▶ YouTube Summaries

• https://www.youtube.com/watch?v=ohCAPyqlJpo
• Original title: The (Overdue) Collapse of Bullsh*t Companies

The Big Four (Deloitte, PwC, EY, KPMG) and their peers in consulting, credit ratings, and law built near-monopolies on reputation. Once safe from competition, their service quality quietly rotted for decades. Now two forces — AI erasing their cheap-labour moat, and governments waking up to their conflicts of interest — are cracking a model that dominated corporate America for a century.

Decay is visible

The rot shows up in headline failures. In December 2024 Deloitte delivered a $290k report to the Australian government (on a broken welfare system) that was AI-written and riddled with fake references, fake court quotes, and citations to work that did not exist. In 2014 PwC overstated Tesco's projected quarterly profit by £250M, nearly 30% too high. These are not one-offs: in 2022 regulators found 43% of EY's audits had a "deficiency" — the auditor did not even gather enough evidence to support the opinion it signed. Across all Big Four the average deficiency rate more than doubled, from 12% in 2020 to 26% by 2022.

How the monopoly formed

1970s multinationals (IBM, Coca-Cola, Ford) needed one trusted outsider to verify books consistently across dozens of tax regimes. That gave rise to the Big Eight, who by the late 1980s controlled ~98% of all public-company audits. Consolidation shrank them 8→5→4. New entrants cannot break in because auditing runs on decades-old reputation — no board picks an unknown startup over the firm that already audits their competitors and half the governments of Europe. The same dynamic concentrated credit ratings (S&P, Moody's, Fitch ~96% of the market) and strategy consulting (McKinsey, BCG, Bain). "We hired McKinsey" is simply easier to defend in a boardroom. With no alternative, buyers tolerate ever-worse service.

Enron, and the lesson ignored

In 2001 Arthur Andersen — one of the Big Five — took $25M in audit fees plus $27M in consulting fees from Enron, then missed roughly $20B in debt hidden through shell companies. Andersen collapsed nine months later. Rather than concluding the sector was too consolidated, the industry consolidated further. Today the four survivors audit 498 of America's 500 largest firms. Clients switch auditor on average once every 23 years; BCE Inc. has been audited by Deloitte for 144 years.

Two threats now

First, AI. The grunt work that justified armies of junior graduates — data extraction, pattern matching, document review — is exactly what AI speeds up, and the technology is not exclusive to the incumbents, so the barrier to entry collapses.

Second, antitrust and conflict of interest. Advisory work now dwarfs audit: the Big Four pulled $95B from advisory in 2023 versus $66B from audit and assurance. But the same firm auditing and consulting for a client is marking its own homework — it has an incentive to sign the books generously and access to confidential data worth a fortune to the consulting arm. In 2013 a PwC partner shared confidential Australian tax-law information with at least 53 partners to help clients including Google get ahead of a law PwC itself helped design. Fallout: PwC was forced to sell its Australian government consulting business for $1; from 2024 UK firms had to operationally separate audit from the rest; the US Senate has weighed a breakup. Splitting hurts because both arms prop each other up — consulting is the profit centre, but it exists only on the credibility and client access audit built over a century.

Consulting output is thin

A leaked 2016 McKinsey deliverable offered phrases like "develop value-creating partnerships" and "build a clear mission" — millions of dollars of essentially meaningless words dressed up in a fancy graph and a complimentary colour tone. By 2024 only 13% of businesses felt consultants did more good than harm. As firms and governments cut these contracts, the arm propping up the whole operation shrinks.

What replaces them

Three non-exclusive outcomes, likely all at once. (1) Companies bring the work in-house, since AI can generate internally what they used to outsource. (2) Small boutique firms compete seriously for the first time in decades, as AI closes the output gap between a 300,000-person firm and a 30-person one. (3) The work stops needing a firm at all — a single experienced ex-Big-Four professional plus AI can replace armies of juniors and undercut dramatically on price with no overhead. The century-old model is no longer guaranteed.