Honest About Our Performance

This page used to open with a headline: our MACD+ADX strategy beat buy-and-hold for BTC by +46.32% over five years. We retracted that number. The backtest behind it filled every trade at the same candle's close — a price you can't actually get, because the crossover is only knowable after the close. When we re-ran the whole study with realistic next-morning fills, BTC's five-year result flipped from +46% to roughly 10 points behind buy-and-hold.

Most apps would have quietly deleted the number. We'd rather show you the correction, because it's a better introduction to how we work than the original headline ever was.

Key Takeaways

• Our BTC signals don't beat buy-and-hold — not per quarter (-1.89% mean alpha) and not over five years (realistic fills put the total slightly behind holding)
• The value is concentrated in crash quarters: Q2 2022 delivered +34.77% alpha while BTC fell 56.8%
• A few assets do show consistently positive alpha (ADA, XRP, ATOM, AVAX) — and in Q4 2023 every asset misfired at once. Both halves of that sentence matter.

The Full Scorecard

Here's how each asset performed with our best strategy configuration. Assets with 5 years of data (20-22 quarters) are shown first; provisional assets with shorter histories are noted separately.

Provisional (limited data):

Five assets with positive mean quarterly alpha. Five negative. Two provisionals that look promising but rest on thin data. That's the honest picture. (One caution on AVAX: a 100% beat-hold rate over 20 quarters is the kind of number that makes us more suspicious, not less — it uses plain MACD with no ADX gate, and we treat it as one five-year sample, not a law of nature.) DOGE's row reflects the same fill-and-compounding correction that flipped BTC's headline: under the corrected math, DOGE trails buy-and-hold at every ADX configuration we tested, and we retracted its old profit-factor headline (1.758 — actually 1.42) along the way.

Where It Actually Works: ADA

ADA is different from everything else we track on the daily timeframe. +8.3% mean quarterly alpha over five years under our refined setup — a deferred-confirmation rule with an ADX>30 gate — beating buy-and-hold in 15 of 20 quarters. And the edge didn't appear only under the refined rule: even our original, stricter-timed gate at ADX>20 showed positive mean quarterly alpha (+1.7%) on realistic fills. The refined rule didn't create ADA's edge, it stopped throwing away legitimate entries.

Why? ADA's big moves follow trend structure. They accelerate, plateau, reverse. MACD can detect that pattern. When ADA crashes — and it crashes hard, -60.5% in Q2 2022, -60.9% in Q4 2025 — the MACD exit signal gets you out before the worst of it. And when it rallies, the ADX filter helps you re-enter during real trends, not false starts.

ADA's worst quarter: -37.25% alpha in Q1 2025. We lead with the worst one on purpose — the best quarters take care of their own publicity.

LINK: The Discipline Play

LINK came in at +0.70% mean quarterly alpha. Essentially breaking even with buy-and-hold — while being in the market only 53% of the time.

That's not an alpha story. It's a discipline story. You'd spend half your time in cash, make about the same return as holding, but with dramatically less overnight exposure and a systematic rule for exits instead of gut feelings. During Q2 2022, that discipline was worth +34.09% alpha as LINK crashed 63.8%.

Worst quarter: -63.93% in Q4 2023, when LINK rallied 85.6% and the strategy was on the wrong side. Real tail risk. We don't hide it.

BTC and ETH: The Quarterly Drag Is Real

BTC at -1.89% and ETH at -0.54% mean quarterly alpha — both negative on a per-quarter basis. Most quarters, following the signals costs you a little compared to just holding. And here's the part we got wrong for a while: the total doesn't rescue it. Under realistic next-morning fills, five years of BTC signals finish behind buy-and-hold too.

So what are the signals for? The crash quarters, which are real and asymmetric.

Q2 2022: BTC dropped 56.8%. Our signal fired. Users who followed it avoided the worst of the crash: +34.77% alpha in a single quarter. ETH dropped 69.0% — signal alpha was +29.25%. Those quarters are enormous. They're just not frequent enough to add up to outperformance.

The system beat buy-and-hold in 50% of quarters for BTC and 55% for ETH. That's roughly a coin flip on any given quarter, with the biggest single wins landing in crashes.

Which is why the emotional reality is the actual product here. If you can hold through a 57% drawdown without panic-selling at the bottom, you probably don't need us for BTC. Most people can't. A systematic exit rule that gets you out of the worst quarters — and costs you a little drag the rest of the time — isn't a trading edge. It's a way to stay sane while holding a violent asset. We'd rather tell you that plainly than show you a headline number we had to walk back.

Q4 2023: When Everything Failed Together

We always disclose this. Q4 2023 was a broad crypto bull quarter. Every asset we track rallied hard. Every strategy we run was poorly positioned.

All strategies underperformed simultaneously. Diversifying across assets didn't help because the failure was systematic — MACD timing was off at the start of a broad rally. When the whole market moves at once, signal-based systems can all be on the wrong side at the same time.

This is a real risk. We can't engineer it away. We can tell you about it.

What We Know and What We Don't

We know this strategy has generated positive alpha for ADA across five years of backtesting. The data is clear and consistent.

We know it detects bearish momentum shifts for BTC and ETH — the smoke detector works when it matters most.

We know it can fail across all assets at the same time during strong bull quarters.

What we don't know: whether these patterns will hold for the next five years. Five years of data is substantial, but crypto markets are young. Regulatory changes, new technologies, shifts in market structure — any of these could change how our signals perform.

We're not going to pretend we have a crystal ball. We have five years of backtested data, all 20 quarters visible, including the ugly ones. Here's the data — judge for yourself.

We don't beat buy-and-hold for BTC — we said we did, we were wrong, and we fixed it. What the signals actually earn is crash avoidance for BTC and ETH, consistent positive alpha for a few assets like ADA, and every quarter shown — not just the good ones.

This is educational content, not financial advice. Past performance does not guarantee future results.