• If you are citizen of an European Union member nation, you may not use this service unless you are at least 16 years old.

  • Buried in cloud files? We can help with Spring cleaning!

    Whether you use Dropbox, Drive, G-Suite, OneDrive, Gmail, Slack, Notion, or all of the above, Dokkio will organize your files for you. Try Dokkio (from the makers of PBworks) for free today.

  • Dokkio (from the makers of PBworks) was #2 on Product Hunt! Check out what people are saying by clicking here.



Page history last edited by PBworks 15 years, 9 months ago

Capital allocation


Summary of topic


This page discusses the topic of capital allocation, including both the why we should (or shouldn't) allocate capital, and also how we should allocate capital for various purposes.


Those working on the topic


Richard Holloway - Richard.Holloway@Kilnplc.com

Declan Lavelle - declanlavelle@hld.ie


Getting started


This paper is probably as good an overall summary as I have found.


this document from Benfields is also a good overall summary and has some good graphs.


This spreadsheet works through sections 5.1 to 5.4 of the Benfields document to demonstrate the calculation of the various allocation methods and risk measures. The spreadsheet uses an approximation to calculate the percentiles, variance and tail conditional expectations for a sum of lognormal variables (instead of simulating the portfolios). This approximation is of interest in its own right.

This paper presented to the Society of Actuaries in Ireland works through the practical aspects of Capital Allocation.



Our findings to date


Key papers include

Myers-Read which is quoted in most later papers

Don Mango's paper and presentation which is similar to the one presented at the 2005 Giro. This includes some novel ideas, but has a very logical approach.

There are also a long list of questions (see below)


If you would like to comment please update this wiki page in the comments section; if appropriate these will be merged in with the main text in due course.




The following are my key questions


1. What is the capital allocation actually used for?


JAMES ORR (Some initial thoughts) - Often, to map the firm's overall objectives, in terms of its profit or return on capital, to the individual business unit level. This can support performance measurement or generate "hurdle rates" that potential policies should exceed before being "allowed" in to the portfolio.


2. Does the allocation only look at the events in the extreme tail (0.5%)?


JAMES ORR - All the examples that I have seen do, but I have heard it suggested that capital allocation (and the ICA process overall) would be of more real life relevance and value to businesses if they also looked at more frequent levels of expected performance, such as every 5 years or so, closer to the average stay for an underwriter or manager.


RICHARD HOLLOWAY - we have some section of the allocation based on the most extreme 20% (i.e. 1 in 5 year). It requires some modification however as many classes are not loss making at that point so you have to reframe it as the capital is the failure to make an arbitary return on capital. This is a recursive definition, but it converges quite quickly.


3. How far does the allocation go (business unit, business line, or individual policy)?


JAMES ORR - This seems to be driven by practicalities. I first saw an application (in a company) at business unit level in the mid-90s, also Lloyd's Risk Based Capital system operated at line of a business level but then determined a syndicate level figure (strictly, capital-setting rather than allocation, but equivalent in the special case of Lloyd's).


Mango's reinsurance pricing examples operate at policy level and I believe that Simon Pollack looked at the incremental effect on a portfolio of adding single policies, but this would be at the more sophisticated end of the spectrum.


4. Is the sum of the allocated capital equal the ICA naturally or is it forced?


JAMES ORR - An ICA is part of the calculation of the regulatory capital requirement and should be naturally (well, in almost all cases) less than the actual capital available.


RICHARD HOLLOWAY - I have changed the question to clarify it somewhat. Some methods of capital allocation are such that the sum of the capital calculated for each business unit is different to that calculated for the organisation as a whole. (this is due to the diversification benefit not being allocated). Some include some allocation of the diversification benefit so that the sum of the parts adds up to the whole without any further scaling.


5. Does the allocation tend to be stable over time?


JAMES ORR - From experience, only with a great deal of effort.


RICHARD HOLLOWAY - This is probably one of the more difficult points in 'selling' any allocation to the ones who will use it. If it changes radically simply due to 'normal' market movements, or worse becasue of things that other divisions are doing this makes it difficult to explain, let alone convince underwriters that it is right!


6. How do you deal with the allocation of capital from multi period ICA models?


JAMES ORR - I'd start by asking Andreas Tsanakas if he had any ideas! I think that Don Mango's approach offers something, in recognising that policies from different years can call on the same capital.

Comments (0)

You don't have permission to comment on this page.