Research you can examine.

Created and published by Paul Ivinskas

Paul Ivinskas created The Financial Current to connect financial developments with the evidence and context behind them. The publication serves bank executives and employees, fintech and mortgage professionals, dealers and sales finance professionals, and government readers. Analysis follows the subject, exploring relevant implications for business, operations, technology, customers, markets and public policy.

Meet Paul Ivinskas and learn more about the person behind the publication. Publication ownership is separate from authorship or independent human review of each article; the research process is described below.

How AI is used

AI assists with finding public material, drafting summaries and analysis, and maintaining the website. Articles are AI-generated research. Publication does not mean that Paul or another human expert has independently reviewed every statement. Readers can inspect the linked evidence and the dates attached to it.

Sources and evidence

Research prioritizes laws and regulations, agency releases and orders, court records, public filings, bank disclosures and technical documentation. News outlets and specialist publications help identify developments and competing interpretations. The Sources & scans directory describes the broader coverage, and the Policy library links to official authorities.

Company claims, proposals, allegations and analysis should be labeled as such. A settlement is not interchangeable with a court finding; a proposed rule is not final law. Public evidence has limits, and articles identify material uncertainty where it affects the conclusion.

What the dates mean

Published records when a story or research revision was added to the site. Source / event date identifies the underlying development. Latest research is the last successful substantive research cutoff. A layout change does not advance that cutoff.

Market observations retain their own timestamps. TradingView quotes update independently and retain their provider delay labels. Calendar dates come from linked schedules and can change after verification.

Updates and corrections

View the public corrections log →

Deep dives have permanent topic pages and dated versions. Substantive updates use a new revision with a change note; earlier versions remain available through Publication details and What changed. Takeaways are excerpts from the selected version, and estimated reading times are based on its length.

To flag a suspected error, contact Paul on LinkedIn with the article link, the statement in question and a supporting source. Corrections should identify what changed and why. Related coverage and topic pages help readers find context; a related link does not independently validate a claim.

Independence and reader privacy

This is a personal publication and does not speak for Paul’s employer or any institution covered. It provides general information and analysis, rather than advice tailored to a reader’s financial or legal circumstances.

Saved articles, saved searches, collections, followed topics and organizations, visit markers, reading positions and display preferences are kept in your browser and do not sync across devices. Reading positions are stored separately for each research version. Aggregate reporting uses Cloudflare Web Analytics and site-owned counters. Site-owned counters include pageviews, broad referral categories (with older referral hostnames preserved in historical counts), broad device/browser/operating-system categories, city/state/country when supplied by the hosting network, source-link, outbound-link and download-link clicks, tool use, search and filter activity, aggregate search outcomes (visible results or no visible results), opens from Top Story, Still in Focus, Featured Research, Policy and general article/research links, share actions, detail-panel expansions, and whether a page remained visible for 10, 30 or 60 seconds. Coarse article-scroll depth thresholds (25%, 50%, 75% and 100%) are counted after scrolling, without sending saved reading positions. Separate hourly pageview totals by canonical, legacy or hosted domain begin with this feature’s publication, have no historical backfill, and are shown for a rolling 90-day window; older hourly rows are removed on the next counted pageview and are not joined to page, location, device or referral dimensions. Future referrals use a fixed set of broad categories; historical daily counts are unchanged. Scroll and visible-time events are not proof that someone read or attended to the page. Engagement actions are totals by page or destination only; they are not joined to location, device or referral data. Site search text is no longer collected by the site-owned analytics counters. Previously collected term totals are no longer displayed; this change does not erase historical data. Keystrokes, live-filter text, IDs, IP addresses, precise locations, calculator inputs and browser-saved reading history are not saved by these counters. Minute-level counts support completed 15-minute through 24-hour windows with page, event and broad domain totals only; they do not join to location, device, referral or search terms. Minute totals are retained for 48 hours and have no historical backfill. Expired minute rows are removed on the next relevant collection or owner-report request. Browser storage failures block collection because the exclusion preference cannot be checked. The counters use no visitor IDs or cookies and honor the browser-local exclusion preference, browser Do Not Track and Global Privacy Control signals. Known automated checks are excluded where detectable; ordinary people arriving from ChatGPT are still counted. The exclusion preference applies on this browser and website origin and does not remove historical counts. This description is not a determination of privacy or consent requirements in every jurisdiction. Counts are private to the site owner and may miss blocked requests or undetected automated traffic. Embedded market widgets and external websites operate under their own privacy policies.