{"id":"2b1e2a67-6bb6-47d6-8367-e9d77867b741","arxiv_id":"2412.07649","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Using Bayesian neural network local projections, the authors find US macro variables respond strongly to adverse financial shocks, weakly to benign shocks, and proportionally to shock size.","lead":"This paper estimates how US inflation, output, and employment respond to financial shocks using Bayesian neural networks inside local projections. It finds that adverse financial shocks hit the economy much harder than benign ones, while small and large shocks appear to generate proportional responses.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The sequential latent-shock estimator in Section 3 is the load-bearing component of the empirical claim, and it has no consistency proof or Monte Carlo validation; a bad-control mechanism can plausibly manufacture sign asymmetry.","rationale":"The reader's weakest_assumption identifies exactly the load-bearing risk: the sequential latent-shock estimator in Section 3 is the only bridge from the BNN regressions to the impulse responses, and it is presented without any formal validation. I reviewed the full text for a more concrete failure mode and found one: the h=1 residual ε_t is itself a function of the treatment ζ_t through the h=0 fitted model, making it a bad control in a nonlinear setting. The paper acknowledges the latent-shock issue in footnote 3 but cites only a forecasting robustness result, which does not address impulse-response bias. The proposed Monte Carlo experiment is a direct falsification check: a symmetric DGP should produce symmetric estimated NLPs if the estimator is sound; if it produces sign asymmetry, the headline result is an artifact of the method. This concern does not change the reader's conditional verdict, but it sharpens the condition under which the paper should be accepted: the authors should supply either a consistency argument or Monte Carlo evidence for the sequential estimator, or reinterpret the empirical results as exploratory. I also noted the absence of formal asymmetry tests and replication materials, but those are secondary to the estimator issue.","tokens_in":5924,"tokens_out":9795,"duration_ms":93055,"concrete_test":"Run a Monte Carlo experiment calibrated to the empirical design (T=732 monthly observations, similar persistence). Generate data from a symmetric linear DGP, for example y_t = A(L)y_{t-1} + B ζ_t + v_t, with the financial shock ζ_t drawn from a symmetric distribution so that the true NLP is symmetric and proportional. Apply the exact sequential BNN-NLP estimator of Section 3 to the simulated data and compute the posterior median of the sign-asymmetry measure, e.g., NLP(h, -1) - |NLP(h, +1)|, for h=1,...,36. If this measure is systematically nonzero, or if the estimated asymmetry is of the magnitude found in Figure 1, in a DGP known to be symmetric, the central empirical claim is not identified. If the estimator recovers symmetry and proportionality, the concern is mitigated.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is produced by the sequential estimator in Section 3, and that estimator is unvalidated. Equation (4) includes the latent vector ε_{t+h} = (ε_t, ..., ε_{t+h-1}) as regressors, and the paper constructs these by saving residuals from the h=0 BNN and plugging posterior draws into subsequent horizons. No consistency theorem, bootstrap, or Monte Carlo evidence is provided; the only supporting citation (footnote 3, Clark et al. 2024) concerns direct forecasts, not impulse-response bias. The problem is more serious than uncertainty propagation. For h=1, ε_t is the fitted residual from a regression that includes ζ_t. In a nonlinear model with shrinkage priors, this residual is not orthogonal to ζ_t in general: part of the contemporaneous effect of ζ_t can remain in ε_t if the first-stage BNN is shrunk or misspecified. The h=1 estimator can then attribute subsequent movements in y_{t+1} to ε_t rather than to ζ_t, attenuating or exaggerating the estimated financial-shock response. Because adverse financial shocks are large and persistent, this bad-control mechanism can plausibly generate the reported sign asymmetry even in a symmetric DGP. For h≥2 the regressors are residuals from previously estimated horizons, so errors accumulate. Throughout, the h≥1 likelihood treats these generated regressors as observed; drawing them from the h=0 posterior propagates first-stage uncertainty but does not correct the errors-in-variables bias. Since every horizon h≥1 in Figure 1 is produced by this pipeline, the sign-asymmetry and proportionality findings are only as reliable as this estimator.