1929 vs 2025: Crashes, Bubbles, and the (Hard-Won) Lessons We Keep Forgetting
1929 vs 2025: Crashes, Bubbles, and the (Hard-Won) Lessons We Keep Forgetting
Author: Zion Zhao Real Estate | 88844623 | 狮家社小赵
Author’s note: This essay is based on Andrew Ross Sorkin about his new book 1929: Inside the Greatest Crash in Wall Street History—and How It Shattered a Nation. I did some checks on the claims against the historical record, and extends the analysis to 2025’s AI-supercharged cycle. As always, not financial advice. Andrew Ross Sorkin is an American journalist and author. He is a financial columnist for The New York Times and a co-anchor of CNBC's Squawk Box. He is also the founder and editor of DealBook, a financial news service published by The New York Times.
Executive thesis
The 1920s and the mid-2020s rhyme—but they do not repeat. Both periods feature: (i) a democratization of finance, (ii) a powerful general-purpose technology (radio then; AI now) that inflames animal spirits, (iii) leverage and regulatory gaps that migrate to the path of least resistance, and (iv) policy choices that can either cushion or amplify the bust. Where today differs is institutional capacity (SEC, FDIC, macroprudential tools), the composition of leverage (more outside the banking core, notably in private credit and real-estate adjacencies), and the speed/scale at which narratives spread and capital reallocates. The lesson is not to suppress speculation—innovation’s unruly twin—but to box it in with guardrails that protect households without choking off risk-taking (Schumpeterian “creative destruction”), while being precise about where regulation actually comes from and what history really shows.
1) The 1929 setup—what actually happened (and what didn’t)
Consumer and margin credit exploded. The 1910s–20s transformed Americans’ relationship with debt. General Motors created GMAC in 1919 to finance cars for dealers and consumers; department stores and catalog giants like Sears used generous installment plans to sell big-ticket appliances and even kit homes. This diffusion of consumer credit normalized borrowing and pulled future consumption forward (Durkin & Elliehausen, 2014; CRS/Treasury summaries on GMAC). Congress.gov+2U.S. Department of the Treasury+2
Stock speculation rode a similar credit wave. With no federal prospectus regime until 1933 and no SEC until 1934, “brokers’ loans” surged; buying on ~10% margin (10-to-1 leverage) was common in the late 1920s, while supervisory tools were embryonic. The Federal Reserve—created in 1913—experimented in 1929 with “direct pressure” guidance urging banks to curb lending to speculators rather than hiking rates aggressively. That moral-suasion approach proved porous in practice. (See FRB and historical scholarship on speculative credit and the Fed’s 1929 policy debate.) Bureau of Labor Statistics
Narratives mattered—radio and the media minted celebrity business leaders. RCA became the “story stock” of its day, as investors extrapolated a world connected by radio; business magazines (Time 1923; Forbes 1917) placed financiers and industrialists on covers, fusing technology, celebrity, and investing into a self-reinforcing loop—what Shiller later calls “narrative economics.” NBER+1
The characters were real—and consequential.
Charles E. “Sunshine Charlie” Mitchell (National City Bank, now Citi) pushed credit availability, including through a powerful securities affiliate. His 1933 grilling in the Pecora hearings exposed practices that led to reforms. SEC Historical Society+1
Carter Glass (Senator) and Henry Steagall (Representative) shepherded the Banking Act of 1933 (Glass–Steagall): separating commercial and investment banking and creating FDIC deposit insurance—core institutional responses to restore trust. Federal Reserve History+1
John J. Raskob, the GM/du Pont executive and Empire State Building developer, evangelized mass participation in equities in his 1929 Ladies’ Home Journal piece, “Everybody Ought to Be Rich,” emblematic of the era’s democratized—and over-optimistic—finance. Joshua Kennon
The crash and its policy aftermath:
The Dow ended calendar 1929 down ~17%, which masked the deeper drawdown and the slow-rolling collapse that followed. Many believed a quick rebound was coming; it wasn’t. Encyclopedia Britannica
Policy errors compounded the slump: President Hoover signed the Smoot–Hawley Tariff (1930) and the Revenue Act of 1932 (broad tax hikes) amid contracting demand—choices economic historians widely judge as pro-cyclical. Unemployment peaked near 25% in 1933 (Lebergott/BLS estimates), while ~9,000 banks failed between 1930–33, culminating in the March 1933 bank holiday and emergency reforms under FDR. Congress.gov+2Federal Reserve History+2
Bottom line: 1929 was not just “stocks went down.” It was credit innovation without adequate guardrails, narrative-charged speculation, and a cascade of policy mistakes that turned a market crash into a social catastrophe.
