Simulated Performance · Rates · Inverse-Vol

Four $100,000 simulated walk-forward portfolios: the same daily directional verdicts, three ETF sizing schemes, plus one that expresses rate views with bond futures. Simulation only - no real money; built to test signal quality.

Universe: 9 fixed-income ETFs — TLT, VGLT, TLH, IEF and VGIT span long, intermediate and short duration; LQD, HYG, IGSB and BNDX add investment-grade credit, high yield and non-US bonds. Benchmarked against AGG rather than SPY.

Sizing · inverse-volatility (causal, unlevered): each position = $10,000 × target vol 15% ÷ its own ~20-day realised vol at entry, clamped to $3,500–$30,000. Low-vol larger, high-vol smaller, so each contributes similar risk. Sizing uses ONLY that position's pre-entry data — nothing about any other trade, past or future — so it is look-ahead-free and reproducible live. Concurrent gross is capped at $100,000 (no leverage): the cash book funds same-day competitors by CONFIDENCE priority; squeezed signals wait and are re-funded at that day's price when cash frees up (since 2026-08-15). The 'Size $' column below is each position's notional market value. ⚠️ This sleeve is all similarly low-vol bonds, so vol-parity barely differentiates — effectively equal-weight; the numbers are driven by deployment level and this rate regime, not a sizing edge. Tiny sample (n≈9, one rate regime) — indicative only.

$100,000 → $102,844.69

Initial capital → Final equity

12 / 4 / 0

Closed / Open / Pending

28.8 d

Avg hold

$1,811

Cash on hand · Initial + realised + interest − open cost

$243

Cash interest earned · Idle balance at the ultra-short rate · 3.85%

Performance by window

Rolling windows. Tier 2 risk-adjusted metrics require ≥30 trading days (else shown as —). Profit factor is an exception requiring only ≥3 closed trades.

	7D	1M	3M	SI since launch · 92d	AGG buy & hold · SI
Tier 1 · Strategic
Portfolio return	-0.42%	+0.86%	+2.12%	+2.84%	-2.55%
Benchmark (AGG) return	+0.48%	-0.85%	-2.35%	-2.55%	—
Value added	-0.90%	+1.71%	+4.46%	+5.40%	—
Win rate	—	50.0%	60.0%	66.7%	—
Closed trades (count)	0	2	10	12	—
Max drawdown · invested	-0.51%	-0.74%	-0.74%	-0.74%	-3.92%
Tier 2 · Risk-adjusted
Sharpe · invested	—	—	1.24	1.12	-2.61
Sortino	—	—	2.16	2.04	-3.17
Calmar	—	—	13.25	12.13	-1.74
Annualised vol	—	—	4.2%	4.0%	4.3%
Beta vs AGG	—	—	-0.68	-0.62	—
Info ratio · vs page BM	—	—	2.42	2.08	—
Profit factor	—	—	3.49	4.02	—

Long/Short breakdown (cash excluded, side-deployed-capital basis)

	7D	1M	3M	SI
↳ Long · Portfolio return	—	-0.10%	-1.12%	-0.14%
↳ Short · Portfolio return	-0.44%	+0.99%	+2.69%	+2.33%
↳ Long · Win rate	—	0.0%	50.0%	66.7%
↳ Short · Win rate	—	100.0%	66.7%	66.7%
↳ Long · Max drawdown · invested	—	—	-1.98%	-1.98%
↳ Short · Max drawdown · invested	-0.51%	-0.74%	-0.87%	-1.57%
↳ Long · Sharpe · invested	—	—	—	-1.12
↳ Short · Sharpe · invested	—	—	1.43	0.52
↳ Long · Sortino	—	—	—	-1.67
↳ Short · Sortino	—	—	2.50	0.82
↳ Long · Annualised vol	—	—	—	4.6%
↳ Short · Annualised vol	—	—	4.4%	4.8%
↳ Long · Beta vs AGG	—	—	—	0.99
↳ Short · Beta vs AGG	—	—	-0.98	-1.05
↳ Long · Profit factor	—	—	0.80	1.51
↳ Short · Profit factor	—	—	11.22	11.22

Weekly P&L

Weekly P&L = week-end NAV − prior week-end NAV (bucketed Mon–Sun, labelled by the last trading day of the week), which inherently includes BOTH realized (closed trades) and unrealized (open positions marked-to-market), since NAV = cash + realized + open MtM. YTD = latest NAV − initial capital.

