Turning data into actionable insights
Nlitn [pronounce: enlighten] is a data consultancy company founded by successful academics. We are experts in econometric modelling and forecasting of time series. Nlitn has outstanding contacts in the academic world which allows us to create value for our clients based on the latest developments in data analysis. The two branches of the company are consultancy and software development.
Consultancy services
Through deep knowledge of econometric time series methodology, Nlitn consultancy works with clients to extract as much information as possible from their time series data sets. We refer to this as signal extraction and we use it to evaluate and forecast time series as accurate as possible. Nlitn has expertise in many areas, among them are finance, sports, and climatology. We refer to the Projects and Research section for more information.
Software development
Nlitn knows how to efficiently implement financial and econometric algorithms in languages like Python, Matlab, R, C, and OxMetrics. A prime example of Nlitn software development is the Time Series Lab software package which allows user to model time series with a variety of dynamic components by the click of a couple of buttons, making it a very suitable program for those without coding expertise.
Solutions
Nlitn your products
We offer full data solution packages to suit your data analysis needs. An example is the Time Series Lab software package which provides many ways of extracting information from your time series. Another example is the Structural Time series Analyser, Modeller and Predictor program or STAMP in short. STAMP is a statistical / econometric software system for time series models with unobserved components such as trend, seasonal, cycle and irregular and is part of the OxMetrics package.
More informationNlitn your knowledge
Nlitn is proud to cooperate with leading academics and professionals to deliver quality training services. We provide time series econometrics courses that range from basic to expert level. Dynamic modelling of time series is at the core of the courses and both theory as well as applied work is discussed. Examples of courses are "Analyzing Time Series with State Space Models" and "Dynamic Models for Volatility and Heavy Tails". Courses are often combined with programming exercises in which time series theory and methodology are utilized.
Nlitn your Excel sheets
Excel sheets and the financial world are closely intertwined. Many companies depend on Excel sheets for their important calculations. However, this dependence becomes problematic when Excel is faced with large datasets and heavy computations. Nlitn often encounters large, slow Excel sheets, possibly built by staff members that long ago left the company, where over time new modules were added to the spreadsheets without regard of structure, efficiency, or error checking capabilities. Nlitn works with finance professionals to transform Excel sheets to highly efficient code modules.
More informationProjects
Timberlake
Nlitn is proud to mention a partnership with Timberlake.
Timberlake has over thirty years of experience providing total and impartial solutions in statistics, econometrics and forecasting. They distribute leading software, deliver quality training courses and provide consultancy services, globally.
Together with Timberlake, Nlitn builds project specific software for the analysis of large time series databases. The software is typically used to discover complex patterns in time series which allows the customer to, for example, improve target advertising or predict future sales more accurately.
Time Series Lab
Nlitn developed the Time Series Lab - Score Edition software package.
Time Series Lab - Score Edition is a software program for analyzing, modelling, and forecasting of time series. The software allows users to specify a wide range of dynamic components and probability distributions to extract the maximum amount of signal from the time series data.
The time series methodology of the Time Series Lab - Score Edition package is well-founded in the academic world and has appeared in highly respected academic journals. For an overview of score-driven publications, we refer to the www.gasmodel.com website. The score-driven methodology was developed independently at VU University Amsterdam and Cambridge University. Currently, the knowledge and experience of both universities have been combined and Professor S.J. Koopman (VU Amsterdam) and Professor A.C. Harvey (Cambridge) are part of the Time Series Lab team.
STAMP
Nlitn is part of the main developing team of STAMP (Structural Time Series Analyser, Modeller and Predictor) software.
STAMP is a statistical / econometric software system for time series models with unobserved components such as trend, seasonal, cycle and irregular. It provides a user-friendly environment for the analysis, modelling and forecasting of time series. Estimation and signal extraction is carried out using state space methods and Kalman filtering. STAMP is set up in an easy-to-use form which enables the user to concentrate on model selection and interpretation. STAMP 8 is an integrated part of the OxMetrics modular software system for data analysis with excellent data manipulation, graphical and batch facilities.
