Nvidia Under Pressure: Is the AI Chip Monopoly Finally Cracking?

Can Claude Replace Investment Bankers? We Graded the Output.

How good is AI at building a DCF?  In this episode, Debs and Graham continue their Claude for Excel series, this time prompting the tool to construct a full discounted cash flow valuation for Lululemon from a single instruction.  The goal is to test what AI can and cannot do in real valuation workflows, and what that means for analysts working in equity research, investment banking and M&A. Graham walks through DCF fundamentals from first principles, covering future cash flow projections, WACC, terminal value and the inputs that genuinely drive valuation outcomes.  He then opens Claude for Excel and gives it a structured prompt — anchored to consensus EPS estimates for stage one, with explicit instructions on modelling best practices including no hardcoded inputs in formulas, standard colour coding, and transparent assumption sourcing. The audit that follows is instructive on both fronts. Claude handles the structural build well — linking assumptions to formulas, applying the Gordon Growth formula correctly for terminal value, and producing a workable enterprise value output.  But the limitations show up in the details that matter most for senior review: the free cash flow build conflates levered and unlevered measures, time period construction is simplistic rather than properly anchored to fiscal year ends and a valuation date, and some formula constructions are opaque enough that auditing them line by line would take longer than rebuilding the section manually. The verdict: a B-minus output.  Workable as a first pass, but not yet at the level where it can be submitted without significant human review.  The broader question the episode closes on is whether AI tools like Claude for Excel are positioned to replace the analyst role or to elevate it — with Graham making the case that the analyst job as historically defined is exactly the workflow these tools are now competent at, while the judgement-heavy associate role remains some distance from being automated. Key Discussion Points: DCF fundamentals: future cash flows, discount rates, terminal value and the inputs that actually drive valuation outcomes.  Prompting strategy: how to structure a Claude for Excel prompt to anchor projections to consensus estimates and enforce modelling best practices.  Where AI delivers: structural build, formula linking, Gordon Growth application, sensitivity analysis output.  Where AI falls short: free cash flow build, time period construction, opaque formulas that resist quick audit.  Sensitivity analysis: long term growth rate versus WACC as the two real swing factors in any DCF.  AI in finance careers: the analyst role versus the associate role and what realistic automation looks like over the next 12 to 24 months. WTBD Newsletter: https://webmail.wallstreetprep.com/whats-the-big-deal [https://webmail.wallstreetprep.com/whats-the-big-deal] Follow Us On Socials: LinkedIn: https://www.linkedin.com/company/wall-street-prep/ Instagram: https://www.instagram.com/wallstreetprep/ Resources: https://linktr.ee/wallstreetprep

Claude for Finance: Building a Live Merger Model with AI

How good is AI at building investment banking models?  In this episode, Debs and Graham put Claude for Excel to the test by prompting it to construct a full merger model from scratch, using GameStop's $56 billion bid for eBay as the live case study, but with the focus squarely on the AI workflow rather than the deal itself. Graham walks through the merger model framework from first principles before opening Claude for Excel and giving it a single instruction: build me a merger model for the proposed acquisition.  What follows is a live demonstration of what AI can and cannot do in a real M&A modelling workflow. The verdict is nuanced. Claude sources factual data quickly, structures the model sensibly, makes a credible first pass at sources and uses, and saves the kind of analyst time that used to go into manual press release scrubbing and 10-K data extraction.  But it also makes errors that anyone trained in proper modelling would catch immediately, hardcoded assumptions buried in cell formulas, fiscal year mismatches between acquirer and target, missing synergy inputs that were publicly disclosed, and modelling practices that would never pass a senior banker's review. The takeaway: Claude for Excel is a powerful first-pass tool that can compress hours of analyst work into minutes, but it is dangerous in the hands of anyone who cannot audit the output.  The fundamentals of modelling, accounting and finance still matter - arguably more than ever, because the cost of accepting AI output without scrutiny is now embedded in every workflow. Key Discussion Points: Merger model framework: accretion, dilution, sources and uses, pro forma adjustments, LTM calendarisation.  Prompting strategy: what a minimal prompt produces versus what structured prompting would deliver.  Where AI saves time: factual data sourcing, model structure, first-pass build.  Where AI fails: modelling best practices, hardcoded inputs, technical errors, judgement calls.  Stress-testing in real time: how to use AI to iterate on synergy, consideration mix and financing assumptions.  AI in finance careers: why the fundamentals matter more than ever in an AI-enabled workflow. WTBD Newsletter: https://webmail.wallstreetprep.com/whats-the-big-deal [https://webmail.wallstreetprep.com/whats-the-big-deal] Follow Us On Socials: LinkedIn: https://www.linkedin.com/company/wall-street-prep/ Instagram: https://www.instagram.com/wallstreetprep/ Resources: https://linktr.ee/wallstreetprep

Nvidia Under Pressure: Is the AI Chip Monopoly Finally Cracking?

