Why Choose unCoded Trading Bot

We only make money when you do.

Self-hosted, non-custodial crypto bot trading with profit-sharing pricing, honest backtests, and real strategy depth.

See Public Backtests • Read the Docs

Disclosure: This page describes the design choices behind unCoded. Operated by ArrowTrade AG, Switzerland. Bot trading involves risk – past performance and backtest data are not indicative of future results. This is not financial advice.

How unCoded differs

The differences between unCoded and typical subscription-based bot platforms are structural, not cosmetic. Here's how the two approaches compare across the dimensions that actually matter.

Pricing. Subscription platforms charge a monthly fee regardless of trading results. unCoded charges 30% of generated profits, dropping permanently to 20% after $2,000 in cumulative fees.

Bad months. Subscription platforms charge their full fee even when your strategies lose money or sit flat. With unCoded, bad months cost zero. No profit means no fee.

Custody. Subscription platforms require your API keys to be stored on their infrastructure. unCoded is self-hosted – your API keys never leave your environment, and even a complete compromise of unCoded's servers cannot affect your exchange account.

Execution. Subscription platforms typically poll the exchange at intervals between 30 seconds and 5 minutes. unCoded processes every order book movement in real time, which makes strategies possible that polling-based platforms simply cannot execute reliably.

Backtesting. Subscription platforms generally backtest on single charts with candle-close evaluation. unCoded backtests multi-chart against the entire Binance Spot market, with intracandle evaluation on 1-second base data – exposing overfitting that single-chart backtests systematically hide.

Distribution shown. Subscription platforms publish cherry-picked highlights from favorable periods. unCoded publishes the full distribution of results, including the years where strategies failed on the majority of tokens. Hiding the bad years would be inconsistent with profit-sharing pricing.

Default settings. Subscription platforms often ship with aggressive defaults that look better in marketing backtests but destroy capital in unexpected conditions. unCoded ships with conservative defaults – canBuyDown=false, mandatory stop losses, maximum exposure caps – because long-term stable users matter more than short-term marketing screenshots.

What this looks like for your portfolio

The pricing structure becomes concrete fast. The numbers below are a hypothetical illustration based on a 15% annual gross return.

Important: 15% is a hypothetical return used for illustration only. Actual results vary significantly and may be negative.

On a $10,000 portfolio, the unCoded fee in year one would be approximately $450 (30% of $1,500 in hypothetical gross profit). The fee remains $450 in subsequent profitable years at the same return level. In any year where the portfolio doesn't generate profit, the fee is $0.

On a $50,000 portfolio, the fee in year one would be approximately $2,167 – calculated as 30% on the first $6,667 of profit, then 20% on the remaining $833 above the threshold. From year two onwards, with the threshold already crossed, the fee drops to approximately $1,500 (20% of $7,500). Bad years cost $0.

On a $200,000 portfolio, the fee in year one would be approximately $6,667 – 30% on the first $6,667 of profit, then 20% on the remaining $23,333. From year two onwards, the fee drops to approximately $6,000 (20% of $30,000). Bad years cost $0.

The mechanic in one line: The 30% rate applies only to your first $6,667 of cumulative lifetime profit. Every dollar of profit above that threshold is capped at 20% permanently.

The point: Subscription platforms charge in losing years, flat years, and great years equally. Profit-sharing platforms charge zero when there's nothing to share. The longer you operate, the more favorable the math becomes relative to fixed monthly drag.

The pricing model is the structural foundation. When our revenue depends entirely on users actually generating profit, every product decision over time aligns with that outcome. Honest backtesting. Conservative defaults. Real strategy depth instead of marketing screenshots. A platform that earns regardless of user outcomes optimizes for retention. A platform that earns only when users earn optimizes for profitability.

The reason unCoded exists

In 2020, I subscribed to Cryptohopper Hero tier at $130/month. Conservative strategies. Multi-coin Spot trading. No leverage.

Five years later, I'd paid roughly $7,800 in subscription fees, around €2,000 in tax software, German income tax on every realized gain, and my net return after all costs was consistently worse than simply holding Bitcoin.

The strategies executed correctly. The platform worked as advertised. The economics didn't.

That experience built unCoded. The structural problem isn't bots – it's the pricing model retail bot platforms have settled on, and every architectural decision that flows from it.

What makes unCoded different (briefly)

Self-hosted, non-custodial

unCoded runs on your VPS. Your CapRover installation. Your API keys, stored exclusively in your infrastructure.

The retail bot industry has seen multiple publicly reported incidents where user credentials on cloud platforms were compromised. With unCoded, even a complete compromise of our servers cannot affect your exchange account – the credentials were never there.

The trade-off: self-hosting requires technical capability cloud platforms don't demand. That friction is the point. It filters for users who take custody seriously.

Tick-based execution

unCoded processes every order book movement in real time. Most retail bots poll at 30-second to 5-minute intervals.

