...

Insights A simple proxy for liquidity premium

by user

on
Category: Documents
83

views

Report

Comments

Transcript

Insights A simple proxy for liquidity premium
Insights
A simple proxy for liquidity premium
In October the CFO Forum revised its MCEV principles to allow the inclusion of the
liquidity premium. They do not, however, say anything on how it should be calculated.
And while CEIOPS doesn’t exactly open the door for such an inclusion, recent advice
does leave it ajar. They also provide advice on how to go about estimating it, suggesting
a proportion of the credit spread as an example method.
Jamie Stark
[email protected]
Given the material impact of any estimate on the solvency position of the balance sheet,
this begs the obvious question - how good can such a proxy be? Here we examine this
and other questions.
Does the liquidity premium exist?
Principle 14, Market Consistent
Embedded Value Principles.
The CFO Forum. October 2009
The existence of an illiquidity
premium is controversial.
In any case, even assuming
its existence, the question
remains whether this
illiquidity premium provides
a suitable solution
B.4. Advice for Level 2
Implementing Measures on
Solvency II: Technical Provisions
- Article 86 b - Risk-free
interest rate term structure.
CEIOPS, October 2009.
While CEIOPS think the existence of the liquidity premium is controversial we don’t1. Nor
do many of the market participants we speak to including bond traders, fund managers,
investment bankers and insurers. Arguably, earning the liquidity premium has been a
cornerstone of the insurance industry’s business model and its ability to deliver cost effective
fixed cash-flow products. The implicit savings of a buy-and-hold investment strategy have
often been shared with consumers.
What are we estimating the liquidity premium of?
To value an insurer’s liabilities we are looking for their transfer value. This is the amount of
cash we would need to pay someone to take on these liabilities. We should assume that this
other party has to play by the same rules, i.e. they have access to the same instruments, are
bound by the same regulation, capital requirements, etc. Depending on the predictability
of the liability (determined by the product structure, policy-holder behaviour, market and
mortality risk) there will be a proportion of backing assets that are not required to meet
expected demands and so can be held in a basket of illiquid assets.2
The premium paid in the market for accessing liquidity relative to this illiquid basket is what
we are trying to estimate. Corporate bonds are less liquid than other assets and in practice
are typical assets insurers use to match fixed cash flow liabilities. Assuming the other party is
an average insurance market participant we are interested in estimating the liquidity premium
of a basket of corporate bonds relative to a liquid one – we use swaps in our analysis below.
CEIOPS suggest using a basket with highest credit quality. We believe a reference portfolio
should include the types of assets an average insurer would use to back its liabilities, which
is typically an investment grade portfolio averaging A rated bonds.
How can you estimate it?
The price of corporate bonds includes discounts for credit risk and illiquidity so estimating
the liquidity premium is not as simple as comparing two prices. In our recent research report3
we considered three different methods to estimate it:
1.
2.
3.
See Liquidity premium: myth or reality? John Hibbert. 2009. Available at: http://www.barrhibb.
com/documents/downloads/LP_MythOrReality.pdf
Note that we do not need to consider the backing assets of the insurer to value its liabilities. The
business model any insurer uses to back its assets does not affect the value of it liabilities.
See Summary of Liquidity Premium estimation methods. Hibbert et al. Available at: http://www.
barrhibb.com/documents/downloads/Barrie_Hibbert_Summary_of_LP_Methods.pdf
CDS negative-basis method: Here we use CDS spreads to give us a market price of credit
risk. Backing this out of the market spread gives us an estimate of the liquidity premium.
An issue with this approach is that when bank liquidity is scarce the CDS spreads will
include counterparty credit risk and so is not necessarily a clean measure, although moves
towards central clearing will reduce this effect.
Structural model method: Here we use a Merton model to estimate the price of credit risk
using equity market volatility. Backing this out of the market spread gives us an estimate
of the liquidity premium. An issue with this method is that the model requires a number of
assumptions to be made.
Covered Bond Method: Here we assume that covered bonds (bonds that have an
actively managed pool of high quality collateral and governed by legalisation, e.g. German
Pfandbriefe) are similar to the highest quality corporate bonds. We assume they have no
credit risk and so simply compute the spread over swaps. Problems with this approach
are that they do not really estimate the liquidity premium of the pool of assets we are
interested in and that it can be hard to find enough estimates outside of the Euro market
to give meaningful results.
Without any observable reference points to verify methods against it can be easy to
criticise each of these methods and the estimates they produce. But by looking at number
