<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by José Ribeiro de Carvalho Carlos on Medium]]></title>
        <description><![CDATA[Stories by José Ribeiro de Carvalho Carlos on Medium]]></description>
        <link>https://medium.com/@jrcc2635?source=rss-b14086792b51------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/proxy/1*TGH72Nnw24QL3iV9IOm4VA.png</url>
            <title>Stories by José Ribeiro de Carvalho Carlos on Medium</title>
            <link>https://medium.com/@jrcc2635?source=rss-b14086792b51------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 21:34:54 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@jrcc2635/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Fault lines and the AI boom]]></title>
            <link>https://medium.com/@jrcc2635/fault-lines-and-the-ai-boom-f6fcae702994?source=rss-b14086792b51------2</link>
            <guid isPermaLink="false">https://medium.com/p/f6fcae702994</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[shadow-banking]]></category>
            <category><![CDATA[dotcom-bubble]]></category>
            <category><![CDATA[2008-financial-crisis]]></category>
            <category><![CDATA[data-center]]></category>
            <dc:creator><![CDATA[José Ribeiro de Carvalho Carlos]]></dc:creator>
            <pubDate>Thu, 06 Aug 2026 12:52:21 GMT</pubDate>
            <atom:updated>2026-08-06T12:52:21.133Z</atom:updated>
            <content:encoded><![CDATA[<p>The last major financial crisis our world encountered in 2008 was the culmination of a long process of faulty supervision, mismanagement of risk, bad political choices and poor governance. What did we learn since then?</p><p>Driven by underlying forces of tectonic movements, fault lines are fractures in the Earth’s crust. When these move or snap, they can shift surface and the crack creates the illusion of a rupture. This analogy was popularized by <a href="https://press.princeton.edu/taxonomy/term/18665?_gl=1*qhr039*_up*MQ..*_ga*MjE0MTkyMzg2MS4xNzg1OTcxMjc1*_ga_N1W9JWKLY3*czE3ODU5NzEyNzQkbzEkZzAkdDE3ODU5NzEyNzQkajYwJGwwJGgxNjgxODUwNDc3">Raghuram G. Rajan</a> in his book<em> Fault Lines</em> released in 2011.</p><blockquote>Think of crisis as an earthquake, what makes the earth shake and buildings fall has little to do with individuals, but something hidden, something deep beneath the earth’s surface. Just like earthquakes, financial crises are caused by deep structural pressures that build silently over years. When those pressures finally release, the shock is sudden and devastating.</blockquote><p>In my analysis, the tectonic plate movements or the rupture causes are analogous to the possible causes of an AI crisis. A supply shock, regulation, geopolitical escalations or even banking and financial setbacks are enough for a big rupture.</p><p>For newcomers, the most simplified way of the AI process can be summarized in: A company makes semiconductors, the next company uses them to produce chips. Hyperscalers then build data centers, using those chips/ GPU’s. Then, AI developers use the large processing power of this data centers to train and run large AI models.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/596/0*gLPztDs_XqNgjlkV" /><figcaption>Data Center Cisco, 2026</figcaption></figure><p>But how important is AI? Exact number of AI users across the world aren’t available. <a href="https://www.sciencedirect.com/science/article/abs/pii/S0305750X25003468">As of March 2024,</a> the top 40 generative artificial intelligence tools attract nearly 3 billion visits per month from hundreds of millions of users. OpenAI reported 1Bn user across their platforms in<a href="https://www.wsj.com/tech/ai/openai-surpasses-one-billion-users-after-cutting-prices-8a9943e4"> 2026</a>, ChatGPT alone commands over 80 percent of the traffic<a href="https://www.wsj.com/tech/ai/openai-surpasses-one-billion-users-after-cutting-prices-8a9943e4">.</a> Antropic declared they reached <a href="https://www.statista.com/statistics/1659723/global-monthly-claude-users/?srsltid=AfmBOori6sA73tCSrOs7fmTlcUEwELVpnU5-y836xdZzBSuJnX7qdVSv">245 million</a> users in Claude as july 2026. Terrifying forecasts from IMF say AI will affect almost <a href="https://www.imf.org/en/Blogs/Articles/2024/01/14/ai-will-transform-the-global-economy-lets-make-sure-it-benefits-humanity">40 percent of jobs</a> around the world. At the same time, It will increase global GDP by $7 trillion or 7% over 10 years, predicts <a href="https://www.goldmansachs.com/insights/articles/generative-ai-could-raise-global-gdp-by-7-percent">Goldman Sachs</a>.