Athena 2 — 3×-Nasdaq + Bitcoin, with crash filters

60% Nasdaq leverage (TQQQ 3× above its 200-day line, QQQ 1× below) + 40% Bitcoin (IBIT) (DBMF below its line) — now with 3 crash filters: a vol breaker (Nasdaq too wild → cash), a 3-day Bitcoin confirmation, and a 2% buffer on the Nasdaq line. Same long-run return, much shallower crashes. Inside HOUSE v2.9 it's the 401K's dialed sleeve.

Paper · high-variance experiment

⚠ High risk. Out-of-sample 2019–26: Athena 2 −38% worst drop vs −59% for Athena 1, same return (full 2011–26 backtest: −45%); the Nasdaq half in the dot-com crash: −3% vs −96%. The catch: it lags fast V-shaped rebounds (2024–26 about 15–20 pts/yr behind Athena 1; 2026 so far —). Still a 3× + Bitcoin strategy: expect deep drops. Paper, no orders.

60/40

TQQQ · Bitcoin (target)

—

Backtest CAGR (2010–26)

—

Max drawdown (backtest)

—

Drawdown this year (live paper)

—

Gross invested now

✓ Seeded $50,000 on Jan 1, 2026 and tracked on real TQQQ + IBIT prices going forward. Backtest (real, 2010–26): —. Crypto sleeve = IBIT (the tradeable spot-BTC ETF), spliced onto spot BTC before 2024.

Live — the paper account right now

Trend gates — each sleeve vs its line

Price above its trend line = that sleeve is ON (invested). Red bands = below the line → that sleeve de-risks (Nasdaq drops from TQQQ 3× to QQQ 1×; crypto to DBMF managed futures). This gate is the only thing between Athena and total ruin.

Athena vs QQQ vs SPY — profit since launch

$— start (Jan 1, 2026). Dashed = a 30%/yr pace line (aggressive by design). Fills in daily.

Athena QQQ SPY 30%/yr pace

Positions — each holding's day change, total P&L and return since start

Allocation vs 60/40 target

What it does

Target book — on $50k

Sleeve	Weight	Instrument	Gate
Nasdaq leverage	60%	TQQQ → QQQ	200-day
Bitcoin	40%	IBIT → DBMF	30-day
DBMF	takes Bitcoin's 40% when it is off	DBMF (managed futures)	Bitcoin 30-day

The idea (and the danger)

Two of the highest-compounding assets in history, levered — 3× Nasdaq for the growth engine, Bitcoin for the asymmetric upside. The trend gate is the whole safety system: above the line it rides the leverage; below, the Nasdaq sleeve steps down to 1× QQQ — except when Nasdaq gets too wild (20-day vol > 30%), when it goes to cash. That breaker is what Athena 1 was missing: holding QQQ all the way down the dot-com crash would have cost −96%.

Filters were picked on 2011–18 and checked on unseen 2019–26, walk-forward, all 27 nearby settings, 4× costs and 5,000 Monte Carlo futures (shallower drop in 96%). Real money runs HOUSE v2.9: the Athena 2 dial is 0/20/40% of the whole household, held inside the tax-free 401K.

The rules

Trend gate

🚪 Per-asset speed

TQQQ switches on above 200-day +2% and off below 200-day −2% (buffer stops wiggle-flips). IBIT gates on the 30-day but only flips after 3 days on the other side. Below the line: Nasdaq → 1× QQQ, crypto → DBMF.

Vol breaker

⚡ Too wild → cash

When Nasdaq's 20-day volatility tops 30%, the Nasdaq sleeve goes to cash until it calms. This is the desk's proven crash signal — it fired almost the whole dot-com crash and most of 2022.

Paper only

🧩 Watch, don't fund

Virtual overlay, no orders. It shows Athena 2 live next to its old self (Athena 1) so you can see what the filters cost in rebounds and save in crashes. Real money rides HOUSE v2.9.

Athena 2 · paper overlay, no real money · 60% Nasdaq leverage (TQQQ above QQQ’s 200-day ±2%, QQQ 1× below, cash if QQQ vol > 30%) + 40% IBIT (30-day gate, 3-day confirm, DBMF below) · record recomputed with Athena 2 rules from launch · seeded $50k 2026-01-01. The desk's high-variance experiment — not investment advice.