Why Portfolio Construction Matters
Portfolio construction is often framed as a question of what assets to include.
But that is only half the story. The other half is how we define risk, and what changes when we optimise through different risk lenses.
In this presentation, I ask a simple question:
If we are comfortable with a benchmark portfolio’s return and volatility, how much more return might be available at the same level of risk when risk is measured differently?
The starting point is the average net return of the Balanced KiwiSaver universe, converted to a gross return for a consistent, apples-to-apples comparison. The Benchmark (BMK) is based on the average Balanced KiwiSaver policy weights through time. Both are mapped to a common universe of NZD asset classes.
For those concerned about implementability, nearly all of these asset-class exposures, with the exception of cash and commodities, are available through the SMART ETF lineup. This makes the analysis not merely theoretical, but both implementable and investable.
To explore the question, I use a new deterministic search heuristic designed for constrained portfolio optimisation. It works directly with real-world allocation limits and tests portfolios across several increasingly demanding views of the return distribution.
These include:
• Volatility
• Mean absolute deviation
• Semivariance and lower partial moments
• Conditional Value at Risk
• Conditional Drawdown at Risk
• Maximum drawdown
• Higher-moment measures incorporating skewness and kurtosis
The analysis then moves beyond the static optimisation result. Each approach is tested through rolling walk-forward optimisation, with turnover and implementation costs deducted, to determine whether the apparent gains survive out of sample.
The purpose is not to declare one risk measure universally superior.It is to show that the choice of risk lens is itself an active portfolio decision, and potentially a material source of return improvement potentially in excess of the full cost structure of the fund.
For more information, contact Peter Urbani : peter.urbani@knowrisk.co.nz
Comments