Thesis: There is a best gap between one evening up and the next / USD v1

run 9532 as portfolio versions hold it (snapshot 85) · started 2026-09-29 09:54:30 · finished 2026-09-29 09:55:02 · peak 1111 MB · the strategy now

held in portfolio versions: Retired couple, 68 v34 (withdrawal, primary, Joint brokerage 7.01%, Rollover IRA 12.82%, Roth IRA 2.40%, Spouse IRA 6.21%, Spouse Roth IRA 1.60%) · Retired, 66 v33 (withdrawal, primary, Brokerage 5.58%, Rollover IRA 20.61%, Roth IRA 3.86%) · Retired couple, 68 v32 (withdrawal, primary, RRSP 0.97%) · Retired, 66 v31 (withdrawal, primary, RRSP 4.69%) · Mid-life couple, 55 v30 (cagr_dd, primary, RRSP 4.02%) · Mid-life saver, 45 v29 (cagr_dd, primary, RRSP 5.66%) · Mid-life couple, 55 v26 (withdrawal, primary, RRSP 0.21%) · Mid-life saver, 45 v25 (withdrawal, primary, RRSP 1.42%) · Mid-life couple, 55 v24 (withdrawal, primary, RRSP 0.21%) · Mid-life saver, 45 v23 (withdrawal, primary, RRSP 1.42%)

How it invests (strategy version 1)

Its universe is every stock and fund in the USD sleeve (every US listing with prices on disk), except those trading under a cent.

At the end of every month, it ranks the names by how many shares it typically trades a day (median over 3 months), and holds the 50 ranked highest in equal amounts.

At each rebalance it only trades if some holding has drifted more than 2% from its target.

No position is larger than 10% of what the stock typically trades in a day.

Trades are placed on the day after the signal, at that day's close, with commissions charged.

Note on the rule: even up only when a weight has drifted 2 points

Source: https://arxiv.org/abs/1009.3753 · rules: --roster universe --select liq:63 --hold 50 --band 0.02 --every M

Key metrics

CAGR

20.67%

2003-01-02 → 2026-09-25 (23.7 years)

Sortino

1.203

daily, target 0

Max drawdown

65.11%

2007-10-31 → 2008-11-20, recovered 2011-01-03

Total return

8529.23%

443,038 → 38,230,789 undefined

Sharpe

0.834

risk-free 0

Volatility

26.88%

annualized, daily returns

Calmar

0.317

CAGR / max drawdown

Longest under water

1,160 days

2007-10-31 → 2011-01-03

Best year

70.30%

2009

Worst year

-49.92%

2008

Positive months

63.16%

of 285

Fills

8,415

fees 343,640 undefined

Quality

B evidence supports an edge: 16.0 years out of sample, alpha +4.8% (t 1.4)

Hypothetical: a backtest of a published rule. The results include commissions and leave out the spread, market impact and taxes; that overstates them. Rated before 2026-10-02: against the index alone, with no Outlook, Risk, Protection or Fit yet.

Evidence: is there an edge beyond what index funds and cash give, since the rule was published?
rule fidelity	A	the rule as stated (the rule as stated)
published	2010-09-30
out of sample from	2010-09-30	16.0 years
alpha before	10.4%	a year against the sleeve's index fund, t 2.7
alpha after (out of sample)	4.8%	t 1.4
alpha after 2021-08-14	5.2%	the fixed 15 + 5 year check, t 0.8
beta	1.27	against the sleeve's index fund, whole run
deflated Sharpe	0.06	chance the Sharpe (0.83) beats the luck of 620 independent tries
decades with positive alpha	3 of 3	2003-09, 2010-19, 2020-
fees a year	3.3%	share of the starting capital
buys in funds, preferreds, bonds	25.0%	for a rule over the whole sleeve

Equity

2003-01-02 443,038 → 2026-09-25 38,230,789 · log scale · 5971 days · hover or use the arrow keys for values

Drawdown: the fall below the highest value reached so far. Worst 65.11% below its peak, from 2007-10-31 to 2008-11-20, back to that peak 2011-01-03 (1160 days from the peak) · 7.59% below its peak on the last day

Fills — 8415

Day ▲	Ticker	Quantity	Price	Fee
2015-05-01	HLSSF-201510	-5,598	0.669	18.73
2015-05-01	HLSSF-201510	-5,598	0.669	18.73
2015-05-01	IWM	60	105.2104	1
2015-05-01	IWM	60	105.2104	1
2015-05-01	JPM	-1,320	47.2041	6.6
2015-05-01	JPM	-1,320	47.2041	6.6
2015-05-01	LRCX	9,472	6.6087	47.36
2015-05-01	LRCX	9,472	6.6087	47.36
2015-05-01	META	90	78.2406	1
2015-05-01	META	90	78.2406	1

Log

(no log)