2025 US Stock Market Annual: New Walls Erected, Fences Collapsed, How to Find Anchors for a New Order in 2026?

Written by: Frank, Maitong MSX Research Institute
All that is past is prologue.
As we approach the end of 2025, looking back at the past 12 months of the U.S. stock market and global financial markets, it is difficult to summarize the year with linear terms like "upward" or "pullback." Instead, it resembles a series of intense, mutually compressing structural changes—the simultaneous occurrence of technological acceleration, capital expansion, political polarization, and institutional loosening, all amplified within the same cycle.
The author has attempted to approach this from a timeline perspective and has also tried to use market fluctuations as the main axis, but soon realized that what truly shaped the market landscape of 2025 was not a few landmark highs or crashes, but a set of recurring, overlapping main narratives. When connected, the underlying tone of 2025 becomes exceptionally clear—it is a year of high contradictions, yet with a strong sense of direction:

• On one side, high walls continue to rise: AI builds new entry barriers with extremely high capital density, tariffs and trade frictions are repeatedly intensified, political polarization deepens, and the 43-day government shutdown directly showcases the "intensification of party strife" in such a straightforward manner;
• On the other side, fences are collapsing: the regulatory stance on AI/Crypto shows systematic loosening, financial infrastructure undergoes comprehensive upgrades, and Wall Street reshapes "trading, clearing, and asset forms" in a more open and engineered manner;

In other words, 2025 is a macro watershed: the old order is reinforcing boundaries, while the new order is dismantling frictions, and the collision of these two forces constitutes the background noise of all market movements and narratives this year (for further reading, see “2025 Global Capital Market Flashback: AI Infrastructure Expansion, Confirmation of Interest Rate Cut Turning Point, and Return of Geopolitical Risk Premium”).
From the results, 2025 is undoubtedly a year worth remembering for investments—if you successfully avoided the violent fluctuations of Crypto, you could feel a long-lost experience of making money amidst the broad rise of U.S. stocks, Hong Kong stocks, A-shares, and major assets like gold and silver. Just like at the recent Xueqiu Carnival opening, Fang Sanwen typically threw out that classic question: "Who made money this year? Please raise your hand?" The audience responded with a nearly simultaneous wave of arms.
Because of this, the author ultimately abandoned the idea of narrating 2025 through a single timeline and no longer attempted to summarize the year with a few crashes and new highs—this article chooses to break it down into ten recurring, overlapping main narratives to review the key turning points of the U.S. stock market and global financial markets in 2025 and to attempt to answer a longer-term question:
What has truly changed this year?