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops Bayesian neural network (BNN) nonlinear local projections for structural impulse responses. Equation (4) regresses y_{t+h} on the shock ζ_t, controls x_t, and latent past shocks ε_t,...,ε_{t+h-1}. Because the latent shocks are unobserved, the paper proposes a sequential procedure: estimate h=0, save posterior draws of ε_t, use them as fixed regressors for h=1, save ε_{t+1} draws, use them for h=2, and so on. The method is applied to monthly US data (1960-2020) with the excess bond premium (EBP) from a VAR ordered first as the financial shock. The main empirical finding is that contractionary financial shocks produce much larger declines in inflation, industrial production, and employment than benign shocks, while one- and three-unit contractionary shocks produce proportional responses.","tokens_in":6285,"tokens_out":6195,"duration_ms":52385,"significance":"The paper addresses an important question and proposes a flexible, prior-regularized alternative to parametric nonlinear local projections. A key strength is that the BNN posterior provides a full distribution of impulse responses rather than point estimates. If the sequential estimator is unbiased, the sign-asymmetry result is a useful contribution consistent with prior evidence. However, the central methodological innovation is not validated with either a consistency argument or Monte Carlo evidence, and the headline conclusions are drawn from informal comparisons of 68% credible intervals rather than formal tests. The significance of the empirical contribution therefore cannot be assessed until these gaps are addressed.","major_comments":[{"comment":"The sequential latent-shock estimator is the load-bearing component of the paper, but the manuscript provides no consistency proof, no asymptotic argument, and no Monte Carlo simulation supporting it. In Eq. (4), for h≥1 the regressor vector ε_{t+h}=(ε_t,...,ε_{t+h-1}) is generated from earlier horizons and then treated as observed. The h=0 residuals are obtained from a BNN with horseshoe shrinkage; they are not guaranteed to be independent of ζ_t conditional on x_t, because regularization and potential misspecification can leave part of the contemporaneous effect of ζ_t in ε_t. At h=1, including such a residual as a control can absorb part of the effect of ζ_t on y_{t+1}, biasing ψ_1. The problem compounds for h≥2, because the shocks used as controls are themselves residuals from biased earlier horizons. Footnote 3 cites Clark et al. (2024), but that reference concerns direct forecasts and does not address impulse-response bias from generated regressors in nonlinear local projections. Since every horizon h≥1 in Figure 1 depends on this procedure, the headline sign-asymmetry result could be an artifact of the bad-control mechanism. The authors should provide either a theorem with primitive conditions or, at a minimum, a Monte Carlo exercise using a DGP with symmetric responses to verify that the estimator does not manufacture asymmetry.","section":"Section 3, Eq. (4) and the sequential latent-shock procedure"},{"comment":"The financial shock ζ_t is obtained from a first-stage structural VAR with EBP ordered first. The estimated shocks are then plugged into Eq. (4) as known regressors. The posterior intervals in Figure 1 therefore condition on the first-stage estimates and do not incorporate VAR estimation uncertainty or identification uncertainty. This can overstate precision, and it is especially relevant for the comparison of positive and negative shocks if the first stage is estimated imprecisely. The authors should propagate first-stage uncertainty, for example by drawing the VAR parameters jointly or by using a Bayesian VAR and passing posterior draws of ζ_t into the BNN estimation.","section":"Section 4, shock construction"},{"comment":"The central claims of sign asymmetry and size proportionality are based on visual inspection of posterior medians and 68% intervals. The paper does not report a posterior probability that the response to a negative one-unit shock differs from the negative of the response to a positive one-unit shock, nor a posterior probability that the rescaled three-unit response equals the one-unit response. Overlapping or non-overlapping intervals are not a formal test. Given that the abstract states these asymmetries as the main findings, the authors should provide explicit posterior summaries for the differences (or ratios) of responses across signs and sizes.","section":"Section 4, Figure 1"}],"minor_comments":[{"comment":"Section 5 contains typos: 'applies the method' should be 'apply the method' and 'ley' should be 'key'.","section":"Section 5"},{"comment":"Footnote 3 states 'show that of ϵt+h only