Ludwig J. Marx.

Working papers

Each study starts with a question and a hypothesis that are written down before the first result. The list shows what is running and what comes next.

Studies
WP 2

Intraday Patterns in E‑mini Nasdaq Futures

in progress

Can a simple intraday rule on Nasdaq futures beat random entries with the same timing, after costs?

Starts from the public trade record of an existing intraday strategy and asks which of its rules hold up under testing.

WP 3

Momentum under Full Review

planned

Does a momentum strategy beat random portfolios with the same number of positions and the same turnover, after costs and out of sample?

WP 4

Relative Weakness, Re-assessed

planned

An earlier version of this short-pair study rested on a faulty trade list. This paper rebuilds the list from scratch and asks whether any edge is left.

WP 5

Opening Gaps in Single Stocks

planned

Does the intraday rule from Working Paper 2 carry over to single stocks, and in which market conditions?

WP 6

Following the Trades of US Senators

planned

Can the disclosed stock purchases of individual senators be followed at a profit from the day they are published, and can the better senators be picked in advance rather than in hindsight?

WP 7

Regime-Switching Pairs in the S&P 500

planned

Does a pairs model that allows for regime changes earn more after costs than the classic distance method on the same pairs?

WP 8

Shorting Weak International Stocks against the S&P 500

planned

Does a market-neutral book of weak non-US stocks short against the S&P 500 earn a return after costs and borrowing fees, and does it come from profitability or only from US outperformance?

Status as of 21 Nov 2026.

Format

What every paper contains

Page one holds the bottom line and the key judgments, each with its level of confidence. Then follow the data and the filter log, the method, the findings, the control tests and the limits. Every figure names its source and date and says whether it shows a backtest or live results.

Working Paper 1 was an earlier study of my own trend indicator. Part of it rested on a faulty trade list. The error and what still holds are documented, and Working Paper 4 rebuilds that part from scratch.