</p><blockquote>“<em>AI-related stocks have accounted for 75 percent of S&amp;P 500 returns, 80 percent of earnings growth, and 90 percent of capital spending growth since ChatGPT</em><a href="https://am.jpmorgan.com/pt/en/asset-management/institutional/insights/market-insights/eye-on-the-market/the-blob-capital-china-chips-chicago-and-chilliwack/?_gl=1*1vvomux*_up*MQ..*_ga*MjE0MzE1MDg0NC4xNzg1OTQwNTE1*_ga_SLLWW8GS4N*czE3ODU5NDA1MTQkbzEkZzEkdDE3ODU5NDA1MTQkajYwJGwwJGgw"><em> launched</em>.</a>”</blockquote><p>Eye on the market, <a href="https://am.jpmorgan.com/pt/en/asset-management/institutional/bios/michael-cembalest/">Michael Cembalest</a> Chairman of Market and Investment strategy for J.P.Morgan Asset &amp; Wealth Management</p><p>On the other side, Daron Acemoglu, no presentantions needed, made more conservative estimates. Over the next 10 years will be roughly 0.7%. This could translate into a maximum growth in GDP of about 1.8%, while a more realistic finding under this scenario is about 1.1%. In the case of US economy, over the next 10 years, only about 5% of tasks will be able to be profitably performed by AI within that time frame, the GDP boost would likely be closer to 1% over that period.</p><p><a href="http://Hannah Rubinton and Bontu Ankit Patro, “Tracking AI’s Contribution to GDP Growth,” St. Louis Fed On the Economy, Jan. 12, 2026.">St. Louis Fed</a> show us that a large and diminishing part of the economic growth in the USA over the first 3 quarters 2025 of has been driven by AI investment, data centers, and related sectors. With 30% of GDP total growth in Q1, Q2: 30% and Q3: 11% of growth. What happens if or when the invesment flattens? Total AI contribution went from 1.30pp in Q1 to 0.48pp in Q3.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/657/1*5CPF2tTsSr8noxPPR2C0fg.png" /><figcaption><a href="https://www.stlouisfed.org/on-the-economy/2026/jan/tracking-ai-contribution-gdp-growth">https://www.stlouisfed.org/on-the-economy/2026/jan/tracking-ai-contribution-gdp-growth</a></figcaption></figure><p>As AI prompted discussions about the similarity with DOT-Com bubble, the data above shows us that in 2000, the US economy grew at 2.94% with tech contributing 28% of growth. In 2025, the economy is growing at 2.51%, yet AI now accounts for 39% of growth. We have a weaker growth being supported by a larger boom.</p><p>On the other side of this boom, we also have environmental consequences. AI requires data centers, where firms use computing power to make billions of algorithmic calculations. Data center electricity consumption in 2024 was 415 TWh, representing 1.5% of global demand. By 2030, projected consumption is 945 TWh approximately 3% of global electricity consumption (IEA base case). Annual AI workload growth is 30%, higher than overall demand growth.</p><blockquote><a href="https://medium.com/data-science/environmental-implications-of-the-ai-boom-279300a24184">“GPU chips involve</a> a number of different <a href="http://www.designlife-cycle.com/nvidia-gpu">specialty raw materials that are somewhat rare and hard to acquire, including tungsten, palladium, cobalt, and tantalum. </a>Other elements might be easier to acquire but have significant health and safety risks, such as mercury and lead.”</blockquote><p>Artificial Intelligence is the largest technological boom our generation is facing. Investment-wise, this surpasses previous eras. For 2026, expected CAPEX for the top 5 tech firms is $674 billion, by 2030, global data center CAPEX is projected at $5.2 trillion. According to the Wall Street Journal (2026), it surpasses the railroad expansion of the 1850s, the Apollo Space program of the 1960s, and the US Interstate Highway System of the 1950s–1970s.</p><p>What could be the tectonic plate movement that could cause a big earthquake in this AI boom? Imagine European Union or even the US ( most unlikely) caps data center power consumption. If power consumption is capped, then the building and normal function of Data Centers is not possible. This itself could cause another problem, what if AI developers don’t generate sufficient economic return relative to CAPEX? This itself could lead to major tech firm missing debt obligations, that would lead to margin calls and a possible spread among the financial system.</p><p>AI companies justify billions in CAPEX by projecting future AI revenue, but that revenue is speculative, almost self-referential. At the same time, a handful of firms (NVIDIA, TSMC, AWS, Google, Microsoft) control big part of the game. One disruption spreads everywhere. Allianz Research puts the growth gap between AI investment and AI sales at nearly 46%, larger than the 32% divergence seen during the 2001 telecom bust.