Glossary term
Time Series
A time series is a sequence of data points recorded in time order, such as daily prices, monthly inflation, or quarterly revenue.
Updated
Read time
What Is a Time Series?
A time series is a sequence of data points recorded in time order, such as daily stock prices, monthly inflation readings, weekly jobless claims, quarterly revenue, or annual GDP. The time order is part of the information. Rearranging the observations would destroy much of the meaning.
Time series data is used in investing, economics, accounting, operations, risk management, and forecasting because many financial questions are about how something changes over time.
Key Takeaways
- A time series records observations in time order.
- Common examples include prices, sales, interest rates, inflation, employment, and production data.
- Time series analysis studies trend, seasonality, cycles, volatility, and structural breaks.
- Forecasts depend on assumptions about whether past patterns remain useful.
- Bad data, changing definitions, outliers, and regime shifts can make time series analysis misleading.
How Time Series Data Works
Each observation in a time series has a time stamp or period. The spacing may be seconds, minutes, days, months, quarters, or years. High-frequency trading data can update many times per second. Economic data may update monthly or quarterly.
The frequency changes what can be seen. Daily prices may show volatility and short-term reactions. Quarterly revenue may show business direction. Annual data may smooth noise but hide turning points.
Common Time Series Features
Feature | Meaning |
|---|---|
Trend | Longer-term direction in the data. |
Seasonality | Recurring patterns tied to the calendar or operating cycle. |
Cycle | Longer expansions and contractions that may not follow a fixed calendar. |
Volatility | Variation around the pattern or expected level. |
Structural break | A change in the relationship or data-generating process. |
Finance and Business Uses
Investors use time series to study returns, volatility, drawdowns, correlations, interest rates, and economic indicators. Businesses use them to monitor sales, churn, inventories, defect rates, cash flows, customer demand, and staffing needs.
Time series analysis can support budgeting, risk management, valuation, inventory planning, capacity decisions, and market timing. The method can be simple, such as comparing year-over-year growth, or technical, such as ARIMA modeling, exponential smoothing, or seasonal decomposition.
Forecasting Context
A forecast uses past and current data to estimate future values. Time series forecasting may rely on trend, seasonality, mean reversion, momentum, or relationships with other variables. The forecast is only as strong as the assumption that the pattern remains relevant.
Forecasts often fail when the underlying system changes. A pandemic, policy change, new competitor, supply shock, accounting change, or interest-rate regime shift can make old relationships less useful.
Data Quality and Interpretation
Time series analysis depends on consistent definitions and clean data. Missing observations, revisions, calendar effects, inflation, stock splits, survivorship bias, and outliers can distort results. Economic time series are often revised after first publication, which can change historical interpretation.
Analysts should ask whether the data is nominal or inflation-adjusted, seasonally adjusted or raw, point-in-time or revised, and comparable across the whole period. Those details can change the conclusion.
Stationarity and Regime Change
Many time series methods work better when the statistical behavior of the series is reasonably stable. If the average level, volatility, seasonality, or relationship with other variables changes, a model fitted to the past can become unreliable.
Financial markets are especially vulnerable to regime change. A period of low rates, low inflation, or calm volatility may produce patterns that do not survive a new policy environment. Analysts should therefore treat model output as evidence, not as certainty.
Seasonal Adjustment
Many economic time series are published in both raw and seasonally adjusted form. Seasonal adjustment tries to remove recurring calendar patterns so underlying changes are easier to see. It can help analysts read month-to-month changes, but it also adds methodology assumptions that should be understood before drawing strong conclusions.
The Bottom Line
A time series is data recorded in time order. It helps analysts study change, but the value comes from understanding trend, seasonality, cycles, volatility, data quality, and whether the past still provides useful evidence about the future.