YTD: +$2,845 (+2.84%) · AGG same period: -2.55% Win

Week ending	Week-end NAV	P&L	%	AGG %	Result
2026-10-14	$102,845	−$159	-0.15%	+0.24%	Loss
2026-10-09	$103,004	−$382	-0.37%	+0.36%	Loss
2026-10-02	$103,386	+$560	+0.54%	-0.56%	Win
2026-09-25	$102,826	+$1,034	+1.02%	-0.87%	Win
2026-09-18	$101,792	−$174	-0.17%	-0.02%	Loss
2026-09-11	$101,966	+$522	+0.51%	-1.04%	Win
2026-09-04	$101,444	−$5	-0.01%	-0.16%	Win
2026-08-28	$101,450	+$151	+0.15%	+0.14%	Win
2026-08-21	$101,299	+$104	+0.10%	-0.13%	Win
2026-08-14	$101,195	+$131	+0.13%	-0.12%	Win
2026-08-07	$101,064	−$556	-0.55%	+0.58%	Loss
2026-07-31	$101,619	+$99	+0.10%	-0.09%	Win
2026-07-24	$101,520	+$814	+0.81%	-0.75%	Win
2026-07-17	$100,706	−$107	-0.11%	+0.12%	Loss
2026-07-10	$100,813	+$457	+0.46%	-0.54%	Win
2026-07-02	$100,356	+$512	+0.51%	-0.40%	Win
2026-06-26	$99,844	−$255	-0.25%	+0.44%	Loss
2026-06-18	$100,099	+$17	+0.02%	+0.14%	Loss
2026-06-12	$100,082	+$39	+0.04%	+0.60%	Loss
2026-06-05	$100,042	+$42	+0.04%	-0.52%	Win

Page 1 / 1

Book scenario (data-only)

All measured numbers — no LLM narration. 2026-06-08→2026-10-14 · 90d

Profit engine: both sides

Bonds/Rates ·14+$2,021

Credit ·3+$580

Other ·2+$0

Sleeves · hedge status

Longs ·8 (avg +$25 · 50%W)+$198

Shorts ·11 (avg +$218 · 57%W)+$2,403

Both sides positive — no hedge cost · sleeve corr -0.46

Implied regime: risk-off / credit contraction

net beta -0.098 · idiosyncratic -2%

Intentional hedges

—

Accidental straddles · 0

none

Latest-day P&L by ETF 2026-10-14

Per-ETF P&L of positions held on 2026-10-14 (4 positions, this page's sizing). Total −$80.

ETF	Direction	Day ret	P&L
HYG	short	-0.36%	+$96
BNDX	short	+0.09%	−$26
IEF	short	+0.19%	−$56
TLT	short	+0.66%	−$95

Factor Attribution · Risk

Return/risk split onto two factor levels — macro (Broad market/Rates/Credit/Dollar/Gold/Oil/Crypto) + sector/theme (Semis/China/Defense/Financials/Energy-eq) — plus idiosyncratic (selection). Window 2026-06-08→2026-10-14 (90d).

