Modern Day ALM – Get in the Mix

RPostBlenderI

Asset Liability Modelling (“ALM”) sure has come a long way since Frank Redington began talking about the interest rate immunization of life office books in the 1950s, or since models such as the Wilkie Model were first developed to deal with need for stochastic projection of economic variables for insurance companies and later pension funds.

Today the ALM picture is far richer than it was in the 1980s or 90s due to the huge spectrum of asset strategies available out there to institutional investors. However a lot of the core challenges are the same.

At Redington we’ve spent a huge amount of time developing what we think is a cutting-edge ALM system, and we’re pretty proud of the results.

 Typically we need to answer the following questions for clients –

  • How can I reduce risk, and increase return?
  • Is the scheme on track to meet its objectives?
  • Which strategies will add diversification?
  • If I invested in [X] would it help the portfolio?
  • What sort of losses am I risking with the current asset/liability portfolio?
  • How does my interest rate hedging risk interact with my assets

What we needed was a tool that delivered the depth and precision of ALM analysis that we were known for, but incorporated the breadth of strategies that are now available to clients, built in latest thinking from Manager research (MRT) and was quick to run so multiple strategies could be tested quickly.

How did we develop BLENDER

It was a real team effort among the different areas of the business. The development team built the initial engine that allowed “blending” of different asset simulations into a portfolio, and a huge library of functions to carry out the analysis. The ALM team refined the output with input from the consulting team so we were producing the most helpful output for clients to make decisions. The ALM team worked with the MRT team and IC output to understand the risk factors in the wide range of strategies and instruments we now employ (from Real Estate Debt to Multi-Class Credit or gilt repo) and incorporate the risk factors into the model. Typically the ALM team will write up a proposed modelling framework for a strategy and take this to the Investment Committee for approval, sometime additional changes will be suggested before the IC is happy. The dev team continued to expand the functionality, moving to a monthly cycle of updating the simulations and data and improving the usability and stability as the tool became bigger and bigger. Other analytical tools were incorporated (eg flightplan) to make sure they were as consistent as possible and make the workflow more efficient. The ALM team worked to transfer each client’s modelling over to blender, and continually identifying and carrying out fixes or improvements to make the tool more helpful for the IC team. For example a recent improvement is to allow the colours in each chart to be explicitly specified – a small thing but one that can save consultants time that can be spent on more valuable work.

What’s the Secret of BLENDER

Breadth – coverage of over 40 different asset classes and strategies all with distinct modelling approaches.

Depth – in each strategy we identify a number of risk and return drivers, we look at the exposures that our preferred managers will take, we ask questions like – how much equity exposure do they have on average, how exposed are they to investment grade vs high yield credit, how much of their risk is relative-value positions vs outright market exposure. We capture the risk level that the manager is looking to target. In LDI we believe that the only way to properly capture these risks is to build up the portfolio on a security level basis, modelling each swap and gilt individually, including any synthetic bond exposures through REPO or TRS.

Blender Key Features

What’s the Benefit?

All our analysis is focused on helping our clients make better decisions. What the blender enables us to do in the case of a new clients is very quickly identify what the major risks are and the levers that a client can pull to really change the risk/return balance. When we are working with a more established client blender allows us to run through a huge variety of possible asset allocation changes very quickly and efficiently, given the number of moving parts in most pension scheme portfolios this is essential in order to make sure clients are making the right decisions to meet their objectives

For more on Blender check out Redington Post.

RPostBlenderII RPostBlenderI

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s