Having looked into options for modelling and forecasting a financial time series based on mixed frequency data, I came across state space models, which seems worth exploring.

I've however been unable to find a clear, intuitive and largely un-mathematical explanation of how a state space model works or what they do to allow mixed frequencies.

As all the info I have found on this topic seems very technical I, and hopefully others, would very much appreciate some clarification on what seems to be a pretty powerful tool.


The good news is that your instincts are right that it would be a useful technique. The bad news is that it's not a technique that you can use without understanding a fair amount of linear algebra. It's all about multiple equations with multiple matrix multiplications.

Some tools like R's bsts package make it more accessible, but it's fundamentally more complex than alternatives. Not that you should be using ARIMA or other methods without some level of technical sophistication, but in my experience most state space (also called dynamic linear model) packages have gaps where you'll need to know what parts of various matrices represent and mean.

Given all of that, as a readable introduction I'd recommend "An Introduction to State Space Time Series Analysis" by Jacques J.F. Commandeur and Siem Jan Koopman, Oxford 2007. It's a short book and used to be pretty expensive, but it appears that it may have been released on the Internet. I don't believe this book mentions mixed-frequency data, though.

And if you use R, you should check out bsts.

  • $\begingroup$ Thanks for the input and for the book recommendation. Looking at it now and it should help explain a lot of key factors. For others that might be interested I found it here $\endgroup$ – youjustreadthis Jun 13 '16 at 11:41

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.