Seventy-five-year-old patterns, under modern testing.

The engine runs once after the close and hands you the whole book on one screen: what triggered today, what is armed and about to, every open position marked to market, and the historical reliability of each pattern family so you know what you are actually holding.

Preliminary research results. Early simulations are optimistic, and we are continuing to fully test and validate performance. Every figure and chart below is a historical simulation — not live or independently validated performance — and may change materially as data, source fidelity, implementation, costs, and out-of-sample behavior are verified. No live capital has traded these rules.

In the current preliminary research output, the recent window models a $10,000 starting account from January 2021 to today, ending at $23,368 — a 16.6% compound annual rate, a deepest drawdown of −24.8%, one losing year out of six, and a profit factor of 1.59 across 474 positions. Twenty-one of those are still open, which is what a multi-week swing book looks like on any given day.

The longer preliminary historical simulation runs the same research configuration back to 1995 — thirty-two years, 2,380 positions, a 19.0% compound annual rate, a deepest drawdown of −33.4%, and five losing years out of thirty-two. Read the annual rate and the drawdown, not the ending balance. Any long-run compounding figure is extraordinarily sensitive to small errors in the annual rate, and the caveats at the bottom of this page are the reason we treat that ending number as a research artifact rather than a target.

The design begins with Edwards & Magee's Technical Analysis of Stock Trends, the 1948 book that defined classical charting. Every candidate must survive implementation, data-quality, source-fidelity, and out-of-sample testing before promotion. Roughly twenty-five candidate improvements have already been rejected, while other rules remain open for correction and retesting. The early results are encouraging, but nothing on this page is a final promotion decision.

See every named variation we have retained, rejected, retired, queued for correction, or marked for a fresh test. The separate tracker includes a paired risk/return map, the full historical configuration archive, and the prospective paper evidence exactly as it exists today—including the places where there is not enough evidence to draw a curve yet.

Open the Variation Tracker →

How fast the milestones arrive

On the backtest sequence, the lump-sum account crosses $100K in year 10, $500K in year 21, and $1M in year 27. Add Roth contributions and the same milestones arrive in years 6, 11, and 19 — with $5M arriving by year 32. At the 401(k) funding rate they arrive in years 3, 9, and 11, with $5M by year 24. Contributions do not just add money; they pull every milestone years closer, because each year's deposit gets the full remaining runway of compounding.

The Roth IRA is the designed home. This book closes nearly every position within the year, so in a taxable account almost everything it earns is short-term capital gain taxed as ordinary income. Inside a Roth, that drag — the single largest cost an active strategy has — is simply gone, and every dollar in the table above would be withdrawable tax-free in retirement.

Be clear about the 401(k) row: most 401(k)s cannot run this. A standard employer plan offers a menu of funds — it cannot trade individual stocks at all. The row applies only where a plan offers a self-directed brokerage window (BrokerageLink or PCRA-style accounts, a minority of plans) or for the self-employed via a solo 401(k). Where it exists, it is the fastest path in the table purely because the contribution limit is the largest, its balance is pre-tax (withdrawals taxed as ordinary income — $13.1M at a ~28% effective rate is roughly $9.4M spendable), and any employer match sits on top of the figures shown. If your plan has no window, your real comparison is the taxable brokerage row — same funding rate, ~30% of each winning year handed to the tax man, and $13.1M becomes $5.16M. That $8M gap between two accounts holding identical trades is the wrapper's value, stated in dollars.

Note what compounding runway does to each deposit. The starting $20,000 rides all thirty-two years and multiplies ~86×; the $7,000 contributed in year one finishes as roughly $600K on its own; a deposit made in year thirty has barely begun. Averaged across the ladder, a contribution dollar gets about sixteen years of runway and multiplies ~17×. That asymmetry is the entire case for starting early — the calendar, not the contribution amount, is the scarce resource.

The lump-sum row is the honest floor. No contributions, no match, no top-ups — $20,000 once, compounded through every losing year the simulation contains. It still ends seven figures on the backtest sequence. But note it is also the slowest: the difference between the rows is not the engine, it is the discipline of funding it.

The age-18 reading — and which account you can actually open

The thirty-two-year table maps exactly onto one working lifetime: start at 18, read the ending column at 50. On the backtest sequence, a single $20,000 stake at eighteen — summer jobs, a used-car fund, a graduation gift — ends near $1.72M by fifty with nothing further added: roughly $69K a year of income at a 4% draw. Add the $7,000 Roth habit on top and fifty arrives near $5.6M tax-free — about $225K a year. Even starting from zero at eighteen with only the annual Roth contribution, the sequence ends near $3.9M. No inheritance, no startup exit, no lottery ticket — an engine, the right wrapper, and three decades of not interrupting it.

