Best Practices¶

Risk management (the only thing that matters long-term)¶

You can't predict markets. You can control how much you lose when wrong. That's the entire skill.

The hierarchy of risk control¶

1. Position sizing — risk a fixed percentage of equity per trade (commonly 0.25%–1%). This is the single most important number.
2. Stop losses — predetermined exit if the trade goes wrong. Set before you enter.
3. Diversification — don't have multiple trades depending on the same factor (sector, theme, macro).
4. Portfolio heat — sum of all open trade risks. Cap at ~5% (you survive a market-wide bad day).
5. Drawdown limits — pre-commit to a level (e.g., -20% account drawdown) where you stop trading and reassess.

Math you must internalize¶

If you lose 50%, you need +100% to break even. Drawdowns are asymmetric:

This is why "small losses, let winners run" beats "I'll make it back." You can't make it back from large drawdowns.

Risk:Reward and win rate together¶

A profitable system needs:

expectancy = (win_rate × avg_win) - (loss_rate × avg_loss)

A 40% win rate is fine if avg_win = 2× avg_loss. A 60% win rate is a disaster if avg_win = 0.3× avg_loss.

Never evaluate your strategy on win rate alone.

Process discipline¶

Always-on rules¶

• No trade without a written plan: entry, stop, target, position size, thesis.
• Place the stop with the entry, not after. Brackets / OCO orders make this physical.
• Don't move stops further away. Ever. A widened stop is a confession.
• Never average down on a losing trade. You're adding to a losing thesis.
• One trade per setup, not three "in case it works." That's emotional sizing.
• End of day: log every trade. Date, ticker, setup, entry, stop, target, exit, P&L, what happened, what you felt.

The trading journal¶

Without a journal, you're not trading; you're guessing and forgetting. Minimum columns:

R = the trade's outcome in units of risk (-1R, +2R, +0.5R, etc.). Aggregate over many trades and you have your real expectancy. Most systems people think are profitable are not, when measured this way.

Weekly review¶

Every Saturday morning (or whatever your "off-market" window is):

1. Sum the past week's R outcomes. What's the running average?
2. Did every trade follow the plan? Mark "yes/no". Track the yes-rate; that's your discipline metric.
3. Did any "rule" need to change because of new information, or because you wanted it to be easier to win? The first is OK; the second is dangerous.
4. Were there setups you saw and didn't take? Why?
5. Are you over-trading? Most retail accounts have too many trades, not too few.

Psychology¶

The classic lessons keep being true because human nature is constant:

• Fear and greed are equally dangerous. Fear makes you cut winners early; greed makes you hold losers and over-leverage.
• The market doesn't owe you a comeback. Losses are sunk cost. Each trade is independent of the last.
• You will have streaks of luck (good and bad) far longer than your intuition expects. A 5-loss streak in a 50%-win-rate system happens 3% of the time. Don't conclude your system is broken from one bad month.
• Process > outcome. A good trade can lose money; a bad trade can make money. Judge yourself by adherence to plan, not by P&L.
• Take a break after a big win OR loss. Both alter your judgment.
• Don't trade on tilt. If you're angry at the market, you've already lost the next trade.

Taxes (US-focused, generic, NOT advice)¶

Account hierarchy¶

• Tax-advantaged first — 401(k), IRA, Roth IRA, HSA. Trading inside these has no annual tax drag.
• Taxable brokerage — every realized gain is taxed. Short-term gains (held <1 year) at ordinary income rates. Long-term gains (≥1 year) at preferential rates (0/15/20%).
• Wash sales — selling at a loss and rebuying within 30 days disallows the loss in the taxable account; the basis is shifted into the new lot. Adds complexity but isn't fatal.

For active traders specifically¶

• Track P&L in real time, not at year-end. Brokers provide 1099-B forms but they have errors; reconcile.
• Consider mark-to-market election (IRS Section 475(f)) if you qualify as a "trader in securities." Lets you treat trading as a business, deduct expenses, but converts gains/losses to ordinary income (no LTCG rate). Talk to a CPA.
• Estimated taxes: if you're profitable, you owe quarterly. Penalties for under-payment.

Brokers and executions¶

• Default to limit orders. A few cents per trade compounds.
• Use a real broker for active trading (IBKR, Schwab, Fidelity). Commission-free brokers route to PFOF venues; for most retail, the price impact is small but real.
• Read your fills. If you're consistently getting filled at the worst price in the spread, your broker's routing is bad.
• Watch fees: data subscriptions (Level 2, real-time futures), inactivity fees, regulatory fees. They add up.

What pros do that retail rarely does¶

• Pre-commit to rules. Then make those rules harder to break (auto orders, brackets, removed UI temptations).
• Pre-mortem every trade: "If this stops out, why?" If you can't articulate a clean reason, the thesis is fuzzy.
• Backtest before trading live. "I think this works" is not data.
• Walk away from the market regularly. Hourly breaks during the session, full days off weekly, months off yearly. Burnout is a real failure mode.
• Maintain a watchlist, not a wish-list. Names to monitor for setups, not "stocks I think will go up."
• Pay attention to base rates. Most strategies you can think of have been tested by quants for decades. Read the academic literature; don't reinvent.

Pitfalls that bite everyone¶

1. Confusing one good year with skill. Bull markets cover a lot of bad strategy.
2. Sizing up after wins ("I'm in the zone"). The math doesn't change because of recent emotion.
3. Sizing up after losses ("I'll make it back"). Worse.
4. Switching strategies after a drawdown. A drawdown of expected magnitude is not evidence of failure.
5. Chasing newsletter / Twitter picks. Anyone confidently sharing trades publicly has a different P&L motive than yours.
6. Trading in retirement accounts as if they're playgrounds. The tax shelter is precious; preserve it.
7. Neglecting the boring core. A great satellite portfolio + mediocre core portfolio still loses to mediocre satellite + great core.