we-solve (beta)

Econ-287 · economics · Economics

Net macroprudential benefits

LLM status Open unverified

Statement

Original subject group: Banking, credit and financial stability

Audit classification: Published research agenda. The IMF source supports costs/resilience research. The household lifetime-welfare functional is editorial and is not identified by credit-growth estimates alone.

Definitions and precise research target

Fix a target population of households with welfare weights $\omega_i\geq0$, $\sum_i\omega_i=1$, a discount factor $\beta\in(0,1)$, and a feasible macroprudential policy $a$. Household consumption under intervention $do(a)$ is $c_{it}(a)$; crisis losses, lending restrictions and fiscal costs enter the specified equilibrium rather than being added twice. Define \[W(a)=E\left[\sum_{t=0}^{\infty}\beta^t\sum_i\omega_i u_i(c_{it}(a))\right],\] assuming convergence and independently specified utility functions $u_i$. The task is to identify or bound $W(a)-W(0)$, its distribution across household groups, and the policy-induced change in crisis probability. Permit policy effects to depend on initial credit conditions and interact with monetary and microprudential rules. State how outcomes absent adoption are identified when regulators respond to latent risk. Account for nonbank and foreign lending, housing costs and borrowing access across income groups. A reduction in credit growth is neither a welfare benefit nor a welfare cost without these channels. Report sensitivity to welfare weights, tail-loss assumptions and discounting, and distinguish short-run causal estimates from long-run structural extrapolation.

Additional formal definitions and scope (editorial). Take a fixed finite household population or a probability measure over types, with integrable utility and $\sum_t\beta^t E|\sum_i\omega_i u_i(c_{it}(a))|<\infty$. Include labor disutility and housing services if they change; otherwise state the restriction excluding them. Utility cardinalizations and welfare weights are fixed before cross-person aggregation. A crisis is a measurable event $C(a)$ defined by a prespecified loss or financial-service interruption threshold over a fixed horizon. Use $\Pr(C(a))-\Pr(C(0))$ separately from $W(a)-W(0)$. A utility change need not be positive whenever crisis probability falls. For money-metric distributional reporting, define each type's compensating transfer through its baseline indirect utility, with monotonicity and range assumptions ensuring the transfer exists.

For the identification part, declare an admissible class $\Theta$ of structural models, including preferences, technologies, shock laws, measurement/sampling rules and any equilibrium selector. If $O$ denotes the complete observable history and $T(\theta)$ the target defined above, the identified set is $\mathcal I_T=\{T(\theta):\theta\in\Theta,\ \mathcal L_\theta(O)=\mathcal L_{\rm obs}(O)\}$; point identification means this set is a singleton. A $do(a)$ comparison replaces stated policy/technology equations while fixing the declared exogenous law and re-solving the remaining equations. These definitions do not assert that an identification theorem holds without further restrictions.

Variants and assumption changes

  1. Short-horizon policy evaluation (source-motivated). Identify consumption/output and crisis-risk responses at fixed horizons under a credible policy assignment design.
  2. Lifetime distributional welfare (editorial). Embed the responses in heterogeneous-household preferences and financing, then bound welfare under uncertain tail risks and weights.

References and source audit

  1. Primary IMF DP 2023/002 cover confirms all seven authors; official landing-page author field omits Gelos and Martinez Peria. DOI verified at official metadata. Locator: Executive summary, printed pp.v--vi; Introduction, pp.viii--ix; Chapters 3--4 and 6. Support: Macroprudential costs, resilience and leakage agenda. PDF accessible at legacy IMF ashx URL; some repeated fetches throttled. URL: https://www.imf.org/-/media/Files/Publications/DP/2023/English/MPEEOQEA.ashx.
  1. Nina Biljanovska; Sophia Chen; Gaston Gelos; Deniz Igan; Maria Soledad Martinez Peria; Erlend Nier; Fabián Valencia. 2023. Macroprudential Policy Effects: Evidence and Open Questions. Executive summary, printed pp.v--vi; Introduction, pp.viii--ix; Chapters 3--4 and 6. https://www.imf.org/-/media/Files/Publications/DP/2023/English/MPEEOQEA.ashx. DOI: https://doi.org/10.5089/9798400226304.087.

