BTC didn't just recover in 2025. It ran.
BTC opened this test on April 15, 2025 at $84,672.79 – climbing out of an earlier-year correction. USDC opened, unsurprisingly, at $1.00.
By September 30, BTC closed at $114,048.93. That’s a gain of $29,376.14 per coin — a 34.70% move in under six months. USDC closed at $0.9996, functionally flat.
For a spot holder, that meant one side of the portfolio did all the work. The question we wanted to answer:
Does actively rebalancing between a rocket and an anchor help, or does it just cap the rocket?
We ran this backtest on real Binance 1-minute OHLCV data across the full 168-day window to find out.
BITCOIN (BTC) — 50% TARGET
USD COIN (USDC) — 50% TARGET
Strategy Parameters
How Each Setting Impacted Performance?
Every parameter had a job. A strong one-directional rally exposed exactly what each one does.
Parameter Impact Summary
| Parameter | Impact | The Logic (Why) |
|---|---|---|
| 50/50 BTC/USDC Allocation | 📉 Capped BTC's upside | Flat USDC couldn't chase BTC's rally |
| 1% Ratio Threshold | 🔄 Triggered 30 rebalances | Tightest threshold = most frequent selling |
| By Coin Ratio Logic | ⚖️ Sold BTC strength repeatedly | Systematic profit-taking on every rally leg |
| No Time Rebalance | 🛡️ No added fee drag | Only 1% drift decided every trigger |
30 swaps. $4.67 in fees. $582.28 in profit — $18 short of just holding.
📝 The math that matters
💰 The Bottom Line
This strategy delivered $582.28 net profit on $3,500 over 168 days – a 16.64% total return. Simple-annualized (ROI × 365/168), that’s roughly 36.2% a year. Treat that number carefully: it assumes this exact rally repeats four more times in a row, which it won’t.
⚡ Activity Level
30 total actions over 5.5 months works out to one trade roughly every 5.6 days. Average fee per trade was just $0.16. Average net contribution per rebalance was around $20 — small, steady, mechanical.
🛡️ The Fee Advantage
Total fees came to $4.67 against gross profit of $586.95 — a fee drag of just 0.8%. At a 0.1% rate with only 30 swaps, cost was never the problem here. Opportunity cost was.
Here comes our A/B/C strategies quick comparison:
| Variant | Threshold | Trades | ROI % | P&L (USDT) |
|---|---|---|---|---|
| CThis Playbook | 2% | 9 | 16.71% | $584.93 |
| B | 5% | 2 | 16.69% | $584.31 |
| A | 1% | 30 | 16.64% | $582.28 |
The table reveals something counterintuitive: the 2% threshold delivered the highest ROI with only 9 trades. Variant A (2%, 9 trades) finished first at 16.71%, narrowly outperforming Variant B (5%, 2 trades) at 16.69% and Variant C (1%, 30 trades) at 16.64%.
Variant B traded the least but fell just short of A, while Variant C traded the most and still finished last. In a market where one asset trends hard in a single direction, every extra rebalance can lock in a sale slightly earlier than the last one – so the 2% threshold struck the best balance between trade frequency and performance. None of the three, however, caught up to simple HODL.
What the results are really telling you.
✅ what worked
The threshold logic did exactly what it was told: 19 of the 30 trades sold BTC strength into USDC as the rally extended, and 10 bought BTC back on dips (e.g., the May 30–June 17 cluster, buying BTC in the $103k–$108k range during a pullback from $108k highs).
The September 22 trade — selling 0.000401 BTC at $112,065.41 for $44.89 — was the single largest rebalance, locking in gains near the test’s final stretch. Fee cost stayed negligible throughout: $4.67 total across 30 swaps, just 0.8% of gross profit.
⚠️What didn't work
Every “sell BTC” trigger capped upside during a rally that kept climbing. The April 21–25 cluster sold BTC at $88,064, $91,654, and $95,513 — all prices BTC would blow past on its way to a July peak near $122,099.
