Investing with rules, evidence and a clear method that is 100% transparent and liquid
The model process below explains the research, portfolio construction and governance behind Arima's rules-based portfolios.
Arima Investment Partners manages client capital using a disciplined, repeatable process grounded in financial science. The method has three layers, and each one sits on top of the one before it.
Layer 1 — Four client protections. The structural conditions any portfolio must satisfy before we discuss returns at all.
Layer 2 — Four components of a robust investment process. The engineering of the portfolio itself: signal generation, optimisation, rebalancing, risk management.
Layer 3 — Modelling, execution and ongoing governance. How the process is tested before capital moves, how it is executed, and how it is reported and refreshed.
The name is deliberate. The firm is ARIMA Investment Partners, and the name derives from ARIMA (Auto Regressive Integrated Moving Average) — a statistical, mathematical model used to forecast time series — applied in fields as varied as weather forecasting, disease spread, and population change. Financial markets are a time series: prices over time, data over time. The name states the method.
These four principles were agreed by the founders before the firm existed, and every Arima portfolio is built to satisfy all four simultaneously.
1. Direct beneficial ownership
Assets are held in the client's beneficial ownership. Client money does not sit in a manager's bank account, and client wealth is not represented by a claim on someone else's structure.
Why it matters: the most damaging failures in Australian wealth management — the Caddick matter among them — were failures of ownership and transparency, not of investment skill.
2. A rules-based approach
Investment decisions are governed by a defined, predefined set of rules applied consistently, rather than by discretion, conviction or market narrative.
Why it matters: a written process can be tested, monitored, audited and improved. Discretion cannot.
3. Full transparency of all investments
Clients can see exactly what they own, line by line, including the banking arrangements behind the portfolio.
Why it matters: transparency is the control that makes the other three protections verifiable rather than promised.
4. Genuine financial flexibility
Portfolios are built from liquid, marked-to-market assets that trade in high volume every business day and settle back to cash in two days.
Why it matters: a stated portfolio value is only meaningful if it can be realised. Many assets carry a value on a statement that cannot be converted at that value on the day it is needed. Investors can discover this at the worst possible moment.
Every investment requires an impetus — a reason, derived from evidence, to commit capital to an asset or asset class. Signal generation is the function that produces it.
Arima draws on academic research, practitioner research and in-house modelling on the factors that drive long-run returns. The factor set includes:
Valuation
Momentum
Quality
Size
Macroeconomic factors
Market Behaviour
Finding a robust combination is a non-trivial problem. Evaluating all combinations of factors requires substantial data science and data analytics work to isolate signals that are robust and scalable over time rather than artefacts of a particular sample.
Method applied: statistical and mathematical modelling, including formal hypothesis testing — the deductive process used to draw conclusions that are grounded in evidence and hold up under scrutiny.
Once signals are established, the portfolio must be optimised — a mathematically non-trivial exercise.
The optimal number of holdings sits somewhere between one and infinity. An investor should never hold one investment, because that is pure concentration. An investor also cannot hold an infinite number. The optimal count lies within a very wide range, and finding it is a modelling problem, not a preference.
Critically, there is no single optimisation function. There is a catalogue of them, including:
Mean-variance optimisation
Black-Litterman
Volatility scaling
Risk parity
Equal weighting
Neural-network-based- optimisation
The key insight: optimisation is an active choice. Even an investor who decides to be entirely passive is still making an active decision about which optimisation model allocates capital across those passive holdings. There is no way to avoid the choice — only to make it deliberately or by default.
Arima applies a range of models rather than committing to one, selecting for robustness and performance together.
There are two types of rebalancing, and both have demonstrated efficacy.
Time-based rebalancing — rebalancing on a schedule: monthly, quarterly, semi-annually or annually, to keep the portfolio aligned with the client's long-term goals, objectives and risk profile.
Threshold-based rebalancing — rebalancing when an allocation drifts beyond a defined band, for example plus or minus 10%.
Which one is optimal, whether both should operate together, and what the optimal interval and threshold actually are, is treated as a data science exercise rather than a matter of convention. Arima runs its models monthly.
The most important component.
Risk and return are two sides of the same coin: some level of risk must be accepted to obtain a given level of return. The work is in understanding precisely which risks are being taken and how each can be mitigated.
