eniwa · Shio · Genki · Modori · Hane · Nagi · updated …

Five legs, one book.
Three years of past, three years of possible future.

Five largely independent trading strategies in one book. The past is each strategy’s own backtest, net of costs and funding, at today’s weights and risk level. The future is thousands of simulated paths, with more cautious return expectations than the backtest. The data is refreshed weekly, and your browser re-runs the simulation.

…over the past 3 years
(model)
…Sharpe ratio
(individually …)
…median multiple
in 3 years
…deepest drawdown
in the worst 5%
01 — Equity curve · past and future

Where the past ends, the future fans out

The white line is the past. After “today”, every faint thread is one possible future; the white continuation is the simulated path that stays closest to the median all the way. Pick a leg, or hover over the curve.

Running the simulation…
02 — The five legs

Five engines, three independent sources

Crypto trend (Shio, Genki), index rebound (Modori, Hane) and market-neutral crypto (Nagi). Click a card to show that leg in the chart above.

Shio

live

ETH · long and short · weight …

Trend-following on Ether: an ensemble of six trend signals of different speeds decides, and the short side only opens in a persistently falling market. The largest risk item in the book.

Past 3 years
…
Sharpe · max DD
… · …
3-year median
…
p10 – p90
… – …

Genki

live

four major cryptos · weight …

Trend and short-term momentum on the four largest cryptos (BTC, ETH, SOL, XRP) under one risk target. Together with Shio it forms one crypto cluster: they lose together in the big crypto sell-offs.

Past 3 years
…
Sharpe · max DD
… · …
3-year median
…
p10 – p90
… – …

Modori

live

Nasdaq-100 · rebound · weight …

Buying Nasdaq-100 down days for the rebound, and closing once the index closes near the top of its daily range. Practically uncorrelated with the crypto legs, which is what makes it valuable in the book.

Past 3 years
…
Sharpe · max DD
… · …
3-year median
…
p10 – p90
… – …

Hane

live

FTSE 100 · rebound · weight …

A long-only rebound strategy on London’s FTSE 100, with an intraday protective stop. The quietest leg: low volatility, shallow drawdowns.

Past 3 years
…
Sharpe · max DD
… · …
3-year median
…
p10 – p90
… – …

Nagi

paper

market-neutral · crypto · weight …

A market-neutral crypto book: it is long and short many smaller cryptos at once on four signals (trend, range, order flow, carry), with the market direction hedged. Today it runs on paper only; this page assumes a planned new account.

Past 3 years
…
Sharpe · max DD
… · …
3-year median
…
p10 – p90
… – …
03 — In numbers

What it gave, and what it may give

weightpast
3 yrs
Sharpemax DDexpected
Sharpe
3-yr
median
p10 – p90chance
of loss

Past: model, at the leg’s current risk level, over the last 3 years. Multiple: value in 3 years relative to today. “Expected Sharpe”: the simulation’s return assumption, varied per path; deliberately below the backtest.

04 — Co-movement

Why does it work together?

ShioGenkiModoriHaneNagi

Correlation of weekly returns, 3 years. Shio and Genki move together (…), so … of the book’s risk comes from them. The other legs are largely independent of crypto and of each other, so the book’s deepest drawdown in the worst 5% is …, while a single leg’s is ….

05 — What Nagi adds

The cost and benefit of a fifth leg

On the same paths without Nagi, the worst-5% drawdown would be … instead of …, while the median multiple would be … instead of …. Nagi improves the low point at a small cost in return. Whether it enters depends on its live paper run (gate: February 2027).

06 — Checkpoints

The gates of the next three years

  • 8 Nov 2026Nagi paper startPaper account, monitored weekly.
  • monthlyMonthly checkBook and leg drawdown, live versus model.
  • Feb 2027Hane gate and Nagi paper gateHane: higher risk level if tracking is tight. Nagi: Sharpe > 0.5 and drawdown above −12% — only then can it get real capital.
  • Oct 2027Genki decisionWhether it stays and gets new capital.
  • Nov 2027Annual reviewAfter 12 months of live data the next risk step is decided.
  • 2028–2029Years 2 and 3Half size or retirement based on 24 and 36 months of live results.
07 — Intervention rules

Written in advance

  • Book drawdown −18%: review, no change. −25%: every leg to half size, back above −10%.
  • 24-month live Sharpe below 0.3: the leg goes to half size; back above 0.5.
  • 36-month live Sharpe below 0: retired, the capital goes to cash.
  • Risk can only be raised at a gate and on live evidence, never during a drawdown.

Method. Past: each strategy’s own backtest (daily net returns, after costs and funding) over the last 3 years, at the leg’s current risk level; the book uses today’s weights, separate accounts, no rebalancing. A few weeks of Nagi before 2024 use substitute price data. Future: … paths, 3 years; stationary block bootstrap (mean 20 days) from the joint daily sample since …, which keeps the legs’ co-movement; Sharpe drawn per path (…). The simulation ignores edge decay, the brake rules and new deposits. Nagi’s weight assumes a planned new account.

Important. This is an internal analysis for information only, not an investment offer, advice or a promise of returns. Past figures come from simulation; past performance does not guarantee future results. Leveraged trading can result in the loss of all capital invested.