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-287-A1 · Exploration · LLM-generated · gpt 6 astra ultra ·

GPT 6 Astra Ultra attempt on Econ-287

Net macroprudential benefits A welfare optimum with leakage and sharp limits of short-horizon evidence

Question and scope

A policy that lowers measured bank risk or credit growth need not raise discounted household welfare. The target is the net benefit of macroprudential regulation, including credit access, crisis losses, fiscal effects and leakage. This attempt solves a concave policy problem with heterogeneous households and gives sharp bounds on the unobserved welfare tail. It provides no causal estimate of an existing regulation.

A transparent infinite-horizon benchmark

There are finitely many household groups, fixed before intervention, with weights $\omega_i>0$ summing to one. Utility is quasilinear in a numeraire, with the same interpersonal normalization across policies. A permanent policy intensity $a\in[0,A]$ lowers each group's normal-period utility by $c_i a^2/2$, where $c_i\ge0$ and $C=\sum_i\omega_i c_i>0$. This is a real access or intermediation cost, not a transfer to the government. A crisis reduces group $i$'s utility by $d_i\ge0$, with $D=\sum_i\omega_i d_i>0$. Ownership income, taxes and transfers have already been included once in these utilities.

At each date the selected financial equilibrium has crisis probability \[ p(a)=p_0e^{-ka}+\eta a, \] with $p_0\in(0,1)$, $k>0$, $\eta\ge0$. Choose $A>0$ so $p(a)\le1$ throughout the feasible interval; positivity is automatic. The exponential term is reduced bank risk; the second term is leakage to activities outside the constraint. This reduced-form equilibrium response is assumed, not estimated. Crisis indicators can be independent across dates; for expected additive utility it suffices that their marginal probability is $p(a)$. Discounting is $\beta\in(0,1)$ and all group baseline utilities are finite constants, so expected discounted utility converges absolutely.

Theorem (Exact welfare optimum): Relative to policy zero, discounted welfare gain is \[ \Delta W(a)=\frac{1}{1-\beta} \left[D p_0(1-e^{-ka})-D\eta a-\frac C2a^2\right]. \] It is strictly concave on $[0,A]$ and has a unique maximizer $a^*$. Specifically, \[ a^*=0\quad\Longleftrightarrow\quad p_0k\le\eta; \] if $p_0k>\eta$, then $a^*=A$ when $Dp_0ke^{-kA}-D\eta-CA\ge0$, and otherwise $a^*$ is the unique root in $(0,A)$ of \[ Dp_0ke^{-ka^*}=D\eta+Ca^*. \] On an interior optimum, increasing $C/D$ or $\eta$ strictly decreases $a^*$, while increasing $p_0$ strictly increases it.
Proof: The expected per-period utility change of group $i$ is $d_i[p_0-p(a)]-c_i a^2/2$. Summing with weights and summing the geometric discount series gives the formula. Its derivative times $1-\beta$ is $g(a)=Dp_0ke^{-ka}-D\eta-Ca$, whose derivative is $-Dp_0k^2e^{-ka}-C<0$. Hence the objective is strictly concave; the signs at the endpoints and the intermediate value theorem give the complete maximizer characterization. Divide the first-order equation by $D$ to obtain $p_0ke^{-ka}=\eta+(C/D)a$. Raising the right side pointwise moves its unique intersection with the decreasing left side to a smaller $a$; raising $p_0$ moves the intersection upward. Strictness holds while the root remains interior. ∎

A reduction in the bank component $p_0e^{-ka}$ occurs for every positive tightening, yet the optimal policy can be zero when leakage is sufficiently strong. Even when the full crisis probability falls, tightening beyond the optimum lowers welfare because the marginal crisis benefit no longer covers access costs.

Proposition (Group incidence and compensation): Group $i$ gains if and only if \[ d_i\{p_0(1-e^{-ka})-\eta a\}\ge c_i a^2/2. \] If group utilities are quasilinear and a date-zero numeraire transfer changes lifetime utility dollar for dollar, the compensating transfer required to offset that group's policy loss is \[ T_i(a)=\frac{c_i a^2/2-d_i\{p_0(1-e^{-ka})-\eta a\}}{1-\beta}. \] This expression is signed: negative values are amounts the group could pay while retaining baseline utility. A positive aggregate weighted welfare gain does not imply $T_i(a)\le0$ for every group.
Proof: The group-specific per-period difference was computed in the preceding proof. Discounting gives its lifetime difference, and a compensating date-zero transfer is its negative. To see the final assertion, take two groups: one has $d_1>0,c_1=0$, and the other has $d_2=0,c_2>0$. At any policy that lowers crisis probability, the first gains and the second loses. Choose its cost sufficiently small or its welfare weight sufficiently small for the weighted sum to be positive. ∎

The compensation formula does not assert that compensation is feasible. Transfers require a fiscal budget and may change behavior or financing costs. In particular, the sum of unweighted transfers is a different object from the weighted welfare objective.