Every AI product you use runs on semiconductors. And for the last several years, the narrative has been almost entirely about Nvidia.  But Q1 2025 results are painting a more nuanced picture and for the first time, the question of whether Nvidia's dominance is structural or temporary feels like a live debate rather than a hypothetical. In this episode, Debs and Graham go inside the semiconductor industry from first principles, mapping out who does what across the AI chip ecosystem before turning to the latest results and what they mean for valuations. Graham explains how GPUs, CPUs and memory chips work together to power AI, covering why the parallel computational demands of AI models require so much chip capacity, why that has driven up the price of consumer memory, and why Nvidia's software ecosystem creates a lock-in that competitors are only now beginning to challenge seriously. Debs then walks through the competitive landscape in detail: Broadcom winning custom chip mandates from Google and Meta on energy efficiency grounds, AMD posting 57% data centre revenue growth, TSMC delivering 41% revenue growth with 66% margins, Samsung flagging memory supply constraints into 2027, and Intel up 150% year to date on the back of a foundry pivot and reported talks with Apple. The valuation discussion unpacks why chip designers like AMD trade at a premium to manufacturers like TSMC despite TSMC's superior margins, the role of CapEx intensity and cash conversion in driving that gap, and the Taiwan geopolitical risk discount embedded in TSMC's 18x multiple. The episode closes with Debs and Graham weighing whether semiconductor valuations reflect genuine AI demand or a market that has run ahead of itself, and flags Nvidia's own results on 20 May as the next major test. Key Discussion Points: Semiconductor ecosystem: GPUs, CPUs, memory and custom chips, who makes what and how they work together.  Nvidia's competitive position: software lock-in, hardware leadership and the first real signs of competitive pressure.  Q1 results: AMD, Broadcom, TSMC, Samsung and Intel, what the numbers say about demand, market share and supply constraints.  Valuation framework: why growth and cash conversion drive the premium for chip designers over foundries, and what geopolitical risk does to TSMC's multiple.  Nvidia's S&P 500 weighting: how index inclusion and passive fund flows affect valuation independent of fundamentals.  Outlook: memory supply constraints into 2027, the Intel/Apple story and Nvidia's results on 20 May as the next major market catalyst. WTBD Newsletter: https://webmail.wallstreetprep.com/whats-the-big-deal [https://webmail.wallstreetprep.com/whats-the-big-deal] Follow Us On Socials: LinkedIn: https://www.linkedin.com/company/wall-street-prep/ Instagram: https://www.instagram.com/wallstreetprep/ Resources: https://linktr.ee/wallstreetprep

