Growth-and-Crash Stock-Flow Model¶

Model — instantiates Overshoot-Crash Load Management

Ties growth, peak, crash-conversion, clearance, and recovery delay into one causal stock-and-flow model, so the size of the coming crash load can be read off the size of the stock.

Most models of a booming system stop at the peak — they tell you how big the stock gets and when growth stalls. Growth-and-Crash Stock-Flow Model keeps running past the peak. It represents the whole causal loop as stocks and flows — the reinforcing inflow that drives growth, the stock that accumulates, the support capacity that sustains it, and then the part every naïve growth model omits: an explicit crash-conversion link that turns accumulated stock into a second load as it unwinds, plus the clearance rate and recovery delay that decide how long that second load lingers. Its defining move is refusing to treat decline as an automatic return to normal: crash load is modeled as a function of peak stock, so a bigger boom mechanically implies a bigger bust. That one structural choice is what separates it from every monitor and controller in the archetype — it is the map the others act on, not an action itself.

Example¶

A lake receives steady nutrient runoff from upstream farms. The model lays it out as stocks and flows: nutrient inflow drives a reinforcing bloom of algal biomass; biomass accumulates well past the point where the water's assimilation capacity can keep up. Then the crash-conversion link does the work no growth curve alone would show — when the bloom dies, decomposition consumes dissolved oxygen in proportion to the peak biomass, and the oxygen stock (the secondary resource) crashes toward the level where fish suffocate. Running the structure forward reveals the trap: cutting nutrient inflow after the peak barely helps, because the biomass is already there and will convert to oxygen demand regardless.

The output isn't a verdict but a trajectory with a crash-load curve attached: "at the current inflow, peak biomass hits a level whose decomposition pulls dissolved oxygen to ≈2 mg/L for roughly three weeks — well into fish-kill territory — and the oxygen tail lags the biomass peak by about a fortnight." That linkage between peak stock and delayed secondary load is exactly what the appraisal needs before it can decide how early to intervene.

How it works¶

The model is built as coupled stocks and flows, in the tradition of system dynamics:

Growth structure — the reinforcing inflow and the stock it feeds, with the benefits that keep the inflow switched on.
Support capacity — the assimilation or carrying capacity that absorbs normal output, modeled as dynamic and uncertain rather than a fixed ceiling.
Crash-conversion link — the distinctive element: a transfer function mapping peak (or unwinding) stock to secondary load, with a delay, so decline is priced as a burden and not a relief.
Clearance and recovery — the rate at which the secondary load is processed and the lag before the secondary resource recovers.

The structure is then run forward to read off peak stock → crash load → secondary-resource trough → recovery time as one linked story.

Tuning parameters¶

Reinforcing-loop gain — how aggressively growth compounds. Higher gain shortens the runway to the peak and shrinks the window in which any intervention is still cheap.
Crash-conversion fraction and delay — what share of the stock becomes secondary load, and how lagged. This is the model's most consequential and least observable dial; small changes swing the whole bust.
Clearance rate — how fast the secondary load is processed. Raising it flattens the tail; it is the lever the clearance-side mechanisms pull.
Capacity margin — how far below the estimated ceiling the model treats as "safe." A thin margin flatters the growth phase and hides how close the crash already is.
Resolution — a single lumped stock or a spatial / cohort-disaggregated one; finer resolution surfaces local hotspots but multiplies uncertain parameters.

When it helps, and when it misleads¶

Its strength is making the crash visible before it happens and exposing the archetype's central illusion — that removing the input after the peak makes the system safe. By pricing decline as a load, it justifies acting before the peak rather than after.

Its failure modes are the failure modes of any model, sharpened by the softness of the crash link. All models are wrong;^[1] here the crash-conversion fraction and delay are the parameters worst constrained by data, so a confident-looking oxygen curve can rest on a guessed coefficient. The tidy trajectory invites false precision, and the structure is easily run backwards — tuned until it says the current boom is fine, to license continued growth. The discipline that keeps it honest is to validate the crash-conversion link against real past unwinds, carry parameter uncertainty through as a band rather than a line, and treat the run as a structured argument whose assumptions must hold.

How it implements the components¶

Growth-and-Crash Stock-Flow Model realizes the archetype's representation layer — the causal picture the rest of the machinery reads from:

growth_stock_and_driver_map — the stock, its reinforcing inflows, and the benefits sustaining them, drawn as an explicit causal structure that separates stock from flow.
carrying_and_assimilation_capacity_model — the support and assimilation capacity, represented as a dynamic, uncertain quantity rather than a fixed target.
crash_load_conversion_model — the archetype's distinctive component: the transfer function linking peak stock to delayed secondary load.

It does not detect anything in real time — the leading-signal and level signals belong to Early Warning Indicator and Saturation Dashboard — and it takes no action: the paced unwind is Controlled Drawdown Schedule, and the protected secondary floor is Secondary-Capacity Reserve Activation.

Instantiates: Overshoot-Crash Load Management — this model is the causal substrate the whole appraisal reasons over.
Sibling mechanisms: Early Warning Indicator · Controlled Drawdown Schedule · Sink Capacity Audit · Cohort Staggering · Clearance Pathway Enhancement · Saturation Dashboard · Threshold-Triggered Input Cap · Source Reduction Program · Secondary-Capacity Reserve Activation · Staged Harvesting or Decommissioning · Hotspot Containment and Removal · Post-Crash Residual-Load Dashboard · Reentry Gate Review

Notes¶

The model's entire value rides on the crash-conversion link. Strip that out and it degrades into an ordinary growth model that says "approaching capacity, slow down" — and misses the die-off entirely, which is the one thing this archetype exists to catch. When reviewing one of these models, audit the crash link first; the growth curve is the easy part.

References¶

[1] "All models are wrong, but some are useful" — George Box's caution that a model earns trust through the decisions it improves, not through fidelity. It applies with force here because the crash-conversion coefficient is typically the least-observed part of the structure and the part the bust depends on most. ↩