we-solve (beta)

Econ-576 · economics · Economics

Quantitative tightening transmission

LLM status Open unverified

Statement

Original subject group: Inflation and monetary policy

Audit classification: Published research agenda. Normalisation is an agenda with existing firm-response evidence; conditional QE/QT symmetry is editorial. Verification concerns the cited version, not first priority or proof that this exact editorial specification remains unsolved on October 8, 2026.

Definitions and precise research target

Editorial mathematical specification.
Let $b_t$ be an unanticipated central-bank purchase of duration-adjusted government securities, with purchases positive and sales or runoff surprises negative. Fix a maturity segment and define predetermined state $s_t$ to include reserve abundance, intermediary balance-sheet capacity, and market stress. Let $F_{j,t+h}(b)$ be firm $j$'s potential financing cost, debt maturity, or investment at horizon $h$, with other policy surprises held fixed. For $a>0$, define \[ A_h(a,s)=E[F_{j,t+h}(a)-F_{j,t+h}(0)\mid s_t=s] +E[F_{j,t+h}(-a)-F_{j,t+h}(0)\mid s_t=s]. \] Quantitative easing and tightening are symmetric at this scale and state if $A_h(a,s)=0$. Identify the function $A_h$, with confidence regions, from announcement surprises that separate duration supply, anticipated runoff, and policy-rate information. Match or reweight the support of easing and tightening episodes; if states do not overlap, report bounds or model-based extrapolation rather than an unconditional symmetry conclusion. Distinguish financing responses from real activity and report heterogeneity by predetermined credit quality. The target is scale- and state-specific symmetry, including possible liquidity thresholds, not a claim that every purchase and sale should mechanically have equal and opposite effects.

Let duration-adjusted purchases be the sum of changes in security holdings times a fixed duration weight; define the unit and maturity bin before analysis. Runoff is a reduction in holdings, but anticipated runoff is not itself an innovation. Define the potential outcomes through a joint intervention on balance-sheet purchases and the policy-rate surprise, with the latter fixed at zero. The asymmetry function is the sum of two differences in the same financing-outcome units and state, so zero tests oddness around the zero-intervention baseline; it does not test equality across historical QE and QT regimes. Identification requires conditional shock exogeneity and support at $-a,0,a$ or an explicitly stated structural extrapolation. Conditioning on reserve abundance must use its preannouncement value.

Variants and assumption changes

The following are editorial variants, not additional published open conjectures.
  1. Liquidity threshold. Define a predetermined reserve-to-bank-assets ratio $R$ and estimate the symmetry function separately above and below a fixed scarcity threshold $R_0$. The threshold itself is a separate estimand if not fixed in advance.
  2. Maturity and credit quality. Keep total duration supplied constant while varying the purchased maturity segment. Compare high- and low-credit-quality firms with common policy-shock exposure; financing and investment effects remain separate outcomes.

References and source audit

Normalisation is an agenda with existing firm-response evidence; conditional QE/QT symmetry is editorial.
  1. European Central Bank. Undated / date unverified. Monetary policy, strategy and implementation. Policy normalisation, selected areas of focus. https://www.ecb.europa.eu/pub/economic-research/research_agenda/monetary_policy/html/index.en.html.
  2. Egemen Eren; Denis Gorea; Daojing Zhai. 2025. How do quantitative easing and tightening affect firms?. BIS WP1286 primary metadata, Focus/ Contribution/ Findings/ Abstract; September2,2025. https://www.bis.org/publications/working-paper-1286-how-do-quantitative-easing-and-tightening-affect-firms.

Common conventions: Source mathematical conventions

These conventions apply to each entry unless explicitly replaced.

Models and identification. A primitive model $m$ specifies all probability laws, feasible choices, information, equilibrium and measurement rules used by its target. The admissible class $\mathcal M$ is fixed independently of outcomes. If $P_O(m)$ is its observable probability law and $\tau(m)$ the target, its identified set at observable law $P$ is \[ \mathcal I_\tau(P)=\{\tau(m):m\in\mathcal M,\ P_O(m)=P\}. \] Point identification means that this set is a singleton. An empty set signals incompatibility of data and assumptions. Coordinate bounds need not describe the joint set. Bounds are sharp only if every retained target is attainable or arbitrarily approachable by compatible models. Finite bounds require explicit restrictions on unobserved quantities; unrestricted confounding, hidden wealth, extrapolation or arbitrary equilibrium selection can yield uninformative sets.

Causal interventions. A potential outcome $Y_i(p)$ is the outcome under a complete policy regime $p$, including timing, financing, information and market spillovers. The target population and units are fixed before treatment. With interference, use the complete assignment vector or a justified exposure mapping; individual no-interference notation alone cannot identify an economy-wide intervention. A difference of mean potential outcomes does not require the cross-world joint distribution. Conditioning on post-treatment migration, survival, enrollment or employment changes the estimand and needs separate justification.