STAMP is part of the OxMetrics package.
Professional courses
Nlitn provides Time Series Econometrics courses to a variety of companies. Recently, Nlitn provided an introductory Matlab coding course to a Commercial bank and an expert time series course to a company in the gambling world.
The Matlab coding course was a four part course that dealt with the basics of Matlab and focussed on coding practices like maintainability of code modules, efficiency of code, and program organization.
The expert time series course was an eight part course and focussed on time series methodology. Dynamic modelling of time series were at the core of the course and both theory as well as applied work is discussed. The use of time series software like STAMP and SsfPack are used to assist the customer in understanding the theory behind time series models. After the course the customer is able to apply time series theory to many data sets.
Research
The analysis and forecasting of tennis matches using a high-dimensional dynamic model
Gorgi, P., Koopman, S.J., and Lit, R.
Published in Journal of the Royal Statistical Society, Series A (2019), 182(4), 1393-1409.
Abstract
We propose a high dimensional dynamic model for tennis match results with time varying player‐specific abilities for different court surface types. Our statistical model can handle high dimensional data sets while the number of parameters remains small. In particular, we analyse 17 years of tennis matches for a panel of over 500 players, which leads to more than 2000 dynamic strength levels. We find that time varying player‐specific abilities for different court surfaces are of key importance for analysing tennis matches. We further consider several other extensions including player‐specific explanatory variables and the match configurations for Grand Slam tournaments. The estimation results can be used to construct rankings of players for different court surface types. We finally show that our proposed model produces accurate forecasts. We provide evidence that our model significantly outperforms existing models in the forecasting of tennis match results.
Long Term Forecasting of El Niño Events via Dynamic Factor Simulations
Mengheng, Li., Koopman, S.J., Lit, R. and Desislava, P.
Accepted to Journal of Econometrics, Series A (2019).
Abstract
We propose a new forecasting procedure for the El Niño 3.4 time series that is linked with the well-known El Niño phenomenon. This important climate time series is subject to an intricate serial correlation structure and is related to many other relevant and related variables. Although the forecasting procedure is valid for all lead times, it is particularly developed for medium to long term forecasting of El Niño.
Intraday Stochastic Volatility in Discrete Price Changes: the Dynamic Skellam Model
Koopman, S.J., Lit, R. and Lucas, A.
Published in Journal of the American Statistical Association (2017), 112, 1490-1503.
Abstract
Volatility in stock prices is a statistical measure of the degree of variation in the stock price process over time. We show that intraday stochastic volatility can be extracted from discrete price changes instead of, commonly used, continuously compounded returns. We model the high-frequency tick-by-tick price changes with the discrete Skellam distribution with time-varying variance parameter. The intraday dynamics of volatility and the high number of trades without price impact require non-trivial adjustments to the basic dynamic Skellam model.
Forecasting football match results in national league competitions using score-driven time series models
Koopman, S.J. and Lit, R.
Accepted to International Journal of Forecasting.
Abstract
We develop a new dynamic multivariate model for the analysis and the forecasting of football match results in national league competitions. Our main interest is to forecast whether the match result is a win, a loss or a draw for each team. To deliver such forecasts, the dynamic model can be based on three different dependent variables: the pairwise count of the number of goals, the difference between the number of goals, or the category of the match result (win, loss, draw). The different dependent variables require different distributional assumptions. In an extensive forecasting study for match results from six large European football competitions, we validate the precision of the forecasts and the success of the forecasts in a betting simulation. We conclude that the dynamic model for pairwise counts delivers the most precise forecasts.
Dynamic Discrete Copula Models for High Frequency Stock Price Changes
Koopman, S.J., Lit, R., Lucas, A. and Opschoor, A.
Published in Journal of Applied Econometrics (2018), 33(7), 966-985.
Abstract
We develop a dynamic model for the intraday dependence between stock price series. The model combines Skellam marginal distributions for one-second positive and negative integer price changes together with discrete copula functions to capture dependence. We consider a range of different copula functions and find strong evidence that dependence between stock price changes varies within a trading day.