fourth Turning, Abundance Shock, Dovish Shock and Application Layer frameworks represent analytical interpretations and conditional investment hypotheses, not guarantees of future economic, political or market outcomes. Election-related analysis, historical seasonality, sentiment indicators and expectations of a potential market recovery are inherently uncertain. The long-duration Treasury exposures discussed may benefit from falling yields but can experience substantial losses if yields rise, including during periods of financial stress. No hedge provides protection under every market scenario.
Model Portfolios are illustrative research portfolios, not client accounts, investable funds or transactions executed by VMF Research. References to portfolio allocations, recommendations and performance are based on published research methodologies and reference prices. Unless otherwise stated, reported returns exclude taxes, transaction costs, custody charges, bid–ask spreads and foreign-exchange effects. The Leaderboard presents selected recommendations rather than a complete performance record. Past performance, Model Portfolio results and forward-looking scenarios are not reliable indicators of future returns. All investments involve risk, including the possible loss of capital.
Disclosure of Interests: the original October publications disclose relevant long positions held by legal entities controlled by the author, including interests in Chinese equities, innovation-related funds, Altius Minerals, Rémy Cointreau, Evolution, Scottish Mortgage Investment Trust, Coinbase Global and other securities referenced in the underlying research. Controlled entities also hold listed ETPs providing economic exposure to Bitcoin, Ether and Solana. These interests may create actual, potential or perceived conflicts of interest and may have changed since the original disclosures. The complete dated disclosures accompany the respective publications. References to the new November recommendation should be read alongside the disclosures in that original research.
The October publications state that neither VMF Research nor the author received compensation from covered issuers in connection with their preparation, and that no covered issuer reviewed, approved or amended their investment conclusions before first dissemination.
Readers should conduct their own due diligence and, where appropriate, consult an authorised financial intermediary or other suitably qualified professional before making investment decisions.


