Bank of England discussion paper: Procyclicality

Michael Henderson contributed to the writing of this blog

The full name of this paper is “Procyclicality and structural trends in investment allocation by insurance companies and pension funds: A Discussion Paper by the Bank of England and the Procyclicality Working Group”. The paper seeks to examine whether, and if so why, insurance companies and pension funds invest procyclically. We shall focus on the outcomes relating to insurance companies, which is primarily applicable to life offices.

What is procyclicality?

Good question! The paper defines procyclicality with reference to both the short term and medium term:

• Short term – the tendency to invest in a way that exacerbates market movements and contributes to asset price volatility.
• Medium term – the tendency to invest in line with asset price and economic cycles, so that willingness to bear risk diminishes in periods of stress and increases in upturns.

Broadly, there is a focus on three different potential underlying causes of procyclicality. A summary of the outcomes from the paper in these areas are given below.

Longer term product trends

The paper recognises that there has been a significant shift in the life insurance industry towards unit-linked products, where the policyholder bears all of the investment risk. There has been a move away from products where the investment risk is either shared between the policyholder and insurer (with-profits products) or entirely borne by the insurer (non-profit non-linked products). The different product types exhibit different asset allocation changes in response to market conditions. The charts below show this, using data between 1997 and 2011 for unit-linked and non-linked products:

It can be seen that non-linked liabilities had particularly large shifts out of equities and into bonds during the dotcom and credit crunch crises, whereas this was not as pronounced for unit-linked liabilities. Indeed, unit-linked liabilities exhibit some countercyclical characteristics, particularly in relation to government bonds. As a result, this long term product trend may help to reduce procyclical behaviour.

Herding

The paper considers that there is a tendency for insurers (and pension funds) to change investment or asset allocation strategies at the same time, causing “aggregate procyclicality”. Potential reasons for this include:

• similar liability characteristics lead to similar investment strategies;
• benchmarks used in investment mandates encourage asset managers to invest in line with the rest of the industry;
• a (relatively) small number of investment consultants influence asset allocation decisions (more relevant for pension funds); and
• regulatory constraints are uniform and influence both strategic asset allocations and reactions to market events (discussed in more detail later).

It also notes a recent trend towards defining risk in terms of volatility rather than amount of potential loss. Assets with low volatility are often viewed as the safest. However, volatility tends to decrease in market upswings before rising during downturns. This reliance on volatility measures may therefore cause procyclical behaviour in response to market events.

Regulatory requirements

The paper acknowledges that risk-based capital requirements and mark-to-market regimes may encourage procyclicality and that this may exacerbate market volatility. In particular it notes that, under the ICAS (Pillar II) regime, selling lower-rated bonds and buying higher rated bonds might reduce capital requirements by more than it increases liability valuations, incentivising a procyclical “flight to safety” during times of stress.

However, many regulators, including the FSA (as was) in the UK, relaxed their approaches in the face of market turmoil to avoid large volume asset sales, as shown by the table below:

The flexibility offered was likely to have lessened procyclical behaviour but, as it has only been applied on an ad-hoc basis, the application has been asymmetric across firms. The paper argues that the lack of a clear policy during times of stress causes uncertainty and that a more formal relaxation of regulation during periods of stress would be preferable. This may be combined with a strengthening of requirements during period of economic stability to improve policyholder security.

Conclusion

The paper’s primary recommendation is that consideration be given to regulatory regimes that are countercyclical and hence build up resilience during benign economic times, enabling constraints to be relaxed during tougher periods. These would form part of a “clearly defined framework” allowing all firms to understand when relaxations may be applied, reducing undesirable uncertainty.

Whilst a regulatory regime with such flexibility would be seen as a very positive step by most in the industry, whether this could happen in practice remains to be seen. Solvency II is just around the corner and consistency with it is clearly a requirement. In addition, EIOPA have been keen to prevent any “gold-plating” by national regulators, keeping a level playing field across the European Union. Bearing this in mind, it would appear that European regulators would also have to be persuaded of the merits of this approach before it could be implemented.

The full paper can be found here: http://www.bankofengland.co.uk/publications/Pages/news/2014/104.aspx