Tack follows trends when Bitcoin's daily trend is positive. When it is not, the strategy switches to buying deeper pullbacks in stages.
The switch is simple: Bitcoin's last completed daily close is compared with its 100-day moving average. Above it, Tack is in bull mode. Below it, Tack is in bear mode. The decision is made from closed daily candles, not from an intraday guess.
In bull mode, each market must also be above its own 50-day average and pass a Kaufman efficiency-ratio filter. In bear mode, the trend positions are closed and a seven-day dip ladder takes over. The system is long-only; it never opens a short position.
Across the test window, bull mode was active 55.8 % of the time, with 75 regime changes.
| Hard gross exposure cap | 0.85× |
| Position leverage | 1.15× isolated |
| Trend allocation per active market | up to 40 % |
| Minimum free-collateral reserve | 15 % |
| Trend rebalance band | 10 % |
| Live markets | 9 |
This run covers 1 October 2020 through 31 March 2026 on native five-minute price data. It uses the current 0.85× gross cap, 1.15× isolated leverage, 40-% trend allocation, 10-% rebalance band and 15-% reserve. Taker fees of 4.5 basis points, seven basis points of slippage and an assumed 10 % annual funding cost on open exposure are included.
These are historical model results, not a return target. Much of the compounding came from strong bull phases, especially 2021 and 2025. The figures include trading costs and funding, but not the vault's 10 % performance fee.
| Phase | Tack | per year | Sharpe | daily drawdown | Buy & hold BTC |
|---|---|---|---|---|---|
| Full cycle | +10,820 % | +134.9 % | 1.693 | −35.6 % | +543 % |
| Bull market 2020–21 | +1,023 % | +699.5 % | 2.565 | −35.6 % | +436 % |
| Bear market 2022 | −6.8 % | −6.3 % | −0.212 | −26.6 % | −71.1 % |
| Recovery 2023–26 | +943 % | +106.0 % | 1.748 | −17.8 % | +311 % |
The difficult stretch is still 2022. Tack lost 6.8 % while Bitcoin lost 71.1 %. That is a useful difference, but it is still a losing period.
| Year | Tack | Buy & hold BTC |
|---|---|---|
| 2020 (from Oct) | +69.7 % | +172.4 % |
| 2021 | +570.5 % | +59.8 % |
| 2022 | −8.0 % | −64.2 % |
| 2023 | +90.1 % | +155.6 % |
| 2024 | +99.9 % | +121.3 % |
| 2025 | +185.0 % | −6.3 % |
| 2026 (Jan–Mar) | −3.2 % | −22.1 % |
The years are uneven. A long-run average hides that: two periods can use the same rules and still produce very different outcomes.
The live vault adds MON and LIT to the seven historical markets below. Neither has enough suitable history for the full-cycle test. In a narrow bull phase, one active market can represent up to 40 % of equity before the portfolio-level cap takes over.
| Live market | Included in published backtest | History used |
|---|---|---|
| BTC | Yes — BTC | Aug 2020 |
| ETH | Yes — ETH | Aug 2020 |
| SOL | Yes — SOL | Aug 2020 |
| ZEC | Yes — ZEC | Aug 2020 |
| SUI | Yes — SUI | May 2023 |
| HYPE | Yes — HYPE | May 2025 |
| XRP | Yes — XRP | Aug 2020 |
| MON | No | Not backtested |
| LIT | No | Not backtested |
Assets enter the calculation only after their price history begins, so the early years contain fewer than seven markets. MON and LIT are live but are not hidden inside the headline backtest number.