Welcome
This is your research desk: test trading rules on decades of real prices, then watch them work (or not) on new data. Fake money only.
Forward tests strategies you're tracking on new prices
Europe desk Rules vs. Claude, live paper trading
Top of the Scoreboard best grades in the Strategy Lab
Honest bottom line so far
- Europe → US open idea: no reliable edge in 2–3 years of real hourly data (details in Europe Lab). One weak lead on S&P futures at 8:00 ET is being studied.
- Classic strategies: most reduce crashes more than they boost returns. Very few beat simply holding the S&P 500 after costs, especially since they were published.
- What that means: the goal is to learn what's real. Forward tests are the honest exam.
Strategy Lab Classic rules tested on decades of daily prices, always against just holding the S&P 500.
Compare strategies pick up to 5; all start on the same date
Forward tests rules frozen on the start date, tracked on prices that came after
Strategy
How sure are we? could this just be luck?
Growth of $10,000
Every 3-year stretch yearly return vs the S&P over each rolling 3 years
The numbers
Year by year
How it decides
Why it might work
When it struggles
Where it comes from
Settings change them and re-run
Careful: tweaking settings until the chart looks great is exactly how overfitting happens.
What research says
How to read a backtest
Glossary tap any ? in the app to see these
Beginner traps why most backtests disappoint live
The five strategy families
Own what's rising, step aside from what's falling. Wins big in long trends, loses small and often in choppy markets.
Own the recent winners, versus their own past or versus other assets. Among the best-documented effects in finance, but prone to sudden crashes.
Bet that sharp moves partly snap back. High win rates, small gains, occasional painful losses.
Patterns tied to the calendar or clock. Fascinating, often small, and easily eaten by costs.
Not timing at all: mixing assets (stocks, bonds) and managing risk. The baseline most professionals are judged against.
Research notes compiled October 2026 from academic papers and practitioner studies, linked on each strategy. Educational only, not investment advice.
Europe Lab your friend's idea, tested
Does what Germany's DAX and France's CAC 40 did before the US open predict the S&P 500's first hours? Pick a rule, run it on ~2–3 years of real hourly prices, and look at the Test (unseen) column.
So far: no setting has shown a reliable edge after costs. That's a real finding, and it saved real money.
Test a rule on history
Trade-by-trade list
Overfitting check tries 24 setting combos
Ranked by how they did on the training years. The test columns are years the ranking never saw. If the top rows fall apart on test, the "edge" was luck.
Right now
Rules vs. Claude
Claude's latest call
Trades
Signal log
Day Trading Desk simulated micro S&P 500 futures · in and out the same day · paper money
- At 9:30 ET the market opens. The desk watches the first 5 minutes.
- If those 5 minutes went up, it buys 1 micro S&P contract at 9:35; if down, it sells short.
- Stop-loss at the other end of those 5 minutes (a small, defined loss). Target at 10× that risk.
- Whatever happens, it's out by 3:55 PM ET. Most days it's stopped out within ~15 minutes.
Two desks run side by side: the plain rule, and your friend's twist that only trades when Europe (DAX/CAC) is moving the same way. Each starts with $2,000 of paper money and trades 1 contract, which controls roughly $38,000 of S&P 500 (about 19× leverage).
Today
Desk accounts
Trades
If it had been running the last ~2 months 5-minute data is only available for about 60 days
Loading…
Two months is a tiny sample. A small number of big winners usually decides the result. An independent replication over many years and five stock indexes found the rule's edge disappears after realistic trading costs. Source ↗
Why "100% in minutes" happens, and why it cuts both ways
Brokers only ask for a small deposit ("day-trade margin") to hold a futures contract intraday, sometimes about $100 for a micro S&P. A routine 0.25% wiggle in the S&P is about $100 on one contract, so that's +100% on the margin in minutes. The same wiggle the other way is −100%. The percentage is real; it's the leverage talking, not the strategy.
Reality check: in a study of Brazilian futures day traders who kept at it for 300+ days, 97% lost money and fewer than 1% earned more than minimum wage. Source ↗ That's why this stays on paper until it proves itself for months.
Ask Claude sees your current backtest, live settings and recent trades
Add a user
Users
Phone alerts get a notification for every signal, trade and strategy change
- Install the free ntfy app (App Store / Google Play).
- Tap Generate below (or type your own long, hard-to-guess name), then Save.
- In the ntfy app, tap + and subscribe to exactly that topic name.
- Tap Send test alert. Your phone should buzz.
Anyone who knows the topic name can read the alerts, so keep it long and random. Alerts never include passwords or keys.