Time Series Analysis and Financial Forecasting in Excel
Expert-defined terms from the Advanced Certificate in Excel for Financial Accounting (translated title from another language to English) course at London School of Business and Administration. Free to read, free to share, paired with a globally recognised certification pathway.
A priori probability #
this term refers to the probability of an event occurring based on prior knowledge or experience, often used in financial forecasting to make predictions about future events. Related terms include a posteriori probability, which is the probability of an event occurring after considering new evidence. In the context of Time Series Analysis, a priori probability can be used to inform the choice of models and parameters.
Additive model #
an additive model is a type of model used in Time Series Analysis, where the effects of different variables are added together to forecast future values. Related terms include multiplicative models, where the effects of different variables are multiplied together. Additive models are often used in financial forecasting to model the relationship between a time series and external variables.
Algorithm #
an algorithm is a set of rules or procedures used to solve a problem or make a prediction, often used in Time Series Analysis and financial forecasting to automate the forecasting process. Related terms include machine learning, which refers to the use of algorithms to learn from data and make predictions.
ARIMA model #
an ARIMA model is a type of model used in Time Series Analysis, which combines the effects of autoregression, differencing, and moving averages to forecast future values. Related terms include ARIMAX models, which incorporate external variables into the forecasting process. ARIMA models are widely used in financial forecasting to model complex time series data.
Auto #
correlation - auto-correlation refers to the correlation between a time series and lagged versions of itself, often used in Time Series Analysis to identify patterns and trends in the data. Related terms include partial auto-correlation, which refers to the correlation between a time series and a lagged version of itself, while controlling for the effects of other lags.
Autoregression #
autoregression refers to the use of past values of a time series to forecast future values, often used in Time Series Analysis and financial forecasting to model the relationship between a time series and its past values. Related terms include vector autoregression, which refers to the use of past values of multiple time series to forecast future values.
Backtesting #
backtesting refers to the process of evaluating the performance of a forecasting model using historical data, often used in financial forecasting to evaluate the accuracy of a model. Related terms include walk-forward optimization, which refers to the process of evaluating the performance of a model using out-of-sample data.
Bayesian methods #
Bayesian methods refer to a set of statistical techniques used to update the probability of a hypothesis based on new evidence, often used in Time Series Analysis and financial forecasting to make predictions about future events. Related terms include Bayesian inference, which refers to the process of updating the probability of a hypothesis based on new evidence.
Box #
Jenkins methodology - the Box-Jenkins methodology is a set of procedures used to identify, estimate, and evaluate ARIMA models, often used in Time Series Analysis and financial forecasting to model complex time series data. Related terms include the Box-Jenkins model, which refers to a specific type of ARIMA model.
Business cycle #
a business cycle refers to the fluctuations in economic activity over time, often used in financial forecasting to model the relationship between economic indicators and financial time series. Related terms include expansion, contraction, and recession, which refer to different phases of the business cycle.
Calendar effects #
calendar effects refer to the effects of calendar-related events, such as holidays and seasonal patterns, on time series data, often used in Time Series Analysis and financial forecasting to model the relationship between a time series and external variables. Related terms include seasonal adjustment, which refers to the process of removing calendar effects from time series data.
Causality #
causality refers to the relationship between two variables, where one variable causes a change in the other variable, often used in Time Series Analysis and financial forecasting to model the relationship between economic indicators and financial time series. Related terms include Granger causality, which refers to a statistical test used to evaluate the causality between two variables.
Coefficient of determination #
the coefficient of determination is a statistical measure used to evaluate the goodness of fit of a forecasting model, often used in financial forecasting to evaluate the accuracy of a model. Related terms include R-squared, which refers to a specific type of coefficient of determination.
Cointegration #
cointegration refers to the long-run relationship between two or more time series, often used in Time Series Analysis and financial forecasting to model the relationship between economic indicators and financial time series. Related terms include cointegrating vector, which refers to a linear combination of time series that is stationary.
Conditional probability #
conditional probability refers to the probability of an event occurring given that another event has occurred, often used in financial forecasting to make predictions about future events. Related terms include unconditional probability, which refers to the probability of an event occurring without considering any other events.
Confidence interval #
a confidence interval is a statistical measure used to evaluate the uncertainty of a forecasting model, often used in financial forecasting to evaluate the accuracy of a model. Related terms include prediction interval, which refers to a statistical measure used to evaluate the uncertainty of a forecasting model.
Correlogram #
a correlogram is a visual representation of the auto-correlation function, often used in Time Series Analysis to identify patterns and trends in the data. Related terms include partial correlogram, which refers to a visual representation of the partial auto-correlation function.