has a small impact'; it is missing a word and should likely read 'show that the inclusion of ϵt+h only has a small impact'.","section":"Footnote 3"},{"comment":"The note to Figure 1 is hard to follow because the negative-shock response is plotted after multiplication by -1; please state explicitly that the plotted green line is -NLP(h,-1) and that asymmetry is assessed by comparing NLP(h,+1) with -NLP(h,-1).","section":"Figure 1 note"},{"comment":"Section 4 investigates size proportionality only for contractionary shocks; the conclusion that 'small and large shocks resulting in almost exactly proportional impulse responses' should be qualified to positive shocks or supplemented with evidence for benign shocks.","section":"Section 4, size asymmetries"}],"recommendation":"major_revision","confidential_remarks":"The stress-test concern raised by the second reader is well placed: the sequential estimator is unpublished and unverified, and the empirical result could be an artifact. I recommend major revision with Monte Carlo validation as a condition. The paper is short and could be turned around quickly if the authors provide the missing evidence. I do not see grounds for rejection if the estimator is validated."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nHere's my read of Hauzenberger et al. (arXiv:2412.07649). The headline: the empirical claims are plausible and consistent with prior work, but the sequential latent-shock estimator in Section 3 is the load-bearing part and it is not validated. I think the paper deserves peer review, but only with the expectation of a serious methods check.\n\nWhat's actually new: applying their BNN framework to nonlinear local projections, and a sequential scheme that propagates posterior draws of residuals forward as controls. The sign-asymmetry result repeats Barnichon et al. (2022) and Forni et al. (2024); the size-proportionality finding is the genuinely new empirical result. The paper is clearly written and the BNN machinery is well documented in a companion paper, so the econometrics is not opaque.\n\nThe soft spot is real. Equation (4) includes ε_{t+h} as regressors, and the paper constructs them by saving residuals from the h=0 BNN and plugging them into later horizons. No consistency proof, no Monte Carlo, no bootstrap. The stress-test concern is legitimate: with horseshoe shrinkage, the h=0 residual is not orthogonal to the financial shock ζ_t in finite samples, so for h=1 the estimated response can be partly attributed to the residual rather than ζ_t. Since adverse shocks are large and persistent, this bad-control mechanism could plausibly generate sign asymmetry even in a symmetric DGP. That is not a narrow point about standard errors; it goes to the center of the empirical claim.\n\nI also note the first-stage VAR that produces the financial shock is estimated separately, and its uncertainty is never propagated. That is a lesser but real issue. No code or data are provided, which makes independent verification impossible. The asymmetry inference is visual, from 68% credible intervals, not a formal test. These are not fatal individually, but together they mean the paper's main result rests on an unproven estimator.\n\nFor all that, the paper is not a throwaway. If the estimator is valid, the size-proportionality result is a useful calibration target, and the BNN-LP framework is likely to be adopted by others. I'd send it to a referee who knows local projections and Bayesian nonparametrics, and ask for a Monte Carlo exercise and a consistency argument for the sequential update.\n\nBottom line: worth a serious referee, but I wouldn't bet the farm on the empirical results until the estimator is checked. I'd bring it to a reading group to discuss the bad-control problem.","headline":"Plausible empirical results built on an unvalidated sequential estimator; deserves peer review but with a demand for validation.","tokens_in":6779,"tokens_out":2101,"would_cite":false,"duration_ms":19720,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["62F15","68T07","91B84"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper claims that US financial shocks transmit to the real economy through strong sign asymmetry — adverse shocks matter much more than benign ones of equal size — while responses scale almost proportionally with shock size.","keywords":["Bayesian neural networks","nonlinear local projections","financial shocks","excess bond premium","sign asymmetries","impulse responses","US macroeconomy","shock size proportionality"],"falsifier":"Run the same BNN local projections on simulated data from a known nonlinear model with known true shocks and known asymmetric responses; if the sequential estimator's