Risk composition (% of variance): Equity family 19% · Other macro 73% · Idiosyncratic 8%

Broad market 18%Rates 70%Credit 3%Idiosyncratic 8%

Factor	Exposure β	Return contrib.	% of return	% of risk
Equity (broad + sectors)	-0.09	-0.03%	-1%	19.0%
Broad market	-0.10	-0.18%	-6%	18.4%
Semis	+0.00	-0.03%	-1%	-0.1%
China	+0.00	-0.04%	-1%	0.1%
Defense	+0.00	+0.00%	0%	0.3%
Financials	+0.01	-0.02%	-1%	-0.0%
Energy eq.	+0.00	+0.03%	1%	0.1%
Real estate	-0.01	+0.04%	2%	0.3%
Tech	+0.00	+0.01%	0%	0.1%
EM	-0.00	+0.14%	5%	-0.2%
Other macro	-0.51	+2.94%	103%	73.0%
Rates	-0.22	+2.52%	89%	70.1%
Credit	-0.31	+0.59%	21%	3.0%
Dollar	+0.02	-0.01%	-0%	0.5%
Gold	-0.00	-0.02%	-1%	0.2%
Oil	+0.00	-0.07%	-2%	-0.6%
Crypto	+0.00	-0.09%	-3%	-0.2%
Idiosyncratic	—	-0.06%	-2%	8.0%

A high idiosyncratic share means return/risk comes mainly from selection, not a factor bet. ⚠️ A factor row aggregates every holding that LOADS on it — not sector membership; short single-regime window — sector-factor βs are noisy, wide error bars.

Per-factor ETF loadings

Broad market

ETF	Net w.	Return contrib.	% of return	% of risk
TLT · top:Rates	-0.062	-0.12%	-4%	3.4%
HYG · top:Credit	-0.175	+0.10%	4%	2.7%
IEF · top:Rates	-0.225	-0.06%	-2%	4.5%
BNDX · top:Credit	+0.020	-0.05%	-2%	3.0%
VGIT · top:Credit	-0.044	-0.03%	-1%	0.8%
LQD · top:Credit	-0.035	-0.02%	-1%	4.0%

Rates

ETF	Net w.	Return contrib.	% of return	% of risk
TLT	-0.062	+0.90%	32%	24.9%
IEF	-0.225	+0.67%	23%	22.5%
VGIT · top:Credit	-0.044	+0.45%	16%	3.1%
LQD · top:Credit	-0.035	+0.32%	11%	9.4%
BNDX · top:Credit	+0.020	+0.24%	9%	8.8%
HYG · top:Credit	-0.175	-0.05%	-2%	1.4%

Credit

ETF	Net w.	Return contrib.	% of return	% of risk
IEF · top:Rates	-0.225	+0.23%	8%	0.8%
LQD	-0.035	+0.19%	7%	1.2%
BNDX	+0.020	+0.14%	5%	0.0%
HYG	-0.175	+0.09%	3%	0.9%
VGIT	-0.044	-0.06%	-2%	0.0%

Dollar

ETF	Net w.	Return contrib.	% of return	% of risk
VGIT · top:Credit	-0.044	-0.05%	-2%	0.3%
BNDX · top:Credit	+0.020	+0.03%	1%	0.1%
LQD · top:Credit	-0.035	+0.02%	1%	-0.1%

Gold

ETF	Net w.	Return contrib.	% of return	% of risk
VGIT · top:Credit	-0.044	-0.04%	-1%	-0.0%
IEF · top:Rates	-0.225	+0.02%	1%	0.1%

Oil

ETF	Net w.	Return contrib.	% of return	% of risk
BNDX · top:Credit	+0.020	-0.04%	-1%	-0.6%
LQD · top:Credit	-0.035	-0.03%	-1%	-0.0%
IEF · top:Rates	-0.225	-0.02%	-1%	0.0%
VGIT · top:Credit	-0.044	+0.02%	1%	0.0%

Crypto

ETF	Net w.	Return contrib.	% of return	% of risk
BNDX · top:Credit	+0.020	-0.05%	-2%	0.0%
VGIT · top:Credit	-0.044	-0.03%	-1%	-0.1%
IEF · top:Rates	-0.225	-0.02%	-1%	-0.1%
LQD · top:Credit	-0.035	+0.01%	0%	-0.0%

