Mürzenvik runs real-time predictive models against market, portfolio and operational data, giving independent investors and business leaders in Australia a quantified view before conditions change.
Interface preview: a single dashboard combining live data ingestion, model output and a plain-language recommendation, updated as new data arrives rather than on a fixed reporting cycle.
Speed to Value
Most analysts who work independently or across multiple engagements do not have the luxury of a dedicated data team. Cleaning spreadsheets, reconciling sources and calibrating a model by hand can consume the better part of a working day before any analysis actually begins.
Mürzenvik removes that preparation step. You connect a data source or select from a pre-built set of Australian market feeds, and the platform structures the data, selects an appropriate model and returns a first output while you are still reviewing your brief.
Core Capabilities
The platform is built around three functions that work together rather than as separate tools: forecasting, risk measurement and ongoing optimisation.
Mürzenvik applies trained time-series and factor models to price, volume and macro data, producing forward-looking estimates with a stated confidence range rather than a single point figure. Outputs update as new data is ingested, not on a weekly or monthly cycle.
Exposure, volatility and correlation across a portfolio or deal are quantified continuously, with thresholds you can set per mandate. The aim is to surface concentration or drawdown risk before it shows up in a standard monthly report.
Recommended adjustments to allocation, pricing or operational parameters are generated as conditions shift, and summarised into a short report suitable for a client update or an internal decision memo.
Methodology
We describe the process openly because professional users need to understand what the model is doing before they rely on its output.
Source data — market feeds, portfolio holdings, operational metrics — is validated, normalised and time-stamped on arrival. Gaps and anomalies are flagged rather than silently filled.
Proprietary models combine the normalised data with relevant historical and sector context, weighting recent signals against longer-run patterns to avoid overreacting to short-term noise.
The system translates model output into a specific recommendation — an allocation change, a risk flag, a pricing adjustment — with the reasoning behind it stated in plain language.
Use Cases
The underlying models are the same; the configuration changes depending on the mandate.
A portfolio manager rebalancing a diversified AU equity mandate uses Mürzenvik to quantify concentration risk across sectors after a market move, and to model the downstream effect of a proposed trade before it is placed. The output is a short, time-stamped note suitable for an investment committee.
A business owner evaluating expansion into a second state uses the platform to forecast demand sensitivity under different pricing and staffing assumptions, drawing on regional economic indicators rather than a single national average.
An analyst working across several retainer clients uses Mürzenvik to produce a volatility forecast for the ASX 200 ahead of a rate decision, then adapts the same model output into client-specific commentary without rebuilding the analysis each time.
Frequently Asked Questions
Client data is held in Australian data centres to support data sovereignty requirements common in financial services. Access is restricted to the account holder and anyone explicitly invited to a workspace; Mürzenvik staff do not access client data as a matter of course, only for support requests you initiate.
The platform accepts standard spreadsheet and CSV uploads, and connects to common market data feeds and portfolio management exports. If your source format is non-standard, our onboarding guidance covers how to map it before the first ingestion run.
Each model reports a confidence interval alongside its forecast rather than a single figure presented as certain. Models are trained on historical AU and relevant global market data and are periodically re-validated against out-of-sample periods; we do not claim a fixed accuracy percentage, because performance varies by asset class and market conditions.