horizon-h responses deviate systematically from the truth, the empirical conclusions about sign asymmetry would be called into question. Alternatively, re-estimate the responses using an external-instrument financial shock series; if the sign asymmetry disappears, the finding is an artifact of the recursive identification.","tokens_in":5732,"feed_emoji":"📉","tokens_out":5179,"duration_ms":45928,"temperature":0.7,"pith_summary":"This paper develops a way to estimate nonlinear impulse responses to structural shocks by replacing the linear regressions of standard local projections with Bayesian neural networks. Applied to US financial shocks measured through the excess bond premium, the method yields a specific empirical claim: adverse financial shocks push inflation, industrial production, and employment down sharply, while positive shocks of the same size produce little or no response. The paper also claims that the response scales almost linearly with shock size: a three-unit contractionary shock generates about three times the response of a one-unit shock. If true, these results would show that sign asymmetry, rather than size nonlinearity, is the dominant form of nonlinearity in the transmission of financial shocks to the real US economy.","feed_headline":"Bad financial shocks hit the US economy far harder than good ones","feed_subtitle":"Neural-network impulse responses show bad shocks hit US inflation, output, and jobs far harder than good ones; size scales proportionally.","key_machinery":"The machinery is a Bayesian neural network (a flexible regression model in which the conditional mean is a neural network with horseshoe shrinkage priors and a mixture of activation functions) embedded in nonlinear local projections. For each horizon h, the h-period-ahead outcome is regressed on the shock, controls, and a learned nonlinear function f_h; responses are computed as the difference between expected outcomes under shock τ and under zero, averaged over R=400 randomly drawn histories. Because the intermediate shocks ϵ_{t+h} are latent, the paper uses a sequential updating scheme: estimate the h=0 regression, save the posterior of the residuals, treat those draws as fixed regressors when estimating h=1, and continue feeding previous-horizon shock draws forward to h=2, ..., H. The sequence of horizon-specific nonlinear functions is what lets the response curves bend differently for different signs and sizes.","core_discovery":"The central discovery, on the paper's own terms, is that the nonlinear impulse responses of US macroeconomic aggregates to financial shocks are strongly asymmetric with respect to sign but approximately proportional with respect to size. A one-unit contractionary shock to the excess bond premium reduces inflation by around one percentage point at its one-month peak, cuts industrial production growth by almost two percentage points around eleven months out, and lowers employment growth for about two years, whereas a benign one-unit shock is mostly insignificant across all three variables. When the contractionary shock is tripled to three units, the responses line up with the one-unit responses after rescaling by one third, indicating no discernible size asymmetry. These patterns are obtained without imposing a particular parametric nonlinear functional form, since the conditional mean is learned by the network.","pith_inferences":["Editorial inference: the sequential shock-updating step is the fragile link; a Monte Carlo study with a known data-generating process would show whether the h=0 residuals remain valid controls once the nonlinear function f_h changes with the horizon.","Editorial inference: because the shock is identified by ordering the excess bond premium first in a recursive VAR, the sign-asymmetry result is conditional on that identification; an external-instrument or high-frequency proxy robustness check would tell whether the asymmetry is an artifact of the ordering.","Editorial inference: the proportionality finding could be tested at more extreme shock sizes (beyond three units), where financial frictions might make large adverse shocks disproportionately damaging.","Editorial inference: the size-proportionality result suggests a separable structure — nonlinearity in sign but not in scale — that a parsimonious semi-parametric model could reproduce; whether the BNN's flexible fit actually matches such a functional form is testable."],"forward_implications":["Policymakers reading the result as structural would expect contractionary financial shocks to be the dominant risk to prices and employment, with symmetric positive shocks providing little offsetting stimulus.","Linear local projections, by construction, would