Semis

ETF	Net w.	Return contrib.	% of return	% of risk
IEF · top:Rates	-0.225	-0.02%	-1%	-0.0%
VGIT · top:Credit	-0.044	-0.02%	-1%	-0.0%
BNDX · top:Credit	+0.020	+0.01%	0%	-0.0%

China

ETF	Net w.	Return contrib.	% of return	% of risk
BNDX · top:Credit	+0.020	-0.04%	-1%	0.1%

Defense

ETF	Net w.	Return contrib.	% of return	% of risk
VGIT · top:Credit	-0.044	+0.01%	0%	0.1%
LQD · top:Credit	-0.035	-0.01%	-0%	0.2%

Financials

ETF	Net w.	Return contrib.	% of return	% of risk
VGIT · top:Credit	-0.044	-0.01%	-0%	0.0%
BNDX · top:Credit	+0.020	-0.01%	-0%	0.0%
LQD · top:Credit	-0.035	+0.01%	0%	-0.1%

Energy eq.

ETF	Net w.	Return contrib.	% of return	% of risk
BNDX · top:Credit	+0.020	+0.03%	1%	0.0%
VGIT · top:Credit	-0.044	+0.02%	1%	0.1%
LQD · top:Credit	-0.035	-0.02%	-1%	0.0%

Real estate

ETF	Net w.	Return contrib.	% of return	% of risk
IEF · top:Rates	-0.225	+0.04%	2%	0.1%
BNDX · top:Credit	+0.020	+0.01%	0%	0.2%
LQD · top:Credit	-0.035	-0.01%	-0%	-0.1%

Tech

ETF	Net w.	Return contrib.	% of return	% of risk
LQD · top:Credit	-0.035	+0.01%	0%	0.0%

EM

ETF	Net w.	Return contrib.	% of return	% of risk
BNDX · top:Credit	+0.020	+0.11%	4%	-0.1%
VGIT · top:Credit	-0.044	+0.02%	1%	0.0%
IEF · top:Rates	-0.225	+0.01%	0%	-0.1%

Per-ETF factor betas (7, the raw sensitivities aggregated above)

ETF	Net w.	Broad market	Rates	Credit	Dollar	Gold	Oil	Crypto	Semis	China	Defense	Financials	Energy eq.	Real estate	Tech	EM
IEF	-0.225	0.14	0.46	0.39	-0.08	0.02	-0.00	-0.01	-0.00	0.00	-0.01	-0.02	-0.01	0.03	-0.01	0.01
HYG	-0.175	0.23	0.18	1.00	0.00	-0.00	0.00	-0.00	-0.00	-0.00	-0.00	-0.00	-0.00	-0.00	-0.00	-0.00
TLT	-0.062	0.23	1.00	-0.00	0.00	0.00	0.00	-0.00	-0.00	-0.00	-0.00	-0.00	-0.00	-0.00	-0.00	0.00
VGIT	-0.044	0.10	0.30	0.41	-0.08	0.01	-0.00	-0.01	-0.00	-0.00	-0.01	-0.02	-0.01	0.02	-0.01	0.01
LQD	-0.035	0.28	0.52	0.90	-0.05	0.00	-0.01	-0.01	-0.01	-0.01	-0.01	0.01	-0.02	-0.01	0.02	-0.01
BNDX	+0.020	0.17	0.26	0.30	-0.07	0.01	-0.02	-0.01	0.00	0.02	0.00	-0.01	-0.01	0.02	-0.02	0.03
VGLT	+0.000	0.30	0.91	0.48	-0.04	0.01	-0.01	0.00	-0.01	-0.00	-0.02	0.01	-0.03	-0.01	-0.01	0.04

β = the ETF's sensitivity to each (orthogonalised) factor (green +/red −, shaded by magnitude); net weight + = book net long the ETF, − = net short. Book exposure = Σ net weight × β.

P&L leaderboard by ETF

Multiple trades on the same ETF are summed, so this ranks tickers — not individual trades. Realized = closed positions; unrealized = still held, marked to the latest close and therefore still moving.

