Advanced Optimizer

Our patented methodology has been peer reviewed by academic institutions and global investment bank quantitative teams

PROPRIETARY OPTIMIZER

The Axona optimizer utilizes a proprietary algorithm that has been in commercial use for over 20 years.

The primary goal of the algorithm is to produce highly diversified portfolios that are more robust and less sensitive to estimation errors.

DIVERSIFIED

The Axona proprietary algorithm ensures that portfolios are well diversified, and generally have 2-3x more assets than typical Industry Standard optimizers which tend to only allocate across a few assets

PREDICTABLE & ROBUST

The optimizer generates predictable portfolio forward expected returns and risks. Even if the input estimates are incorrect, the portfolio will still behave in a predictable manner

OPTIMAL

The algorithm ensures the portfolio will exhibit the best possible tradeoff between a robust portfolio with known risks, whilst achieving the best possible probability of outperformance

MULTI VARIATE ANALYSIS

Multi variate representation of risks, returns & correlations leads to well diversified, intuitive and robust portfolios.

ASYMMETRICAL RETURNS

Asset returns are not symmetrical or normally distributed as represented by most models.

Real world representation is messy and does not fit neatly into traditional optimization models.

The Axona optimizer takes into account this asymmetry.

RISK

Volatility is not a good measure of Risk - it is unintuitive, and does not encapsulate the full danger to the portfolio. Risk is the probability of achieving underperformance, and is a multivariate parameter.

Axona supports multiple risk models, including semi-variance, VaR, and CVaR.

CORRELATIONS

Axona takes the full range of rolling correlations into account, thus ensuring our portfolios are more resilient to significant market events.

CORRELATIONS

Correlations are integral to a portfolios behaviour, and key in the ability to reduce the overall portfolio risk.

Axona uses in depth analysis on correlations to ensure no hidden risks are incorporated in the correlations used

DYNAMIC CORRELATIONS

The behaviour between asset pairs does not remain constant, and any model that is attempting to reflect the future reality needs to incorporate this non-static nature of asset pair behaviour. Axona uses non-static correlation coefficients in its algorithm

ROLLINGS CORRELATIONS

Axona uses a rolling correlation coefficient over the allocation term to quantify the changing correlation

CALIBRATE

The Axona model allows the user to calibrate the correlation to a conservative or aggressive setting, time weight and even over-ride the default correlation

RULES & CONSTRAINTS

The investment world we operate in is subject to complex rules and regulations which need to be reflected in the risk models.

The Axona optimiser supports numerous complex pre-defined and ad-hoc constraints

INDIVIDUAL & GROUP ASSET CONSTRAINTS

Minimum Weight
Maximum Weight
Specific Weight
Range

PORTFOLIO
CONSTRAINTS

Statistical
- eg Sharpe > 1.5
Fundamental
- eg PE > 8.0
Classifications
- eg Consumer Cyclical < 30%

ARBITRARY & COMPLEX CONSTRAINTS

Ratios between asset weights
- eg AAPL = 3 x GOOG
Factors
- eg 1xAAPL + 2xGOOG < 30%