How AI Data Centres Are Funded — And What Happens When the Money Stops

OpenAI has missed a revenue target in the run-up to what is expected to be one of the largest IPOs in history. Sam Altman and the company's CFO have been publicly at odds.  And behind all of this sits close to $700 billion of committed CapEx across the major hyperscalers, much of it financed through project finance structures that were built on the assumption of hyper-aggressive AI revenue growth. In this episode, Debs and Graham use the OpenAI revenue miss as a lens to examine how AI infrastructure financing actually works, who is exposed when things wobble, and how a shortfall at the end of the chain could propagate upward. Debs walks through the mechanics of project finance as it has been adapted for data centre construction. SPVs are set up to construct and operate individual facilities, with construction contracts and take or pay revenue agreements signed in advance to create predictable cash flows.  That predictability is what allows the SPV to finance itself at up to 90% debt, significantly more leveraged than a typical LBO, and on 15 year lease terms.  The financing is bankruptcy remote, meaning SPV investors have no direct recourse to the hyperscalers themselves. That structure works cleanly until one of the counterparties at the end of the chain stops performing.  Oracle, which handles two thirds of OpenAI's compute commitments and carries the weakest credit rating among the major hyperscalers, is identified as the most exposed party.  A sustained revenue miss from OpenAI puts Oracle under pressure on its own SPV contract obligations, raising the prospect of a credit downgrade from just above investment grade to junk, with potential covenant implications that would compound the problem further. The episode closes with the broader question of whether the AI infrastructure build-out is entering its first genuine stress test, and what the next 12 months of investor reporting might finally reveal about the numbers behind the narrative. Key Discussion Points: > OpenAI pre-IPO: what the revenue miss and exec conflict signal about the state of the business.  > Hyperscaler CapEx commitments: the scale of spending committed for 2026 and how it is being financed across public and private markets.  > Project finance mechanics: SPV structure, construction contracts, take or pay agreements, and the debt waterfall.  > Leverage and risk: why data centre project finance operates at 90% leverage and why that is only sustainable with locked-in cash flows.  > Oracle's position: credit rating, exposure to OpenAI and the domino risk within the financing chain. Why Wall Street Prep?  Wall Street Prep is the trusted training provider for the world's top investment banks, private equity firms, Fortune 1000 companies and business schools.  Our online training and instructor-led boot camps are direct adaptations of our corporate training, making Wall Street Prep the ideal choice for those looking to break into finance. DISCLAIMER:  The information provided in this video is for educational and entertainment purposes only and does not constitute financial, investment, tax, or legal advice. Investing involves risk, and you may lose some or all of your capital.  Past performance is not indicative of future results. Please conduct your own due diligence or consult with a certified professional before making any financial decisions. WTBD Newsletter: https://webmail.wallstreetprep.com/whats-the-big-deal [https://webmail.wallstreetprep.com/whats-the-big-deal] Follow Us On Socials: LinkedIn: https://www.linkedin.com/company/wall-street-prep/ Instagram: https://www.instagram.com/wallstreetprep/ Resources: https://linktr.ee/wallstreetprep

Private Equity: Leveraged Buyouts Explained (How to Analyze Deals Like a Pro)

This week Graham and Debs try something different.  Rather than dissecting a single deal, they go back to basics with one of the most important concepts in finance — the leveraged buyout — and build up from first principles using two of the biggest real-world examples in the market right now: the $18B acquisition of Hologic and the $55B acquisition of Electronic Arts. Graham walks through the core LBO framework using an accessible house purchase analogy, explaining how leverage turns a 1.5x equity return into a 3x return, what drives that amplification, and what the key variables in any LBO analysis actually are.  From there the conversation covers what makes a good LBO candidate, the concept of cash conversion, how loan-to-value has evolved since the early days of private equity, and the three main value creation levers available to a private equity owner. The second half of the episode puts theory into practice.  Graham runs a live napkin LBO on both the Hologic and EA deals — walking through sources and uses, entry multiples, debt paydown assumptions and return calculations — and asks the central question: do the numbers actually make sense? The episode closes with a broader conversation about the evolution of private equity — from the generalist, high-leverage model of the early 90s to today's specialist, operationally-focused landscape — and what record levels of dry powder mean for returns going forward. Key Discussion Points: LBO fundamentals — what a leveraged buyout is, how leverage amplifies equity returns, and the key variables that drive an LBO model.  LBO candidates — what makes a business suitable for a leveraged buyout: cash conversion, recurring revenues, predictable cash flows.  Sources and uses — how deals get financed, what refinancing existing debt means and why a target's cash is a legitimate source of transaction funding.  Money multiple vs. IRR — what each metric measures and why you need both to evaluate a deal properly.  The Hologic LBO walkthrough — entry and exit multiples, debt structure, return sensitivity and the revolving credit facility question.  The EA deal — 30x entry multiple, $20 billion of debt, and why the base case numbers require a significant EBITDA growth story.  Co-investment and sovereign wealth — why mega-deals increasingly rely on structures beyond the traditional GP/LP fund.  The evolution of private equity — dry powder, multiple expansion and why operational improvement matters more than ever. Why Wall Street Prep?  Wall Street Prep is the trusted training provider for the world's top investment banks, private equity firms, Fortune 1000 companies and business schools.  Our online training and instructor-led boot camps are direct adaptations of our corporate training, making Wall Street Prep the ideal choice for those looking to break into finance. DISCLAIMER:  The information provided in this video is for educational and entertainment purposes only and does not constitute financial, investment, tax, or legal advice. Investing involves risk, and you may lose some or all of your capital.  Past performance is not indicative of future results.  Please conduct your own due diligence or consult with a certified professional before making any financial decisions. WTBD Newsletter: https://webmail.wallstreetprep.com/whats-the-big-deal [https://webmail.wallstreetprep.com/whats-the-big-deal] Follow Us On Socials: LinkedIn: https://www.linkedin.com/company/wall-street-prep/ Instagram: https://www.instagram.com/wallstreetprep/ Resources: https://linktr.ee/wallstreetprep