This matters most in recessions, crises and periods of acute market stress. Two questions drive the analysis:
1. What kinds of risk is the portfolio actually carrying, and which of them can be mitigated?
2. Which asset classes could benefit during a financial crisis or acute market stress?
Before capital is committed, the strategy is modelled across risk profiles from conservative through to high growth, and simulated over time.
The simulations confirm the fundamental relationship: moving from conservative toward high growth should produce higher expected returns, and also higher volatility and higher risk over time. Risk can be mitigated, but the relationship holds.
A single back-test proves very little. A proper scientific process runs out-of-sample tests, repeatedly, to establish that the strategy is robust rather than fitted to history.
Strategies are assessed on a full metric set, not a single headline number:
Returns
Risk
Winning Percentage
Risk-adjusted returns
Drawdown Profile
Implementation is performed using technology, so the program that was tested is the program that runs.
Portfolios are monitored and rebalanced on a regular basis. Research is continuous: when credible new research reaches the market, Arima obtains it, evaluates it, and where it holds up, implements it for client benefit.
Because the approach is rules-based, the rules themselves can change — but only when the evidence says they should.
Clients receive a monthly report covering:
Any changes made to the portfolio, and the reasoning behind each
Performance data updates
Economic events that may be affecting the portfolio, positively or negatively
Clients who want to go further can call and have the detail explained. Ongoing education is treated as part of the service, on the principle that clients should understand what they hold and why.
Arima deliberately removes reliance on forecasting variables that are extremely difficult to forecast consistently — interest rates, oil prices, macroeconomic events.
The alternative is to be adaptive and dynamic rather than predictive: run the models monthly, measure, and adapt to what the numbers are showing. A process that responds does not need to be right about next year.
This is why the method is deliberately not short-term trading. It is long-term investing, appropriate only for investors with a long-term timeframe.
Diversification at Arima means ensuring the portfolio does not depend on a single return driver.
The technical requirement is that signal generation is orthogonal — genuinely independent, not correlated with or modelled off return drivers that behave similarly. A portfolio holding many positions that respond to the same underlying driver is concentrated, regardless of how many line items it shows.
Diversification across multiple investments and asset classes produces a steadier return path over time, is more robust than dependence on any single asset class, and keeps portfolios fit for the compliance obligations the firm carries for its clients.
The method is not novel in its philosophy, and that is a point in its favour. Some of the most respected investment managers in the world are scientific and quantitative — Renaissance Technologies, AQR, Two Sigma, Citadel among them — using scientific discovery in one form or another to drive better outcomes and reduce the influence of speculation and cognitive bias.
Three options are available, matched to client risk profile. Figures to 30 June 2026.
Past performance is not a reliable indicator of future performance. Investments can fall in value, and returns are not guaranteed. Performance should be assessed with its fee basis, inception dates and methodology, not in isolation.
Portfolio Option
Balanced
Growth
High Growth
1 Year return
10.94%
14.19%
19.18%
Since-Inception (p.a.)
12.2%
15.99%
18.52%
Balanced
1 Year Return: 10.94%
Since Inception (p.a.): 12.2%
Growth
1 Year Return: 14.19%
Since Inception (p.a.): 15.99%
High Growth
1 Year Return: 19.18%
Since Inception (p.a.): 18.52%
Asset Class
Australian Shares
Global Shares
Defensive assets
Cash
Balanced
31.3%
31.6%
34.1%
3.2%
Growth
39.1%
41.8%
15.4%
3.7%
High Growth
43.2%
46.4%
7.7%
2.7%
Balanced Asset Allocation
Australian Shares: 31.3%
Global Shares: 31.6%
Defensive assets: 34.1%
Cash: 3.2%
Growth Asset Allocation
Australian Shares: 39.1%
Global Shares: 41.8%
Defensive assets: 15.4%
Cash: 3.7%
High Growth Asset Allocation
Australian Shares: 43.2%
Global Shares: 46.4%
Defensive assets: 7.7%
Cash: 2.7%
Past performance is not an indicator of future performance.
Stage
Structural protections
Signal generation
Optimisation
Rebalancing
Risk management
Validation
Execution
Governance
Question Answered
Does the client own it, see it, and can they access it?