</p><p>Another possible cause of rupture is directly related to the enormous risk created by the the concentration of Chips in TSMC. If we observe geopolitical escalation of China vs Taiwan, there is a possibility of full rupture. Formosa Island TSMC produces: 90%+ of advanced chips (5nm, 3nm and below), 100% of NVIDIA’s flagship chips (H100, H200, GB200). As august 2026, there are no public known alternatives in semiconductor production with similar capacity as TSMC. The utilization rates of GPUs for AI workloads in a datacenter run by cloud service providers is between 60% and 70%. With such utilization rates, a GPU will typically survive between one and two years, three years at the most. How can a firm use an AI processor with a 3 to 5-year estimated lifetime as collateral for a 50-year loan? The main problem is not the lifetime but the queue to order newer and more powerful GPU’s and chips. Currently, the queue for NVIDIA chips is around 2 years.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/862/0*5A23o-_f9XZueAVm" /><figcaption>China — Taiwan Map</figcaption></figure><p>There are sound similarities with 2008 financial engineering that could pose some threat, for example credit transformation in 2008, banks pooled low-quality mortgages and repackaged them as AAA-rated senior tranches. Close to trash assets became attractive through financial engineering Nowadays, Hyperscaler credit ratings (NVIDIA, Google, AWS) transfer to weaker borrowers through bilateral contracts. The borrowers are less creditworthy, but they borrow on the hyperscaler’s reputation.</p><p>Second is the way maturity is transformed. In 2008, long — term mortgage assets (30years) were funded by short-term CDOs (commercial debt obligations). Banks were betting mortgages would stay stable, but funded them with short-term debt that had to be rolled over constantly.</p><p>As mentioned before, rapidly depreciating GPU assets (3 to 5 year useful life) are funded by long-term bonds (50-year infrastructure loans). Short-term lease revenues from the GPUs are supposed to collateralize long-term infrastructure debt.</p><p>Power asymetry is a core part of this boom. Information is not shared symmetrically and this is not accidental. Insiders, are fully aware of contracts, risk and have control of the infrastructure. They can leave earlier with lower to minor losses. Outsiders are structurally excluded, supervision does not have proper mechanisms to monitor and most of the billateral contracts are hidden. This way, its close to impossible to evaluate risk.</p><p>At the end of the day, everyone acted at forefront of their own interests and rational actions. Technological companies needed capital to grow, so they borrowed it from banks, who thought they were lending to clients with profitable projects. Pension funds who were only looking for “safe” investments, chose AI tech. Lastly, people who were looking for jobs in the sector of the future weren’t expecting this much instability.</p><h3><strong>During the 2008 Financial Crisis, everyone was asking “ Why didn’t anyone see this coming?” Even Queen Elizabeth II. Are we seeing this time?</strong></h3><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f6fcae702994" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Angola has left OPEC, and now ?]]></title>
            <link>https://medium.com/@jrcc2635/angola-has-left-opec-and-now-b927baebf4a8?source=rss-b14086792b51------2</link>
            <guid isPermaLink="false">https://medium.com/p/b927baebf4a8</guid>
            <category><![CDATA[economy]]></category>
            <category><![CDATA[opec]]></category>
            <category><![CDATA[oil]]></category>
            <category><![CDATA[angola]]></category>
            <dc:creator><![CDATA[José Ribeiro de Carvalho Carlos]]></dc:creator>
            <pubDate>Sun, 24 Dec 2023 01:18:05 GMT</pubDate>
            <atom:updated>2023-12-24T01:18:05.337Z</atom:updated>
            <content:encoded><![CDATA[<h3>Angola has left OPEC, and now ?</h3><p>The Organization of the Petroleum Exporting Countries (OPEC) was founded in 1960 with the goal of coordinate, unify countries and its policies on oil production, exportation and economic cooperation (investment in the oil sector, market share and price stability). In 2021, OPEC had 12 members that accounted, in 2021, for 36% of oil production (barrels per day) on Earth.</p><p>In December 2023, after 16 years of being a member of OPEC, the Angolan Government, during a Council of Ministers session, put an end to the participation of the second largest oil producer in Africa and joins the group of lapsed members: Equator, Indonesia and Qatar.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/886/1*BhVCfTbg20chkN94dmomwA.png" /><figcaption>Source:<a href="https://ieefa.org/resources/ieefa-africa-angola-cautionary-tale-guyanas-oil-wealth-hopes"> IEEFA</a></figcaption></figure><p>Last month, OPEC established oil production quotas to sustain oil prices until 2024. If Angola had remained in OPEC, it would have had to reduce its oil production and would have been in breach of some contracts made previously.