TimeCycleLab @timecyclelab1

“The Phase That Comes Before the Rally” This week I’m publishing a case study on a reader-submitted T — a live forecast that went into the audit as a pre-registered test. It’s one of the cleaner examples in the dataset, and it raises a question I’ll keep coming back to: when a T works exactly as projected, what does that actually prove? After that post, the series shifts into something I’ve been building toward since the beginning: the Cash Build Up phase. In T Theory, no rally is valid without one. The CBU is the left side of every T — the period when the volume oscillator makes a sequence of descending peaks, often while prices drift or grind sideways. Terry called it a “cash build up” because the theory holds that money normally committed to equities gets withheld during this phase, accumulating on the sidelines. When that pattern breaks — when the oscillator finally reverses — the pent-up demand releases, and the right side of the T begins. That’s the logic. Whether it holds up statistically is exactly what this project is testing. The CBU post establishes why this matters beyond theory: identifying the CBU start date is the hardest mechanical problem in the entire framework. Get it wrong and the center post shifts, the target date shifts, and the whole construct drifts. It also turns out to be the piece Terry was most opaque about in his public notes — which is saying something for a man who explained everything at length and still left the hardest part undefined. I’ve been working through 22 confirmed T-events. Here’s where that work stands and what’s coming: 22 Ts verified end-to-end, spanning 2003–2008 Time-symmetry: the correlation between left-side duration and eventual peak date is significant across the full dataset, trending toward significance in the realtime subset Four rule tests completed — bear rally duration, magnitude taxonomy, moving average behavior, and center post sequencing — all confirmed at the preliminary level Next phase: extending the dataset into 2009–2011, the crash recovery period, where the most interesting regime questions live The CBU Framework post goes up the week after Steve’s T. It’s the one I’ve been most cautious about writing, because it’s where the methodology either holds together or it doesn’t.

1) The cash build-up in the S&P 500 here still looks like it’s forming. In T Theory a build-up isn’t finished until the volume oscillator stops making descending peaks — and right now I’d read this one as incomplete, still laying down the left side of the T. 2) My expectation: it resolves, the S&P spikes higher to end at the July volume-oscillator T. That’s the construction’s natural right-end date — the cash accumulated on the left side fuels the rally across the matching span on the right. 3) After that is where it gets interesting. In T Theory, once a T’s time span is set, there are only two ways its right side can resolve. The bullish T uses the built-up cash to carry the S&P higher. The bearish T takes the same build-up, but the money never comes back in — investors stay in a selling mood — and the right side collapses. 4) The character of the Ts looks like it’s shifting toward that second kind. So — unless something news-driven emerges that changes the picture — I’d expect a more meaningful decline in the S&P to follow the July peak, not another leg up. 5) We’re approaching a very interesting window. Summertime first, winter eventually. When the character of the Ts starts to turn, the calendar tends to start mattering more. Watching closely.

#5. Case Study: The March 2003 T open.substack.com/pub/timecyclel…

This week at Timecyclelab: the first case study. The March 2003 T marked the bottom of the bear market. By the numerical record, the framework’s call landed within tolerance. Terry himself, looking back ten months later, counted it as a near-miss. That gap — between what the data says and what the analyst was willing to claim — is the texture of the case study.

What I’m working on at Timecyclelab over the next few weeks: — Posted this week: How do you actually test a forecaster’s record after he’s gone? timecyclelab.substack.com/p/how-do-you-t… — A reader spotted a pattern Terry hadn’t drawn. The market peaked when Terry projected. What does that one call tell us? — Why long-range Ts work and short-range ones don’t — What the 2007-2009 crash period adds to the picture — I’ll also be looking at a couple of case studies One post per week.

A Price T just resolved. Target May 29th → market topped June 2nd → sold off right to the 55-day MA. One of Terry Laundry’s channel techniques. (More on those soon.) Next: a short-range T resolving in the coming weeks. I’ll post the date and the result live — watch it play out. Beyond that, a larger Advance-Decline T points higher into late this year. I have a precise date for the turn. Skill or luck? That’s the whole experiment.

IV. How do you test a forecaster's record after he's gone? open.substack.com/pub/timecyclel…

Terry Laundry left behind eight years of public market calls. I tested them. Same archive, three numbers: — +92pp alpha vs buy-and-hold at perfect timing — +83pp at his own mechanical signal (1-2 days after each low) — +30pp acting only on the formal newsletter confirmation The framework worked. The newsletter ran weeks behind its own signal. Which number a subscriber received depended on how closely they were reading. timecyclelab.substack.com/p/the-skill-wa…

A reader named Steve emailed Terry Laundry in early 2004 about a pattern he’d spotted in Terry’s charts. Terry took it seriously enough to name the pattern after him and project an S&P peak three weeks out. The S&P peaked the day Terry projected, give or take twenty-four hours. That’s the kind of thing that, if you saw it once, you’d assume was a coincidence. So I went back and tested all of Terry’s projections from 2003 to 2008. timecyclelab.substack.com/p/the-lost-arc…

Markets spend roughly half their time going up and half going down. Yet they produce positive returns over the long run. Symmetry in time. Asymmetry in return. Terry Laundry spent fifty years saying this had a shape. I tested his published record. Same archive, three findings: — The framework’s signal was real — The newsletter that delivered it ran weeks behind — Subscribers’ returns depended on how closely they were watching Introducing Timecyclelab. A research letter on T-Theory — quantified, verifiable. timecyclelab.substack.com/p/on-the-subop…