Monetary Policy ABM Model
How do interest rate changes propagate through an economy of heterogeneous households and firms with different balance sheets, credit access, and expectations - and why does this transmission break down at the zero lower bound?
Graph workspace
Multi-panel time-series with distributional sidebar
Monetary policy ABMs produce stochastic time series (output, inflation, interest rates) alongside cross-sectional distributions (income distribution, MPC distribution, debt-service ratios). The canonical visualization pairs a time-series panel showing the policy rate and output response with a histogram of consumption responses across the income distribution. This makes the heterogeneous transmission mechanism visible in one view.
Current results
Output Gap
Percentage deviation of aggregate output from potential. Negative after rate hikes, positive after rate cuts - with the magnitude depending on the MPC distribution.
-0.50%
percent
Inflation Rate
Annualized inflation rate. Responds to the output gap with a lag; the Phillips curve slope () determines how tight the link is.
1.96%
percent
Policy Rate
Current central bank rate. Floored at zero (ZLB). When the ZLB binds, this readout stays at 0 even as the Taylor rule prescription goes negative.
2.89%
percent
Quintile Response Ratio
Ratio of the bottom-quintile percentage consumption response to the top-quintile response. This replaces the earlier 'Consumption Gini', which was applied to a SIGNED variable: a Gini is defined for non-negative quantities, and under the tightening preset every is negative, so the statistic is undefined rather than merely large. The ratio handles signed changes cleanly - both quintiles move the same direction, and the ratio measures how much harder the low-wealth, high-DTI end is hit. Each quintile's response is at the lending rate in force before the move, so MPC cancels and so does a common , leaving at the seeded quintile DTIs of 2.4 and 0.8: about 3.27 at the seeded 5% lending rate, for a cut and a hike alike, and higher where lending was already dearer. This page's sketch computes exactly that ratio, with equal to the 3% neutral policy rate plus the Base bank spread setting, so the readout moves with the spread and not with the policy-rate or MPC-dispersion settings; it does not model households crossing into the constrained branch.
--
ratio
Controls
Policy rate: 0.03
The central bank's opening rate, entered as a lasting move away from 3% that then fades.
Higher rate reduces output (output gap goes negative), cools inflation, and increases debt-service burden for borrowers. The shock decays with a half-life of about three quarters, so the rate path shows a visible hump rather than a step. The consumption histogram shifts left, with high-DTI households hit hardest.
Taylor rule inflation weight: 1.5
How aggressively the central bank responds when inflation misses target; here that sets speed and stability.
A low is stable but slow - inflation drifts far from target and returns monotonically. A high returns faster but overshoots. At the seeded Phillips slope () and a gap semi-elasticity near 2.0, the inflation root is : monotone below 1.7, damped oscillation between 1.7 and 3.3, explosive above 3.3. The slider tops out at 3.0, so every setting on it is stable.
How much the spending rate varies with wealth; zero means every household spends the same share.
Higher dispersion widens the LEVEL of consumption across quintiles and the currency size of each quintile's response. Note what it does NOT do: the PERCENTAGE consumption response to a rate change is , in which MPC cancels, so the percentage gradient on the histogram is produced by the DTI schedule and not by this slider. At zero the MPC bars collapse to one height while the percentage-response bars keep their slope.
Base bank spread: 0.02
The banks' opening markup, which then adjusts toward a target margin and absorbs part of policy moves.
Higher spread raises the level of borrowing costs and dampens the output and inflation response. It does not by itself change the pass-through coefficient - that is governed by the bank-specific adjustment speeds .
QE bond purchases (% of GDP): 0
How many bonds the central bank buys, as a share of GDP; it moves long rates and spending.
QE compresses the long rate by about 2.5 bp per percentage point of GDP purchased, raising the 10-year bond price and delivering a wealth gain proportional to each household's bond share (0.3 at the bottom of the wealth distribution, 0.7 at the top). Because is only 0.03 and the asset-rich hold more bonds, the effect on output is small and its incidence is regressive. The bond price is a finite-maturity annuity, so it stays finite and positive as the long rate approaches its 0.5% floor - the perpetuity formula diverges there, which is precisely the ZLB regime this preset is meant to illustrate.