What it means
A straight-line trend assumes growth of the same amount every period. Real business data often curves: sales may speed up, plateau and then fall.
A polynomial trend fits a curve to such data using terms like x squared (x multiplied by itself). The order of the polynomial sets how many bends are allowed.
A second-order (quadratic) trend has one bend, a third-order trend can have two, and so on. Spreadsheet programs can fit these curves in a chart with a few clicks and can display the equation.
The temptation is to use a high order because it hugs the past data closely. That usually backfires, since the curve follows random noise and then swings wildly when extended into the future.
Most forecasters stay with second or third order and test the fit on data the model has not seen. A polynomial trend is a description of the past, not an explanation of why things happened.
A curve that fits well is not evidence that the pattern will continue, and extending it far beyond the data is risky. Always sanity-check forecasts against what is physically or commercially possible.
In practice, finance teams use polynomial trends for demand, cost curves and seasonal adjustments. They are quick to build and easy to explain, provided the limits are made clear.
Choosing the order can be done by comparing a few candidate curves. Fit order 1, order 2 and order 3, hold back the last few data points, and see which curve predicts them best.
If a higher order barely improves the result, stay with the simpler curve.
In practice
Real-world examples.
Example
A software company plots monthly sign-ups after a product launch. A quadratic trend line shows growth slowing, which tells the team the launch boost is fading. The team shifts its marketing spend from acquisition to retention.
Example
A manufacturer fits a curve to unit cost against production volume. The curve falls and then rises, suggesting there is an efficient output level beyond which costs climb. Management sets a target output near the low point of the curve.
Example
A retailer fits a cubic trend to five years of sales to separate the long-term path from seasonal swings. The result is used as a baseline for budgeting. Seasonal peaks are then added on top of this baseline.
Formula
Calculation
Quadratic trend: y = a + b x + c x^2
Suppose a product's monthly sales (in dollars) fit the trend y = 100,000 + 4,000 x - 200 x^2, where x is the month number.
At month 5: y = 100,000 + 4,000 x 5 - 200 x 25 = 100,000 + 20,000 - 5,000 = $115,000.
At month 10: y = 100,000 + 4,000 x 10 - 200 x 100 = 100,000 + 40,000 - 20,000 = $120,000.
At month 12: y = 100,000 + 4,000 x 12 - 200 x 144 = 100,000 + 48,000 - 28,800 = $119,200.
Sales peak at month 10, since the peak is at x = 4,000 / (2 x 200) = 10, and then begin to decline. By month 20 the curve gives 100,000 + 80,000 - 80,000 = $100,000, which shows how quickly a quadratic can fall once it turns.Case study
Seen in the real world.
Clearwater Beverages is a fictional drinks company that plots two years of monthly sales of a new sparkling water. A straight-line trend suggests steady growth, but the data points clearly bend. An analyst fits an illustrative quadratic curve and finds that sales are expected to peak around month 10 at roughly $120,000 a month.
The finance director is cautious about extending the curve past month 12 because the product is new and the data set is short. She asks the analyst to compare the curve with market data and customer feedback before using it in the budget.
In the end the company plans for flat sales after the peak and invests in a new flavour to prompt fresh growth. The curve helped flag the slowdown early, giving the company time to respond rather than react.
Watch out
Common mistakes.
- Using a high-order polynomial because it fits the past data perfectly. It will often forecast badly, because it has learned the noise as well as the pattern.
- Extending the curve far into the future. Polynomials can swing to impossible values outside the data range, such as negative sales or costs that fall below zero.
- Assuming a good fit proves the cause. Fitting a curve describes a pattern but does not explain it.
Questions
People also ask.
What does the order of a polynomial mean?
It is the highest power of x in the equation. Order 2 gives a single bend and order 3 gives up to two bends.
When should I use a linear trend instead?
When growth is steady and a straight line fits well. Simpler models are easier to explain to a board and often forecast better.
How can I judge the fit?
Look at R-squared and test the curve on data it has not seen, then compare with a simpler model, and prefer the simpler one when the results are similar.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%Related