Cross #
validation - cross-validation refers to the process of evaluating the performance of a forecasting model using multiple subsets of the data, often used in financial forecasting to evaluate the accuracy of a model.
Cycle #
a cycle refers to a periodic pattern in time series data, often used in Time Series Analysis and financial forecasting to model the relationship between a time series and external variables. Related terms include trend, which refers to a long-term pattern in time series data.
Damping factor #
a damping factor is a parameter used in exponential smoothing models to control the rate of smoothing, often used in Time Series Analysis and financial forecasting to model the relationship between a time series and its past values. Related terms include smoothing parameter, which refers to a parameter used to control the rate of smoothing.
Data mining #
data mining refers to the process of discovering patterns and relationships in large datasets, often used in Time Series Analysis and financial forecasting to identify opportunities and risks.
Decomposition #
decomposition refers to the process of separating a time series into its component parts, such as trend, seasonality, and residuals, often used in Time Series Analysis and financial forecasting to model the relationship between a time series and external variables. Related terms include seasonal decomposition, which refers to the process of separating a time series into its seasonal component.
Degrees of freedom #
degrees of freedom refer to the number of independent observations in a dataset, often used in Time Series Analysis and financial forecasting to evaluate the accuracy of a model. Related terms include chi-squared distribution, which refers to a statistical distribution used to evaluate the goodness of fit of a model.
Differencing #
differencing refers to the process of subtracting each value in a time series from its previous value, often used in Time Series Analysis and financial forecasting to make a time series stationary. Related terms include first-order differencing, which refers to the process of subtracting each value in a time series from its previous value.
Dummy variable #
a dummy variable is a binary variable used to represent a categorical variable, often used in Time Series Analysis and financial forecasting to model the relationship between a time series and external variables. Related terms include indicator variable, which refers to a binary variable used to represent a categorical variable.
Econometrics #
econometrics refers to the application of statistical techniques to economic data, often used in Time Series Analysis and financial forecasting to model the relationship between economic indicators and financial time series. Related terms include econometric model, which refers to a statistical model used to analyze economic data.
Efficient market hypothesis #
the efficient market hypothesis refers to the idea that financial markets reflect all available information, often used in financial forecasting to evaluate the accuracy of a model. Related terms include random walk, which refers to a type of time series model that assumes that the future value of a time series is unpredictable.
Error correction model #
an error correction model is a type of model used in Time Series Analysis, which models the relationship between two or more time series and their errors, often used in financial forecasting to model the relationship between economic indicators and financial time series. Related terms include vector error correction model, which refers to a type of model that models the relationship between multiple time series and their errors.
Exogenous variable #
an exogenous variable is a variable that is external to the system being modeled, often used in Time Series Analysis and financial forecasting to model the relationship between a time series and external variables. Related terms include endogenous variable, which refers to a variable that is internal to the system being modeled.
Exponential smoothing #
exponential smoothing is! A type of model used in Time Series Analysis, which weights recent observations more heavily than past observations, often used in financial forecasting to model the relationship between a time series and its past values. Related terms include simple exponential smoothing, which refers to a type of model that weights all observations equally.
Forecasting #
forecasting refers to the process of predicting future values of a time series, often used in Time Series Analysis and financial forecasting to make predictions about future events. Related terms include prediction, which refers to the process of predicting future values of a time series.
GARCH model #
a GARCH model is a type of model used in Time Series Analysis, which models the volatility of a time series, often used in financial forecasting to model the relationship between a time series and its volatility. Related terms include EGARCH model, which refers to a type of model that models the volatility of a time series and its asymmetry.
Heteroskedasticity #
heteroskedasticity refers to the phenomenon of non-constant variance in a time series, often used in Time Series Analysis and financial forecasting to evaluate the accuracy of a model. Related terms include homoskedasticity, which refers to the phenomenon of constant variance in a time series.
Impulse response function #
an impulse response function is a function that describes the response of a time series to a shock, often used in Time Series Analysis and financial forecasting to model the relationship between a time series and external variables. Related terms include variance decomposition, which refers to a technique used to decompose the variance of a time series into its component parts.
Information criterion #
an information criterion is a statistical measure used to evaluate the goodness of fit of a forecasting model, often used in financial forecasting to evaluate the accuracy of a model. Related terms include Akaike information criterion, which refers to a specific type of information criterion.
