What it means
Imagine a chart of monthly sales that rises and falls in a bumpy pattern. Fourier analysis treats that pattern as the sum of several smooth waves, each with its own length, height and timing.
By measuring how strong each wave is, you can see which cycles matter most, such as a 12-month seasonal cycle or a 3-month quarterly effect. The three key ideas are the period, the amplitude and the phase.
The period is how long one cycle takes, the amplitude is the size of the swing above and below the average, and the phase is where in the cycle the series is at the start. Together these describe any regular wave.
In finance, analysts use the method to find seasonal patterns in sales, commodity prices and energy demand, and to filter noise from price series in technical analysis. Some traders look for dominant cycle lengths in prices.
Risk teams and economists also use frequency methods to separate slow trends from fast fluctuations. The technique has limits.
It assumes that the patterns in the data repeat in a stable way, and financial markets often change their behaviour, so a cycle found in the past may vanish. With short or noisy data, it is easy to find apparent cycles that are just random chance, and a model built on them will fail.
A good way for a non-specialist to use the idea is to ask whether the data shows a regular rhythm. Seasonal adjustment, which many statistical agencies apply, relies on this same logic.
If a business sees a strong 12-month cycle, it should plan staffing and cash around it.
In practice
Real-world examples.
Example
A beverage company examines five years of monthly sales and finds a very strong 12-month cycle with peaks in summer. The finance director uses the size of the wave to plan stock purchases and temporary staff. She also arranges a short-term credit line to cover the spring build-up.
Example
An energy trader studies hourly power prices and finds regular daily and weekly waves. He uses them to schedule purchases for the hours when prices are lowest. The result is a lower average cost per unit of power bought.
Example
A financial analyst tests a trading rule built on a supposed 40-day price cycle. After checking the result on new data, she finds that the cycle disappears and that the earlier success was due to chance. She drops the rule from the firm's models.
Formula
Calculation
Monthly value = Average + Amplitude x sin(2 x pi x t / Period)
Suppose a retailer's monthly sales follow a yearly cycle with an average of $100,000 and an amplitude of $20,000, with a period of 12 months. In month 3 the sine of 2 x pi x 3 / 12, which is the sine of 90 degrees, equals 1, so sales are 100,000 + 20,000 x 1 = $120,000.
In month 9 the angle is 270 degrees, the sine is -1 and sales are 100,000 - 20,000 = $80,000. In months 0 and 6 the sine is 0, so sales sit at the average of $100,000. The wave therefore swings between $80,000 and $120,000 each year.Case study
Seen in the real world.
Harvest Table is a fictional chain of restaurants, and its finance team noticed that cash balances swung widely over the year. The analyst, Carlos, applied a frequency analysis to three years of weekly takings.
He found a dominant yearly wave with an amplitude of about $60,000 around an average of $400,000 per week, with peaks near the end of the year and troughs in late winter. A smaller four-week wave was linked to the pay-day cycle of customers.
In this illustrative case the company used the findings to build a cash reserve of $240,000, equal to four weeks of the low-season shortfall, and to schedule maintenance and promotions in the quiet weeks. Carlos warned that the model needed to be refreshed every year to check that the pattern still held.
Watch out
Common mistakes.
- Assuming that a cycle found in past data will continue, when markets often change their behaviour.
- Reading random noise as a meaningful cycle, especially in short data sets.
- Using the method on data with a strong trend without removing the trend first, which distorts the results.
Questions
People also ask.
What is Fourier analysis in simple terms?
It is a way to split a changing series into regular waves of different lengths, so you can see which rhythms matter.
Is it used to predict share prices?
Some traders try, but evidence is mixed because market patterns are unstable, so it is better used to understand seasonality and structure than to forecast prices.
Do I need advanced maths to benefit from it?
Not to use the results, since software does the calculations, but you need to understand the assumptions so that you do not over-interpret the findings.
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