Realized

Top 5 gainers

TLT 2 trades+$918

LQD 1 trades+$519

VGIT 5 trades+$105

IEF 2 trades+$98

Top 5 losers

HYG 1 trades−$71

BNDX 2 trades−$1

Unrealized (open)

Top 5 gainers

IEF 1 trades+$757

TLT 1 trades+$294

HYG 1 trades+$132

Top 5 losers

BNDX 1 trades−$150

Total

Top 5 gainers

TLT 3 trades+$1,212

IEF 3 trades+$855

LQD 1 trades+$519

VGIT 5 trades+$105

HYG 2 trades+$61

Top 5 losers

BNDX 3 trades−$151

Structure diagnostics (incl. open, funded basis)

Long/short books and entry-confidence buckets (all trades incl. open MTM) — which side loses money, and whether confidence is trustworthy.

+$198

Long book · 6 trades · win% (incl. open MTM) 67%

+$2,403

Short book · 10 trades · win% (incl. open MTM) 70%

Entry-confidence buckets

Bucket	N	Win rate	Avg P&L
0.60–0.64	10	70%	+1.02%
0.65–0.69	5	60%	+0.25%
0.70+	1	100%	+1.44%

Strategy rules (v2)

Signal source = consolidated ETF verdict (every slot the reasoning layer synthesises all active narrative opportunities for an ETF into one direction + confidence). Max 1 position per ETF.

Capital / sizing

$100,000 initial (v2 walk-forward starts 2026-06-01)
$5,000 fixed nominal per position (NOT 5% of equity)
Max 20 simultaneous positions ($100k / $5k)
Signal source: the daily consolidated ETF verdict's confidence

↗ Entry rule (ALL of)

direction ∈ {lean_positive, lean_negative}
confidence ≥ 0.60 + persistence (≥2 same-side verdict-days in 5d window, no opposite day)
not in active blacklist
ETF has no open position
Capacity available (incumbents priority; contested slots ranked by persistence days, NOT confidence — 2026-07-09, conf measured anti-predictive)

↘ Exit triggers (any-fires)

1. unrealized PnL ≤ −30% → close + add to blacklist (persistent, manual unblock)
2. Direction flip: ≥3 opposite verdict-days in 5d window (2→3, 2026-07-09) → close + 1-slot cooldown
3. Direction cleared: NO same-side lean in window (any confidence counts as support — 2026-07-09 fix) → close
4. Min hold 14 days (7→14, 2026-07-09; hold-forever peaks median ~17d): only the stop-loss can close earlier
5. Held ≥ 90 days → close

Per-trade Alpha computed against each ETF's natural peer benchmark (SPY default / EFA DM country / EEM EM country / AGG bonds) — see Alpha vs BM column in the trade table. Portfolio-level Alpha vs SPY kept for the marketing-card view.

⚠️ Preliminary sample. Ratio-style stats (Sharpe, Sortino, beta, IR) come from daily equity returns and need ≈30 trading days to be reliable (currently 93); and even at 30, with only a few overlapping positions the daily returns are autocorrelated, so the effective sample is smaller still. Flow stats (win rate, profit factor) rest on 12 closed trades. Read both as directional, not definitive.

Equity curve

Three lines: Portfolio = the $100k account's actual NAV (idle cash included); Benchmark = buy-and-hold over the same span; Deployed return = return on invested money only (P&L ÷ invested cost), rescaled to $100k notional. Because its denominator is invested capital only it is amplified vs the portfolio: above it when winning, below when losing — the gap is idle-cash dilution.

Drawdown curve (underwater)

How far the current portfolio is below its all-time high. 0% = at a new high.