In #
sample - in-sample refers to the data used to estimate the parameters of a forecasting model, often used in Time Series Analysis and financial forecasting to evaluate the accuracy of a model. Related terms include out-of-sample, which refers to the data used to evaluate the performance of a forecasting model.
Integrated process #
an integrated process is a type of time series model that assumes that the time series is non-stationary, often used in Time Series Analysis and financial forecasting to model the relationship between a time series and its past values. Related terms include differencing, which refers to the process of subtracting each value in a time series from its previous value.
Kurtosis #
kurtosis refers to the fourth moment of a distribution, often used in Time Series Analysis and financial forecasting to evaluate the shape of a distribution. Related terms include leptokurtosis, which refers to a distribution with heavy tails.
Lag #
a lag refers to the time delay between two time series, often used in Time Series Analysis and financial forecasting to model the relationship between a time series and its past values. Related terms include lead, which refers to the time delay between two time series in the opposite direction.
Likelihood function #
a likelihood function is a statistical function used to evaluate the probability of a forecasting model, often used in financial forecasting to evaluate the accuracy of a model. Related terms include log-likelihood function, which refers to the logarithm of the likelihood function.
Linear model #
a linear model is a type of model used in Time Series Analysis, which assumes a linear relationship between the variables, often used in financial forecasting to model the relationship between a time series and external variables. Related terms include non-linear model, which refers to a type of model that assumes a non-linear relationship between the variables.
Mean absolute error #
the mean absolute error is a statistical measure used to evaluate the accuracy of a forecasting model, often used in financial forecasting to evaluate the accuracy of a model. Related terms include mean squared error, which refers to a statistical measure used to evaluate the accuracy of a forecasting model.
Mean reversion #
mean reversion refers to the phenomenon of a time series returning to its mean value over time, often used in Time Series Analysis and financial forecasting to model the relationship between a time series and its past values. Related terms include trend reversion, which refers to the phenomenon of a time series returning to its trend value over time.
Model selection #
model selection refers to the process of choosing the best forecasting model for a given dataset, often used in Time Series Analysis and financial forecasting to evaluate the accuracy of a model. Related terms include model evaluation, which refers to the process of evaluating the performance of a forecasting model.
Moving average #
a moving average is a type of model used in Time Series Analysis, which weights recent observations more heavily than past observations, often used in financial forecasting to model the relationship between a time series and its past values. Related terms include exponential moving average, which refers to a type of model that weights recent observations more heavily than past observations.
Multi #
step forecasting - multi-step forecasting refers to the process of predicting multiple future values of a time series, often used in Time Series Analysis and financial forecasting to make predictions about future events. Related terms include single-step forecasting, which refers to the process of predicting a single future value of a time series.
Non #
linearity - non-linearity refers to the phenomenon of a non-linear relationship between the variables, often used in Time Series Analysis and financial forecasting to model the relationship between a time series and external variables.
Non #
stationarity - non-stationarity refers to the phenomenon of a time series that is not stationary, often used in Time Series Analysis and financial forecasting to evaluate the accuracy of a model. Related terms include stationarity, which refers to the phenomenon of a time series that is stationary.
Normalization #
normalization refers to the process of scaling a time series to have a mean of zero and a variance of one, often used in Time Series Analysis and financial forecasting to evaluate the accuracy of a model. Related terms include standardization, which refers to the process of scaling a time series to have a mean of zero and a variance of one.
Out #
of-sample - out-of-sample refers to the data used to evaluate the performance of a forecasting model, often used in Time Series Analysis and financial forecasting to evaluate the accuracy of a model. Related terms include in-sample, which refers to the data used to estimate the parameters of a forecasting model.
Overfitting #
overfitting refers to the phenomenon of a forecasting model that is too complex and fits the noise in the data, often used in Time Series Analysis and financial forecasting to evaluate the accuracy of a model. Related terms include underfitting, which refers to the phenomenon of a forecasting model that is too simple and does not capture the underlying patterns in the data.
Parameter estimation #
parameter estimation refers to the process of estimating the parameters of a forecasting model, often used in Time Series Analysis and financial forecasting to evaluate the accuracy of a model. Related terms include maximum likelihood estimation, which refers to a technique used to estimate the parameters of a forecasting model.
Partial auto #
correlation - partial auto-correlation refers to the correlation between a time series and a lagged version of itself, while controlling for the effects of other lags, often used in Time Series Analysis to identify patterns and trends in the data. Related terms include auto-correlation, which refers to the correlation between a time series and lagged versions of itself.
