Research

Schularick and Taylor (2012), American Economic Review

Reproduces. On the current JST vintage, with twelve more years of data than the paper had, lagged real loan growth still predicts banking-crisis onsets: the lag-two coefficient is large and significant, the five-lag block is jointly significant, and the credit model beats the money model on fit and discrimination in both samples.

Baron, Verner and Xiong (2021), Quarterly Journal of Economics

Reproduces, on the authors' own replication panel, and the note says so rather than claiming an out-of-sample test. Bank-equity crashes without panics are followed by real output losses; crashes with panics by substantially larger ones.

Shiller (1981), American Economic Review

The direction survives, the magnitude does not. On the data through the present, price volatility still exceeds the bound implied by realized dividends, but by far less than the factor of five to thirteen the paper reported, and the note shows how much of that depends on the terminal condition.

Reinhart and Rogoff (2010) vs Herndon, Ash and Pollin (2014)

Herndon, Ash and Pollin reproduce; Reinhart and Rogoff do not. On the JST panel over the papers' own window, growth above 90% of GDP is positive under both weighting schemes, growth declines with debt across the whole range, and there is no cliff at 90%.

Cruces and Trebesch (2013), AEJ: Macroeconomics

The headline moment reproduces almost exactly on the original cases. The paper's central price-of-haircuts regressions are untestable on this estate, because it carries no market-reaccess dates and no sovereign spread panel, and the note confines itself to what the file can support.

Borio and Lowe (2002), BIS Working Papers

Half reproduces. On the BIS's own published credit-to-GDP gap scored against the Laeven-Valencia chronology, the gap does lead banking crises, but the low noise-to-signal ratio the paper reported does not reproduce at its own thresholds.

Kaminsky and Reinhart (1999), American Economic Review

The direction reproduces, the coupling does not. On the atlas's annual chronology a banking onset still raises the near-term odds of a currency crisis, and more than the reverse, but only about one banking crisis in five is a twin, and on the authors' own 1970-1995 window the signal nearly vanishes.

Jordà, Schularick and Taylor (2015), Journal of Monetary Economics

Reproduces, as pure counting. On the JST panel, house-price booms with above-median credit growth are followed by a banking crisis far more often than booms without it, while on the default cut booms without the credit sit at or below the unconditional rate, and the credit-fuelled bucket leads at every boom threshold tested.

Fama and French (2015, 2016), Journal of Financial Economics and Review of Financial Studies

Reproduces. Most industry alphas are individually small, the significant ones are mostly negative, and yet the joint test still rejects the model on the balanced panel: industries remain the hard test the authors said they were.

Drehmann and Juselius (2014), International Journal of Forecasting

Does not reproduce, and that is the headline. On the horizons and crises this panel can test, all from one crisis wave, the credit-to-GDP gap beats the debt service ratio at every horizon; only the paper's short-fuse shape, the DSR improving as the crisis nears, survives.

Mian, Sufi and Verner (2017), Quarterly Journal of Economics

Both directions reproduce, as pure counting. On the JST panel, the growth gap that opens between the middle and the top quartile of household debt expansions is far wider than the same gap on business debt, and the quartile view adds a hump shape: heavy deleveraging is also followed by weak growth, while the middle of the distribution grows fastest.

Frankel and Rose (1996), Journal of International Economics

Both tested claims survive, as pure counting. Outside the advanced economies, crash onsets are most frequent after the deepest quartile of prior-year reserve losses and rarest after the fastest reserve growth, in the full panel and inside the paper's own era, and the famous null largely holds: the deepest current-account deficits are not the riskiest bucket, and only the surplus quartile is clearly protected.

Correia, Luck and Verner (2026), Quarterly Journal of Economics

Reproduces out of sample where the data test it hardest, then weakens. A logit on eight public FDIC ratios trained only through 2006 ranks the 2007-2012 failure wave almost perfectly, the median failing bank walks the paper's exact path into failure, and the in-sample fit lands on the paper's own modern-sample number. Trained through 2019 and tested on 2020-2025 the discrimination drops sharply, and Silicon Valley Bank's high model rank comes from size and thin equity, not the historical failing-bank markers.

Reinhart and Rogoff (2013), Journal of Banking and Finance

Reproduces on its own era, then inverts. On the Laeven-Valencia chronology alone, advanced and other economies spend a similar share of years in banking crisis before 2007, but across 2007-2017 the advanced group's crisis-year share and episode-onset rate pull far clear, and the same split holds on the Global Macro Database variant. Pooling all five chronologies, two of which cover only rich countries, manufactures an advanced-economy excess even over a window where the single-source test shows parity.

Jordà, Knoll, Kuvshinov, Schularick and Taylor (2019), Quarterly Journal of Economics

Reproduces, as pure counting. On the JST panel the real return on wealth beats real GDP growth in the large majority of country-decades since 1870, and pooled by decade it wins in every completed decade bar the two of the world wars, with the recomputed gap landing inside the paper's own sanity band. The note measures only that r has exceeded g, not that inequality followed.

Federal Reserve Distributional Financial Accounts, quarterly since 1989

An event study on the estate, in timing language only. The bottom half of US households held the same share of net worth at the 2007:Q4 and 2019:Q4 business-cycle peaks. After the first the share fell hard and took forty-seven quarters to return to where it began; after the second it never fell and kept rising. The note lays the two paths on one clock and compares the recoveries as timing; it does not say why they differ and makes no causal or predictive claim.

SEC Form N-MFP Data Sets (DERA), monthly since 2022

An event study on the estate, in timing language only. US money market funds run government-heavy through the whole Form N-MFP panel, and in the March 2023 report month, the one the Silicon Valley Bank failure falls in, the government share of total fund assets jumped two points in a single month as cash left prime funds and the complex took in more than a third of a trillion dollars. That is the panel's largest genuine one-month move, distinct from a later reclassification step the note holds outside its window. The note reports the timing and nothing more, and makes no causal claim.