1. Convergence of Power: The Right in Silicon Valley, New Crypto Elites, and New Washington

On January 20, 2025, with the official inauguration of the new U.S. government, the right in Silicon Valley and the new crypto elites completed a rare convergence of power and quickly launched a "lightning war" against the traditional establishment. It can be said that this convergence directly manifested in a series of disruptive personnel arrangements, shifts in policy priorities, and changes in regulatory attitudes.
The first to be pushed to the forefront was, of course, Elon Musk, wielding the surgical knife of D.O.G.E (Department of Government Efficiency). During his brief but high-pressure intervention, he targeted the long-overlapping regulatory system, especially in the AI-related fields, promoting the withdrawal or merger of regulatory functions related to AI within the Federal Communications Commission (FCC) and the Federal Trade Commission (FTC). This move, to some extent, broke the traditional intervention paths of the bureaucratic system on technological boundaries, significantly reducing institutional friction in emerging technology fields like AI.
Following this, the crypto industry welcomed an epic vindication. With Gary Gensler stepping down as chairman of the U.S. Securities and Exchange Commission (SEC), the "enforcement-style regulation" that had long hovered over the crypto market began to loosen. The new SEC chairman, Paul Atkins, quickly pushed for the release of a statement on "Securities Issuance and Registration in the Crypto Asset Market," shifting the regulatory logic towards rule-making and project compliance.
Under this systematic shift, several unresolved cases reached a phase of resolution, including the long-term investigations and accusations against projects like Coinbase (COIN.M), Ripple, and Ondo Finance being gradually withdrawn or downgraded. Crypto officially returned from being an enforcement target to the policy discussion table.
What is even more intriguing is the deep connections between the new government's core team members and tech capital and crypto capital: from the Trump family's TRUMP/MELANIA token to direct involvement in projects like WIFI (USD1), to Treasury Secretary Scott Bessent, Commerce Secretary Howard Lutnick, Director of National Intelligence Tulsi Gabbard, and Secretary of Health and Human Services Robert F. Kennedy Jr. It can be said that from the president and his core aides to a group of cabinet officials, a batch of decision-makers deeply embracing AI, the Silicon Valley tech right, and even Crypto are systematically entering the power center.
At the same time, the attitude of New Washington towards AI also underwent a fundamental change, shifting from "Eliminating Barriers to U.S. AI Leadership," "U.S. AI Action Plan," to the "AI Custodianship Act," with the policy narrative switching from "preventing risks" to "ensuring (absolute lead over China)," which has allowed many vertical AI companies to thrive in the secondary market.
The most representative examples are companies like Palantir (PLTR.M) and Anduril (not yet listed), which carry strong right-wing colors and "tech patriotism" labels, becoming some of the hottest targets on New Wall Street in 2025, with their market capitalization and valuations rapidly rising.
Objectively speaking, over the past decade, new technology fields represented by AI and Crypto have become engines of wealth growth, emphasizing efficiency, innovation, and decentralization, with the call for "less regulation" becoming a common demand. Thus, the convergence of the Silicon Valley right and the new crypto elites is essentially a phase of revelry centered around "technological freedom, capital efficiency, and regulatory loosening," with the market even momentarily viewing deregulation and technological supremacy as the only path to prosperity.
However, the problem is that the stability of this narrative itself is worth caution, as the paths of liberalization and deregulation will inevitably further strengthen the advantages of tech giants and capital giants—technologies like AI and Crypto will more efficiently concentrate wealth, accelerating the widening wealth gap and marginalizing the interests of Trump’s base, such as Rust Belt workers, social conservatives, and anti-globalization middle-class individuals (for further reading, see “The Clash of Old and New Money Behind the 'Trump Trade': Peter Thiel and the Power Games of Crypto”).
As the 2025-2029 political cycle progresses, especially with the 2026 midterm elections approaching, electoral pressures and macro and fiscal constraints will gradually return to the policy core. The originally seemingly solid capital alliance is likely to show signs of differentiation. The author has always believed that how to find a balance between the "new money" pursuing efficiency and the "old money" maintaining stability in the next four years and beyond will determine the final direction of this power reshuffle.
This not only concerns the political and economic structure of the United States itself but will also, on a deeper level, influence the evolution direction of global capitalism in the new technological era.

2. AI: When Capital Builds High Walls, CapEx Drama Peaks

If the competition focus of AI in 2023-2024 was still on "whose model parameters are larger, whose exam scores are higher," then 2025 is a year that returns to common sense and enters deep waters: the moat of AI is redefined, no longer a single model breakthrough, but who has the ability to continuously bear the pressure of CapEx (capital expenditure) over a sufficiently long time dimension.
Looking back on the timeline, the comparison of facts at the beginning and end is quite humorous. After all, at the very start of 2025, the Chinese AI startup DeepSeek released DeepSeek-R1, which indeed greatly impacted the global AI market's pricing logic with its low cost, high efficiency, and open-source approach, shaking the long-held "compute power" myth in Silicon Valley for the first time and sparking a global re-discussion on "whether compute power really needs to be so expensive."
It was precisely due to this impact that market skepticism about the huge input-output ratio of AI reached its peak. On January 27, NVIDIA (NVDA.M) saw its stock price plummet 18% in a single day, causing the "small model + engineering optimization" route to re-enter mainstream view.
However, the irony lies in the fact that although the efficiency revolution brought by DeepSeek was widely referred to as the "Sputnik moment" in the AI world, in the competition among leading players, the ultimate battleground of AI competition actually shifted this year from model architecture further down to power, infrastructure, and the sustained cash flow of investment.
This is specifically reflected in two sub-dimensions:

• On one hand, the physical world foundation carrying the models has become unprecedentedly expensive. Giants like OpenAI, Meta (META.M), and Google (GOOGL.M) have almost simultaneously upgraded their arms race, continuously raising CapEx expectations, with market predictions that the cumulative CapEx of these giants will soar to $2-3 trillion between 2025 and 2030 (for further reading, see “OpenAI 2025 Review: $500 Billion Alchemy, Is the Shape of Capital-Built AGI Wall Emerging?”);
• On the other hand, OpenAI, Google, and Chinese AI giants like Alibaba (BABA.M) are also engaging in comprehensive confrontations based on their respective advantages in technology, ecology, and commercialization, attempting to build a complete closed loop covering entry—cloud—compute power—applications, with AI competition entering a system engineering game (for further reading, see “Entry, Cloud, and Compute Power: Insights into OpenAI's 'AI Empire' Landscape from DevDay”);

It can only be said that within a brief window, the old order was indeed impacted. However, through repeated pulls, the market gradually formed a new consensus—AI competition is still a marathon with no end in sight. The real moat does not lie in whether the model itself is smarter, but in who can endure higher capital expenditure intensity and sustained investment capability.
In other words, the shock brought by DeepSeek at the beginning of 2025 did not end the "expensive" nature of AI; rather, it pushed it into a more brutal and realistic phase: a high wall built by capital, energy, and time is slowly closing at the entrance of the AI world.
However, entering the fourth quarter of 2025, the market's pricing logic for AI began to show a subtle yet crucial turning point, such as the "differentiation" in trends based on financial reports—companies like Oracle (ORCL.M) and Broadcom (AVGO.M) saw significant stock price retreats after releasing their latest quarterly reports, not due to a slowdown in AI-related revenues, but because the market began to reassess a question: when CapEx has been fully drawn in advance, does the next phase of growth still possess linear extrapolation certainty?
In contrast, Micron Technology (MU.M) became a new anchor for capital during the same time window, with HBM order visibility, price improvements, and profit release rhythm allowing it to transition from being a beneficiary of the AI narrative to a direct recipient of profit certainty, quickly gaining a repricing of funds.
The contrasting cold and hot dynamics indicate that the market is no longer indiscriminately rewarding AI relevance but is beginning to distinguish who is burning CapEx and who is harvesting CapEx. From a more macro perspective, this also marks that the AI investment paradigm is shifting from an infrastructure arms race to a new phase of cash flow and return rate auditing.
Capital has not wavered in its long-term faith in AI, but it is no longer willing to pay a premium for every high wall—and this reordering of monetization capability and profit paths may very well be the core focus that the market will repeatedly price in 2026.

3. Tariff Storm: The Violent Collision of New and Old Geopolitical Orders

In 2025, tariffs are no longer just a macroeconomic variable; they have officially evolved into the "number one killer" of risk appetite in the U.S. stock market.
Against the backdrop of valuations at historical highs and liquidity being extremely sensitive to policy, April 2, 2025, was dubbed "Liberation Day" by the White House, as Trump signed an executive order announcing a 10% baseline tariff on all imports into the U.S. and implementing precise "reciprocal tariffs" on countries with significant trade deficits.
This policy instantly ignited the most intense structural shock in global financial markets since the pandemic in 2020.
The consecutive crashes on April 3 and 4 became one of the most representative "stress tests" in recent years, with major U.S. stock indices recording their largest declines since 2020, evaporating about $6.5 trillion in market value, and the Nasdaq Composite Index and Russell 2000 Index even temporarily entering a technical bear market (i.e., down over 20% from their peaks).

Subsequently, the market fell into a prolonged period of policy negotiation. Although a rebound driven by AI occurred in May due to the "90-day negotiation buffer," in October, with the resurgence of the government shutdown crisis and tariff uncertainties, the market once again experienced a deep correction reminiscent of April.
From a higher dimension, the essence of this round of tariff storm is not merely the short-term fluctuations of trade policy but the last counterattack of the old trade order under the new industrial structure. After all, over the past few decades, the globalization dividend has been built on three premises: low tariffs, cross-border supply chain efficiency, and a relatively stable geopolitical framework.
However, entering a new phase where AI, semiconductors, energy, and security are highly intertwined, trade is no longer just an efficiency issue but has become an extension battlefield for national security, industrial control, and technological sovereignty. It is precisely for this reason that tariffs were re-priced in 2025; they are no longer just a cyclical policy tool but are viewed by the market as a structural friction cost in the process of geopolitical order restructuring, becoming a source of uncertainty that cannot be simply hedged but must be incorporated into the long-term pricing system.
This change also marks that the global capital market has officially entered a new phase, where any enterprise and any profit must account for a high "geopolitical security cost" in addition to operating/profit.

