An investment journal is a structured record of your decisions, reasoning, and outcomes. It turns market noise into data you can learn from.
What you'll learn
How to structure an investment journal that links thesis, process, and outcomes
Advanced analysis tools: Expectancy, R-multiples, Kelly fraction, and drawdown metrics
How to perform decision and performance attribution (process vs. luck, selection vs. allocation)
Step-by-step calculations to evaluate edge, position sizing, and risk-adjusted results
How to run pre- and post-mortems, track base rates, and update probabilities
Practical review routines for continuous improvement and bias mitigation
Concept explanation
An investment journal is more than notes. It is a disciplined system for capturing your investment hypotheses, pre-trade checklists, position sizing, risk controls, and post-trade reflections. Think of it as a personal laboratory notebook: you write down your experiment setup, the conditions, the result, and what to change next time.
Because markets are noisy, short-term outcomes often mask decision quality. A good journal separates luck from skill by recording what you knew, what you assumed, and how you sized the position at the time. Over multiple decisions, patterns emerge: whether your thesis types perform, which setups produce favorable payoff distributions, and where your blind spots lie.
Finally, the journal becomes a feedback loop. By using consistent fields and periodic reviews, you can compute expectancy, analyze drawdowns, adjust position sizing, and refine your process. The goal is not perfection; it is continuous, measurable improvement.
Why it matters
Professional investors treat decisions as a repeatable process. If you cannot measure it, you cannot improve it. Journaling transforms subjective impressions into a dataset: entry and exit rationales, valuation assumptions, catalysts, alternative scenarios, and risk management actions. With these, you can quantify edge, understand variance, and recalibrate.
Over time, a journal helps you avoid common pitfalls: conviction creep, thesis drift, and cherry-picking memories. It also improves team communication if you invest with others: your journal makes your thinking auditable and facilitates pre-mortems and devil's advocate reviews.
Most importantly, journaling supports resilience. During drawdowns, objective records counteract panic by reminding you of base rates, your process, and risk limits. During winning streaks, they temper overconfidence with data on variance and expected outcomes.
Calculation method
Below are core calculations you can embed into your journal templates. You can compute these per trade and as rolling aggregates.
R-multiple and expectancy
Define 1R as the initial risk per position: entry price to stop-loss distance times position size. Express outcomes in R to normalize different trade sizes.
Expectancy is the average R per trade.
R = (Exit Price - Entry Price) / (Entry Price - Stop Loss) for long positions
Expectancy = (Win Rate × Avg Win in R) - (Loss Rate × Avg Loss in R)
Example:
Entry 50, stop 45, exit 57 for a long. 1R = 50 - 45 = 5. Gain = 57 - 50 = 7. R = 7 / 5 = 1.4R.
If over 100 trades, win rate 45%, average win 2.1R, average loss 0.9R:
Position sizing with fixed fraction and Kelly fraction
Fixed fraction: risk a constant fraction of equity per trade (e.g., 0.5% to 1%).
Kelly fraction (for frequent, independent opportunities) uses edge and odds. For R-based trading, approximate Kelly using expectancy and variance.
Kelly (binary payoff) = p - q / b where p = win probability, q = 1 - p, b = payoff odds
Kelly (R framework, approximation) = Expectancy / Variance of R
Example (binary): win prob p = 0.45, average win 2R, average loss 1R ⇒ odds b ≈ 2/1 = 2. Then Kelly ≈ 0.45 - 0.55 / 2 = 0.45 - 0.275 = 0.175 or 17.5% of capital. In practice, many use one-quarter to one-half Kelly to reduce volatility.
Maximum adverse excursion (MAE) and maximum favorable excursion (MFE)
MAE: worst unrealized loss during the trade measured in R.
MFE: best unrealized gain during the trade measured in R.
These help evaluate stop placement and profit-taking rules.
Drawdown statistics
Track peak-to-trough drawdown, length, and recovery time.
Drawdown (t) = (Equity Peak up to t - Equity at t) / Equity Peak up to t
Compute average drawdown, max drawdown, and the Ulcer Index to understand pain and persistence.
Risk-adjusted returns
Use return over max drawdown (RoMaD), Sharpe ratio, and Sortino ratio on your equity curve from the journal.
Sharpe = (Mean Portfolio Return - Risk-free Rate) / Std Dev of Returns
Sortino = (Mean Return - Risk-free Rate) / Downside Deviation
RoMaD = CAGR / Max Drawdown
Attribution and base rates
Selection vs. allocation: Did you pick outperformers, or did your asset mix drive results?
Base rates: Track hit rates by setup type, sector, market regime. Use these to weight future priors.
