Andrew Ross Sorkin Is Right About 1929 – And Wrong About 2026

A counterpoint to the confidence in post-Depression regulatory protections

Published:

Disclaimer This analysis is not investment advice, financial guidance, or market timing recommendations. No content should be interpreted as suggesting specific investment actions or positioning strategies. Readers should consult qualified financial professionals before making investment decisions.

Andrew Ross Sorkin’s new book on the 1929 crash carries a comforting message. In recent interviews, he systematically catalogs why another 1929-style collapse can’t happen: the SEC prevents manipulation, the FDIC stops bank runs, capital requirements prevent insolvency, and the Federal Reserve has learned the Bernanke playbook. “The 1920s, there was no regulations. Zero,” he explains. “There were no insider trading laws, there were no capital requirements for banks.”

He’s right about all of it. The traditional banking system is protected.

And he’s missing the actual crisis.

While Sorkin focuses on traditional banks within the regulatory perimeter, $48.1 trillion sits in retirement accounts—with roughly $33 trillion in tax-deferred accounts now subject to forced withdrawals under Required Minimum Distribution rules. This isn’t money being moved by investor sentiment or market panic. It’s being moved by the Internal Revenue Code, on a schedule set by demographics, outside any regulatory protection Sorkin describes. More importantly, these forced liquidations are peaking right now—and their transmission through shadow banking, insurance companies, and into broader markets creates the exact kind of structural fragility that brought down 1929, just in a completely different form.

This isn’t speculation about whether history repeats. It’s arithmetic about what’s happening now.

The Architecture Sorkin Trusts

To be fair, Sorkin’s confidence has a foundation. The post-1929 regulatory architecture represents one of the most successful policy achievements in American history. The Glass-Steagall Act of 1933 separated commercial and investment banking. The FDIC, created the same year, insured bank deposits and ended the bank run problem. The Securities and Exchange Commission, established in 1934, brought transparency and enforcement to markets where manipulation had been legal. Bank capital requirements mean institutions have buffers to absorb losses.

These protections are real. They worked in 2008, when the government deployed extraordinary measures—TARP, Fed liquidity facilities, coordinated central bank action—to prevent a second Great Depression. Ben Bernanke, who studied 1929 extensively, had the playbook ready. “The playbook is to throw money at the problem,” as Sorkin notes. “We now have the playbook.”

When Sorkin worries about the future, he focuses on government debt constraints—whether bond markets will tolerate the next round of crisis intervention when the national debt exceeds $38 trillion. It’s a legitimate concern, if that’s where you think the next crisis originates.

But notice what’s missing from this list of protections.

The $700 Billion Annual Forced Liquidation Nobody Regulates

Every day in 2026, more than 11,200 Americans turn 65. This isn’t a projection—it’s locked in by birth records from 1961. These are baby boomers, the 67 million Americans born between 1946 and 1964, now reaching what actuaries call “Peak 65.” It’s the largest surge of new retirees in American history.

Eight years from now, in 2034, these same individuals will turn 73—the age at which the IRS requires them to begin withdrawing money from their retirement accounts. It’s called a Required Minimum Distribution, or RMD. The government let you defer taxes for decades. Now it wants those taxes.

The math is straightforward. Americans held $18.9 trillion in IRAs and $13.9 trillion in defined contribution plans as of September 2025. Not all of that is subject to RMDs—Roth accounts are excluded—but conservatively, $25-30 trillion in tax-deferred retirement assets will face forced withdrawals over the next decade. At age 73, you must withdraw roughly 3.8% of your account balance. At 75, about 4.1%. At 80, 5.3%. The percentages rise as you age because your remaining life expectancy shortens.

Here’s what that looks like in aggregate: with roughly 4 million people per year hitting RMD age, and another 20-25 million already taking RMDs at higher ages, the annual forced withdrawal from tax-deferred accounts approaches $600-800 billion. Every year. By law. Non-negotiable.

This is the demographic time bomb nobody’s pricing in.

And here’s the part that matters for 2026: those withdrawals don’t come from liquid Treasury bonds or money market funds sitting in retail brokerage accounts. They increasingly come from alternative assets—private credit, private equity, real estate partnerships—held in pension plans and self-directed IRAs. The baby boomers spent decades building up these accounts, and advisors spent decades convincing them to “diversify” into higher-yielding alternatives during a zero-rate environment.

CalPERS, the nation’s largest public pension, has 33% in alternatives. Many other major pensions run 25-40% alternative allocations. When individuals reach 73 and request their RMD, pension plans can’t simply mark-to-model and send a check. They need actual cash. If they can’t sell the illiquid alternatives fast enough, they sell what they can: public equities and Treasury bonds.

Now scale that to $600-800 billion annually, with the peak hitting between 2024 and 2030.

This forced liquidation operates entirely outside the regulatory perimeter Sorkin describes. The SEC doesn’t regulate private credit valuations. The FDIC doesn’t insure pension alternative assets. Bank capital requirements don’t apply to these holdings. And the Fed’s playbook assumes crisis arrives as a shock, giving authorities time to deploy liquidity—but RMDs operate on a calendar, not a crisis timeline.