Trade-level detail

Win rates and breakdowns below count closed trades only; open positions are marked-to-market and excluded.(closed 12 · open 4)

+$261

Avg win

−$130

Avg loss

+$666

Best trade

−$194

Worst trade

+0.67%

Long win rate · n=6

+0.67%

Short win rate · n=6

Exit reason mix

flip: 7 direction_clear: 5 open: 4

Confidence calibration + breakdowns

Win rates and breakdowns below count closed trades only; open positions are marked-to-market and excluded.(closed 12 · open 4)

By confidence bucket

Bucket	N	Win rate	Avg P&L
60%–70%	11	64%	+0.62%
70%–80%	1	100%	+1.44%

By ETF

ETF	N	Win rate	Avg P&L
VGIT	4	50%	-0.19%
BNDX	2	50%	-0.00%
IEF	2	100%	+0.31%
TLT	2	100%	+3.32%
HYG	1	0%	-0.24%
LQD	1	100%	+1.99%

Blacklisted ETFs (auto stop-loss −30%)

Hitting −30% unrealized auto-closes the position and lists the ETF here; it stays out until manually unblocked.

Empty — no ETFs have hit the 30% stop-loss.

All trades (no cherry-pick)

Every closed and open trade listed in full (nothing cherry-picked); open trades marked-to-market at the latest close. Alpha vs BM = excess return vs each ETF's natural peer benchmark. Newest first.

19 rows · newest first · 10/page

ETF	Direction	Entry	Exit	Reason	Hold days	Size $	Carry	P&L %	P&L $	BM	BM return	Alpha vs BM	Conf
VGIT	long	2026-09-30 $56.66	2026-10-14 $56.80	direction_clear	14	$0	+0.33%	0.58%	+$0	—	—	—	62%
HYG	short	2026-09-28 $77.54	2026-10-14 $76.81	open	16	$26,379	-0.44%	0.50%	+$132	SPY	-0.52%	+1.02%	63%
VGLT	short	2026-09-21 $52.42	2026-10-13 $50.41	direction_clear	22	$0	-0.39%	3.44%	+$0	—	—	—	65%
BNDX	short	2026-09-14 $46.93	2026-10-14 $47.05	open	30	$30,000	-0.24%	-0.50%	−$150	—	+0.00%	-0.50%	60%
TLT	short	2026-09-11 $80.87	2026-10-14 $78.89	open	33	$14,274	-0.39%	2.06%	+$294	—	+0.00%	+2.06%	63%
LQD	short	2026-09-09 $105.31	2026-09-25 $103.21	direction_clear	16	$26,028	—	1.99%	+$519	AGG	+1.61%	+0.38%	60%
VGLT	long	2026-09-01 $52.56	2026-09-15 $51.81	direction_clear	14	$0	—	-1.43%	+$0	—	—	—	60%
IEF	short	2026-08-31 $92.74	2026-10-14 $89.71	open	44	$29,347	-0.69%	2.58%	+$757	AGG	+1.97%	+0.61%	68%
VGIT	long	2026-08-19 $58.54	2026-09-14 $57.28	direction_clear	26	$10,653	+0.35%	-1.80%	−$192	TLT	-2.14%	+0.34%	66%
TLT	long	2026-08-14 $82.04 funded 2026-08-17	2026-08-31 $82.52	flip	17	$17,503	—	0.59%	+$252	—	+0.00%	+1.44%	72%
BNDX	long	2026-08-11 $47.82	2026-09-08 $47.39	direction_clear	28	$30,000	+0.25%	-0.65%	−$194	—	+0.00%	-0.65%	68%
IEF	long	2026-08-11 $92.87	2026-08-26 $93.32	direction_clear	15	$11,844	—	0.48%	+$57	AGG	+0.55%	-0.07%	62%
VGIT	short	2026-08-03 $58.25	2026-08-17 $58.38	flip	14	$28,156	—	-0.22%	−$63	TLT	+1.02%	-1.25%	62%
HYG	short	2026-07-09 $79.75	2026-09-08 $79.12	direction_clear	61	$30,000	-1.03%	-0.24%	−$71	SPY	-1.90%	+1.66%	60%
TLT	short	2026-06-24 $87.38	2026-08-11 $82.19	flip	48	$12,810	-0.74%	5.20%	+$666	—	+0.00%	+5.20%	60%
VGIT	short	2026-06-24 $59.00	2026-07-23 $58.23	flip	29	$28,156	-0.32%	0.99%	+$278	TLT	+4.45%	-3.47%	65%
BNDX	long	2026-06-11 $48.18	2026-06-25 $48.49	flip	14	$30,000	—	0.64%	+$193	—	+0.00%	+0.64%	62%
IEF	short	2026-06-08 $93.52	2026-08-10 $92.76	flip	63	$29,034	-0.67%	0.14%	+$41	AGG	+0.27%	-0.13%	66%
VGIT	long	2026-06-08 $58.53	2026-06-22 $58.69	flip	14	$30,000	—	0.27%	+$82	TLT	+1.74%	-1.46%	60%