Point forecasting #
point forecasting refers to the process of predicting a single future value of a time series, often used in Time Series Analysis and financial forecasting to make predictions about future events. Related terms include interval forecasting, which refers to the process of predicting a range of future values of a time series.
Portmanteau test #
a Portmanteau test is a statistical test used to evaluate the goodness of fit of a forecasting model, often used in financial forecasting to evaluate the accuracy of a model. Related terms include Ljung-Box test, which refers to a statistical test used to evaluate the goodness of fit of a forecasting model.
Prediction interval #
a prediction interval is a statistical measure used to evaluate the uncertainty of a forecasting model, often used in financial forecasting to evaluate the accuracy of a model. Related terms include confidence interval, which refers to a statistical measure used to evaluate the uncertainty of a forecasting model.
Random walk #
a random walk is a type of time series model that assumes that the future value of a time series is unpredictable, often used in Time Series Analysis and financial forecasting to model the relationship between a time series and its past values. Related terms include drift, which refers to the phenomenon of a time series that has a constant rate of change.
Regression analysis #
regression analysis is a statistical technique used to model the relationship between a time series and external variables, often used in Time Series Analysis and financial forecasting to model the relationship between a time series and external variables. Related terms include linear regression, which refers to a type of regression analysis that assumes a linear relationship between the variables.
Residual #
a residual is the difference between the observed value of a time series and the predicted value, often used in Time Series Analysis and financial forecasting to evaluate the accuracy of a model. Related terms include residual plot, which refers to a visual representation of the residuals.
Seasonal decomposition #
seasonal decomposition refers to the process of separating a time series into its seasonal component, often used in Time Series Analysis and financial forecasting to model the relationship between a time series and external variables. Related terms include trend decomposition, which refers to the process of separating a time series into its trend component.
Seasonality #
seasonality refers to the phenomenon of a time series that has a periodic pattern, often used in Time Series Analysis and financial forecasting to model the relationship between a time series and external variables. Related terms include seasonal adjustment, which refers to the process of removing seasonality from time series data.
Spectral analysis #
spectral analysis is a statistical technique used to analyze the frequency components of a time series, often used in Time Series Analysis and financial forecasting to identify patterns and trends in the data. Related terms include Fourier analysis, which refers to a statistical technique used to analyze the frequency components of a time series.
Stationarity #
stationarity refers to the phenomenon of a time series that has a constant mean and variance over time, often used in Time Series Analysis and financial forecasting to evaluate the accuracy of a model. Related terms include non-stationarity, which refers to the phenomenon of a time series that is not stationary.
Stepwise regression #
stepwise regression is a statistical technique used to select the best subset of variables to include in a regression model, often used in Time Series Analysis and financial forecasting to model the relationship between a time series and external variables. Related terms include forward selection, which refers to a statistical technique used to select the best subset of variables to include in a regression model.
Time series #
a time series is a sequence of observations measured at regular time intervals, often used in Time Series Analysis and financial forecasting to model the relationship between a time series and external variables. Related terms include cross-sectional data, which refers to a sequence of observations measured at a single point in time.
Trend #
a trend refers to the long-term pattern in a time series, often used in Time Series Analysis and financial forecasting to model the relationship between a time series and external variables. Related terms include seasonality, which refers to the phenomenon of a time series that has a periodic pattern.
Unit root #
a unit root is a statistical test used to evaluate whether a time series is stationary or non-stationary, often used in Time Series Analysis and financial forecasting to evaluate the accuracy of a model. Related terms include augmented Dickey-Fuller test, which refers to a statistical test used to evaluate whether a time series is stationary or non-stationary.
Vector autoregression #
a vector autoregression is a type of model used in Time Series Analysis, which models the relationship between multiple time series and their past values, often used in financial forecasting to model the relationship between economic indicators and financial time series.
Volatility #
volatility refers to the phenomenon of a time series that has a high degree of uncertainty, often used in Time Series Analysis and financial forecasting to model the relationship between a time series and its volatility. Related terms include standard deviation, which refers to a statistical measure used to evaluate the uncertainty of a time series.
Walk #
forward optimization - walk-forward optimization refers to the process of evaluating the performance of a forecasting model using out-of-sample data, often used in financial forecasting to evaluate the accuracy of a model. Related terms include backtesting, which refers to the process of evaluating the performance of a forecasting model using historical data.
White noise #
white noise refers to a type of time series that has a constant mean and variance over time, and is uncorrelated with its past values, often used in Time Series Analysis and financial forecasting to model the relationship between a time series and its past values. Related terms include Gaussian white noise, which refers to a type of white noise that has a Gaussian distribution.