The Transmission Mechanism Through Insurance

Here’s where demographics meet finance in a dangerous way. The same baby boomers facing RMDs also bought permanent life insurance policies over the past 30-40 years. In recent years, private equity firms discovered they could buy these life insurance companies, invest policyholder premiums in private credit and alternative assets, and earn the spread between what they promise policyholders and what they earn on investments.

Apollo Global Management’s Athene platform holds $430 billion in total assets, with over 50% in illiquid credit allocations. The model works brilliantly in normal times: policyholder obligations are long-dated and predictable, so you can invest in illiquid assets with higher yields.

But insurance companies have a structural vulnerability. They need to maintain Risk-Based Capital ratios—regulatory thresholds that ensure they can meet obligations. When private credit marks down 10-15% (the level many analysts expect if we enter recession), those RBC ratios fall.

The U.S. insurance industry holds approximately $362 billion in U.S. government bonds across all carriers as of year-end 2024, representing just 6.7% of the industry’s $5.4 trillion bond portfolio. Critically, PE-backed insurers like Athene (51% alternatives), Global Atlantic (45% alternatives), and Fortitude Re (>50% alternatives) hold disproportionately fewer Treasuries than the sector average due to their aggressive alternative allocations. Their Treasury buffers are thinner precisely when their private credit exposures create the greatest liquidity stress.

This creates a fatal imbalance: a 30% mass lapse scenario across the sector would generate approximately $400 billion in liquidity needs, requiring liquidation of nearly 90% of the industry’s Treasury holdings plus substantial sales of other liquid investment-grade bonds. For PE-backed carriers with minimal Treasury allocations and maximum alternative exposures, even modest private credit impairment (10-15%) leaves them with insufficient liquid assets to meet accelerated surrenders.

Now combine two forces:

  1. RMD-driven pension liquidations: $600-800B annually in forced withdrawals, with $250B+ coming from pension alternatives that require selling liquid holdings
  2. Insurance sector stress: If private credit impairment reaches 10-15%, insurance companies face forced liquidation of nearly their entire $362B Treasury portfolio to maintain capital ratios, with PE-backed carriers hitting liquidity constraints first

October 2026 represents a particular stress point. Private credit deals from 2021—the peak of zero-rate-driven lending—face their first major refinancing wave. CLO overcollateralization tests, which have been deteriorating since Q4 2025, likely fail in significant numbers. The private credit secondary market faces accelerating forced selling: LP volume surged 40% year-over-year in H1 2025, with public pensions driving 48% of activity as they manage -2% to -4% annual cash flow deficits—and these distribution pressures intensify through Q1 2026 marking season and Q2 2026 maturity walls.

When insurance companies liquidate hundreds of billions in Treasuries to meet RBC requirements and policy surrenders, yields spike. This isn’t monetary policy or inflation expectations—it’s forced supply hitting the market. And higher Treasury yields mean higher discount rates for equity valuations.

The Equity Cascade: When 60% of the Market Is on Autopilot

This is where market structure amplifies the crisis. In recent analysis, investor Steve Eisman documented that 60% of equity market flows are now passive—index funds and ETFs with zero human discretion about valuation. These funds must buy or sell based purely on flows in and out. When Treasury yields spike from forced insurance liquidations, equity valuations drop (higher discount rate). When equities drop, retail investors hit the sell button on their index funds. And 60% automated selling overwhelms the 40% active buying capacity.

“Fast and very ugly,” in Eisman’s words.

Between February and April 2025, the market fell 19% in seven weeks—without a recession, without an AI spending slowdown, without any fundamental economic deterioration. That was a warning shot. Eisman notes that if there is “ever an actual recession or AI bubble bursts, the decline will almost certainly be steeper.”

Now imagine the same market structure facing three simultaneous pressures:

The passive market structure Eisman describes makes cascade risk mechanical, not probabilistic.

Why Policy Can’t Stop This

Sorkin’s confidence rests on the government’s ability to “throw money at the problem” when crisis arrives. The 2008 playbook: Fed liquidity facilities, TARP, coordinated central bank action, whatever it takes.

There’s a timing problem.

The 2008 crisis arrived suddenly. Lehman Brothers collapsed on a weekend. Markets seized up. By Monday morning, policymakers knew they had an emergency. The intervention could begin immediately because the crisis announced itself dramatically.

RMDs don’t work that way. They operate on a calendar. People turn 73 every day. They request withdrawals throughout the year. Pensions gradually increase their selling to meet RMD obligations. Insurance companies watch their RBC ratios deteriorate slowly as private credit marks drift lower. There’s no single crisis moment that triggers “break glass” intervention.

By the time official statistics show recession—which typically lag real economic deterioration by 6-9 months—the forced liquidation will already be underway. Consumer savings have been depleting since late 2024. Q1 2026 earnings reports will likely reveal the AI spending slowdown. The insurance sector Treasury liquidation follows as private credit stress builds. The equity cascade is triggered by these mechanical forces, not by a sudden shock.