2) Correcting the record on today’s regulatory foundations
A point raised in the discussion deserves precision. The “accredited investor” concept sits in Regulation D (1982) under the Securities Act of 1933, not the 1940 Acts. The 1940 Acts (Investment Company/Investment Advisers) govern funds and advisers; the “qualified purchaser” standard lives in the 1940 Company Act. The SEC updated accredited-investor criteria in 2020 to include knowledge-based pathways—not just wealth/income. Conflating these regimes obscures why private-market access remains contested in 2025. Bureau of Labor Statistics+1
3) 2025’s “AI everywhere” moment—bubble or backbone?
Real investment is large and rising. Goldman Sachs estimates that AI-related capex (data centers, chips, software) could approach a few percent of U.S. GDP in coming years; the White House projects AI-driven compute demands could push U.S. data-center electricity use to 7–12% of national consumption by 2028, a material macro tailwind with clear infrastructure bottlenecks. BEA-based commentary suggests that recent GDP prints would have been meaningfully lower absent data-center construction, though exact contributions depend on methodology. Wikipedia+1
But leverage is migrating. Inside the S&P’s “AI core,” spending is mostly cash-funded. On the periphery (real estate power/land, structured leases, private credit, yieldcos), leverage and duration risk can accumulate outside bank balance sheets—harder for regulators to see in real time (echoes of 2008’s shadow credit structure, but with different plumbing). Regulators today have better tools than in 1929, yet they must look where risk moves, not where it was.
Narratives are again powerful. Prices of “picks-and-shovels” enablers, GPU vendors, and AI landlords embed heroic growth paths. That’s not intrinsically unhealthy—speculation finances innovation—but left unbounded it can misallocate capital. As in the 1920s, the story (then radio; now AI agents) can outrun cash flows.
Macro cross-currents complicate the picture. Despite record gold prices in 2025 and renewed “debasement-trade” chatter, U.S. rates and the dollar have not behaved like classic debasement episodes, reminding us that multi-asset signals can conflict. Treat simple narratives with care. Reuters+1
4) Labor markets, productivity, and the AI unemployment debate
A fair fear is whether AI’s productivity boom will rhyme with the 1930s’ mass joblessness. The historical record and current research point to a mixed effect: technology both displaces tasks and creates complementary roles; net outcomes hinge on diffusion speed, retraining, and demand elasticities. Studies by Acemoglu & Restrepo and others show large transitional frictions when automation targets routine tasks, but also sizable gains when new tasks and sectors appear. The policy lever is timing: speed up diffusion where complements (skills, compute, power, data quality) bottleneck, and cushion adjustment where displacement bites. (See peer-reviewed literature on automation and employment; not every gain is a “layoff.”)
5) Tariffs, resilience, and the “Smoot–Hawley problem”
Sorkin raises a crucial distinction: Smoot–Hawley (1930) was an indiscriminate tariff shock that invited retaliation during a collapse in global demand; today’s “national security” trade policy—e.g., targeted curbs on EVs/chips and CHIPS Act-style industrial policy—frames tariffs as a resilience premium. The academic baseline remains: broad, across-the-board tariffs depress welfare and trade (Irwin’s canonical history). The actionable question is not “tariffs good/bad,” but how narrow and how temporary—and whether the security externality is credibly priced. Congress.gov+1
6) What 1929 teaches 2025 (so we don’t relearn it the hard way)
A. Keep speculation, fence the fallout. We need risk-capital markets to fund generational technologies. The job of policy is to channel speculation—through disclosure (Securities Act 1933), market surveillance (SEC 1934), and bank/affiliate separation (Glass–Steagall’s original logic), updated for today’s non-bank and platform risks. Don’t criminalize animal spirits; box them in. Bureau of Labor Statistics+1
B. Protect households at the leverage edge. Most damage in 1929–33 fell on small savers and the newly enfranchised investor. That is why FDIC deposit insurance and orderly resolution were created—and why retail margin rules and suitability standards matter. Extend the same logic to retail access to private markets: broaden access carefully(knowledge-based routes, guardrails against predatory terms), not by diluting disclosures. FDIC
C. Be honest about deficits and cycles. The CBO projects persistent U.S. federal deficits >5% of GDP through the decade, an unusual stance in peacetime expansions; fiscal settings condition the “debasement” debate more than slogans do. Avoid pro-cyclical shifts (Smoot–Hawley-like or 1932-style tax hikes) in downturns; pair industrial policy with offsetting consolidation over the cycle. EH.net
D. Don’t mistake the story for the system. In 1929, RCA’s radio story looked like destiny; in 2025, “AI will eat GDP” sounds similar. Long-only narratives can be right in direction yet wrong in valuation and timeline. Make scenario ranges, not point guesses. Diversify across the stack (compute, power, software, applied AI), stress-test for power constraints, export controls, and policy risk.