force positive and negative shocks to have mirror-image effects; the paper's finding implies such models understate the cost of adverse financial conditions.","The proportionality result implies that, within the range studied, scaling a shock is equivalent to scaling the response, so no separate nonlinearity in shock size needs to be built into forecasting or stress-testing models.","The method offers a template for estimating nonlinear impulse responses without choosing a parametric nonlinear functional form in advance."],"supporting_citations":[{"why":"Supplies the Bayesian neural network specification, priors, and MCMC sampler that the paper's local projections build on.","marker":"Hauzenberger et al. (2024)"},{"why":"Introduces local projections, the estimation framework the paper extends to nonlinearity.","marker":"Jordà (2005)"},{"why":"Provides the flexible local projection approach for impulse responses that the paper adapts to BNNs.","marker":"Mumtaz and Piffer (2022)"},{"why":"Defines the excess bond premium used to measure financial shocks.","marker":"Gilchrist and Zakrajšek (2012)"},{"why":"Supplies the FRED-MD data from which the macro variables are drawn.","marker":"McCracken and Ng (2016)"},{"why":"Documents sign-dependent effects of financial market disruptions, the benchmark the paper's results are compared with.","marker":"Barnichon et al. (2022)"},{"why":"Shows how estimated residuals can be treated as fixed regressors in linear local projections, the idea the sequential estimator extends to the nonlinear case.","marker":"Lusompa (2023)"},{"why":"Frames the two-step conditional/unconditional nonlinear local projection construction used in the paper.","marker":"Kilian and Lütkepohl (2017)"}],"fun_headline_variants":["Financial shocks: direction matters, magnitude scales linearly","Bad shocks bite, good shocks barely move US economy","Neural nets show financial shock effects flip with sign","Asymmetric financial shock responses: negative hits, positive misses"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The results stand or fall on the sequential shock-updating estimator: the paper treats the residuals from the zero-horizon network as the true shocks and feeds them forward as fixed regressors at every later horizon, an assumption it asserts but does not prove or validate by simulation.","fun_headline_variants_meta":{"raw":{"variants":["Financial shocks: direction matters, magnitude scales linearly","Bad shocks bite, good shocks barely move US economy","Neural nets show financial shock effects flip with sign","Asymmetric financial shock responses: negative hits, positive misses"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000226,"raw_usage":{"total_tokens":1372,"prompt_tokens":753,"completion_tokens":619,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":369,"completion_tokens_details":{"reasoning_tokens":556}},"tokens_in":369,"tokens_out":619,"duration_ms":6316,"temperature":1.0,"reasoning_tokens":556,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T18:38:47.288922+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the same BNN local projections on simulated data from a known nonlinear model with known true shocks and known asymmetric responses; if the sequential estimator's horizon-h responses deviate systematically from the truth, the empirical conclusions about sign asymmetry would be called into question. Alternatively, re-estimate the responses using an external-instrument financial shock series; if the sign asymmetry disappears, the finding is an artifact of the recursive identification.","supporting_citations":[{"cited_title":"Bayesian neural networks for macroeconomic analysis,","cited_arxiv_id":null,"evidence_quote":"Supplies the Bayesian neural network specification, priors, and MCMC sampler that the paper's local projections build on."},{"cited_title":"Credit spreads and business cycle fluctuations,","cited_arxiv_id":null,"evidence_quote":"Defines the excess bond premium used to measure financial shocks."},{"cited_title":"FRED-MD: A monthly database for macroeconomic research,","cited_arxiv_id":null,"evidence_quote":"Supplies the FRED-MD data from which the macro variables are drawn."},{"cited_title":"Local projections, autocorrelation, and efficiency,","cited_arxiv_id":null,"evidence_quote":"Shows how estimated residuals can be treated as fixed regressors in linear local projections, the idea the sequential estimator extends to the nonlinear case."},{"cited_title":"L ¨utkepohl (2017): Structural vector autoregressive analysis","cited_arxiv_id":null,"evidence_quote":"Frames the two-step conditional/unconditional nonlinear local projection construction used in the paper."}],"review_version":1}