Page 1 / 1

Exit-reason legend:

stop_loss — Stop-loss — unrealised loss hit −30%; force-closed AND the ETF is blacklisted until manually unblocked. The only exit that can fire inside the 14-day minimum hold

flip — Direction flip — closed only when the opposite-side verdict PERSISTED (≥3 opposite-lean days with conf ≥0.45 within a 5-day window), not a single flipped verdict. Only checked after the 14-day minimum hold

direction_clear — Direction cleared — closed only when same-side leans fade to zero across the 5-day window (went neutral/mixed and stayed); a single neutral day does NOT close it. Only checked after the 14-day minimum hold

conf_low — Confidence exit (RETIRED) — legacy rule: closed when confidence fell below 0.45. Since 2026-07-09 exits are persistence-based; only historical rows may still show this reason

max_hold — Max hold — force-closed after 90 days

open — Still open — the exit column shows mark-to-market at the current price, not a real exit

Re-entry: after a close (except stop_loss blacklist), the same ETF re-enters as a NEW trade once its verdict re-passes the entry gate (conf ≥0.6 with direction, that direction persisting ≥2 days within a 5-day window — not a single slot) — a new row is appended; closed rows are never rewritten. A chase gate also applies: entry is skipped when the ETF has already moved more than 3.0% in the signal's direction over the 10 trading days before the signal.

Rate-sleeve tradable ETFs · 9 total (7 ever held)

BNDX ●HYG ●IEF ●IGSBLQD ●TLHTLT ●VGIT ●VGLT ●

● = held at some point in the backtest; the rest are whitelisted but not yet triggered.

ℹ️ How buy / sell signals are emitted

Pipeline: News → event → narrative (arc) → the reasoning layer reasons per-ETF and assigns a direction → hard gates → opportunity / daily verdict. Direction is inferred by the reasoner, not derived from price moves (a blueprint's direction is only a prior).

Buy (long): All of: ① an active narrative routes to this ETF (blueprint tag match) ② the reasoner calls direction = long ③ the direction has non-price evidence (policy / positioning / flows / macro — never just 'it's rising') ④ not blocked by the priced-in gate ⑤ clears the red line (probe evidence exists). Medium-term positions also need support confirmation and the ≥14-day hold horizon.

Sell (short): Fully symmetric to buy, with direction = short. Extra: low-confidence shorts are damped by the verdict layer (calibration shows low-conviction shorts are the worst cohort); once a 'ceasefire'-type bearish catalyst is realised, a veto gate blocks continuing to short.

Hard gates (any failure = no signal): blueprint tag · non-price grounding · priced-in · red line (no data → no conclusion) · region ISO match · confidence calibration. Buy and sell pass the same gates — only the direction differs.

Not triggers (don't confuse): regime (Strong / at-high / off-low — observation only) · single-day reads · pure price momentum · inverse-vol sizing (sets position size only; it neither creates nor changes direction).

Note: this page replays those signals under fixed / inverse-vol sizing — it measures the signal-plus-sizing combination, and does not alter the signal itself.

Methodology

(1) $100,000 simulated portfolio, walk-forward, fills at the signal day's close; (2) no fees, slippage or taxes, except the rates-futures portfolio; (3) benchmarks are total-return (dividends included); (4) simulated results are not live trading and do not predict future performance.

← Simulated Performance