E. When the wind shifts, move fast. The 1933 bank holiday, Emergency Banking Act, and subsequent Glass–Steagall/FDIC moves rebuilt confidence quickly. Crises are won on speed and clarity. If a non-bank funding run emerges (e.g., in private credit or structured real-estate tied to AI capex), be ready with standing facilities and resolution playbooks that reach beyond insured banks. Federal Reserve History
7) Are we “in 1929” again?
My humble opinion would be No. The architecture built in 1933–34—and expanded after 2008—means the U.S. now has playbooks unimaginable a century ago. But we are repeating the human parts of the cycle: optimism chasing a general-purpose technology; credit that oozes into the least-regulated cracks; and policy that risks being late either to rein exuberance or to support demand in a downturn. The best guardrail is sober, historical memory—precisely the contribution Sorkin’s 1929 offers by putting people and incentives at the center of the story (HarperCollins 2025; Reuters review). FRASER
Practical checklist for investors & policymakers (2025)
Follow the plumbing: Track leverage migration (private credit, real-estate adjacencies to AI, structured utilities). If you can’t map counterparty chains, size positions smaller.
Underwrite power risk: Data-center economics hinge on power availability/costs and grid timelines—more than on chip counts alone. Factor national and local constraints explicitly.
Demand real disclosure: In public and private markets, insist on audited metrics (utilization, pre-leases, contracted power, take-or-pay). The 1920s lacked prospectuses; you don’t. Bureau of Labor Statistics
Diversify narratives: Own some “picks-and-shovels,” some profitable adopters, and some “boring” cash-yielders that benefit from AI indirectly. Avoid single-story exposure.
Policy: precision over posture. Use scalpel-not-sledgehammer tariffs; keep industrial policy time-bound; safeguard households at the margin; keep crisis facilities broad enough to reach non-banks.
Closing take
The 1929 crash was the culmination of exuberant credit innovation, porous guardrails, and policy error. Today’s cycle rides on unprecedented AI investment and powerful narratives—but with stronger institutions and better (if incomplete) visibility into risk. If we remember why Glass–Steagall, the SEC, and FDIC were built—and update their logic for non-bank leverage and platform finance—we can welcome the right kind of speculation, protect households, and avoid turning the next correction into a social calamity. That is the sober gift of 1929’s history for 2025.
Build Resilience in a 1929-vs-2025 World—Own Singapore, Smartly
For International / China Chinese / SE Asia / Singapore investors, UHNW families & institutions
In a cycle where narratives move faster than fundamentals, you deserve a real-estate advisor who can read both the tape and the title deed. I’m a Singapore-based realtor and SAF (Army) Officer Commanding (Captain) with deep training in macroeconomics, cross-asset investing (equities & crypto), portfolio construction, and Singapore Land & Business Law. Every day, I dedicate hours to research—writing long-form essays, stress-testing data, and doing real due diligence—so my clients can make decisions that stand up to volatility.
Why work with me (humbly stated, results-focused)
Cross-asset lens, not tunnel vision: I translate lessons from 1929 vs 2025—credit cycles, leverage migration, policy risk—into entry timing, financing choices, and exit strategy for property.
Legal & risk discipline: Contracts aligned to Singapore statutes, structured to protect you (and your family office) on covenants, timelines, and deliverables.
Macro → Micro selection: From URA Master Plan and GLS land cost to rental comps and yield-on-cost, I map top-down signals to bottom-up assets.
