Broad trend following (long only)
Did not hold up as a ruleA simple trend rule is assumed to carry across many markets at once.
- What was tested
- Whether simple trend following holds up across markets.
- Data
- Futures + ETFs, ~9 years
- Where you will meet this idea
- Common in managed-futures material and the "just follow the trend across markets" pitch.
Spreading across markets carried it, the signal did not.
Currency interest-rate differential (carry)
Did not hold up as a ruleHigh-yield currencies are assumed to pay more than low-yield ones, tested with a crisis-protection rule applied.
- What was tested
- Whether high-yield against low-yield currencies holds up with a crisis-protection rule.
- Data
- ~20 years of FX rates
- Where you will meet this idea
- The carry trade, a staple of FX commentary, and a recurring headline after unwinds.
Timing bonds off the yield curve
Did not hold up as a ruleThe steepness of the curve is assumed to time phases in bonds.
- What was tested
- Whether curve steepness times bond phases.
- Data
- ~24 years of bond ETFs
- Where you will meet this idea
- Yield-curve timing, the bond version of "the chart tells you when".
Reaction to quarterly results (drift)
Test not finishedPrices are assumed to keep drifting in the direction of an earnings surprise.
- What was tested
- Whether prices drift on after earnings.
- Data
- US equities (several datasets)
- Where you will meet this idea
- Post-earnings drift, a well-documented anomaly in the academic literature.
Not cleanly testable with the data available to us.
The Treasury auction cycle
Did not hold up as a ruleDealers forced to absorb supply around auctions are assumed to create a rhythm.
- What was tested
- Whether forced dealer buying around auctions creates a rhythm.
- Data
- ~four years of 10-year Treasury futures
- Where you will meet this idea
- The auction-cycle argument you meet in fixed-income commentary.
Volume-zone breakout (“auction market”)
Did not hold up as a ruleA move out of the previous day's volume zone is assumed to run.
- What was tested
- Whether breakouts from the prior day's volume zone hold up.
- Data
- 9 years of Nasdaq futures
- Where you will meet this idea
- Auction market theory, the framework behind most value-area and market-profile teaching.
Passive bond risk premium (duration)
Test not finishedA calmly sized, permanent duration position is assumed to harvest a genuine premium.
- What was tested
- Whether a quietly dosed duration position harvests a real premium.
- Data
- 10-year Treasury futures
- Where you will meet this idea
- The "own duration, collect the premium" argument in bond-allocation debates.
Buying the dip in a bull market (equity index)
Test not finished“Buy the dip” is assumed to work as long as the longer trend points up.
- What was tested
- Whether buying the dip only holds up in an uptrend.
- Data
- ~20 years of a Nasdaq ETF
- Where you will meet this idea
- "Buy the dip, but only in a bull market", the rule that sounds too sensible to test.
Reaction to earnings guidance
Test not finishedDeviations from analyst expectations are assumed to keep working through the price.
- What was tested
- Whether deviations from analyst expectations drift on.
- Data
- US equities
- Where you will meet this idea
- Guidance surprises, the reason earnings previews focus on the outlook, not the quarter.
Needs data we currently do not have.