Polling creates blind spots. Stop losses miss wicks. Take profits trigger late. Strategies that require precise execution – per-split trailing stops, time-based profit curves, micro-trading with hard stops – simply don't work on polling architectures.

Honest backtesting

Multi-chart testing against the entire Binance Spot market. 1-second base data with intracandle evaluation. Correct fee calculation in the actual asset fees are paid in.

The full distribution is published publicly at uncoded.ch/backtesting, including years where strategies failed on the majority of tokens. We publish the bad years because hiding them would be inconsistent with profit-sharing pricing.

Strategy depth

Buy Splits across up to 7 independent levels. Sell Time Curves that adjust profit targets based on position age. Per-split trailing stops. Micro-Trading with hard stop losses. Signal Editor with 152 indicators and full boolean logic.

This is what becomes possible when the foundation supports it.

Conservative defaults

canBuyDown=false by default. Mandatory stop losses. Maximum exposure caps. Order minimums checked against exchange filters.

Aggressive defaults produce better marketing backtests and destroy capital in unexpected conditions. We choose long-term stable defaults because our business model depends on long-term stable users.

Who unCoded is built for

Serious Spot traders, portfolios $1,000–$500,000, who have outgrown DCA and Grid templates and want:

• Real strategy depth

• Honest backtesting

• Custody control over their own capital

• Pricing aligned with outcomes, not fixed monthly drag

Who unCoded is NOT for

This matters more than the previous section.

Not for leverage traders. Spot only. We won't give up that safety property.

Not for beginners. No trading experience = don't start with a bot. Bots automate execution, they don't replace strategy development.

Not for set-and-forget. Bot trading requires active monitoring and intervention. Platforms promising autopilot wealth are lying – we say so even though it costs us users who'd prefer the lie.

Not for users who don't want to manage a VPS. Self-hosting is real operational responsibility. If you don't want it, cloud platforms exist for a reason.

Not for copy trading or signal following. No marketplace, no copy features. Strategy understanding is a prerequisite in our model.

If any of these describe you, you're probably in the wrong place – and we'll tell you directly rather than try to sell you something that doesn't fit.

Where to go instead, if unCoded isn't right

We recommend specific alternatives because honest comparisons build trust better than dishonest ones save users.

• Multi-exchange hosted convenience: 3Commas

• Marketplace strategy access: Cryptohopper

• Grid trading specifically: Bitsgap

• Copy trading / signals: Wundertrading or 3Commas

• Beginners testing automated trading with zero commitment: Pionex

Get started

The practical path forward:

1. Look at the actual backtest data first. uncoded.ch/backtesting. See the full distribution across multiple years before anything else.

2. Read the documentation. uncoded.ch/docs. If the technical level feels accessible, you're in range.

3. Join the Telegram community. Active since the Aureum Victoria YouTube days. Real users, real questions, no AI chatbots filtering inquiries.

4. Start small. Deploy with 10-25% of intended capital for the first 60 days. Validate live behavior against backtest expectations before scaling.

5. Stay engaged. Bots that users actively monitor outperform passive deployments significantly. The work isn't optional.

Technical foundation

For users who want to know what's under the hood. Skip if not relevant.

Stack: Node.js with PostgreSQL. ACID guarantees on all state transitions. fulfilledOrders table is INSERT-only, enforced by database triggers – immutable audit trail for tax reporting.

Execution: Tick-based via WebSocket with polling fallback. Reconciliation loop runs every 2 seconds, comparing internal state to actual exchange state and correcting drift before next trading decision.

Backtesting: 1-second base candle data with intracandle evaluation. Multi-chart testing against full Binance Spot market. Realistic fee calculation in actual fee asset.

Deployment: Self-hosted via CapRover one-click install. Compatible with most VPS providers (Hetzner, DigitalOcean, Contabo). Zeabur template available.

Exchange support: Binance Spot live in production. 16 additional major exchanges (Coinbase, Kraken, KuCoin, OKX, Bybit, MiCA-compliant EU alternatives) in internal alpha. Each ships when production-ready, not when marketing pressure demands.

The bottom line

unCoded is built on one principle: a platform should only make money when its users do. Everything else flows from there.

This isn't a universal best platform. It's the right platform for traders who want pricing aligned with outcomes, custody control over their own capital, and strategy depth that subscription-driven competitors can't match because their business model doesn't reward it.

If that sounds like you, we built this for you.

If it doesn't, we've named the alternatives. Honest mismatch beats convinced fit every time.

The bot is a tool. The thesis is yours.

Get started:

• 📚 Documentation: uncoded.ch/docs

• 📊 Public backtest data: uncoded.ch/backtesting

• 💬 Telegram community: uncoded.ch

• 🛠 Deploy via Zeabur: zeabur.com/templates/FX6Y6O

ArrowTrade AG, Switzerland. The "unCoded" trademark is registered through EUIPO. Operating under the Swiss DLT regulatory framework.