Initialize Analysis
Choice Dividex predictive analytics dashboard visualising market entry data
AI-Optimised Entry Timing

Predictive modelling for disciplined entries into GBP, USD and EUR-denominated assets

Choice Dividex applies automated dollar-cost averaging optimisation to irregular, multi-currency income, reducing the exposure that comes from manual timing decisions and emotional reaction to short-term volatility.

Initialize Analysis View Documentation
3Base currencies supported at launch
24/7Continuous data ingestion
0Manual entry decisions required
GBPReporting currency default
The Friction

Irregular income and manual timing compound each other

Remote professionals paid in multiple currencies face a dual problem. Income arrives on inconsistent schedules, often converted at inconsistent rates, while market entry decisions are made manually, under time pressure, and frequently influenced by recent price movement rather than underlying conditions.

The result is a pattern of reactive buying: entries clustered around emotionally salient price points rather than distributed according to any defined strategy. This is a measurable source of avoidable drawdown, independent of whether the broader asset allocation itself is sound.

Core Technology

How predictive entry logic and DCA optimisation work together

01 / Signal Layer

Predictive Modelling of Short-Term Price Dispersion

The platform ingests historical and streaming price data for supported assets and currency pairs, then models the probable range of near-term price dispersion rather than attempting to forecast a single future price. This produces a confidence band used to classify current conditions as favourable, neutral, or unfavourable for a scheduled entry.

input: price series, volatility index, FX spread
process: rolling volatility model → confidence band
output: entry_favourability_score (0–100)
02 / Execution Layer

Dollar-Cost Averaging Optimisation

Instead of fixed-interval purchases, Choice Dividex adjusts the size and timing of each scheduled contribution within a bounded tolerance, shifting marginal weight toward windows with a higher entry-favourability score. The total committed capital over a period stays the same; only its distribution across entries changes.

rule: total_contribution fixed per cycle
adjustment: ± weighting within tolerance band
objective: lower average cost basis, bounded deviation
03 / Risk Layer

Asymmetric Risk Filtering

Before any adjusted entry executes, a separate filter checks for conditions associated with asymmetric risk — situations where potential downside materially outweighs potential upside given current volatility. When detected, the system defaults to the standard scheduled contribution rather than an increased one.

check: volatility spike, liquidity gap, correlation break
default: standard contribution on flag
log: decision recorded for later review
Methodology

A transparent process, without claims that cannot be shown

Choice Dividex does not publish testimonials or performance guarantees, since past model output cannot establish future results. What follows is a description of the three operational stages that every scheduled entry passes through.

Step 01

Data Ingestion

Market price feeds, volatility measures, and FX rates for GBP, USD and EUR pairs are collected continuously and normalised into a common time series before any modelling begins.

Step 02

Predictive Modelling

The normalised data is scored for entry favourability using a rolling statistical model, producing a band of expected short-term dispersion rather than a single predicted price.

Step 03

Automated Execution

Scheduled contributions are executed with weighting adjusted inside a pre-agreed tolerance range, and every decision — including when the system defers to a standard contribution — is logged for review.

Application

Built around the realities of multi-currency, remote income

Portfolio Diversification

Scheduled contributions can be split across asset classes and currency exposures, with entry-favourability scoring applied separately to each, so diversification does not depend on manually timing several markets at once.

Currency Risk Hedging

For income earned in USD or EUR and ultimately reported in GBP, the platform accounts for FX spread and volatility when scoring entries, reducing the compounding effect of poor conversion timing on top of poor market timing.

Long-Term Wealth Accumulation

Because the model adjusts distribution rather than total contribution, the approach is designed for a multi-year horizon where consistency of process matters more than any single entry's outcome.

Choice Dividex analyst reviewing predictive model output on a data terminal
About the Platform

Data-driven autonomy for people managing their own capital

Choice Dividex was built for a specific situation: income that arrives irregularly, in more than one currency, managed by someone without the time to monitor markets throughout the working day. The platform does not replace financial advice and does not promise a particular return. Its function is narrower and more specific — to apply a consistent, logged, rules-based process to the timing of scheduled contributions, reducing the influence of ad-hoc manual decisions.

Every adjustment the system makes stays inside a tolerance range agreed in advance, and every decision, including deferrals triggered by the risk filter, is recorded and available for review.

Technical FAQ

Data security, latency, and model accuracy

How is account and market data secured?

Data in transit is encrypted, and stored records are held on access-controlled infrastructure. No credentials required for execution are stored in plain text, and the platform's decision log is retained separately from personal account data.

What latency exists between signal generation and execution?

Entry-favourability scoring runs on a rolling basis rather than tick-by-tick, so there is an inherent short delay between a market condition and a scheduled contribution being weighted accordingly. This delay is a deliberate design choice: the model targets short-term dispersion, not microsecond price movement, so sub-second latency is not the relevant benchmark.

How accurate is the predictive model, and how is that measured?

The model is evaluated on its ability to classify entry conditions correctly against historical data, not on predicting exact future prices. No fixed accuracy percentage is published, because performance varies with market regime; the decision log exists specifically so users can review how scoring aligned with subsequent price behaviour over their own chosen period.

What happens if the model is uncertain or data feeds are interrupted?

If a data feed is degraded or the confidence band cannot be calculated reliably, the system defaults to the standard, unweighted contribution for that cycle rather than acting on incomplete information.

Can the tolerance range for DCA adjustment be changed?

The tolerance band that bounds how far a contribution can be weighted up or down is set during onboarding and can be revised, but changes take effect from the next scheduled cycle rather than retroactively.

Is Choice Dividex regulated as a financial adviser?

Choice Dividex provides a data-driven execution process, not personalised financial advice. Users remain responsible for their own asset allocation and investment decisions; the platform's role is limited to the timing and weighting of pre-agreed scheduled contributions.

Next Step

Review the methodology in detail before committing any capital

A platform demonstration walks through the entry-favourability model, the DCA tolerance mechanism, and the decision log, using historical data rather than a live account.

Explore the Model Request a Demo

Choice Dividex does not guarantee investment returns. Predictive models describe probable ranges based on historical data and cannot eliminate market risk; all figures and scores are decision-support tools, not forecasts of certain outcomes.