Those early trims locked in modest gains that look small in hindsight. This is rebalancing’s core limitation in a trending market: it can’t tell the difference between “this rally is done” and “this rally has 40% further to go” — it sells into both the same way.
💡 The key insight
In a strong uptrend, discipline can cost you money.
Rebalancing is built to sell strength and buy weakness — a genuinely useful habit when assets oscillate. But BTC didn’t oscillate here; it trended.
Every one of the 19 “sell BTC” triggers took capital out of the exact asset producing all the portfolio’s returns, and handed it to an asset (USDC) that returned essentially nothing. The bot didn’t fail. It just did its job in a market where that job worked against it.
🚩 Watch out for - a potential red flag
Pairing BTC with a stablecoin means one side of the portfolio is designed to never move.
That’s a feature in a crash — USDC would have protected half the capital. But in a one-sided 34.70% rally, that same stability became dead weight, and every rebalance transferred upside into it.
The mandatory end-date USDT conversion also locks in whatever the split happens to be on September 30, regardless of where BTC goes next. Before running this exact 50/50 BTC/USDC config, ask whether you’re trying to protect against a crash or capture a rally — this structure is built for the former, not the latter.
🏆 Score: 6.8 / 10 — Solid Return, Capped by Design
Solid return, but capped by design.
16.64% over 5.5 months is a real, positive result with clean fee management. But it trailed the simplest possible alternative — doing nothing — by half a point, in a market this strategy is structurally not built for.
🧭 What this strategy does well
- Fee drag was negligible (0.8% of gross profit)
- Captured profit methodically on 19 separate rally legs
- Bought 10 dips automatically without emotion
- Only 30 total actions across 168 days — low maintenance
🚫 What went wrong this month
- Underperformed HODL by −0.52% (~$18 opportunity cost)
- Every rally leg triggered a sell, capping BTC's full 34.70% move
- The tightest threshold tested (this config) traded the most and earned the least
- USDC ballast returned effectively 0% (−0.04%), acting as dead weight in a trending market
Quick Takeaways
- Tighter thresholds don’t mean better returns in strong trends
- This config capped BTC’s rally rather than amplifying it
- A stablecoin ballast underperforms in one-sided bull runs
- Wider thresholds (2–5%) matched returns with far fewer trades
- Fees were never the real cost here – opportunity cost was
How did passive HODL compare?
If you had simply bought $1,750 of BTC at $84,672.79 and $1,750 of USDC at $1.00 on April 15, 2025, and held until September 30, here’s how it compares:
Running this rebalance config cost about $18 versus simply holding: $582.28 vs. an estimated $600.25. In a genuine trending bull market, doing nothing beat active rebalancing here.
Before you run this playbook, check these off.
Use this as your go/no-go checklist before deploying this exact parameter set.
🧠 Market Suitability Matrix
| Market Condition | Performance (Rating) | Strategy Notes |
|---|---|---|
| Both assets sideways / choppy | ★★★★★ Excellent | Harvests spreads without giving up trend upside |
| One asset dips, then recovers | ★★★★★ Ideal | Buys dips automatically — 10 of 29 trades did exactly this |
| Both assets in a mild bull market | ★★★★☆ Good | Both sides contribute, less one-sided capping |
| One asset strongly outperforms (this test) | ★★★☆☆ Moderate | Capped BTC's 34.7% rally, trailed HODL by −0.52% |
| Both assets in steep decline | ★★☆☆☆ Risky | Redistributes capital between two losers |
| One asset in structural breakdown | ★☆☆☆☆ Poor | Forces buying into a collapsing asset |
| Highly correlated assets (same direction) | ★☆☆☆☆ Poor | No spread to harvest between synced assets |
How to tune this playbook for different scenarios.
Disclaimer: All data sourced from CryptoGates Rebalance Backtest Bot. Results are historical simulations using Binance 1-minute OHLCV data. Past backtest performance does not guarantee future live trading results. DYOR.
Battle-Test Your Strategy
Before the Market Does.
Eliminate guesswork with institutional-grade backtesting for DCA, Grid, and Rebalance bots. Real historical data. Real-world results.


