March 2025 Fintwit Book Club: Diary of a Very Bad Year with Byne Hobart from The Diff
Summary
The book’s core lesson is that genuine expertise provides a probabilistic edge, not immunity from consequential error. The anonymous hedge fund manager is incisive across markets yet says in March 2008 that “the worst has passed,” “subprime looks contained,” and Bear Stearns lacks a solvency problem. Byrne Hobart’s corrective to hindsight bias: “on a dollar-weighted basis almost nobody saw a financial crisis like that coming”—otherwise positioning would have defused it before it became a crisis.
The decisive 2008 failure was not merely mortgage losses but an information shock that seized the financial plumbing. Once AAA could mean either “money good” or “probably worth 95 cents on the dollar,” paper financed with 3 cents of collateral suddenly demanded far more, forcing deleveraging across apparently unrelated strategies. The book’s tap-water analogy carries the mechanism: systemic failure begins when a foundational assumption stops being true.
Credit bubbles can grow from tiny pricing errors because scalable funding attracts precisely the borrowers who should pay more. Equities can trade at 10 times fair value or 30 times forward revenue; credit may only be mispriced from 7% to 6.5%, but scalable funding turns that half-point error into a concentrated book. As Walker puts it, the scariest financial institution is a fast-growing one: losses appear years later or all at once in a downturn.
The most clarifying underwriting question is “what economic activity is being funded here?” Hobart’s specimen is 2021 DeFi yield farming: a dollar-pegged asset paying 20% was ultimately the riskiest layer of a leveraged margin-lending stack, not a productive source of 20% returns. In housing, actual credit funded construction, wages, remittances, and consumption abroad—so unlike vanished equity market capitalization, the money went somewhere real and often irrecoverable.
Uncertainty can damage an economy before conventional fundamentals reveal the break. Home prices roughly stalled in 2005, the first Bear Stearns hedge fund collapsed in 2007, and the full crisis arrived later; Walker wonders whether the March “tariffs on, tariffs off” regime could similarly freeze investment with effects visible only 15 months afterward. Hobart refuses false precision: Q4 2018 also felt ominous, yet going entirely to cash would have been “catastrophically bad.”
The AI boom has obvious misallocation risk, but its asset duration is materially shorter than housing or WeWork’s long-lease structure. Walker says CoreWeave used to depreciate GPUs over three years and, he thinks, was using five years in its IPO, while Hobart expects almost all peak capex to be fully depreciated by 2030 or 2031. Extra generation and cheaper power could remain useful even if demand forecasts disappoint, though Walker warns that prolonged exuberance could create follow-on distortions. Hobart’s balanced verdict is blunt: “current economics do not support current capex,” yet AI research already compresses hours of source-finding into roughly 10 minutes.
AI adoption in investing may divide flexible thinkers from rigid ones more than young analysts from older portfolio managers. A 23-year-old may resist LLM transcript summaries because meticulous manual reading is the only process he knows, while a 52-year-old accustomed to delegation may adopt them immediately. The investable edge remains judgment: recognize when to verify deeply, when a shortcut is enough, and when “something has switched” and it is time to “go for the jugular.”
Deep dive
1. Expertise can be real and still fail at the decisive moment
Byrne Hobart reads Diary of a Bad Year: Confessions of an Anonymous Hedge Fund Manager as a meditation on expertise, its limits, and action under uncertainty. The anonymous manager is “clearly very smart,” intellectually wide-ranging, and consistently focused on underlying systems—yet his accurate calls coexist with conspicuous errors.
Walker’s sharpest evidence comes from March 2008: the manager thinks “the worst has passed,” expects things to be fine, calls subprime contained, and says Bear Stearns has no solvency problem. The diary format preserves those mistakes before hindsight can sanitize them.
Walker asks whether readers are unfairly Monday-morning-quarterbacking Bear. Hobart’s market-based response is that sophisticated counterparties, including firms with explicit crisis-risk policies, still entrusted Bear with money and prime-broker relationships. Walker notes that plenty of people were predicting failures across Bear, Lehman, Bank of America, and other major institutions, but that does not make the manager’s call obviously irrational in the moment.
2. The crisis lived in financial plumbing, not just mortgage losses
Hobart argues that an obvious, gradually recognized housing slowdown would not have produced the same crisis. Ordinarily, weakening credit causes capital inflows to slow and risk to come off incrementally; a crisis requires a feedback loop that keeps financing alive until the system breaks abruptly.
The second-order shock was informational: holders no longer knew what complicated securities contained. AAA might still mean full repayment, or it might mean “probably worth 95 cents on the dollar”—a distinction that becomes existential when financing requires only 3 cents of collateral per dollar.