</p><p>The Mineral Resources and Petroleum Minister Diamantino Azevedo said:</p><blockquote>“We feel that at this moment Angola gains nothing by remaining in the organization and, in defense of its interests, it decided to leave.”</blockquote><h3><strong><em>What does Angola gains and loses from leaving OPEC?</em></strong></h3><p>By leaving OPEC, Angola now has more power to control the production and export of oil and its derivatives, thus being able to fulfill previously signed contracts and promises.</p><p>Angola avoids the financial commitment of an annual participation fee of 2 million dollars but loses access to a niche of technological collaboration, political influence and a stable flow of oil at fixed prices. Nonetheless, it loses its negotiating power with OPEC, the organization that includes the world’s major oil producers (the United Arab Emirates, Saudi Arabia and Nigeria).</p><p>In terms of participation and decision-making power, each State has one vote and all pay the same participation fee. On the other hand, for example, Saudi Arabia, which produces ten times more oil than Angola, has greater influence. As analysts say, countries that export petroleum at relatively low volume would have limited negotiating power as OPEC members and would not necessarily justify the burdens imposed by OPEC (quotas, fees,…).</p><p>Analyzing the country’s oil production output, the exit is unlikely to have a significant impact on the group’s operations, according to Helima Croft, a former CIA analyst and head of commodities research at RBC Capital Markets.</p><h3>International effects</h3><p>Following Angola’s intention to exit OPEC, oil prices shortly fell on Thursday, Brent crude futures took a 31-cent decrease, settling at $79.39 per barrel. U.S. West Texas Intermediate crude futures also experienced a decline of 33 cents, reaching $73.89 per barrel. At the beginning of the session, both benchmarks had initially dropped by over $1 following Angola’s announcement of its intention to exit the group. This raised questions about the producer group’s ability to keep prices stable by limiting global supply.</p><p>After Angola’s departure, Nigeria solidified its position in OPEC. Heineken Lokpobiri, the Minister of State for Petroleum, affirmed Nigeria’s dedication to OPEC, emphasizing its role in maintaining the stability and sustainability of the oil market.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*CW2tE-X-P7xfATQO.jpg" /><figcaption>Sonangol’s Deep ocean oil platform (Source: Energy Capital Power)</figcaption></figure><h3>What can we expect?</h3><p>Angola’s exit from OPEC is a major shift in the global oil scene. As a key African oil producer, this move adds a new dimension to the organization, affecting both regional and worldwide energy dynamics. It has triggered talks about how it might influence the overall stability of OPEC. Angola’s decision highlights the changing and intricate geopolitical and economic factors guiding choices in crucial oil-producing nations.</p><p>Is leaving the OPEC a well thought decision? Since one of the main reasons is a missing consensus over the quotas imposed on oil production it is already expected a rise in Oil production output, is strategic to increase dependence on oil? We can not forget that Angola is already heavily dependent on oil with 90% of its exports being made of Oil and its derivatives and Gas. But, as a developing economy, net cash flow it is important to balance State Budget.</p><p>According to OPEC, Angola has proven natural gas reserves of 343 billion cubic meters and oil resources of 7.8 billion barrels, which offer enormous potential for corporate prospects and economic growth. In response to a probable significant boost in oil production, Angola may witness a remarkable surge in cash revenue. This expected increase would not only fortify economic resilience but also underscores the strategic importance of the energy sector. The additional revenue would serves as a catalyst for broader economic development, allowing for crucial investments in infrastructure and social programs.</p><p>Not forgetting that, recently, Fitch Ratings <a href="https://www.fitchratings.com/research/sovereigns/fitch-downgrades-angola-to-ccc-04-09-2020">downgraded Angola</a>, indicating that its economy “continues to be constrained by its high level of commodity dependence, which contributes to low growth and increased macroeconomic instability.”</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b927baebf4a8" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>