How AKSJENET calculates and presents analysis

1. Data basis and coverage

Market data is synchronized with Oslo Bors-related datasets and stored with the date of the latest available observation. Stock pages may include closing prices, price changes, history, company data, sector, dividends, and calculated fields. Data is normally delayed or end-of-day unless explicitly stated otherwise.

A stock is included in the automated sitemap only when it is active, has a valid positive price, and has recent historical data within the system’s permitted freshness window. The dates displayed on a page matter more than the time at which the page is opened.

2. Scores, signals, and ranking

The system calculates separate model fields for short, medium, and long horizons. The market overview uses the medium-horizon score as the main basis for relative ranking. A rank is the stock’s position against other active Oslo Bors stocks in the same dataset.

BUY, HOLD, and SELL are categories derived from scores and model rules. They describe the model’s current state, not the probability of making a profit. Users should review price action, liquidity, company events, valuation, their own risk limits, and the data date before acting.

3. Model portfolios

A model portfolio follows a defined mandate, risk profile, selection model, rebalancing rule, and benchmark. New runs may keep, increase, reduce, buy, or sell positions according to the rules for that model.

Holdings, weights, transactions, and run history are displayed so that users can examine what changed. Model portfolios are examples and research, not personal recommendations.

4. Return, costs, and benchmarks

Returns are calculated from recorded portfolio values, transactions, cash flows, and available market prices. When a page says performance is gross and pre-tax, personal taxes, individual trading terms, and all of a user’s actual costs have not been deducted.

A benchmark is a comparison series, not an investable alternative with identical risk. Currency, dividend handling, start dates, cash allocation, and data availability can affect the comparison.

5. Automation and written summaries

Data points, tables, charts, signal labels, and some explanatory text may be generated automatically. Automated summaries describe structured fields and may not contain new information that is not yet present in the dataset.

Structured facts and visible dates should take priority if a written summary appears to conflict with a number. Report discrepancies so they can be investigated.

6. Freshness and quality controls

• Active tickers are checked against prices and the latest historical observation before sitemap publication.
• Pages show relevant score, data, or run dates when available.
• Invalid, halted, or stale data series can be excluded automatically.
• Automated tests check canonical URLs, language annotations, XML output, and key presentation rules before publication.

7. Limitations

• Past performance does not determine future performance.
• Models can react late, overstate patterns, or fail during rapid regime changes.
• Company news, issuance, trading halts, currency, and provider errors can make data temporarily incomplete.
• Small and illiquid stocks can have larger differences between calculated and achievable prices.

8. Corrections and method changes

Confirmed material errors are corrected in the data, code, or content according to their cause. The method may change when data quality, models, or product features improve. The date above records the latest material update to this explanation.