of different methods together we can have a better informed view to judge the suitability
of a simple proxy.
How good can a simple proxy be?
So is a simple proxy such as 50%*(credit spread – 40bps) any good?
Figure 1: Comparison of liquidity premium estimates for GBP using Merrill Lynch credit spread
indices
As you can see in Figure 1, the results look surprisingly good for the UK bond market, over
the period considered. The proxy (green line) follows the rise and subsequent fall of the
measure and gives a reasonable level. We should note that the period considered here is
far from typical – we have an extremely benign period followed by a highly volatile period.
We have also looked at the EUR and USD markets and get similar results. We see no
reason to expect that bond market liquidity should be the same in different economies,
so we would also warn about using this proxy in other economies without some sort of
verification to other measures, if available.
Insights December 2009
www.barrhibb.com
Where next?
Estimating the liquidity premium is one thing, but applying it in valuation or using it to inform
investment decisions is another matter.
Firstly, one needs to understand the predictability of liability cash flows. We have developed
a methodology to help answer this question. Our methodology projects liabilities together
with backing assets and calculates net cash flows over a number of different scenarios. If
these are negative it means some of the assets need to be sold to finance the liabilities and
so require liquidity. We project over a large number of scenarios to produce a distribution
that includes extreme events – giving a better understanding of the predictability of liabilities
over their life and under different economic conditions. More details can be found in a
forthcoming research note.4
Secondly, you need to consider how to apply the liquidity premium during valuation. For
predictable liabilities with fixed cash flows (e.g. annuities) it is fairly obvious how to apply it it is added to the risk-free rate and the cash flows are discounted at this new illiquid risk-free
rate. However, it is not as obvious how to apply for less predictable market-contingent cash
flows, especially when stochastic approaches are required. Here you need to consider how
the liquidity premium should be applied in projecting future profits and the choice of the
appropriate discount rate. In our experience there a number of firm-specific considerations
that need to be taken into account.
4.
Insights December 2009
http://www.barrhibb.com/documents/downloads/Research_Note_Measure_of_Liquidity_of_
Insurance_Liabilities_DP.pdf
www.barrhibb.com
Disclaimer
Copyright 2011 Barrie & Hibbert Limited. All rights reserved. Reproduction in whole or in part is prohibited
except by prior written permission of Barrie & Hibbert Limited (SC157210) registered in Scotland at 7 Exchange
Crescent, Conference Square, Edinburgh EH3 8RD.
The information in this document is believed to be correct but cannot be guaranteed. All opinions and estimates
included in this document constitute our judgment as of the date indicated and are subject to change without
notice. Any opinions expressed do not constitute any form of advice (including legal, tax and/or investment
advice).
This document is intended for information purposes only and is not intended as an offer or recommendation
to buy or sell securities. The Barrie & Hibbert group excludes all liability howsoever arising (other than liability
which may not be limited or excluded at law) to any party for any loss resulting from any action taken as a result
of the information provided in this document. The Barrie & Hibbert group, its clients and officers may have a
position or engage in transactions in any of the securities mentioned.
Barrie & Hibbert Inc. and Barrie & Hibbert Asia Limited (company number 1240846) are both wholly owned
subsidiaries of Barrie & Hibbert Limited.
Insights December 2009
www.barrhibb.com
Disclaimer
Copyright 2009 Barrie & Hibbert Limited. All rights reserved. Reproduction in whole or in part is prohibited except by prior written permission of
Barrie & Hibbert Limited (SC157210) registered in Scotland at 7 Exchange Crescent, Conference Square, Edinburgh EH3 8RD.
The information in this document is believed to be correct but cannot be guaranteed. All opinions and estimates included in this document constitute our judgment as of the date indicated and are subject to change without notice. Any opinions expressed do not constitute any form of advice
(including legal, tax and/or investment advice).
This document is intended for information purposes only and is not intended as an offer or recommendation to buy or sell securities. The Barrie &
Hibbert group excludes all liability howsoever arising (other than liability which may not be limited or excluded at law) to any party for any loss resulting from any action taken as a result of the information provided in this document. The Barrie & Hibbert group, its clients and officers may have a
position or engage in transactions in any of the securities mentioned.
Barrie & Hibbert Inc. and Barrie & Hibbert Asia Limited (company number 1240846) are both wholly owned subsidiaries of Barrie & Hibbert Limited.
www.barrhibb.com
Fly UP