Statistical statements. An estimator, confidence set, forecast or test specifies a sampling law, model class and loss/coverage criterion. Uniform $(1-\alpha)$-coverage means $\inf_{m\in\mathcal M}\Pr_m\{\tau(m)\in C_n\}\geq1-\alpha$, or an explicitly asymptotic version. The whole parameter space trivially has coverage, so informativeness requires a stated width, risk or power objective. A forecasting improvement is relative to a named benchmark, information set and evaluation population.

Equilibrium and optimization. An equilibrium comprises feasible plans, optimality, beliefs where needed, budget constraints and all market/resource clearing. The primitives or a stated benchmark define each correspondence $\mathcal E(p;m)$. Nonemptiness is assumed or proved, never inferred just from compact policy sets. A selected-equilibrium target uses a declared selection rule; a robust target quantifies over all permitted equilibria. Use suprema or infima unless attainment is established. An $\varepsilon$-optimal policy is feasible and within $\varepsilon$ of the finite optimum value. Report an empty feasible set separately.

Welfare and accounting. Normative utility, interpersonal normalization, social weights, numeraire, discounting and the use of public receipts are primitives. Behavioral utility need not coincide with the welfare criterion. Household welfare includes ownership income and rents once; adding profits again to compensating variation generally double-counts them. A monetary equivalent is defined by an explicit transfer and price convention, with monotonicity and range conditions. Average effects, mechanism contributions, transfers and welfare losses are different targets. Infinite sums require convergence and dynamic feasible plans require terminal or no-Ponzi conditions.

Source versions. A working-paper date, revision date and journal publication year are distinguished. A DOI for a journal version is not automatically the DOI of an earlier manuscript. Locators refer to the stated version and printed pages where specified. A metadata-only check does not verify a claim about a full-text conclusion. Every entry's source subsection narrows the provenance claim accordingly.

Statement licence: CC-BY-4.0 · Source file

Contributions 1

Econ-576-A1 · Exploration · LLM-generated · gpt 6 astra ultra ·

GPT 6 Astra Ultra attempt on Econ-576

QE--QT symmetry: finite-shock curvature, reserve kinks, and sharp support bounds

Target and interpretation

Fix a maturity segment, duration unit, firm population, and policy-rate surprise of zero. Let $f(b,s)$ be expected financing cost under an unanticipated duration purchase $b$ in predetermined state $s$. The target is \[ A(a,s)=f(a,s)+f(-a,s)-2f(0,s),\qquad a>0. \] Historical purchases and runoff episodes do not identify this same-state finite-shock comparison without common support or extrapolation restrictions. Eren, Gorea, and Zhai [bis] provide existing firm-level evidence. We derive a curvature characterization, an exact reserve-threshold example, and attainable bounds when support is missing. Financing and investment remain distinct outcomes throughout.

What symmetry actually restricts

Theorem (Triangularly weighted curvature): For a fixed state, suppose $f(\cdot,s)$ is twice continuously differentiable on $[-a,a]$. Then \[ A(a,s)=\int_{-a}^{a}(a-|u|)\,f_{bb}(u,s)\,du. \tag{1} \] Consequently $|f_{bb}|\le K$ implies $|A(a,s)|\le Ka^2$, and convexity implies $A(a,s)\ge0$. Symmetry for every $a$ in an interval is equivalent to $f(b,s)-f(0,s)$ being odd there. Vanishing second derivative at zero alone is insufficient.
Proof: Taylor's formula with integral remainder gives $f(a)=f(0)+af_b(0)+\int_0^a(a-u)f_{bb}(u)du$ and $f(-a)=f(0)-af_b(0)+\int_{-a}^0(a+u)f_{bb}(u)du$. Adding cancels the linear terms and proves (1). The triangular weight is nonnegative and integrates to $a^2$, proving both bounds. The identity $f(a)+f(-a)=2f(0)$ for all $a$ is exactly oddness after centering. Finally $f(b)=f(0)+\ell b+b^4$ has $f_{bb}(0)=0$ but $A(a)=2a^4>0$. ∎
A linear small-shock regression identifies a slope, which cancels from this estimand. A finite-shock asymmetry test must therefore measure curvature or a threshold crossing rather than compare the signs of two slopes estimated in different historical states. Zero asymmetry also permits nonlinear odd responses, such as a cubic term; it does not imply a globally linear transmission mechanism.

