Controlling risk with ILGs
In this conference for institutional fund managers and trustees, Stuart provided insights from private wealth management, where most investment objectives are defined in terms of purchasing power at future known or uncertain horizons. This being the case, inflation risk is one of the most important sources of risk to manage. Traditional ‘balanced management’, which is focused above all on short-period nominal return volatility, does not address this risk well and the main instrument of risk control, bond exposure, involves an unrecognised or unstated substitution of equity risk by inflation risk which is highly inefficient. Stuart explained how Fowler Drew uses two assets, diversified equity exposure and horizon-matched index linked gilts, to control the range of outcomes in purchasing power at the planning horizon.
This audience was familiar with the concept of ‘portfolio separation’ and using hedges, rather than constantly trying to find more (and less correlated) risky asset classes or strategies, to control risk. It is the basis of Liability Driven Investing (LDI) that has so successfully challenged traditional diversification, or ‘balanced management’, in UK occuaptional pension funds. But Stuart suggested that applying LDI successfully assumes you have correctly defined the nature of the liabilities, the hedging assets and the right priorities for different risk sources. Whereas for private wealth there is a clear definition for purchasing power objectives, and a clear risk free asset in the form of index linked gilts, the typical LDI hedging solution relies excessively on non-indexed bonds (because of the prescribed ways they measure liabilities) and the typical risky portfolio solution focuses excessively on smoothing return paths (which sponsors care about) at the risk of poor matching of purchasing-power outcomes (which is a problem for the next generation of trustees or managers). Whereas individual plans may be acting rationally, the society-wide effects are not ;at all rational.
Stuart also challenged this institutional audience to consider why, when most actuarial consultants believe in mean reversion of equity returns (which is implicit in their assumptions about the persistence of common, ‘normal’ means for real returns) , they do not make their future return assumptions specific to current market conditions and asset values or indeed specific to the planning time horizon. Both dependencies are part of a mean reverting return model whereas so-called ‘normative’ assumptions, blind to time and deviations from trend, clearly are not. Fowler Drew’s own model of real returns, also built on reversion to sustainable trend returns, does not rely on normative assumptions.
Though we have learnt much from the institutional market, perhaps we also have something to teach it!