Why invest in this at all?
How much of each, and how many holdings?
When do we adjust?
What can go wrong, and what protects us?
Does it actually work?
Does the live portfolio match the tested one?
How does it stay current and accountable?
Method
Direct beneficial ownership, rules-based process, full transparency, liquid mark-to-market assets
Multi-factor research — valuation, momentum, quality, size, macro, market behaviour; hypothesis testing
A catalogue of optimisation models, selected for robustness; treated as an active choice
Time-based and threshold-based rebalancing, intervals set by data science; models run monthly
Risk identification and mitigation, volatility controls, position sizing, rotating defensive sleeve including gold
Out-of-sample testing, full metric suite including drawdown and risk-adjusted returns
Technology-driven systematic execution
Monthly reporting with reasoning, continuous research, ongoing client education, transparent aligned fees
Investment values and income can fall. Diversification, rebalancing and a rules-based process cannot prevent losses or ensure outperformance. Models depend on assumptions and data, and relationships observed in the past may not continue.
Back-testing and simulations are not the same as actual client returns. Fees, taxes, transaction costs and the timing of investments can affect individual outcomes. Liquidity, custody and access also depend on the asset, account and platform; superannuation access rules apply where relevant.
Consider your objectives, timeframe, capacity for loss and the relevant advice and account documents before deciding to invest.
Understand the approach, talk with the Arima team, then consider a recommendation based on your circumstances. You do not need to choose a model or account structure before the first call.
The firm is Arima Investment Partners. The name derives from ARIMA, (Auto Regressive Integrated Moving Average) a statistical, mathematical model used to forecast time series into the future. It is applied in many fields — forecasting weather patterns, the potential spread of disease, the growth or decline of a species — anything that generates data over time.
Financial markets are a time series: prices over time, data over time. That makes the model a natural fit for the firm's purpose, which is using a systematic, scientific, evidence-based approach to support disciplined investment and retirement decisions for clients.
The three founders worked under the same roof at a large private wealth advising firm, running their own businesses. After sitting through a product pitch in a morning meeting that neither of them wanted any part of, they had their first real conversation on the way out — and discovered they already agreed on four fundamentals of investing: direct beneficial ownership, a rules-based approach, full transparency of all investments including the banking, and the financial flexibility that genuinely liquid mark-to-market assets provide.
Their shared assistant, Renato, then made the introduction that added the quantitative side, and identified Openmarkets Group as a licensee with the technology stack to run Managed Discretionary Accounts. Arima was effectively born that afternoon.
Daniel Lalabalavu holds a Master of Applied Finance, a Bachelor of Business Administration in Economics, a BPhil in Statistics and a Graduate Diploma in Financial Planning. His experience and formal training inform the investment process.
Read Daniel's professional biography and Michael Holmes' biography for more about the Arima team's background.
Arima Investment Partners Pty Ltd is an Authorised Representative of Openmarkets Australia Limited (ABN 38 090 472 012, AFSL 246705), a Market Participant of ASX, Cboe Australia and NSX, and an ASX Clear and Settlement Participant (Arima Investment Partners).
An MDA allows investment decisions to be made on your behalf within an agreed investment programme. It is not a trust type or a guarantee of performance. Arima's models are delivered through an adviser-assisted MDA process, with the relevant advice and agreement considered before implementation.
Understand the scope, costs, risks, reporting and termination terms before deciding whether to proceed. Read more about managed discretionary accounts at Arima or schedule a call with the Arima team.
Level 15, 388 George Street, Sydney NSW 2000. +61 489 088 819.
Video meetings available nationally or for Australians overseas..
© 2026 Arima Investment Partners Pty Ltd. All rights reserved.
Authorised Representative of Openmarkets. Level 15, 388 George Street Sydney NSW 2000. Openmarkets Australia Limited (ABN 38 090 472 012, AFSL 246705) is a Market Participant of ASX, Cboe Australia and NSX, and an ASX Clear and Settlement Participant. Legals. Privacy policy. The information provided on this website is general in nature and does not take into account your personal objectives, financial situation, or needs. You should consider whether the information is appropriate to your circumstances before acting on it. We recommend seeking professional advice tailored to your individual situation before making any financial decisions.