This site walks you through a data-driven investigation, step by step, so you can evaluate the evidence and reach your own conclusion. Every chart is interactive — hover for details, click legends to filter, drag to zoom.

The question

Every payday, millions of Americans have money automatically deducted from their paychecks and invested in the stock market through their 401(k) retirement plans. This creates a massive, predictable wave of buying — roughly $1 trillion per year flowing into equities on a known calendar.

Does this predictable flow leave a measurable fingerprint in stock prices? And if so, does someone profit by trading ahead of it?

What we found

65
years of S&P 500 data analyzed (1960–2026)
16,680
trading days tested
4
independent statistical methods applied
6
pay schedules tested

Three headline findings

1. The pattern exists — but it's shifted by a week

At first, we found nothing: stock returns around paydays looked random. But when we shifted the calendar forward by 7-8 trading days — accounting for the time it takes 401(k) money to actually clear and settle into the market — a statistically significant pattern emerged.

The S&P 500 shows elevated returns in the 5 trading days before the settlement date (not the paycheck date), peaking at +0.055% per day across 65 years (p = 0.019). In the peak decade (2000–2019), the effect doubles to +0.115% per day.

2. It didn't exist before 401(k) plans were created

The clearing-lag pattern is completely absent in the pre-1980 data (when 401(k) plans didn't exist). It emerged in the 1990s alongside 401(k) asset growth, peaked in the 2000s when 401(k) assets surpassed $2 trillion, faded in the 2010s (possibly arbitraged away by quant funds), and partially returned in the 2020s.

3. It's strongest where the money is most concentrated

We tested four different pay schedules: weekly, biweekly, semi-monthly, and monthly. The pattern appears only in semi-monthly pay (15th + end of month) — the schedule used by higher-paid salaried workers who contribute disproportionately to 401(k) plans. Weekly payroll (which spreads flow across all 52 weeks) shows no signal at all. Social Security payments (which don't flow into stocks) show no signal — a clean negative control.

How to read this site

The investigation follows a logical progression. Each page builds on the previous one:

StepPageWhat you'll learn
1The HypothesisThe original theory and how we translated it into testable predictions
2MethodologyHow we tested it: the data, models, controls, and assumptions
3Initial FindingsThe naive test — and why it initially looks like the theory fails
4Clearing-Lag DiscoveryThe turning point: what happens when you account for settlement time
5Historical EmergenceDid this pattern always exist, or did it appear with 401(k) adoption?
6Pay Schedule AnalysisTesting weekly, biweekly, semi-monthly, monthly, military, and Social Security
7Alternative ExplanationsCould FOMC, options expiration, or CPI explain this instead?
8Manipulation InvestigationWho has the means and motive? What does the microstructure show?
9Investor ImpactWhat does this cost 401(k) investors? When is the optimal time to buy?

Interactive tools

Technical terms are highlighted with dotted underlines — hover over them for a definition and a plain-English translation. Each section includes an "In plain English" callout box that restates the finding without jargon.

Reproducibility: Every number and chart on this site is generated from Python scripts that read raw S&P 500 price data from Yahoo Finance. The scripts, data, and methodology are fully documented and reproducible. No data was hand-edited.