Which account, in order. First: if an employer offers a 401(k) match, take the match before anything else — a typical 50–100% match is an instant return no trading system can promise, even if that plan only holds index funds. Second: fund the Roth IRA (up to $7,500 in 2026, subject to earned-income and income-eligibility rules) — the designed home for this system. Third: if you have any self-employment income — freelancing, 1099 work, a side business — a solo 401(k) with Roth deferrals can open the big-limit row with no outside employer required, subject to plan and compensation rules: the 2026 employee-deferral limit is $24,500 and is shared with any day-job 401(k), with eligible employer-side contributions potentially available on top. Fourth: the taxable account — last, for the reasons the table just showed you.

The wrapper stack is also a withdrawal sequence. Retiring at fifty is a three-phase plan, not a single account. Phase one (50–55): the taxable account is the bridge — that is what it is for; the tax drag it pays during accumulation buys unrestricted access at any age. Phase two (55–59½): the 401(k) unlocks penalty-free under the rule of 55 (separate from that employer in or after the year you turn 55) — or as early as fifty via substantially-equal-payment elections or a Roth-conversion ladder started five years ahead. Phase three (59½ on): the Roth IRA comes online fully untaxed — and it goes last on purpose, because tax-free money should compound the longest and a Roth never forces required distributions. Design the exit ramp in your forties, not at forty-nine.

How the two accounts divide the work

The Trump Account is the from-birth account. Created in the 2025 tax law: an eligible child under 18 with a Social Security number can have one, and eligible U.S.-citizen children born in 2025–2028 can receive a one-time $1,000 federal pilot contribution after the required election. Other contributions are generally capped at $5,000 a year during the growth period, including up to $2,500 from an employer under a qualifying program. Its defining restriction: until 18 the money must sit in qualifying low-cost U.S. equity index investments. A stock-picking system like ours cannot run inside one during that period, and that is fine: the index is a perfectly good engine for the opening leg. After the growth period, ordinary IRA distribution rules generally apply; verify the current rules before acting.

The custodial Roth IRA is the from-first-job account. A Roth legally requires the child's own earned income — a newborn cannot have one, but a teenager with a W-2, a lifeguard summer, or documented, market-rate wages from the family business can. A parent opens and manages it as custodian; for 2026 the limit is the lesser of what the child actually earned or $7,500, subject to eligibility rules. The table retains the original fixed $7,000 illustration, so four teenage summers contribute the $28,000 shown there. A custodial Roth can be self-directed subject to the custodian's brokerage rules, and qualified distributions are tax-free.

The custodial brokerage (UGMA/UTMA) is the unlimited account — with a catch worth stating plainly. A parent can open one at birth, fund it with no cap beyond the ~$19,000-per-donor annual gift exclusion, and trade anything in it — it is the only childhood account where this system could run from day one. But the kiddie tax sends a child's investment gains above ~$2,700 a year to the parents' tax rate, and our math says that changes the answer: $7,000 a year from birth compounds to roughly $406K at 18 on this engine after kiddie tax — versus ~$386K just buying and holding an index fund, whose deferred gains dodge most of the tax entirely. An active system's edge nearly vanishes in a taxable custodial account, so the honest play is: index fund in the UTMA, and save this engine for the tax-free wrappers. The UTMA's real jobs are capacity (it takes money the capped accounts cannot), and destiny — at the handoff it becomes exactly the phase-one taxable bridge the retire-at-fifty sequence above requires. (A 529 is the education sibling: fund-menu only, so this system cannot run there either, though up to $35,000 of leftover 529 money can now roll into the child's Roth IRA over a lifetime — a nice tailwind for the Roth row.) Remember on all custodial accounts: at the age of majority it is irrevocably the child's money, and student-owned assets weigh heavier against financial aid.

The handoff at 18 is the point — and the Trump Account is the biggest baton. At eighteen the index-only restriction lifts: the account becomes ordinary traditional-IRA treatment, fully self-directable at any brokerage — which means this system can take it over. A maxed account's $206K–$286K, run on the backtest sequence from 18 to 50, finishes near $17.7M–$24.6M — the largest number on this page, and it needs its labels: those are pre-tax dollars (roughly $12.8M–$17.7M after ordinary-income tax at withdrawal), produced by the most caveat-laden column we publish. The sharper move is quieter: convert it to Roth gradually during the low-income college years, paying conversion tax at the cheapest rates of a lifetime, after which the same compounding runs entirely tax-free. And the first row deserves a second look either way: the government's own $1,000 seed, never added to by anyone, rides the index to ~$6,500 by 18 — and if the young adult then self-directs it on this system's sequence, that free thousand dollars finishes near $563,000 by age 50. The earlier sections said the calendar is the scarce resource; a birth-year account is eighteen extra years of the scarcest thing there is.

Honesty notes, as always. Trump Accounts are brand-new law — contribution limits, the seed window (currently births 2025–2028), and administrative details will evolve, so verify the current rules before acting. Family-business wages funding a child's Roth must be legitimate: real work, market-rate pay, payroll records — the IRS looks. The at-50 columns inherit every caveat on this page, including that the adult-years engine numbers are a survivorship-flattered simulation with no live track record. None of this is tax advice; a custodial account is the child's money, irrevocably. What survives all the caveats is the structure: seed early, wrap it right, and let the calendar do the heavy lifting.