Notes on trading.

31.4% of 1,076 human traders on Hyperliquid made money from May 2024 to October 2026. Half lost 90% at some point, small accounts did worst, and the top is mostly bots.

Botsfolio AI against 1,076 real-money human traders, same money, same dates. 31.4% of the traders made money, the AI 88.6%. Only 2.4% beat it on return and risk.

The same AI crypto portfolio returned 33.6% a year at a 0.10% fee and 17.1% at 0.90%, below holding. Which exchanges charge what, and how to pay less.

Six AI models traded $10,000 of real crypto each and four lost money. Our own test of a frontier model picking trades found no skill. What works instead.

Overfitting, look-ahead, missing fees and tiny samples make backtests lie. How to spot a curve-fit bot, and the tests a record should pass before real money.

Forget $ per day. What an AI crypto portfolio returned from 94 start dates since 2018: median, best and worst, the drops along the way, against holding.

Paper trading skips fees, slippage and fear. What changes with real money, and a checklist with numbers for deciding when an AI or a strategy is ready.

Yes, an AI trading bot can lose money, and with leverage it can lose all of it. What the worst cases look like across 94 start dates, and how to limit them.

Subscription bots cost $29 to $140 a month whether they trade or not. Per-trade pricing costs far less on small accounts. The full math from $500 to $50K.

Some AI trading bots are real tools and many are scams. The red flags regulators warn about, and five checks that tell a real record from a fake one.

Yes, AI can trade crypto for you on your own exchange account. What it decides, what stays with you, what it costs, and the record to check first.

We put seven AI-managed crypto portfolios against simply holding the same coins in 2026. Returns, the biggest drops, and when holding still wins.

Copy trading follows a person. AI trading follows a process. How each works, where copiers lose even when the leader wins, and how to check any record.

Crypto portfolio trackers show what you own. Which type fits you, what to check before connecting anything, and when you want an AI to manage it instead.

AI crypto trading bots buy and sell for you. How they work, the claims to ignore, and the one test that matters before you trust one: a public record.

Every guide admits paper trading overstates performance, then does nothing about it. The four reasons your paper results lie, and how to get a forward number honest enough to build a real edge on.

Forget "30 days" or "100 trades." You're ready when your strategy has a measured edge that survives out-of-sample and a second coin, and a forward paper record confirms it. How to test both, with real data.

The rule most crypto challenges use to disqualify a trader who was already in profit. What the 30% single-day cap actually blocks, the math on a $50k account, and which firms let you skip it.

Revenge trades look exactly like normal trades until the equity curve rolls over. What a companion measures per-trade to catch tilt before the firm's daily drawdown limit does, and what to do when it flags you.

Manual trade journals capture what you were thinking. An auto-journal captures what you actually did, tagged with structural context. Here is what changes about your review when the log writes itself.

The seven daily-PnL shapes that show up in a losing trader's calendar before the account gets wiped. What each pattern looks like, why it forms, and how to catch it in your own history.

Staking, lending, yield farming, systematic trading, or letting an AI trade for you. Real yields, real risks, and which playbooks survive a bear market.

From what Bitcoin actually is to how modern DeFi works, in a single sitting. No filler, no fluff. The foundation you need before you touch a trade.

Highs, lows, breaks, and character shifts. The framework every SMC trader uses to read the tape, from higher timeframes down to the entry candle.

Spot vs perp, order types, taker vs maker, funding, liquidity, and how a trade actually executes on a modern exchange. The plain-language explainer.

The first 30 days of trading crypto without blowing up. Wallets, exchanges, position sizing, journaling, and what to actually read on a chart before you touch a trade.

Mean-reversion, breakout, grid, SMC. The strategies bots actually run, how each one behaves in a real market, and why a companion can trade the same edge without unsupervised risk.

Not the tired cliches. Real, quantifiable investor mistakes: buying tops, panic exits, position-size drift, ignoring correlation, chasing narrative. What they cost, and how to catch them early.

Charting, execution, journaling, backtesting, alerts. What each category is actually good for, and which tools deserve a spot in a modern trader's stack.

Full-auto bots and pure-manual trading are both flawed. The AI trading companion is a third category that keeps the trader in the loop while removing the invisible mistakes.

Over-concentration, correlation blindness, and rebalancing on emotion. The portfolio-level mistakes most active traders never audit, and how to see them in your own record.

The three behavioral patterns that show up in almost every losing trader's history. What they look like on your PnL, why the brain does them, and how to see them in your own record.

One SMC confluence has outperformed every standalone pattern we backtested on BTC, ETH, SOL. Real data, real chart, and why it fails in the 3% of setups that go wrong.

The four measurable habits that separate the top-decile of crypto traders from everyone else. Not chart patterns. Not indicators. Behavior you can audit in your own trade history in an hour.