The anonymous manager’s memorable rule is that major blowups begin when an assumption proves false. The interviewer’s analogy is ordinary tap water: people organize life around water appearing when the tap turns, so discovering it is unavailable disrupts far more than the immediate transaction.
Hobart’s pushback on crisis revisionism: many celebrated subprime bears, including people discussed in The Big Short, identified bad mortgages without fully tracing the dollar-liquidity crunch that followed. What converted periodic bank write-downs into systemic failure was that “the financial plumbing seized up,” making short-term dollars and adequate collateral suddenly difficult to source.
3. Tiny credit-pricing errors scale into dangerous books
Hobart contrasts equity excess with credit excess. A company might trade at ten times what it should be worth or at 30 times revenue two years forward; in lending, the initiating mistake may be merely charging 6.5% when the risk deserved 7%.
Cheap credit scales that modest error. A borrower suddenly able to finance an otherwise marginal asset has strong incentives to do a lot more of it, while a bond fund naturally attracts issuers amazed they can borrow so cheaply. In consumer finance, the analogous borrower sees a $5,000 credit limit as “free money.”
Walker’s formulation is worth keeping: “the scariest thing in finance is a fast-growing financial institution.” Mispriced loans can look excellent for three years because defaults lag origination; only a downturn reveals that 9%, rather than 6%, was the appropriate price.
Recent parallels are mechanisms, not repetitions. Walker points to SVB, First Republic, and Flagstar/NYCB: securities assumed to be money-good can still create balance-sheet trouble if held in size and prices fall, while regulated rents and rising operating costs can undermine loans that once appeared durable.
4. Asking where the money went exposes the real misallocation
The anonymous manager repeatedly asks what underlying activity a boom finances. Hobart applies that test to friends earning 20% on dollar-pegged DeFi assets in 2021: the yield ultimately represented the highest-risk slice of a giant decentralized margin-lending stack, not productive activity capable of turning $1,000 into $1,200.
“Where did the money go?” is often wrong for equities because market capitalization is last price multiplied by shares; much of a collapse is vanished paper wealth. Credit is different: proceeds bought assets, paid workers, or caused assets to be constructed, so actual cash necessarily traveled through the economy.
Housing credit therefore became lumber, houses, construction wages, remittances, and consumption elsewhere. Hobart’s vivid endpoint is a recent immigrant sending earnings home, where parents finally buy a used car; recovering the defaulted bond would notionally require seizing that car in rural Mexico—neither practical nor morally persuasive.
This lens also explains the manager’s interest in emerging-market sovereign credit. Buying a Brazilian bond requires asking whether government spending will expand GDP and the future tax base enough to service it, or whether something else will happen to the borrowed money.
5. The manager’s breadth revealed correlations, sequencing, and creditor power
Both speakers speculate, without identifying him, about the anonymous manager. Walker initially suspects a macro trader, then considers someone from prop trading or a value-oriented emerging-markets background. Hobart guesses the fund may have started in the 1980s or 1990s with convertible arbitrage before expanding into a sprawling mix of strategies, including emerging markets, sovereign credit, black-box trading, and private lending.
Hobart treats that breadth as risk management: an apparently unrelated market may become a bottleneck for the strategy paying one’s bonus. In August 2007, statistical arbitrage suffered “a nightmarish couple weeks” when funds losing money in mortgage securities de-grossed other books, revealing correlations created by shared ownership and leverage.
The manager was early to name-check Huawei and correctly anticipated a U.S. credit-rating downgrade, but thought the downgrade would be a much bigger event than it was. Walker’s sequencing explanation is that it happened after recapitalization, when Europe had become more of the problem area and the U.S. remained a safe haven; he suggests it may even have been bullish for the dollar as investors sought dollar liquidity.
Hobart adds a counterfactual: had the downgrade occurred before 2008, regulators might have been slow to adjust risk weights, forcing banks to raise capital against Treasuries while also covering subprime losses and potentially worsening the crisis.
When tradable-credit spreads became “pennies in front of a steamroller,” the manager moved toward private credit and used every available pressure point. He called a delinquent borrower’s customers and suppliers—“Did you know that this company doesn’t pay its bills?”—and used the resulting pressure to force repayment.
6. Capital allocation can magnify talent and rationalize self-interest
The manager raises the concern that PhDs, doctors, and other scarce specialists migrated into pre-crisis finance. Walker updates the tension: a drug researcher might change lives inside Pfizer, yet potentially earn 100 times more investing in the companies developing the best drugs.
Hobart holds the libertarian counterargument alongside the discomfort. Complex economies need information routed to the right decision-makers, and financial prices might guide capital indirectly—even telling a CFO whose stock rose from 20 to 50 times earnings to stop repurchasing shares and reinvest.
Promotion turns specialists into allocators everywhere. Mark Zuckerberg mostly directs layers of people rather than writing code; an exceptional FDA reviewer may come to supervise reviewers; a McKinsey analyst doing quantitative work can become a partner focused on clients, sales, and team direction.