4. Pullbacks, Liquidations, and Recoveries: The U.S. Stock Market Steadily Holds the "Anchor of Global Risk Assets"

At the same time, if the tariff storm in April was an extreme stress test, the subsequent market performance actually tested the true "quality" of the U.S. stock market: the pullback was fierce, but the recovery was equally swift; funds did not leave for long but quickly flowed back to the core market after a brief deleveraging.
This textbook-like resilience is not only reflected in the speed of price recovery but also in its status as the ultimate safe haven for global liquidity—amid rising global uncertainties, the U.S. stock market remains the place where capital is most willing to "return" (for further reading, see “Pullbacks, Liquidations, and Recoveries: The 4-Year Cycle Revelation of U.S. Stocks, A-Shares, and Crypto”).

Looking at the entire year, this resilience is not coincidental. On February 19, 2025, the S&P 500 reached a historic high. Although it subsequently experienced repeated pulls from AI bubble concerns and tariff shocks, the index did not trend toward destructive patterns but continuously completed structural re-evaluations amid volatility. This structural "stabilizing force" was further reinforced by the end of the year:
As of the time of writing, the Nasdaq 100 (QQQ.M) rose 21.2% for the year, with the tech narrative remaining the growth backdrop; the S&P 500 (SPY.M) increased by 16.9%, steadily refreshing its range in high-frequency trading; the Dow Jones and Russell 2000 (IWM.M) rose 14.5% and 11.8%, respectively, completing the structural puzzle from "value recovery" to "small-cap repair."
Although precious metals like gold (GLD.M) and silver (SIVR.M) exhibited more dazzling performances in absolute return rates in 2025 (for further reading, see “Piercing the 'Silver Frenzy': Behind the $60 New High, Is the Market Snatching the Last Batch of 'Liquidity Silver'?”), the value of the U.S. stock market lies not in being the fastest runner but in its irreplaceable structural profit-making effect—it is both a deep-water port under complex geopolitical games and a certainty that global capital repeatedly anchors in high-volatility environments.
When tariffs raise frictions, geopolitical tensions amplify noise, and technological revolutions reshape industrial structures, the U.S. stock market does not avoid risks but absorbs risks, re-prices risks, and ultimately bears risks.
It is precisely for this reason that in the violently colliding year of 2025, the U.S. stock market was able to firmly hold its position as the "anchor of global risk assets."