Global clientele, seamless execution: Discreet support for 移民/留学/家办 needs (immigration, study, family offices), with a trusted network of bankers, lawyers, tax and relocation partners.
Service mindset: I’m courteous, professional, and humble—your goals come first.
The portfolio case—why include Singapore real estate now
In a world of AI booms and shifting rates, Singapore property adds a lower-volatility, income-producing core: historically resilient prices, dividend-like rental yield, and transparent rule-of-law governance. It complements equities/crypto by smoothing drawdowns, while still offering long-run capital appreciation driven by population, infrastructure, and prudent policy. (No guarantees—just disciplined positioning.)
What you’ll get in our first strategy session (complimentary)
Portfolio fit: How property can stabilise your broader holdings (risk budget, FX, liquidity).
Target map: Shortlist by district, tenure, quantum, rental yield, and exit pathways.
Deal hygiene: Red-flag scan on titles, timelines, and financing structure.
Action plan: Next steps for acquisition, leasing, and asset management.
Call to Action
Add a resilient Singapore property sleeve to your portfolio—thoughtfully, legally, and data-driven.
Message me to book a confidential strategy consult (English / 中文).
面向国际/中国/东南亚/新加坡投资者与家办的专业置业服务（简体中文）
在“1929 与 2025 交错”的时代，您需要的不只是房产中介，而是懂宏观、懂资产配置、懂新加坡法律的全能顾问。我每天投入大量时间写作与研究，严谨尽调，用跨资产视角（股票、加密资产、利率与政策）为您把关选筹、定价、融资与退出。
为什么现在配置信加坡房地产？ 历史波动更低、租金像股息般稳定，且受益于人口与基建驱动的长期增值；与股市/加密形成互补，平滑组合回撤。
欢迎预约一对一保密咨询（英文/中文），获取您的专属投资清单与执行方案。
Compliance & care: This is general information, not financial advice.
References (APA)
(Where relevant, statements above also include inline web citations.)
Acemoglu, D., & Restrepo, P. (2019). Automation and new tasks: How technology displaces and reinstates labor. Journal of Economic Perspectives, 33(2), 3–30.
Board of Governors of the Federal Reserve System. (n.d.). Securities Act of 1933 and Securities Exchange Act of 1934 (SEC historical overviews). Retrieved 2025. Bureau of Labor Statistics
Bureau of Labor Statistics. (n.d.). Great Depression unemployment estimates. Retrieved 2025. Bureau of Labor Statistics
CBO. (2024). The Budget and Economic Outlook: 2024–2034. Washington, DC: Congressional Budget Office. EH.net
Durkin, T. A., & Elliehausen, G. (2014). Consumer Credit and the American Economy. Oxford University Press. See overview/summaries. Google Books+1
FDIC. (n.d.). A Brief History of Deposit Insurance in the United States. Washington, DC. FDIC+1
Federal Reserve History. (n.d.). Banking Act of 1933 (Glass–Steagall). Federal Reserve History+1
HarperCollins. (2025). 1929: Inside the Greatest Crash in Wall Street History—and How It Shattered a Nation (A. R. Sorkin). FRASER
Irwin, D. A. (2011). Peddling Protectionism: Smoot-Hawley and the Great Depression. Princeton University Press. (See also Britannica overview.) Congress.gov
Macrotrends. (n.d.). Dow Jones—historical annual returns. Retrieved 2025. Encyclopedia Britannica
Pecora, F. (1933). Stock Exchange Practices (U.S. Senate hearings). (Mitchell testimony and final report). SEC Historical Society+1
Reuters. (2025, Oct). Gold rallies/pulls back—debasing debate; records and forecasts. Reuters+2Reuters+2
SEC. (2020). Amending the “Accredited Investor” Definition. Washington, DC. FRASER
Shiller, R. J. (2019). Narrative Economics: How Stories Go Viral and Drive Major Economic Events. Princeton University Press.
Smithsonian Magazine. (2011). The man who busted the ‘banksters’ (on the Pecora Commission). Smithsonian Magazine
Time | Forbes. (n.d.). Founding histories (Time 1923; Forbes 1917). Yahoo Finance
U.S. Treasury (2010). Written testimony on GMAC/Ally—history and role. U.S. Department of the Treasury
White House (2025). Assessment of the AI Revolution and the Grid. (Data-center electricity share projections). SEC
Comments
Post a Comment