The Portable MBA in Finance and Accounting, 3rd Edition

Forecasts and Budgets 193

Notice how the estimated trend line reacts to changes in quarterly sales.
This technique weights recent observations more heavily than those in the dis-
tant past. The result is a trend line whose slope changes over time to ref lect
changes in sales growth. Our projections for the next 12 quarters, then, begin
from the last estimate of the underlying trend and at the most recent slope in-
dicated by the data.
Many other statistical techniques can also be brought to bear on this
problem. These provide an objective estimate of future sales from the data it-
self. Their advantage is that they are not prone to biases from wishful thinking
or undue pessimism. Their drawback is that they cannot take into account all
of the variables witnessed by our sales personnel and therefore, do not have as
much of a “feel” for the market. Companies must utilize a variety of inputs
into the projection process, and they derive some level of comfort when several
different approaches yield similar results.
Projection is a critical part of the budgeting process. It follows from our
SWOT analysis and the resulting strategic and tactical plan. Once these are
formulated, sales projections and the subsequent budgeting process outlined
above provide an evaluation of the effectiveness of the business plan.

FIXED VERSUS FLEXIBLE BUDGETS

Many organizations operate in an environment where they can predict with
great accuracy the volume of business they will experience during the upcom-
ing budget period. In such cases, budgets prepared for a single level of activity
typically are very useful in planning and controlling business activities. Bud-
gets prepared for a single level of activity are called fixed budgets.

EXHIBIT 6.8 Double exponential smoothing of Kellogg company’s
quarterly sales (1990 –2000).

0 10 20 30 40 50

Quarters

Gamma (trend):

Alpha (level):

Smoothing Constants

Double Exponential Smoothing

Sales ($)

2,150,000

1,900,000

1,650,000

1,400,000

Actual

Predicted

Forecasts

MAPE:

MAD:

MSD:

5

79,481

0.200

9.08E+09

0.200