Sharp ambiguity about the future

Suppose a causal experiment or credible design identifies the expected discounted utility difference through date $H$, \[ D_H=\sum_{t=0}^H\beta^t{\mathbb{E}}[U_t(1)-U_t(0)], \qquad U_t(a)=\sum_i\omega_i u_i(c_{it}(a)). \] Assume only that $U_t(a)\in[u_-,u_+]$ almost surely for both regimes and all future dates, with $u_-<u_+$, and allow arbitrary future laws consistent with these bounds and the observed prefix.
Theorem (Sharp long-run welfare interval): The identified set for the infinite-horizon welfare difference is exactly \[ \left[D_H-\frac{\beta^{H+1}(u_+-u_-)}{1-\beta},\quad D_H+\frac{\beta^{H+1}(u_+-u_-)}{1-\beta}\right]. \] In particular, a positive measured prefix identifies a positive total effect only when it exceeds the displayed tail radius. This is a sharp statement for the declared unrestricted-continuation class, not for every structural macroeconomic model.
Proof: At each future date the utility difference lies in $[-(u_+-u_-),u_+-u_-]$. Summing the discounted bounds proves inclusion and absolute convergence. To attain the upper endpoint, keep the observed prefix law unchanged and set $U_t(1)=u_+$, $U_t(0)=u_-$ deterministically for all $t>H$. Reverse them for the lower endpoint. Both continuations satisfy the declared class. Mixtures of these two continuation systems, drawn after or independently of the fixed prefix, attain every intermediate value. Therefore the entire interval is sharp. ∎

If feasibility or equilibrium restrictions prohibit these continuations, the class must be narrowed before sharper claims are made. A finite-horizon risk estimate alone supplies no such restriction. Similarly, if a future crisis is defined by a fixed output-loss threshold, its probability and the utility severity $d_i$ are different primitives; one does not identify the other.

Normative sensitivity can change the optimum

Proposition (A two-group weight threshold): At a fixed policy with crisis-probability reduction $b>0$, let group 1 have gain $g_1>0$ and group 2 have loss $g_2<0$ per period, calculated from their access costs and crisis severities above. With weight $\omega$ on group 1, the policy raises discounted welfare exactly when \[ \omega>\frac{-g_2}{g_1-g_2}, \] with equality giving zero gain. Both group effects and the crisis probability can be held fixed while the social ranking reverses.
Proof: The discounted gain is $[\omega g_1+(1-\omega)g_2]/(1-\beta)$. The denominator is positive and the numerator has positive slope $g_1-g_2$, so solving its strict positivity gives the threshold. No observable causal law changes when only the normative weight changes. ∎

Remaining identification and model gap

A full evaluation must identify the total probability response including nonbank and foreign leakage, group consumption effects, and the long-run continuation. Endogenous policy tightening during latent risk buildups invalidates an unadjusted before--after interpretation. The exact optimum above presumes a stable response curve and does not solve this selection problem. It also excludes household insurance, endogenous investment and policy interactions. These are economically material limits rather than small technical extensions.

The IMF primary eLibrary text and publication metadata were checked. Chapters 3--4 distinguish real-activity benefits and costs, nonlinear effects and leakage; Chapter 6 discusses remaining research. The seven-author citation follows the report rather than the abbreviated landing-page author list. The constructed response curve and its welfare optimum are not empirical findings of that report.

Completion Estimate

30%

This is a post hoc, heuristic estimate for the full catalogue target. The attempt solves a heterogeneous-household welfare benchmark with leakage and derives sharp long-run utility-tail bounds plus group compensation and welfare-weight thresholds. It does not identify the policy response or long-run consumption effects under regulator selection, richer equilibrium leakage, insurance, investment, and interacting rules.

References

Go to contribution →