The page to read before the first dollar moves

1. Introduction and general warning

Crypto trading is speculation, plainly: prices swing wide and fast, and losses ride the same express as gains. Before trading, write down your objective, your horizon and your maximum acceptable loss. Automation executes rules; it does not convert uncertainty into certainty, however tidy the engine. The money split that saves accounts is three layers, emergency reserve, goals for the year, risk capital, with only the third ever trading, and the effect is psychological as much as financial, because with the bills unreachable a red dashboard becomes information instead of an emergency, and calm decisions cost a fraction of panicked ones.

2. Market risk: volatility

Volatility is the amplitude and speed of price movement: an asset can add ten percent before lunch and hand it back by dinner, and losses become real at the moment of a bad exit, which engine rules do not waive, they follow a script in a market that signs nothing.

Practical recommendation: fix the share of your savings assigned to this class and honour the ceiling through winning streaks too. Weigh price risk against time risk, since assets may recover in months while capital with a deadline cannot wait. Treat correlation with respect, because diversification fails exactly when needed, when "uncorrelated" crypto assets fall as one and everyone sprints for the same exit, and treat previous highs as history rather than obligation, as sizing from old peaks is the most reliable known method of quietly oversizing a position.

3. Liquidity risk and slippage

Liquidity is leaving a position without pushing the price, and in thin markets or panics orders fill worse than the screen promised, the difference being slippage, which grows with your order relative to the market.

Practical recommendation: the tell-tale of a shallow market is your own order turning the quote, so if it happens, shrink the size or change assets, and stay out of the first minutes of violent moves while prices jump whole levels. The sizing corollary follows straight from slippage: the exit deserves more planning than the entry, since buying quietly in a calm market is easy while leaving a crowded one is where size punishes hardest, exactly why per-position limits are engineered into the strategies instead of left to in-the-moment judgement.

4. API and integration risk

Automated trading speaks to exchanges through API keys, and a misconfiguration, an exposed key or instability on the far side interrupts the operation or opens a gap, mitigated here by minimal scopes, IP pinning and instant revocation as detailed on the security page. Treat key hygiene as routine rather than event: a monthly read-through of active keys, a revocation test on one old key and a permission check on the exchange side, ten minutes that remove the commonest incident family in automated accounts, the forgotten wide key. The external dependency cuts the other way as well, since an exchange pausing withdrawals for maintenance leaves the strategy waiting even with our platform fully operational.

5. Counterparty and custody risk

Funds normally sit with exchanges and third-party providers, whose financial, technical or legal troubles become your access troubles. Spreading across providers trims the concentration, and each provider's record should weigh on the choice. An objective checklist for counterparties: order-book depth, response times on busy days and a public incident history, because an exchange untested by stress is an unknown on precisely the worst day, which is when the quality of the choice reveals itself.

6. Operational risk

Software defects, infrastructure outages, network drops and provider failures delay execution at the exact moment it matters most, and while the platform runs redundancy, no system is immune to failure. Distinguish brief from long outages: minutes rarely move an outcome, hours can leave a position without its protective order, incidents are announced with estimated duration, and the account history lets you rebuild what executed inside the affected window.

7. Cybersecurity and phishing

The preferred target remains the person: platform-lookalike emails, cloned sites, fake social offers, and a reused password is the cheapest house key an attacker can copy. The base defence that works is short: 2FA on, unique passwords per service, official domain checked before typing anything.

Practical recommendation: an urgent message asking for a password or code is a scam by definition, as the platform never asks, and it belongs forwarded to [email protected] rather than answered.

8. Model and automation limits

The engine recognises historical patterns and markets break patterns unannounced; AI does not foresee the unforeseeable and no bot assures results, while the volatility brake trims exposure in storms but cannot refund what the violent stretch already took. Limitation two, unvarnished: models err in series when the regime changes until the new context is absorbed, history supplying frequencies rather than promises that quietly expire when the regime turns, months of sideways ended by a macro headline doubling volatility in two days and invalidating old patterns for weeks. A strategy blind to that trades the wrong regime, the per-strategy performance panel exposes the divergence in days, and the pause decision stays yours, informed.

9. Service availability

Crypto trades 24/7 while platforms need maintenance, and updates or external events can take the service down for windows at a time, during which the dashboard is unreachable and settings frozen while active strategies keep following their rules. Recommendation: scheduled maintenance is announced in advance inside the platform, and reading the notices costs seconds, because surprises and money make poor housemates.

10. Before you start

Four actions before the first deposit: understand the strategy, with your manager explaining what each does and where it trades; fix your maximum loss, the number going on paper before the money moves; enable 2FA, day one, no exceptions; and supervise, dashboard plus weekly report, adjusting when results drift from plan as routine rather than failure. Revisit this page every six months as well: with real miles on the account the theory gains weight, and stale limits get refreshed.