5. Compute Power Equals Power: From NVIDIA's $5 Trillion to the "Heavy Infrastructure" Echoes of Various Sub-sectors

If the U.S. stock market was able to firmly hold its position as the "anchor of global risk assets" in 2025, then the heaviest link in the anchor chain undoubtedly points to compute power.
On October 29, 2025, the global capital market witnessed a historic moment as NVIDIA (NVDA.M) saw its market capitalization surpass $5 trillion, becoming the first company in capital market history to achieve this milestone, with a market value exceeding the total market capitalization of several developed countries, including Germany, France, the UK, Canada, and South Korea.
More symbolically, the non-linear acceleration trajectory presented by its market value leap: it took 410 days to cross from $3 trillion to $4 trillion, while it only took 113 days to go from $4 trillion to $5 trillion. This change itself is difficult to explain solely by performance growth, marking that the market has begun to price core assets with a new metric of "compute power hub" (for further reading, see “Impacting $5 Trillion? A Comprehensive Interpretation of GTC 2025, NVIDIA's 'AI Factory' Prototype Emerges”).
Objectively speaking, NVIDIA's significance has long surpassed the growth narrative at the individual stock level. With its high binding to the GPU and CUDA ecosystem, it occupies 80%-90% of the key share in the AI chip market, and the extreme dependence of large model training and inference on scaled compute power makes it an unavoidable infrastructure node in the entire AI industry chain.
However, it is also at this stage that the market gradually realizes that the end of compute power is colliding with the boundaries of the physical world. Therefore, the speculation logic of the AI sector has undergone profound shifts, with bottlenecks no longer solely existing in the GPU itself but continuously transmitting downstream along the industry chain: compute power → memory → electricity → energy → infrastructure.
This transmission chain has also directly triggered a round of capital linkage across multiple sub-sectors.
The first to ignite was memory and storage. As the scale of AI training and inference continues to expand, the compute power bottleneck began to shift from GPUs to HBM (High Bandwidth Memory) and storage systems themselves. In 2025, HBM remained in a state of supply shortage, and NAND flash prices entered a new upward cycle, leading to impressive performances of 48%-68% for companies like Micron Technology (MU.M), Western Digital (WDC.M), and Seagate Technology (STX.M) throughout the year.
At the same time, data centers, typical "electricity-hungry beasts," have made companies with nuclear energy assets and independent power grids start to master hard currency in the AI era. Thus, in 2025, many companies originally viewed as defensive assets in the energy and utilities sector exhibited tech stock-like trends: Vistra Corp (VST.M) +105%, Constellation +78%, GE Vernova +62%.
This spillover effect even further transmitted to Bitcoin mining companies, which were originally seen as old-cycle assets. With AI's occupation and redistribution of electricity resources, former mining companies like IREN (IREN.M), Cipher Mining (CIFR.M), Riot Platforms (RIOT.M), Core Scientific (CORZ.M), Marathon Digital (MARA.M), Hut 8 (HUT.M), CleanSpark (CLSK.M), Bitdeer (BTDR.M), and Hive Digital (HIVE.M) were also re-incorporated into the new valuation framework of "compute power---energy" (for further reading, see “AI Kills Miners: 'Energy Run' Tears Open a New Cycle, What Fate Do Mining Companies Stand At?”).

Of course, as the year drew to a close, the AI chip market was not without its ripples. In November, Google released Gemini 3, which surpassed OpenAI's GPT-5.1 in several benchmark tests. At the same time, reports emerged that Google plans to sell its self-developed TPU chips on a large scale and double TPU production to 7 million units before 2028, a 120% increase from previous expectations.
Even more impactful was the pricing strategy, with Morgan Stanley predicting that Google's TPU costs are only one-third of NVIDIA's, which undoubtedly has the potential to force the AI chip market back to a commercial game of "cost-effectiveness" rather than pure "scarcity" monopoly, posing a structural challenge to NVIDIA's market share and ultra-high profit margins.

6. Trump-style Capitalism: From Intensified Political Polarization to State Intervention in Capital

If AI and tokenization are the "technological manifestations" of 2025, then the institutional fluctuations caused by political polarization and the deep shift in U.S. industrial policy constitute the most complex "underlying background" of 2025.
This year, the market witnessed an unprecedented 43-day federal government shutdown. Widespread flight delays, interruptions in food assistance programs, stagnation of public services, and hundreds of thousands of federal employees forced into unpaid leave… It can be said that this month-long deadlock impacted the livelihood and economic operation of American society, penetrating into every capillary (for further reading, see “Shutdown Ends, Flood Comes: After 43 Days of 'Information Vacuum,' How Will Backlogged Data Impact the Market?”).
However, more concerning than the economic losses is the institutional signal conveyed by this shutdown. Political uncertainty is shifting from a "predictable event" to a systemic risk source. In traditional financial frameworks, risks can be priced, hedged, or deferred; but when the institution itself frequently malfunctions, the space for market choices suddenly shrinks, either raising overall risk premiums or leading to phased withdrawals.
This also explains why the U.S. stock market experienced multiple severe pullbacks in 2025, not due to the deterioration of any macro data but more like a series of brutal stress tests on institutional reliability.