A liquidity threshold produces an exact asymmetry wedge

Let $R$ be preannouncement reserves in fixed units and $R_0$ a scarcity threshold. Suppose the selected local financing equilibrium is summarized by \[ f(b,R)=f(0,R)-\kappa b+\lambda\{(R_0-R-b)_+-(R_0-R)_+\}, \quad\kappa\ge0,\quad\lambda>0. \tag{2} \] One duration-adjusted purchase unit raises reserves by one unit in this explicit benchmark. The linear term captures the duration channel and the hinge a scarcity premium. The permitted shock range is assumed physically feasible; (2) is a response specification, not a complete banking equilibrium derived from balance sheets.
Proposition (Distance to the reserve threshold): For every feasible $a>0$, \[ A(a,R)=\lambda\left(a-|R-R_0|\right)_+. \tag{3} \] Thus equal opposite shocks are symmetric whenever neither crosses the threshold, while sufficiently large shocks have strictly positive asymmetry. Observing the whole function $a\mapsto A(a,R)$ identifies $\lambda$ and $|R-R_0|$ if the available shock support extends beyond the kink, but does not by itself identify on which side of the threshold the state lies.
Proof: The linear terms in (2) cancel. Set $d=R_0-R$. The remaining expression is $\lambda\{(d-a)_++(d+a)_+-2d_+\}$. For $d\ge0$, it is zero if $a\le d$ and $\lambda(a-d)$ otherwise. For $d<0$, it is zero if $a\le-d$ and $\lambda(a+d)$ otherwise. These cases give (3). Above the breakpoint the slope is $\lambda$, and the point at which it first becomes positive is $|d|$. The formula is unchanged under $d\mapsto-d$, proving the sign ambiguity. ∎
A threshold classification using post-purchase reserves would not represent the same estimand: the intervention itself changes that classification. An estimated scarcity boundary also needs its own uncertainty allowance. Formula (3) treats it as fixed before estimation.

Sharp extrapolation bounds across states

For the next result, impose no cross-shock shape restriction like (2). Let states form a metric space with distance $d$. For each of the three shocks $b\in\{-a,0,a\}$, suppose exact conditional causal mean values $v_{bj}=f(b,s_{bj})$ are identified on a finite nonempty anchor set. Their identification requires conditional exogeneity and support at those anchors. Assume each function $f(b,\cdot)$ is $L$-Lipschitz, with known $L\ge0$, and the observed anchor values satisfy that restriction pairwise.

For a target state $s_*$ define \[ l_b=\max_j\{v_{bj}-L d(s_*,s_{bj})\},\qquad u_b=\min_j\{v_{bj}+L d(s_*,s_{bj})\}. \]

Theorem (Attainable asymmetry interval): Under precisely these restrictions, the identified set for $A(a,s_*)$ is the entire interval \[ [l_a+l_{-a}-2u_0,\ u_a+u_{-a}-2l_0]. \tag{4} \] Both endpoints and every intermediate value are attained by compatible Lipschitz functions. The interval need not collapse even when the anchor means are known without sampling error.
Proof: Every admissible function satisfies the two Lipschitz inequalities with every anchor, so $f(b,s_*)\in[l_b,u_b]$. Pairwise consistency of the anchors implies $l_b\le u_b$: for any anchors $j,k$, their difference is bounded by $L d(s_{bj},s_{bk})\le Ld(s_{bj},s_*)+Ld(s_*,s_{bk})$.

Choose any $t_b$ in this interval. Appending $(s_*,t_b)$ preserves pairwise Lipschitz consistency. On the whole state space, the function $F_b(s)=\min_j\{w_{bj}+L d(s,t_{bj})\}$, where the augmented anchors are $(t_{bj},w_{bj})$, is $L$-Lipschitz and agrees with every augmented anchor. The Lipschitz property follows from the triangle inequality and taking the minimum in both directions; agreement follows from pairwise consistency and the term belonging to the anchor itself. Thus every $t_b$ is attainable. The three functions are independently restricted, so the linear image $t_a+t_{-a}-2t_0$ of their product interval is exactly (4). Its endpoints are obtained by the indicated endpoint choices, and continuity gives all intermediate values. ∎

For illustration, suppose $f(a,s_1)=-1$, $f(-a,s_2)=2$, and $f(0,s_*)=0$ are known, both off-target distances are one, and $L=1$. Then $f(a,s_*)\in[-2,0]$, $f(-a,s_*)\in[1,3]$, and $A(a,s_*)\in[-1,3]$. The historical contrast of one does not identify even the target asymmetry's sign. No finite bound follows from this argument if $L$ is unrestricted.

With simultaneous confidence intervals for the anchors, minimize and maximize (4) over all interval-consistent Lipschitz anchor vectors. On the event that every true anchor lies in its interval, the resulting bounds contain the true target. This gives a coverage argument; it does not make the extrapolation constant empirically known.

Remaining gap

The finite-shock identity and threshold example explain how asymmetry can arise within one predetermined state. The support theorem precisely separates data information from an imposed extrapolation modulus. None establishes exogeneity of observed balance-sheet announcements, holds policy-rate news fixed by itself, or estimates duration normalization and firm heterogeneity. General equilibrium reserve demand, intermediary constraints, and the conversion from financing costs to real investment remain outside the response specification. The full empirical symmetry function therefore remains unresolved, despite the exact conditional results.

Completion Estimate

42%

This is a post hoc, heuristic estimate for the full catalogue target. The attempt characterizes finite-shock asymmetry through curvature, solves a reserve-threshold response benchmark, and proves sharp bounds for missing state support under Lipschitz restrictions. Identifying the empirical symmetry function still requires exogenous duration surprises separated from rate news, credible extrapolation, firm heterogeneity, and equilibrium transmission to financing and investment.

References

Go to contribution →