Coffee Price Dynamics – Part 1
- Igor Bragato
- 3 hours ago
- 7 min read
Do you want to know how coffee prices really work?
Many coffee professionals spend years in the market without ever developing a deep understanding of how coffee prices actually work. And that’s perfectly fine: you don’t need to be a mathematician to have a successful career in coffee.
However, there is a statistical framework that describes the way coffee price dynamics work: time series decomposition.
This is not your normal casual social media reading. We’re going deep into the statistics that actually move coffee prices. This article is not for the faint of heart; it's only for those interested in a deep understanding of coffee price dynamics. So, if you’re still here, you have the intellectual curiosity and 15-20 minutes to use, grab a coffee, turn off your phone notifications and let’s dive in.
Introduction
Coffee price moves are often linked to narratives. A frost in Brazil, a headline about stocks, a shift in fund positioning, a major currency swing – each price move can be linked to a story. While these narratives matter and it’s important to stay on top of them, prices follow patterns that go beyond these individual events.
That’s because coffee prices evolve as a time series; a sequence where long-term trends, broader cycles, seasonality and short-term noise all intertwine, and this creates complexity: is a 30c price rally a random move, the start of a new trend, a seasonal pattern, or the early phase of a bullish cycle?
[Stay informed as the coffee market evolves. Sign up for a free trial of our Premium Coffee Research today.]
Without time-series context, isolated price movements are hard to interpret and can lead to costly mistakes. In this article, we teach the statistical framework you need to understand and anticipate coffee prices.
We’ll start by defining time-series decomposition, then examine its individual components, and finally show how to apply this framework to the coffee market. Each section includes real examples from the coffee market to demonstrate how this applies in the real world.
Coffee Prices as a Time Series
A time series is a set of observations ordered over time, usually at regular intervals (daily, weekly, monthly), which means that coffee futures prices, export volumes, stocks, or differentials are all time series.
This distinction matters because time series behave differently from other data: time series are path-dependent, which means that they have a “memory”. Yesterday’s price directly influences today’s, and today’s will shape tomorrow’s.
As a result, we can't understand a single data point in isolation. The true story—whether a persistent trend, a predictable cycle, seasonality or just fleeting noise—only reveals itself when we step back and observe the sequence over time.
Treating price movements without this context is like trying to predict a symphony's finale by listening to a single, out-of-context note: you miss the structure, the buildup, and the themes that give the music its meaning.
Disaggregating Coffee Prices
Returning to our original question — is a 30-cent rally random noise, the beginning of a trend, a seasonal move, or the early stage of a bullish cycle? To answer this question, we need to decompose coffee prices.
As a time series, coffee prices can be decomposed into four components, according to classical statistics and econometrics: trend, seasonality, cycle and noise.
Y(t) = T (t) × S(t) × C(t) × R(t)
Where:
T (Trend) = long-term direction
S (Seasonality) = regular, calendar-based patterns
C (Cycle) = medium- to long-term fluctuations without fixed timing
R (Noise) = random, unpredictable external shocks
A simple and practical working rule for coffee markets is to link the time horizon to the type of price behavior observed. Moves over days and weeks are usually dominated by noise, driven by headlines, short-term weather scares, fund flows, and technical triggers.
[Want better context around coffee price moves? Explore our Coffee Research with a free trial.]
Over months, price action more often reflects seasonality or positioning, shaped by harvest pressure, export flows, crop calendars, and fund rebalancing. Over years, prices tend to express true trends and cycles, rooted in structural supply. This hierarchy is not a law, but it aligns well with the logic of time-series decomposition.
In the following sections, we will look at each of these four components individually, examining how they appear in the coffee market and illustrating them with real examples.
Trend: The Structural Price Direction
The trend is a persistent directional movement that survives the removal of short-term noise and seasonality. In practice, a trend exists when price moves in the same direction across multiple time horizons, when pullbacks fail to disrupt the broader direction, and when the move can’t be explained by one single new event or calendar effect.