Posterior Odds ∝ Prior Odds × Likelihood Ratio
Example: If your breakout setup historically wins 40% with 2.2R wins and 1R losses, and the current market regime historically reduces its win rate by 20%, adjust p to 0.32 before sizing.
Case study
Imagine a swing trader who logs 50 trades over six months. Her journal fields include: date, ticker, market regime, setup type, thesis summary, catalysts, entry, stop, target, position size, 1R, exit, R, MAE, MFE, notes, checklist score, and an after-action review (AAR).
Sample trade record:
Date: 2026-01-15
Ticker: ALFA
Regime: Uptrend with elevated volatility
Setup: Earnings breakout with strong guidance
Thesis: Positive surprise, accelerating revenue, short interest could fuel squeeze
Entry: 50.00; Stop: 46.00; Target: 60.00
Position size: 500 shares
1R per share: 4.00; Total 1R: 500 × 4.00 = 2,000
Exit: 59.20
R result: (59.20 - 50.00) / 4.00 = 2.30R
MAE: -0.5R; MFE: +2.6R
Notes: Trailing stop would have captured more; partial at 58 might reduce variance
Breakout setups underperform when the VIX is high; pullback setups do better.
Exits are the main leak: MAE is small relative to MFE; a trailing stop after 1.5R would raise Avg win.
Position sizes exceed comfort during drawdowns; partial Kelly at 0.25× would reduce peak stress.
Action plan derived from journal:
Reduce breakout exposure during high volatility regimes.
Implement tiered profit-taking: scale 25% at 1.5R, trail remainder at 2× ATR.
Cap per-trade risk at 0.6% of equity, with a daily loss limit of 1.2%.
Practical applications
Build a standardized template: Include thesis, variant perception, base rate table by setup, checklist score, entry/stop/target, 1R, sizing method, pre-mortem, and post-mortem.
Schedule reviews: Weekly micro-review for process adherence; monthly performance review with expectancy, variance, MAE/MFE; quarterly deep-dive with attribution and regime analysis.
Decision tagging: Tag trades by setup, sector, catalyst, time of day, and regime. Use pivot tables to find what actually works.
Pre-mortem: Before entering, imagine the position fails. List top three failure modes and mitigation steps. This becomes your contingency plan in the journal.
Bayesian updates: Record prior probabilities from base rates. After receiving new evidence (e.g., guidance change), update your belief and document why.
Risk governance: Embed rules like maximum concurrent correlated exposure, daily and weekly loss limits, and a stop-trading threshold after N losing trades.
Playbooks: For each setup, attach a mini playbook with entry triggers, invalidation points, sizing ranges, and exit rules. Link outcomes back to the playbook to refine it.
Attribution: Split P&L by asset allocation decision (beta exposure) vs. security selection (alpha). If most returns come from beta, consider more passive exposure and be selective on active bets.
Make journaling low-friction. Use drop-down tags, formulas, and checklists so you can log a trade in under two minutes. The easier it is, the more consistently you will do it.
Common misconceptions
よくある誤解
- A journal is just a diary of prices. Reality: it is a process tool capturing thesis, risk, and decisions, not just outcomes.
- Good outcomes prove good decisions. In noisy markets, luck can dominate single results; judge decisions by process quality and long-run stats.
- More detail is always better. Excessive text without structure reduces usability; standardized fields and tags are key.
- Kelly sizing always maximizes growth. Full Kelly is highly volatile and sensitive to estimation error; fractional Kelly is safer.
- Post-mortems are for losses only. Wins need analysis too to ensure the process, not luck, drove the result.
Summary
まとめ
- Treat your journal as a measurement system linking thesis, risk, and outcomes.
- Normalize results with R-multiples and compute expectancy to quantify edge.
- Use drawdown and risk-adjusted metrics to assess comfort and sustainability.
- Apply fractional Kelly or fixed-fraction sizing based on realistic base rates.
- Tag decisions for attribution by setup, sector, and regime to find true strengths.
- Run pre- and post-mortems to improve before and after each decision.
- Make journaling easy, consistent, and review it on a set schedule.
Glossary
Expectancy: The average profit or loss per trade measured in R, computed from win rate and average win/loss size.
R-multiple: Outcome normalized by initial risk per trade; 1R equals the distance from entry to stop.
Kelly fraction: Position sizing approach that maximizes long-term growth based on edge and odds; often used fractionally.
MAE: Maximum Adverse Excursion; the largest unrealized loss experienced during a trade.
MFE: Maximum Favorable Excursion; the largest unrealized gain experienced during a trade.
Drawdown: Peak-to-trough decline in portfolio equity, often expressed as a percentage.
Attribution: Analysis separating returns from allocation (beta) and selection (alpha) decisions.
Base rate: The historical frequency or success rate of a setup or outcome, used as a prior probability.
