FUND RISK RATINGS

What's in a name?

FUND RISK RATINGS

Alex White
Head of Research
Asset & Liability Management

“What’s in a name? That which we call [risk] by any other name would [cause a headache].”

Admittedly, I don’t think Shakespeare could ever have imagined his famous words being shoehorned into an article about types of fund risk ratings. But in this case, there’s a lot in a name. Does ‘risk’ mean volatility? Is it backward-looking or forward-looking? Is it Value at Risk? Or is it based on historical stress tests? You get the picture; there’s much at play beneath the umbrella term. For LGPS funds, risk is a crucial consideration in investment decision-making. But the type of risk metric used and the extent to which it drives decision-making can make a material difference to outcomes. It’s essential to consider – and manage – fund risks to avoid being blown off track. Suppose the chosen risk metric is 3-year rolling volatility, and a manager is buying high-yield bonds. The overall volatility since 1999 has been 10-11%, so let’s suppose the manager will reduce their allocation when rolling volatility reaches 12% – what would that have done to performance? Left alone, excess returns on the index were 4.1%; but by reducing allocations after a fall (i.e., when spreads were higher), the returns fall to 3.2%, losing 0.9% a year, or c.20% of the returns. And while the risk is lower (9.4% against 10.3%), the difference in the risk reduction is smaller than the returns lost and the Sharpe ratio is worse.

FUND RISK RATINGS

What's in a name?

The following table highlights some of the trade-offs involved with different metrics. What jumps out is that all metrics have plenty of flaws, so the best way to build a robust portfolio is to use multiple.

At Redington, we typically start with Value at Risk (VaR). This is a statistic that quantifies the possible losses within a portfolio or position over a specific timeframe. Our calculation measures the loss that would be incurred in a worst-case scenario over the next year; the worst case is often assumed to be a 1-in-20 event (VaR 95). We think it’s critical to look at this alongside other risk lenses like stress tests, sensitivity analysis and long-term projections.

Look deeper at your chosen methodology, challenge the assumptions and, if you need any help determining “what’s in a name” when it comes to your portfolios and outcomes, we’re here to help.

FUND RISK RATINGS

What's in a name?

	Advantages	Disadvantages
Stochastic Measures	Captures the whole fund risk rather than an isolated element Can draw out key risk factors Generally assumes what goes up can come down	Requires judgment, so is subjective Sensitive to time horizon used for calibration Sensitive to underlying assumptions, especially correlation Harder to calculate
Historical Volatility	Objective Assumes what goes up can come down Easy to calculate	Can lead to poor timing decisions (e.g. selling credit just after a fall) Can easily understate risk in a period of calm before a crash Short horizons do not provide meaningful information, especially for tail risk measures
Historical VaR and Max Drawdown	Broadly objective Easy to calculate	As above, but even more sensitive to horizon Doesn’t assume gains can be reversed, so a lack of crashes in the period investigated can give a false sense of security
Historical Stress Tests	Broadly objective Easy to calculate	Only tests for specified scenario – doesn’t account for other possible combinations of risks Doesn’t account for how likely scenarios are
Sensitivity Tests	Broadly objective Isolates key factors which may be hidden or diversified in other metrics (e.g. FX risk) Useful for limiting single bets	Only tests for specified risks Does not account for how likely scenarios are