Sizing your monthly DCA by where Bitcoin sits in its cycle accumulated more Bitcoin per dollar than buying a fixed amount. Here is the result, and exactly how we tested it.
What if your DCA responded to market conditions instead of buying the same amount every month? We tested exactly that — and measured it the honest way.
Cycle-smart DCA vs. fixed monthly DCA over the Jul 2022 – present window (~one Bitcoin cycle).
The most common way a backtest misleads you is hindsight. Here's how we ruled it out.
It's easy to quietly tweak a strategy against past data until it looks perfect — that's called curve-fitting, and it makes almost anything look brilliant in a backtest. We ruled it out by writing down the exact method and timestamping it publicly before running the test, so we couldn't peek at the results and adjust to fit them.
The signal was calibrated on one stretch of history and then measured on a completely separate, later stretch it was never tuned on (this is called “out-of-sample data”). The ~27%-more-Bitcoin result is from that unseen window — not the data it learned from.
When you try several ideas, one can beat a benchmark by chance. We ran a standard statistical test that accounts for exactly that — the edge held up with a very high degree of confidence, and it stayed positive across every individual year we removed. It didn't depend on one lucky year.
The Cycle Factor is a multiplier that scales your buy amount based on the signal — higher in historically depressed conditions (Winter), lower when valuation is historically stretched (Summer). You pick how strongly it swings in Account Settings (sign up to enable).
Gentler swings. Lower variance around your base amount.
The published signal. This is the setting the validated result was measured on.
Wider swings. Leans harder into depressed conditions, backs off more when stretched.
BitcoinIQ reads 20 indicators and distills them into a single Cycle Position — then maps that to your Cycle Factor.
Every indicator is published and refreshed daily, each with its own AI analysis. How they combine into the single Cycle Position is the proprietary part — the ingredients are open; the recipe is ours.
On-chain valuation historically depressed (Winter)
Market internals leaning supportive
Mixed signals, no directional bias
Valuation in historically extended territory
Valuation at historically extreme levels (Summer)
Cycle Factor applied to your custom base amount. Conservative (0.5x–1.5x), Standard (0.25x–2.0x), Aggressive (0.1x–3.0x). The signal never drops to zero — you never stop stacking, you just stack smarter.
Start your free trial to see your personalized Cycle Factor based on current market conditions.
Current market conditions and your personalized Cycle Factor.
What we measured: cost basis — the average price paid per Bitcoin — for signal-sized DCA versus fixed-amount DCA. Cost basis is measured per dollar invested, so the comparison is fair even though the two approaches invest different total amounts (more when valuation is depressed, less when it's stretched).
Out-of-sample: the signal was calibrated on an earlier period, then evaluated on a separate later window (Jul 2022 – present) it was never tuned on. The ~27%-more-Bitcoin result comes from that unseen window.
No lookahead bias: each buy decision uses only the most recently finalized monthly signal — exactly what a real user would have seen at the time.
Single-cycle limit: the out-of-sample window spans roughly one Bitcoin cycle. That is genuine evidence, not a guarantee across all future regimes — which is why the live, real-money track record continues the test going forward.
Past performance does not guarantee future results. This is educational content, not investment advice.
NOT INVESTMENT ADVICE
BitcoinIQ provides educational content and analysis tools for informational purposes only. This is not investment, financial, or trading advice. Cryptocurrency investments are highly volatile and risky. Always do your own research and consult with qualified financial advisors before making investment decisions. Past performance does not guarantee future results.
v1.0.0 (build 20260724.1802 • d55d58c)