Dynamic Models of Industry Cycles
Objective
Asset investment decisions and other long term strategic choices are more often than not based on "gut feel" instead of deductive reasoning - simply because the latter is not an option when making judgements regarding the future. Executives can improve their strategic sight with scenario-based simulation analyses, in particular where the business in question involves one or more of the following:
- Cyclical supply and/or demand
- Regulation
- Significant delays (e.g. from the inception of a project through regulatory approval to construction and completion)
- Game-changers such as rapid growth or ground-breaking technological achievements.
Developed tool
Applying System Dynamics - a modelling technique based on a research tradition at MIT - our consultants have built a number of dynamic simulation models of industries and markets characterized by expensive assets with steep price curves - from supertankers to inner city condominiums.
Depending on the complexity, such models can be built as an advanced spreadsheet or using bespoke tools such as Vensim or Powersim. Very often, there is significant data processing involved outside the actual simulation model.
Dynamic simulation models are based on an explicit understanding of market drivers and their interdependencies, including feed back loops.
Result
A strategic decision support tool capable of quantifying the impact of the client's strategy under a set of plausible scenarios. In addition, the involvement of the client's own experts builds understanding, confidence and consensus.
- Industry
- Shipping
- Tool type
- Dynamic simulations
- Project duration
- Two or more phases of 4 weeks each
- d2i effort
- 60+ man days
Simple diagram showing (a part of) a generic dynamic model for asset intensive industries (shipping, commercial aircraft, housing). Completing the full diagram in collaboration with client experts (internal and external) is a major early milestone in any dynamic modelling and simulation project.
Dynamic simulation models are always calibrated to fit observed behaviour for all relevant parameters where data exist. The observed behaviours of such endogenous variables (the red dotted lines) are used as "yardsticks" for the simulation (blue lines), but never as direct input.