The investment accelerator is a model in which changes in expected output alter the desired capital stock and therefore investment spending.
The investment accelerator is a model in which changes in expected output or sales alter the desired capital stock and therefore investment spending. If firms need a roughly stable amount of capital to support each unit of output, even a modest increase in expected output can require a large addition to capital.
The accelerator is a framework, not a mechanical law. Existing capacity, financing, uncertainty, technology, adjustment costs, and delivery lags determine whether and when firms actually invest.
The desired capital stock is often written as:
where K_t^* is desired capital, Y_t^e is expected output, and v is a capital-output coefficient. If firms adjust immediately:
Substituting the desired-capital equation shows why a change in expected output induces investment. In the most restrictive version, v is fixed and the existing stock begins at the previous desired level.
Real capital cannot usually be adjusted instantly. A partial-adjustment model is:
where lambda is the fraction of the desired gap closed in the current period. Gross investment adds replacement for depreciation:
The adjustment rate can reflect project lead times, installation capacity, financing, permits, irreversibility, and uncertainty.
Suppose expected output rises from 100 million to 108 million and the desired capital-output ratio is 2. Existing capital is 200 million, so desired capital rises to 216 million.
If the firm closes half the gap this period:
If depreciation and retirement require replacement equal to 5% of existing capital, replacement investment is 10 million. Gross investment is therefore 18 million.
Output expectations increased by 8%, while gross spending moved sharply because investment is the flow used to change and maintain a much larger capital stock. If expected output later stops rising, expansion investment can slow even though output remains high.
Capital stock is large relative to annual investment. A small change in the desired stock can therefore be large relative to the normal investment flow. This is the accelerator mechanism.
Consider an economy needing 3 units of capital per unit of annual output. An expected output increase of 2 units raises desired capital by 6 units. If normal replacement investment was 4 units, the additional 6 can produce a large percentage increase in total investment even though output changes modestly.
The mechanism also works in reverse. Slower expected growth can sharply reduce expansion investment without requiring the desired capital stock to fall. An outright expected-output decline can create excess capacity and negative net investment through retirements or disposals.
| Concept | Main channel | Key distinction |
|---|---|---|
| Investment accelerator | Expected output changes desired capital | Demand and capacity mechanism |
| Induced investment | Investment responds to output or income | Broader category represented by accelerator models |
| Investment multiplier | Initial autonomous spending changes equilibrium income | Spending propagation rather than capital adjustment |
| Financial accelerator | Balance sheets and credit spreads amplify shocks | Financing-friction mechanism |
| Capital deepening | Capital services rise relative to labor | Productivity/input-ratio outcome |
Confusing the accelerator and multiplier is common. The accelerator asks how output changes investment; the multiplier asks how an autonomous spending change can affect output. Some cycle models combine both feedback directions.
Unused capacity: Firms can raise production before adding capital.
Temporary demand: Management may not invest if the sales increase is unlikely to persist.
Adjustment costs: Construction, installation, training, and disruption make rapid capital changes expensive.
Financing constraints: A positive capital gap does not ensure funding is available at an acceptable cost.
Uncertainty and irreversibility: Waiting can preserve the option to avoid a costly mistake.
Technology and asset prices: New technology can change the capital-output ratio, while capital-goods inflation changes nominal spending.
Global sourcing: Imported equipment can satisfy domestic capital demand while affecting domestic production and trade measures differently.
Business-cycle analysis: The accelerator helps explain why business investment can amplify changes in expected demand.
Forecasting: An explicit desired-capital equation exposes assumptions about output, utilization, adjustment speed, and replacement. It is more informative than applying a constant investment-growth rate.
Credit and valuation: Expansion can increase leverage and execution risk before revenue arrives. A slowdown in capital spending can reflect weaker demand, adequate capacity, or completion of a project cycle rather than distress alone.
Policy analysis: Fiscal or monetary changes may affect investment through demand and financing channels. The model does not guarantee a particular response or justify a policy without empirical evidence.
The investment accelerator is an educational model. It does not forecast a specific company, project, business cycle, security, or policy outcome.