Every number on this page is computed by the Chapter 5 notebook's own code. Your notes stay on this page until you leave it; copy the memo before you go.
Before choosing among its three options, Meridian's board has asked the strategy office for a one-page “state of the company”: where Meridian stands, which walls—covenants, liquidity floors, risk limits—are closest, how much room each option leaves, and where each would settle. You are in the strategy office.
Use AXIOM to declare Meridian's state, map its feasible region, follow each option's path against the walls, and find where the turnaround settles and how fast.
Table 5.2 declares the eight readings with their units, types, measuring offices and error bands; Table 5.4 lists the walls.
In the state constructor, review Meridian's declaration (Table 5.2). Switch the attrition rate from parameter to state variable and rerun the digital program to see what the switch changes.
Table 5.1 classes the attrition rate as a parameter; Table 4.3 declares its range, 6 to 11 percent a year.
In the constraint sculptor, add the constraints of Table 5.4 one at a time and watch the feasible region shrink. Probe today's state and the digital program's low point, and record the binding set and the slack on each wall.
The projection shows liquidity against strategic risk, the two readings with walls the options approach; leverage is held at 3.4 by the Chapter 1 model, so the covenant's slack stays 1.1 turns on every path. Paths are the three options, month by month over five years.
In the trajectory sampler, run the digital program in continuous time, take quarterly readings, and join its two phases. Check that half-yearly readings equal two quarterly steps.
The build runs years 0–2 and the scale-up years 2–5; the continuous path is the Chapter 1 model written as a linear system with an exact flow map (Definition 5.6; Propositions 5.3–5.4).
In the equilibrium bench, run the turnaround's annual map from today and from two other starting states, with the day-one bound overlaid. Then run the start-up map and watch it leave zero for the market wall.
Each reading closes a fixed share of its gap to the turnaround's targets (106, 33, 79, 67) each year: factors 0.638, 0.819, 0.549 and 0.741 (Table 5.3; Worked Example 5.1). The day-one bound: largest first move × 0.819k / (1 − 0.819) (Equation 5.1; Theorem 5.4).
A user base growing 22.5 percent a quarter at first, slowing as it nears the market wall of 100 (Online Appendix 5.B).
In the profile theater, display today's state and the three Year-5 states in parallel coordinates, with the distance readout alongside.
The eight readings on Chapter 3's declared scale (Figure 5.5); distances are Euclidean in that scale (Section 5.8).
Every starting point of the turnaround converges to the same steady state, inside the day-one bound (Theorem 5.4); the binding set is empty today, but the liquidity floor comes close on the digital path; and profiles that look alike on the page can be far apart in the declared distance.
Meridian's binding and near-binding constraints today and along each option's path, the slack at the tightest point, and where and when the turnaround settles.