The scale argument is powerful: a doctor treats perhaps eight patients daily, while financing a drug might affect 800,000. Hobart’s warning is sharper: be suspicious whenever careful reflection reveals that “the most lucrative thing” is also conveniently the most moral thing one could do.
7. Uncertainty can freeze activity before fundamentals visibly break
On March 26, Walker connects the diary to markets enduring “tariffs on, tariffs off” and shifting definitions of friend and enemy. Earlier in the month, the Russell had fallen for “like 12 weeks in a row,” turning friends’ opportunity emails into what he calls therapy sessions.
The book’s cleanest self-inflicted uncertainty comes from automakers and dealers warning Congress that nobody would buy a car from a bankrupt manufacturer. The manager thinks the industry created the fear it described: consumers already bought airline tickets from bankrupt companies, but public rhetoric taught them cars were different.
Hobart resists turning present anxiety into a mechanical crash call. Q4 2018’s Nasdaq decline made worsening sentiment and slower corporate spending plausible, yet an investor who declared the growth cycle finished and moved entirely into cash would have made a catastrophically bad call.
His underwriting approach is to examine the final activity. Lending is safer when the marginal dollar funds an asset whose income services the debt; danger begins when new liquidity merely raises collateral values supporting older loans, with “income plus expected price appreciation” disguising the circularity.
8. AI’s short asset duration limits losses, but infrastructure can outlast hype
Walker frames the allegedly March 27 CoreWeave IPO as a possible WeWork-style comparison: bulls see technological growth, bears see fragile structure. Hobart’s distinction is duration—WeWork took long-term leases and resold office space on the spot market, making it first to absorb a collapse in demand.
CoreWeave’s GPUs depreciate much faster. Walker says the company moved from roughly three-year to five-year depreciation, while Hobart expects almost all peak AI capex to be fully depreciated by 2030 or 2031. Short duration cannot prevent mistakes, but it can cover more of them than a decades-long asset structure.
Walker’s pushback is that the boom is reshaping longer-lived infrastructure: Microsoft and Constellation were entering a deal to restart Three Mile Island, while the investment case contemplated U.S. power-demand growth rising from roughly 0% over the prior 20 years toward 5% annually. If AI disappoints after two more exuberant years, excess generation and related infrastructure could create follow-on distortions, even if consumers ultimately benefit from surplus power.
Hobart accepts that “current economics do not support current capex,” but argues cheap electricity is broadly useful, particularly if manufacturing returns to a high-labor-cost United States. His Y2K analogy: the dot-com boom funded waste, yet also financed the unglamorous software replacement that helped keep systems running after January 1, 2000.
9. AI rewards flexible judgment, not youth by default
Existing utility is already visible. Hobart gave Deep Research questions he had previously investigated and received largely the same sources and summaries in 10 minutes rather than many hours. That outsources the initial screen—whether something merits deeper human work—and expands the range intelligence can examine.
Walker keeps asking people he discusses finance and AI with how a 45-year-old portfolio manager differs from a 25-year-old analyst, without yet receiving satisfying answers. He expects clarity when the current cohort of college seniors reaches a second year on the desk or a first year at a private-equity firm after treating LLMs as naturally as earlier cohorts treated Google.
Hobart complicates the age stereotype. New entrants can harden instantly around one method—such as defining value as six times earnings—while veterans who survived multiple regimes learned to replace frameworks. A 23-year-old may insist on reading every transcript; a 52-year-old already comfortable delegating may summarize first.
The extreme shortcut is Druckenmiller reportedly asking ChatGPT for the five largest U.S.-traded Argentine ADRs and buying all five without further research. Hobart’s own libertarian enthusiasm trade used an Argentine ETF, but shallow work produced shallow conviction: “It was quick. It was lazy,” and he exited early.
10. Newspapers show how structural decline becomes cyclical collapse
N+1’s chapter introductions mark the crisis through newspaper layoffs and closures, with Michael Jackson’s death as the only celebrity landmark Walker recalls. Walker finds the choice jarring only 17 years later: institutions once central enough to mark social and economic deterioration are now largely absent from daily life.
Hobart reconstructs the economics. From the 1950s through the 1980s, two-paper towns consolidated into one-paper monopolies, especially in classifieds; cable television and AM talk radio weakened local advertising power before Craigslist separated classifieds into businesses resembling Zillow, eBay, and Backpage.
Newspapers protected near-term cash flow by serving older, higher-spending readers rather than cultivating younger ones. That made the audience “literally dying”; when durable-goods advertising collapsed in the recession, the structural decline became terminal. Hobart’s grim metric: the more valuable the obituary section, the more subscribers the product is documenting on their way out.