QDRA Pty Ltd : QDRA Dynamic Macro

Investment Strategy

The strategy targets a beta of negative 0.5 to the S&P500 and a volatility of 16 to 18%.

Dynamically allocate risk to 50+ CTA models, using 4 key components: 1. A slow moving single variable econometric model that allocates risk based on where we are in the economic cycle for 8 countries and how the CTA models have performed in relation to this 2. A faster moving volatility of equity market volatility model that allocates based on where we are in the equity volatility cycle 3. A bear bias model that allocates risk based on where we are in the US equity market cycle 4. These 3 models, are combined to propose a range of risk allocation for each model. This is then an input to a minimum variance portfolio construction that seeks to force diversification, to avoid the concentration risk that can come from dynamically allocating risk. The minimum variance portfolio that is calculated is scaled to the volatility of a "risk parity" portfolio of the unscaled models. If the underlying models have stronger than average signals, that are also more correlated than average, then the portfolio is scaled to a relatively higher volatility. The volatility scaling allows more information from the underlying models in final portfolio.

Risk Management

Stop losses and profit take levels are set at the model level, not at the individual position level. Stop losses are also related to the amount of risk allocated to a model at the start of the month, so that a stop for a model with a large risk allocation is wider than the stop for a model with a small risk allocation. If a model trades a spread position it isn't a concern that one side of the spread is losing if the other side of the spread is making more money. What matters is the model level profitability (or loss) relative to the risk being taken, hence basing stops on the models, not on the individual positions.

Trading futures is inherently risky and there is also risk of loss.