Ränthetsen combines AI-powered predictive models with community-verified performance logs, so capital allocation decisions can be grounded in validated data rather than assumptions.
Most smaller companies manage excess liquidity based on quarterly reconciliations and gut feeling. Between these occasions, an opportunity cost arises: capital that is neither actively working nor protected against volatility. The decision-making material has time to become out of date before it is even used.
The problem is rarely a lack of data, but a lack of structure to interpret it in time. The more variables that have to be weighed together manually, the greater the risk of important signals being lost in the crowd.
Ränthetsen was developed based on a simple assumption: capital allocation recommendations should be verifiable in retrospect, not just accepted in advance. Therefore, the platform is structured around two interconnected parts – an analysis engine that processes financial data, and a public log that shows how well previous recommendations have actually held up.
The methodology is aimed at companies that manage their own excess liquidity and that want to make decisions based on validated data rather than individual forecasts. The platform does not replace a financial manager, but functions as a structured decision support in the ongoing work.
The engine processes large volumes of historical and ongoing financial data to identify patterns that are statistically significant, not coincidental. The result is recommendations adapted to each company's liquidity profile and risk framework.
Historical cash flows and market data are combined in predictive models that are updated continuously to identify trends before they become apparent in quarterly reports.
Algorithmic optimization is used to balance return potential against exposure, with clear limit values adapted to the company's risk appetite.
Recommendations are continuously adjusted as new data comes in, allowing decisions to keep up with changes in liquidity rather than lag behind them.
Each recommendation generated by the platform is time-stamped before the outcome is known. The result is then published in an open log, where deviations between forecast and actual outcome can be followed over time.
Existing flows for accounting and liquidity are connected without replacing current systems. Historical data is used to calibrate the starting position.
The predictive model is adapted to the company's specific goals, time horizon and risk tolerance before any recommendations are made.
Recommendations are presented together with underlying reasoning, so that the responsible person can make the final decision with full transparency.
All financial data is processed in accordance with GDPR, with encryption both during transmission and storage. Only the data points required for the analysis are collected, and the company retains full control over which sources are connected. No data is passed on to third parties without express consent.
The models are based on a combination of time series analysis and machine learning, validated against historical data before deployment. Accuracy is continuously tested against real outcomes, which is why the result log is updated regardless of whether the forecast was correct or not.
The cost of the platform should be set against the opportunity cost that arises when capital remains idle or misallocated for extended periods. Because each recommendation and its outcome is documented in the public log, it is possible to assess return on cost based on actual history rather than assumptions.