Information Matrix
Filter
Posts feedConcentrate!
One of the challenges in modelling financial portfolios is the concentration of risk arising from the fact wealthier people will usually have significantly higher benefits than the less well-off.
Winter mortality
In previous posts we looked at seasonal fluctuations in mortality. Since the UK is about to experience some particularly cold weather again, we will look at winter mortality in more detail.
Logistical nightmares
A common Generalised Linear Model (GLM) for mortality modelling is logistic regression, also sometimes described as a Bernoulli GLM with a logistic link function. This models mortality at the level of the individual, and models the rate of mortality over a single year.
Double trouble
Scientists strongly prefer ideas and processes which have undergone anonymous peer review in published, refereed journals.
Interesting times
The Bank of England has reduced its current bank rate to 1.5%, the lowest since it was founded in 1694. Whilst this is good news for borrowers, it is bad news for those in retirement who are living off the interest on their savings.
Table talk
When valuing a portfolio, an actuary must often make a decision as to what tables to use for the risk. The ideal is to use tables which have been created from the portfolio's own experience, preferably using a statistical model to account for the various risk factors.
Size isn't everything
In an earlier post we discussed the correct way of using postcodes for analysing mortality, and also how this works in countries outside the UK. It is worth re-iterating why insurers use so-called geodemographic profiling.
Great Expectations
When fitting statistical models, a number of features are commonly assumed by users. Chief amongst these assumptions is that the expected number of events according to the model will equal the actual number in the data. This strikes most people as a thoroughly reasonable expectation. Reasonable, but often wrong.
Confounding compounding
Earlier posts discussed the importance of deduplication in annuity portfolios and pension schemes and some of the issues around the deduplication of names, specifically the use of double metaphone to look through common variant spellings of the surname or family name.
A likely story
The foundation for most modern statistical inference is the log-likelihood function. By maximising the value of this function, we find the maximum-likelihood estimate (MLE) for a given parameter, i.e. the most likely value given the model and data. For models with more than one parameter, we find the set of values which jointly maximise the log-likelihood.