Some items in a model may be believed to vary in direct proportion to other items. That is, the dependent item is to be derived as a multiple of another (independent or precedent) item. For example, quite commonly some items are assumed to be linked to the level of sales revenue, such as:

- The value of invoices owned by customers (at any point in time) is derived as a proportion of the annual sales revenue.
- The cost of materials bought from suppliers varies in proportion to the sales revenue.
- The value of capital equipment needed to produce the goods varies in proportion to sales revenue (or future sales revenue).

These methods are often known as a “ratio-based” (or “multiples”) method. That is, the independent item would be forecast first, with the dependent item calculated from it (by multiplying the value of the independent item by the ratio or multiple). The value of the ratio (or proportion or multiple) that is used in the forecast would typically be derived from historical information (for example, by dividing the historical value of the dependent item by that of the independent item).