November 2020–April 2023: The U.S. Property Boom That Defied Every Prediction

By November 2020, American residential property had entered a phase that left veteran analysts searching for historical analogies. The Federal Reserve held its benchmark rate near zero; mortgage applications were climbing even as pandemic uncertainty lingered. On trading desks tracking housing futures and REIT indices, the conversation had shifted from "is this a rebound?" to "how far does this run?" The material conditions were unambiguous—liquidity was abundant, borrowing costs were historically suppressed, and supply across most metros remained structurally constrained. The question was not whether markets would react, but which geographies would diverge most sharply once the initial surge settled into a more complex pattern.

The causal mechanism was textbook monetary transmission: ultra-low rates compressed mortgage payments, expanding purchasing power for buyers who could access credit while locking out would-be buyers priced out by the arithmetic of affordability. Demand rotated toward mid-sized metros and Sun Belt cities as remote work dissolved the geographic constraints that had historically anchored property values to coastal employment centers. Supply, meanwhile, lagged—a function of construction backlogs, labor shortages, and a decade of underbuilding that preceded the pandemic. The result was a sustained imbalance between buyer interest and available inventory that pushed valuations higher across nearly every monitored metro, though the magnitude of price appreciation varied dramatically by city.

FinFable's scenario data for this window shows the divergence in raw performance. The US 10-City Composite Historical Replay captured a 25.6% gain over the period, but that aggregate figure concealed the spread between Atlanta's 39.1% surge and New York's comparatively restrained 26.1% advance. The Atlanta Historical Replay recorded the highest city-level return in the member set, with a maximum drawdown of only 3.5%—the kind of ratio that makes momentum strategies look trivially easy in hindsight. Yet that same member's data warns against the impression: a 3.2% single-bar move, while modest in isolation, arrived at decision points where conviction was being tested against a market that had already climbed substantially.

Volatility profiles diverged even more sharply than returns. The San Francisco Historical Replay delivered 19.6% total gains but carried a maximum drawdown of 17.2%—the steepest reversal in the member set. The New York Historical Replay, by contrast, wrapped the same broad macro forces in a far tighter band: 2.0% maximum drawdown and a largest single-bar move of just 1.9%. These are not measurement artifacts. They reflect structural differences in local supply elasticity, migration patterns, and the degree to which pandemic-era demand shocks actually penetrated each metro's housing stock. A practitioner replaying these windows side by side confronts the uncomfortable truth that geographic selection is not a neutral variable—it determines the variance profile as much as the direction.

The Boston Historical Replay illustrates how even a market delivering solid nominal returns can punish process failures. Boston's 25.1% close over the window came with a 5.8% maximum drawdown and a 2.9% largest monthly move. In a market where the dominant narrative was persistent appreciation, the episodes of reversal demanded the same disciplined response as any explicit trend-following strategy. The San Diego Historical Replay pushes this further: 36.1% total gain against an 11.4% drawdown—numbers that reward conviction but extract a real psychological cost during the troughs. The core tension replay participants report is not identifying the trend; it is sizing positions to survive the reversals that trend-following generates by definition.

These scenarios collectively reveal why low-volatility trending environments are among the most pedagogically dangerous replay conditions. When the US National (Case-Shiller, 2000+) Historical Replay shows a 29.8% close with a maximum drawdown of just 5.0% and a largest single-bar move of 2.7%, the implied lesson for a decision trainee is dangerously easy to misread: position sizing feels less urgent, stop-loss discipline feels premature, and the cost of chasing feels negligible. The actual lesson is the inverse. The same low-rate conditions that produced this performance also produced the leverage and speculation that made the eventual correction more violent than the sustained calm suggested.

In institutional terms, this window reshaped how mortgage lenders, real estate investment trusts, and policy analysts calibrated risk. The Federal Reserve's own communications had framed near-zero rates as a temporary bridge measure; the sustained duration of that stimulus caught portions of the regulatory apparatus unprepared for the asset price inflation that followed. Rating agencies and systemic risk monitors learned—slowly—that property markets can synchronize across geographies in ways that challenge diversification assumptions baked into portfolio models. The regulatory record shows a gradual revision of stress-test scenarios to incorporate low-rate extensions longer than prior simulations had assumed.

For practitioners training decisions today, this archive window offers three layered lessons. At the market level, it demonstrates that directional bets on any asset class are incomplete without a view on the monetary policy horizon—the direction of the Federal Reserve's next move determined whether low rates were tailwind or headwind for property positions entered in November 2020. At the institutional level, it shows how a macro regime shift redistributes risk across geographies: cities with inelastic supply saw sharper drawdowns when the regime reversed, while more flexible markets absorbed the same rate shock with lower variance. At the decision-training level, the core reconstruction challenge is positional: sizing for the drawdowns that occur, not the drawdowns that feel acceptable in a trending environment.

The data coverage for all member scenarios extends through April 2023; the narrative in this article focuses on the cluster year 2020 and the conditions that defined the initial phase of the boom, while the full replay windows capture subsequent market evolution through early 2023.