In this highly polarized political environment, the economic governance logic of the new U.S. government has also begun to show a more distinct characteristic: the state’s will is no longer satisfied with traditional subsidies and tax incentives but chooses to directly intervene in capital structures.
Unlike past industrial policies primarily focused on subsidies, tax incentives, and government procurement, 2025 saw the emergence of a more controversial and symbolic shift: moving from "grant subsidies" to "direct equity stakes," constructing a supportive system with "financial participatory" equity. The author has also observed numerous discussions on paths resembling "Trump-style capitalism" or a variant of state capitalism.
The Intel (INTC.M) agreement serves as the first shot fired, with the federal government reaching a milestone agreement to acquire a 10% equity stake in Intel, marking the beginning of the U.S. federal government playing the role of a long-term shareholder in key strategic industries.
From the supporters' perspective, this shift is not without logic, as for many fields at the technological frontier but still early in commercialization (such as quantum computing), direct equity stakes from the government can theoretically significantly reduce financing uncertainty, extend the cash flow runway for companies, and provide stable expectations for long-term R&D. Compared to one-time subsidies, equity-based support is also seen as more aligned with "long-termism" policy goals.
For this reason, there were rumors that the U.S. government might consider exchanging federal funds for equity stakes in quantum computing companies like IonQ (IONQ.M), Rigetti Computing (RGTI.M), and others (for further reading, see “The Final Battle for Compute Power: Will Quantum Computing Be the Next 'AI Moment'?”, “Is the U.S. Government Going to 'Invest' in Quantum Computing? Understanding the First Shot of the 'Policy Equity' Era”), but the U.S. Department of Commerce quickly stepped in to deny this, clearly stating that no formal negotiations had been initiated regarding equity stakes in the aforementioned quantum computing companies.
This clarification reflects that the policy boundaries are still in a state of repeated negotiation and highlights the high sensitivity of this issue. In fact, the core of the matter is not whether the government will truly invest in a specific quantum company, but how the market will reprice the boundaries of policy, capital, and risk when the state begins to intervene as a shareholder in cutting-edge technology industries.
This resource allocation driven by state will has long been the original sin criticized and questioned by Western public opinion and capital markets regarding China's photovoltaic, new energy, and other industrial policies over the past few decades. Now, this boomerang has circled halfway around the globe and struck right at the heart of the U.S. itself.

7. Decoupling of Major Currency Policies: The Fed Moves Left, the BOJ Moves Right

Beyond industrial policy, the changes in monetary policy in 2025 further expose the systematic contraction of macro-control space.
Amid the ongoing tug-of-war between inflation and employment, the Federal Reserve officially restarted its interest rate cut cycle in September 2025, subsequently cutting rates by 25 basis points in both October and December, for a total reduction of 75 basis points over the year.
However, in the current macro environment, the market's understanding of this round of rate cuts has already changed. Observant individuals know that this does not signal a return to a loosening cycle but rather resembles a "pain management" approach to exert pressure on the economic system and even politics. This also explains why multiple rate cuts have not alleviated market uncertainties as expected; the U.S. stock market did not experience an indiscriminate liquidity frenzy but instead further differentiated structurally.
Ultimately, everyone is increasingly aware of a reality: the available space for monetary policy is becoming increasingly limited, especially under the constraints of high debt, high fiscal deficits, and structural inflation. The Federal Reserve can no longer provide market support through significant loosening as it did in the past.
In simple terms, each rate cut today feels more like drinking poison to quench thirst rather than creating new growth momentum.
In a contrasting manner, while the Federal Reserve is shifting towards rate cuts, the Bank of Japan is firmly advancing monetary policy normalization. On December 19, the Bank of Japan announced a 25 basis point rate hike, raising the policy rate to 0.75%, the highest level since 1995, and marking the fourth rate hike since the end of its prolonged negative interest rate policy in March 2024.
As of December 26, according to Jin10 data, although inflation in Tokyo has cooled more than expected, with pressures on food and energy prices easing, the market generally believes this is not enough to prevent the Bank of Japan from continuing its rate hike process. The divergence between one central bank cutting rates and another raising rates has brought global monetary policy differentiation to the forefront, severely squeezing the space for yen carry trades that have been maintained for years, forcing global capital to reassess the risk structures across currencies and markets.
Objectively speaking, by 2025, monetary policy has gradually lost the aura of a "magic wand." In the context of deeper state capital will intervention and the rising geopolitical walls, interest rates are no longer the universal lever for adjusting the economy but rather resemble a painkiller to prevent acute systemic collapse.
Among the major central banks globally, Japan is becoming the "last bastion" of tightening global liquidity, which is likely to become one of the fiercest risk sources in 2026.