When policymakers recognize the need to intervene, the intervention arrives after the cascade has begun—when it’s least effective and most expensive. The fiscal multiplier for government spending is currently around 0.69, meaning each dollar spent generates only 69 cents in GDP growth. With $38 trillion in national debt, the political will for massive intervention is weaker than in 2008.

More fundamentally, there’s no policy tool to stop RMDs. The IRS could theoretically waive the requirement, as it did for one year during the 2009 crisis and again in 2020 for COVID. But that only delays forced selling by a year—it doesn’t eliminate it. And it removes $600-800 billion in annual tax revenue at a time when Social Security and Medicare face their own funding crises from the same demographic wave.

The Fed could buy private credit directly, backstopping insurance companies’ illiquid holdings. But the legal authority is questionable under Dodd-Frank, the political optics are terrible (Wall Street bailout 2.0), and the moral hazard is extreme. More likely, the Fed waits until Treasury liquidations spike yields, then steps in to buy Treasuries—which solves the symptom but not the cause.

The Difference Between 1929 Protection and 2026 Vulnerability

Sorkin writes that “each wave seduces us into thinking that we’ve learned from history and, this time, we can’t be fooled.” He’s warning against complacency. But the complacency isn’t about believing markets are rational—it’s about believing our regulatory protections cover the actual vulnerabilities.

The 1929 crash taught us to protect banks, regulate securities, and maintain lender-of-last-resort capability. We built those protections, and they work. But they protect the form of crisis we had in 1929: bank runs, stock manipulation, monetary policy errors.

The 2026 vulnerability is different. It’s not about banks failing—it’s about forced demographic liquidation transmitting through shadow banking into market structure designed for passive flows. It’s not about the absence of regulation—it’s about regulation (RMD requirements) forcing selling that other regulations can’t stop.

Consider the parallels that matter:

1929: Mass psychology drove selling. Once confidence broke, there was no natural buyer.
2026: Forced liquidation drives selling. RMDs + insurance stress create structural selling regardless of sentiment.

1929: No circuit breakers or market structure to prevent cascade.
2026: Circuit breakers exist, but 60% passive flows overwhelm them in multi-day declines.

1929: Government intervention came too late and was too small.
2026: Government intervention will come after forced liquidation is underway, when it’s least effective.

1929: Recovery took a decade because structural reforms came slowly.
2026: Recovery mechanism unclear—active managers burned by underperformance, passive flows can reverse but not stabilize, tax lock-in broken by forced selling can’t be rebuilt.

The question isn’t whether we have better protections than 1929. We do. The question is whether those protections address the actual vulnerability. They don’t.

The Calendar Is Running

Sorkin spent eight years researching his book. It’s an important contribution to understanding financial history. But history’s value lies in recognizing patterns, not in assuming the same crisis repeats in the same form.

We’re not going to have 1929 again. We’re going to have something different that produces similar outcomes despite different safeguards.

The demographic wave is locked in. The RMD calendar operates mechanically. The insurance transmission path is documented. The passive market structure creates cascade risk. These aren’t speculative concerns—they’re observable facts about how the system is structured today.

Three independent forcing functions, each with substantial individual probability, converging on 2026:

  1. Boomer RMD peak: 11,200 people daily hitting forced distribution age, $600-800B annual liquidation
  2. Shadow banking stress: Private credit 2021 vintages refinancing into higher rates, insurance sector RBC pressure
  3. AI capex correction: $602 billion planned 2026 spend increasingly disconnected from revenue realization

If these converge as structural analysis suggests, the cascade follows mechanically from the arithmetic

Sorkin’s confidence in post-1929 protections isn’t wrong. It’s misdirected. The fortress is secure, but the siege operates outside the walls. And the calendar says the siege begins in months, not years.

These mechanisms operate on calendars, not sentiment. RMD season runs January through April for most taxpayers. Q1 earnings reports come in mid-April. Private credit marking season happens in Q1 and Q2. Insurance companies report their capital ratios quarterly.

By the time authorities recognize this as a coordinated crisis rather than isolated stresses, the transmission will already be underway. That’s not pessimism. That’s just how calendar-driven crises work.

Andrew Ross Sorkin wrote an excellent history of 1929. But when he says we can’t have another crisis of that magnitude because of the protections we built, he’s defending the fortress while the actual threat operates on completely different terrain.

History doesn’t repeat. It rhymes. And if you listen carefully, the rhythm building in 2026 sounds nothing like 1929—but it might produce remarkably similar outcomes.

The math isn’t speculative. It’s arithmetic. And arithmetic doesn’t care about regulatory confidence.

This analysis identifies structural vulnerabilities based on observable facts. Actual outcomes depend on policy responses, market adaptations, and timing variables that remain uncertain.

Further Reading:

— Free to share, translate, use with attribution: D.T. Frankly (dtfrankly.com)

§