Settings · Data
Fetching price data, editing the symbol list, the missing days report, and the company information that supplies the mass and link data of the propagation model.
Price data
What it is: Incremental and full fetch buttons, CSV import of symbols and prices, fetch progress and errors.
What data feeds it: The primary source is İş Yatırım, the backup source Yahoo Finance. Because the XU100 level and the USD/TRY rate come with every stock response, they need no separate request. The result is written idempotently to the price_daily table.
What it shows: An incremental fetch continues from each stock's last date, overlapping 5 days back; a full fetch starts from 2015. Progress is shown with the number of symbols and rows.
What can be concluded from it: If the source cannot be reached, no data is made up; the symbol is marked as failed and can be retried. Four symbols (KERVT, KOZAL, KOZAA, IPEKE) return no data at all from the source.
Symbols
What it is: The editable list of stocks included in the analysis: code, name, sector index, row count, last date, active flag.
What data feeds it: The symbol table; row count and last date from price_daily.
What it shows: Cells are edited by double-clicking; the active switch removes the stock from the analysis.
What can be concluded from it: For a stock with no sector, factor cleaning is done without the sector term. The list is today's composition; analysing the past with this list creates survivorship bias, so adding stocks that have left the index as well makes the results more honest.
Missing days report
What it is: For each stock, the number of trading days that are in the market calendar but missing for that stock.
What data feeds it: /api/data/coverage: the trading days of the XU100 series are compared with each stock's price_daily rows.
What it shows: Stock, row count and number of missing days; the one with the most missing days at the top.
What can be concluded from it: Missing days are not filled in the computations; they stay empty. A stock with many missing days is one that was not traded or went public later; it carries less weight in the results.
Company information
What it is: Each stock's size, detailed sector, free float ratio, paid-in capital history and shareholding structure. In the propagation model, a company's mass comes from its size, and its links from its sector and shareholders.
What data feeds it: İş Yatırım's public pages: the Fundamentals and Ratios page (today's market capitalisation, free float, capital and sector of all stocks in a single request), the financial statements (paid-in capital at the end of each quarter since the end of 2014) and the company card (shareholding structure). Written to the company_profile, capital_history and shareholder tables.
What it shows: Code, sector, market capitalisation, free float, capital, how many quarters of capital history there are, and the largest shareholders. Fetching is an administrator operation; progress and errors appear in the panel. The table is sorted by market capitalisation, largest first.
What can be concluded from it: The market capitalisation on a past day is the product of the latest paid-in capital that could have been published by that day and that day's close; information learned later is not used. The first fetch takes about half an hour, because it needs a dozen requests per stock. The shareholding structure is as it is today; it has no history.
Exchanges
What it is: The eleven exchanges the app can follow, which of them are open, each one's index, number of companies, closing time and time zone; switches to open and close them.
What data feeds it: The market table (V20, V23) and the company lists shipped with the app (seed/markets/CODE.csv): about a hundred of the most traded members of each main index, from public constituent lists. The price source is in the market.price_source column (V26, V27): İş Yatırım for Borsa İstanbul, Nasdaq for the US exchanges, Tencent for Shanghai, Shenzhen and Hong Kong, Euronext's own download service (companies by ISIN and venue), onvista for Frankfurt (the Xetra price), Moneycontrol for India, TMX for Toronto, Kabutan for Tokyo (waiting three seconds a page). All are free and need no key. Where a source rescales its history after dividends, the stored history is brought to the same scale by the ratio on the days both hold; where a source gives raw prices, splits are found from the price jumps. Only the US exchanges and Borsa İstanbul have a currency series; elsewhere the currency term of returns is dropped.
What it shows: For an open exchange, the companies followed; for a closed one, the companies that will be added when it is opened. Borsa İstanbul cannot be closed.
What can be concluded from it: When an exchange is opened its companies are added, and the hourly "Other exchanges" job fetches its history, returns, signals and links one after another, an hour after its close; its first full analysis starts by itself once two years of history have built up. A closed exchange is not listed to visitors and not processed; its data is kept.
The investment information, comments and recommendations given here are not within the scope of investment advisory services. Investment advisory services are provided under an investment advisory agreement to be signed between a client and brokerage houses, portfolio management companies, or banks that do not accept deposits. The signals here are produced from historical data with statistical models, are shown the same to everyone and are not personalised; they may not suit your financial situation or your risk and return preferences. Therefore, making investment decisions based solely on the information given here may not produce results that meet your expectations.