8. The Fed's "Balancing Act": Rate Cut Cycle and New Chair Prospects

At the same time, the invasion of political pressure is gradually causing the "altar" of the Federal Reserve to collapse.
In 2025, Trump's criticisms of Powell extended from 𝕏 to the White House. As Powell's term is set to expire in May 2026, the market has already begun to trade in advance the policy orientation of the "next Federal Reserve chair" or even the "current shadow Federal Reserve chair."
If the eventual successor is a "loyal dove" represented by Kevin Hassett, then the Federal Reserve, following the White House's baton, is likely to release more aggressive liquidity signals in the short term, with the Nasdaq and Bitcoin potentially experiencing an emotion-driven frenzy. However, the cost may be a renewed loss of control over inflation expectations and further erosion of the dollar's credit (for further reading, see “Prospects for the New Fed Chair: Hassett, Coinbase Holdings, and Trump's 'Loyal Dove'”).

Conversely, if the successor aligns more with a "reformist" path like Kevin Walsh, the market may experience a painful period due to liquidity tightening. However, under deregulation and a stable monetary framework, long-term capital and traditional financial institutions may actually gain greater institutional security (for further reading, see “The Fed's 'Successor' Reversal: From 'Loyal Dove' to 'Reformist', Has the Market Script Changed?”).
We cannot predict who will be "the last person to talk to Trump." This unpredictable state may continue until the moment a candidate is finalized.
However, regardless of who ultimately prevails, one fact will not change: interest rates themselves are gradually evolving from an economic variable to a part of political games. After all, in 2020, Trump could only criticize Powell on Twitter; by 2025, Trump, returning with an overwhelming victory, is no longer satisfied with merely being an observer.
Whether the leading actor is Hassett or Walsh may determine the direction of the plot, but the overall director of this drama has firmly become Trump.

9. Financial Infrastructure Revolution: From 5×16 to 5×23, and Then to 7×24?

If there is one change in 2025 that is most easily underestimated yet likely to produce long-term chain reactions, the answer lies not in a specific star stock or sector but in the trading system itself.
This is also the most far-reaching yet easily overshadowed invisible transformation of 2025, as Wall Street has officially decided to actively dismantle barriers and move towards tokenization and 7×24-hour liquidity.
If we extend our view and connect the recent intensive actions of Nasdaq, we will be more convinced that this is a strategic puzzle that is gradually unfolding, with the core goal aimed at enabling stocks to ultimately possess the capabilities of circulation, settlement, and pricing like tokens. To achieve this, Nasdaq has chosen a path of gentle reform with a strong traditional financial style, with a roadmap advancing step by step:

• The first step occurred in May 2024, when the U.S. stock settlement system officially shortened from T+2 to T+1, a seemingly conservative yet crucial infrastructure upgrade;
• Following this, at the beginning of 2025, Nasdaq began signaling its intention for "around-the-clock trading," hinting at plans to launch uninterrupted trading services five days a week in the second half of 2026;
• Subsequently, Nasdaq integrated blockchain technology into the Calypso system to achieve 7×24-hour automated margin and collateral management. This step did not create significant visible changes for ordinary investors but sent a very clear signal to institutions;
• By the second half of 2025, Nasdaq began to push forward on institutional and regulatory fronts. First, in September, it formally submitted an application to the SEC for "tokenized" stock trading, and in November, it explicitly stated that tokenizing U.S. stocks was a primary strategy to be "advanced as quickly as possible";
• Almost simultaneously, SEC Chairman Paul Atkins stated in an interview with Fox Business that tokenization is the future direction of capital markets, and that by putting securities assets on-chain, clearer ownership verification can be achieved. He expects that "within the next two years, all U.S. markets will migrate to on-chain operations, achieving on-chain settlement";

It is against this backdrop that Nasdaq submitted an application to the SEC for a 5×23-hour trading system in December 2025 (for further reading, see “U.S. Stocks Sprinting Towards 'Never Closing': Why Did Nasdaq Launch the '5×23 Hour' Trading Experiment?”), and in this context, The Economist also published an article discussing how RWA tokenization could change finance, drawing a highly symbolic analogy: if history serves as a reference, the current stage of tokenization is roughly equivalent to the internet in 1996—at that time, Amazon sold only $16 million worth of books, and today, among the "Magnificent 7" dominating the U.S. stock market, three companies had not even been born yet.
From yellowed paper certificates to the electronicization of the SWIFT system in 1977, and now to the atomic settlement of blockchain, the evolution curve of financial infrastructure is replicating and even surpassing the speed of the internet.
For Nasdaq, this is a gamble of "being revolutionized if it does not self-revolutionize"; for the Crypto industry and new RWA players, this is not only a brutal reshuffle of survival of the fittest but also a historic opportunity comparable to betting on the next "Amazon" or "NVIDIA" in the 1990s (for further reading, see “Nasdaq Hits the Gas: From 'Drinking Soup' to 'Eating Meat', Is U.S. Stock Tokenization Entering the Decisive Stage?”).