This means that if the move disappears once we zoom out or de-seasonalize, it’s not a trend. As such, a very reasonable heuristic rule (yet, not a strict law) that we can apply is that a trend usually reflects structural forces (supply balance, acreage, productivity, demand growth, macroeconomics), and those unfold over years - or at the very least, months - not weeks or days.
The clearest recent example of a trend in the coffee market is the bullish move triggered by the major global deficits of the 2021/22 and 2023/24 seasons.
This advance was driven by structural forces and unfolded over multiple years, remaining intact despite short-term noise and seasonal pressures. Note that even though prices experienced corrections — such as during 2022–2023 — these moves did not break the broader upward trend.
Trends can unfold across different time horizons. Structural trends such as the one we just covered are typically multi-year in nature, while shorter-lived trends can last only several months — they commonly emerge during harvest phases, or positioning shifts.
This happened last year, when the Brazilian crop reached the market in April and stocks began to be replenished, triggering a three-month bearish trend. While at its core, this episode was indirectly rooted in seasonality; it had a downward trend as its market expression.
Ultimately, one way to think about a price trend is to ask: “what is driving the market right now, and is that force getting stronger or starting to fade?”
Staying with last year’s example, by September the Brazil crop flow had largely been absorbed and the harvest was over, so it was no longer fresh information and the bearish trend lost momentum. At the same time, US–Brazil tariff risk emerged as a new bullish driver, meaning the original force was fading while a new one was starting to strengthen.
[Follow coffee fundamentals with confidence. Try our Premium Coffee Research free and get timely market insights.]
So, what is driving the market at that point? It was no longer Brazil’s crop flow, as that force had largely faded after the harvest was absorbed. Instead, tariff risk had become the new driver, and unlike the old bearish force, it was starting to strengthen.
Seasonality: The Calendar Effect
Seasonality describes the patterns that repeat at regular, calendar-based intervals. Typical “real-life” examples are the increased energy consumption in summer and stronger sales during the Christmas period.
Coffee markets are deeply seasonal. Anchored in the agricultural calendar, they exhibit recurring, time-bound patterns that tend to repeat every year. The key distinction here is that seasonality is not a cycle, nor a trend. It is predictable, regular, and calendar driven.
Hence, one way to think about seasonality is to ask: "what time of the year is it, and what does the harvest/weather calendar dictate?"
For example, supply and export flows usually peak during harvest, leading to stock builds, while the off-season (Brazil’s, in special) tends to see inventory drawdowns. Prices often react not to the level of exports or the size of the stock increase or drawdown, but to whether flows are above or below the seasonal norm.
While many types of coffee market data can be analyzed seasonally, in coffee the term “seasonals” most often refers to price seasonals. These aim to capture how futures prices tend to behave at different times of the year (based on factors we have just discussed).
[Looking for clearer coffee market signals? Start a free trial of our Coffee Research and see what you’re missing.]
Price seasonals are constructed in a very straightforward way: they are simply the average coffee prices for each day (or week or month) of the calendar year.
The time frame used to construct seasonality matters because it reflects a trade-off between consistency and recency. Using a long history helps smooth out anomalies and capture what typically happens across many cycles, but markets evolve, and patterns that mattered decades ago may no longer be relevant.
Focusing only on recent years, on the other hand, risks missing longer-term tendencies that still shape behavior. To balance this, a practical approach is to use a front-weighted average. This method incorporates a broad data set (for example, the past 20 years) while giving progressively more weight to recent periods. The result preserves historical consistency but emphasizes the price behavior that is most relevant to today’s market.
This is the approach we use in our own seasonal analysis. In our weekly seasonality report, which we publish for free, we balance consistency and recency by incorporating long-term history while placing greater weight on the most recent years, so the seasonal patterns remain both statistically robust and relevant to the current coffee market.
A recent example of this was our December–January seasonality reports, which consistently pointed to a bearish setup and suggested downside of roughly 20–25c. Prices did fall during that period. While the move wasn’t driven by seasonality alone, the seasonal framework was one of the inputs that helped shape our broader analysis.
In Part 2, we will continue by examining the remaining components of a time series, focusing on cycles and noise and how they interact with the underlying structure of the data.
[Start a free trial of our Premium Coffee Market Research today!]
Comments