10. The "Year of AI Agents": It Exploded, but It Seems Not Fully Exploded

The most frequently heard term in 2025, yet one that seems to feel somewhat lacking, is undoubtedly "the Year of AI Agents."
One word to describe this year's AI Agent market is "explosive."
The market has formed a clear consensus that AI is transforming from a passive response dialogue box into an Agent form capable of autonomously calling APIs, handling complex task flows, executing operations across systems, and even participating in decision-making in the physical world. The explosive success of Manus at the beginning of the year indeed fired the first shot (as of the time of writing, it was reported that Meta acquired Manus for billions of dollars, with Xiao Hong set to become Meta's vice president). Subsequently, products like Lovart and Fellou emerged, further creating the illusion in the market that "the application layer is about to explode."
However, to be fair, while Agents have validated the direction, they have not yet scaled effectively. Early hit products quickly encountered issues of declining user engagement and reduced usage frequency; they have demonstrated "what Agents can do," but have not yet answered "why they should be used long-term."
This is not a failure but a necessary stage in the technology diffusion cycle.
In fact, whether it is OpenAI's CUA (Computer-Using Agent) or Anthropic's MCP (Model Context Protocol), they do not point to a specific application but rather indicate a longer-term judgment: the AI capability curve will be exceptionally steep over the next two years, but the true value release relies on system-level integration rather than single-point functional innovation.
Thus, the AI Agents of 2025 resemble a directional tuning, following the diffusion pattern of innovative technologies. Transitioning from a "year one" to large-scale implementation will take at least three years, so 2025 is merely about completing the consensus shift from 0 to 1.
Of course, it is worth noting that as the year draws to a close, a highly vital new variable has begun to emerge—the exploration of AI terminal forms by ByteDance, pulling the Agent concept back to the issues of hardware entry and scene binding. This does not necessarily mean that AI phones will succeed immediately, but it reminds the market once again: the ultimate fate of Agents may not lie in a specific app but in becoming actors within the system.
This time, capital may have run ahead of applications, but once the direction is set, there is no turning back in 2026.

Conclusion | What Did 2025 Leave Behind?

In a sense, 2025 was not a year that provided answers but a "year one" of collective turning.
Looking back at this year, the global capital market seems to be in a maze constructed of paradoxes:

• On one side, high walls continue to rise: global trade frictions heat up, tariff barriers return, political polarization intensifies, the shadow of government shutdowns looms, and great power games move from backstage to the forefront;
• On the other side, fences are collapsing: regulatory attitudes towards new technologies undergo a major shift (re-evaluation of SEC/CFTC policies), financial infrastructure accelerates dismantling barriers (comprehensive on-chain/tokenization of assets), and AI brings cross-temporal and cross-spatial productivity leaps;

This extremely absurd yet opposing landscape essentially reflects the continuous erection of new boundaries by political and geopolitical structures, while Washington and Wall Street are attempting to dismantle the old barriers of finance and technology.
In fact, the alarm has long been sounded.
When gold, silver, and other precious metals lead this year's major asset classes and even outperform the vast majority of tech stocks, we should realize that the assertion of a "great change" is not a prophecy. After all, the capital expenditure games of hundreds of billions to trillions of dollars in AI are destined to be unsustainable, and the geopolitical games looming over the global capital market are pushing us toward the "Minsky moment" that has been warned about for years—namely, the collapse point after excessive expansion.
Shakespeare wrote in "Romeo and Juliet": "These violent delights have violent ends." As the shadow of 2025 gradually recedes and the bell of 2026 is about to toll, what we may have to face is likely not the result of a single event but rather a natural extension of this structural state.
The real change may not lie in "what will happen," but in—the market no longer allows participants to pretend that nothing will happen.
Goodbye, 2025; hello, 2026.