Risk Architecture & Market Regimes
Principles
Dalio's book is a manual for deciding under uncertainty: learn from painful mistakes, run a fixed five-step loop, and weight opinions by track record. His firm's All Weather portfolio applies the same humility to markets: spread risk evenly across four growth and inflation environments instead of betting on one.
The big picture
Ray Dalio founded Bridgewater Associates in 1975 from his New York apartment and built it into one of the largest hedge funds in the world. Principles (2017) is his account of how he makes decisions and how he ran the firm: part autobiography, part list of life rules, part list of work rules. The turning point he describes is 1982, when he publicly forecast a debt-driven depression, was wrong, and lost so much that he had to borrow money from his father to pay bills. His conclusion was not to forecast better but to assume he could be wrong at any time, and to build decisions and portfolios that survive that.
A plain note on sources: the book itself contains no portfolio formula. The investment version of the same idea is Bridgewater's All Weather approach, launched in 1996 for Dalio's family trust and described in the firm's public paper The All Weather Story and in Dalio's interviews: markets move on surprises in growth and in inflation, each of the four combinations favours different assets, and because no one knows which comes next, each environment gets an equal share of the portfolio's risk. This entry keeps the two apart. The pillars summarise the book; the pack implements the textbook stand-ins for All Weather (inverse-volatility and equal-risk-contribution weights) on the reader's own data.
Why it matters now: a classic 60/40 portfolio puts most of its dollars in bonds but takes most of its risk from equities, and 2022 showed that stocks and bonds can fall together when inflation surprises upward. Knowing which environment a portfolio really depends on is the first check.
The 3 strategic pillars
-
Pain plus reflection is how you improve
Dalio treats a painful mistake as the most useful data point available: the cost is already paid, so the only question left is whether you extract the lesson. Avoiding the pain, or explaining it away, wastes it.
The practice is a written log: what went wrong, what it cost, the root cause (a habit or blind spot, not the one-off symptom), and the rule you will follow next time. Over years the rules become a personal rulebook, which is literally what the book is. Bridgewater formalised this as an issue log, where staff record mistakes, their own included, so the causes can be traced.
-
The five-step loop
Getting what you want is a repeatable process with five distinct steps, and most people are good at only some of them. Knowing which step you are weak at matters more than working harder at all of them.
1) Set clear goals. 2) Identify the problems that stand in the way, and do not tolerate them. 3) Diagnose each problem down to its root cause. 4) Design a plan that removes the cause. 5) Execute the plan. The loop repeats. Dalio's advice for the steps you are weak at: get help from people who are strong at them, rather than skipping the step.
-
Believability-weighted decisions
Not every opinion deserves the same weight. Weight each view by the holder's record on this kind of question and by whether they can explain the cause-and-effect behind it.
Dalio's rule of thumb: a believable person has succeeded at the thing at least three times and can give a logical account of how. At Bridgewater, meeting participants rated each other's views in real time with an app (the Dot Collector), and decisions could be taken on the believability-weighted tally instead of by the boss or by head count. It came with radical transparency: many meetings were recorded and open to staff. When the weighted answer and the simple majority disagree, that disagreement is the signal to keep talking.
What a Closelooknet reader does with it
The working use is an audit of where a portfolio's risk really sits. Take monthly returns for each holding or asset class, compute each line's volatility and the correlations, and measure what share of total portfolio variance each line carries. For a 60/40 mix with equity volatility of 15%, bond volatility of 5% and zero correlation, the equity line carries about 95% of the risk. Inverse-volatility weights (25% equity, 75% bonds in that example) split the risk 50/50; the pack also solves the equal-risk mix with correlations included, and shows the leverage a lower-risk mix would need to reach a chosen volatility. A second sheet sorts your own history into the four growth and inflation quadrants, so you see which environment each line has depended on.
The mistake it prevents: believing a portfolio is diversified because the dollars are spread out, while almost all of the risk sits in one environment. The book side of the pack is a believability-weighted vote and a mistake log, for decisions that are not about numbers.
The bridge to the Closelooknet approach
Closelooknet measures the two axes of the quadrant directly: the Growth & Recession monitor and the Structural Inflation monitor read the direction of each, which is the reading the pack's quadrant sheet asks you to note. Money Temperature scores regime intensity across eight instruments, and Market Regime 101 explains the house's green, yellow and red classification, defined in the glossary under market regime. The risk arithmetic uses realized volatility and correlation; breakeven inflation is the market's own inflation expectation, the baseline an inflation surprise is measured against. On the shelf, Trend Following is the other documented way to hold up across regimes, and Fortune's Formula covers sizing a single bet rather than balancing a portfolio.
Action-Kit — from theory to practice
Tooling & data
| What you need | Where to get it | Cost |
|---|---|---|
| Monthly total-return series for YOUR holdings or asset classes The only input the risk-parity sheet needs; five series for the spreadsheet, any number for the script | Your broker's statement export, or fund factsheet data The pack ships no market data; the sample file holds EXAMPLE_ rows made up for the format only. | Free |
| The All Weather Story Bridgewater's own account of the four environments and of balancing risk across them; the source of the quadrant logic | Bridgewater Associates | Free |
| Growth and inflation data Series to label each month of your history as growth up or down and inflation up or down (industrial production, payrolls, CPI, breakevens) | FRED, Federal Reserve Bank of St. Louis | Free |
| Portfolio backtesting and risk-parity optimisation A second tool to cross-check risk contributions and equal-risk weights on your own asset mix | Portfolio Visualizer | Freemium |
The formulas
-
Inverse-volatility weights
w_i = (1 / σ_i) / Σ_j (1 / σ_j)- σ_i — annualized volatility of asset i (monthly standard deviation × √12)
Ignores correlation. Exact equal risk only when all correlations are equal.
-
Risk contribution
RC_i = w_i × (Σ w)_i / (wᵀ Σ w)- w — weight vector
- Σ — covariance matrix of asset returns
- (Σ w)_i — the i-th entry of Σ times w
The RC_i of one portfolio add up to 1. For 60/40 with σ 15% and 5% and zero correlation: 0.36 × 225 = 81 vs. 0.16 × 25 = 4, so equity carries 81/85 ≈ 95%.
-
Equal risk contribution (iteration)
w_i ← w_i × √((1/n) / RC_i), then divide by Σ w- n — number of assets
- start — inverse-volatility weights
Converges in a few dozen steps for ordinary covariance matrices; the spreadsheet runs 30, the script 500.
-
Believability-weighted score
score_k = Σ_v b_v × s_v,k / Σ_v b_v- b_v — believability weight of voice v (track record + explainable reasoning)
- s_v,k — score voice v gives option k
Compare with the plain average; a gap between the two is worth a conversation before deciding.
Applied Pack · free members
Dalio Regime Allocator Pack
Your own return series in, the risk audit out: where each line's share of portfolio risk really sits, inverse-volatility and equal-risk weights, and your history sorted into the four growth and inflation quadrants. Plus the book part: a believability-weighted vote. Software for your own research, never signals.
- Dalio_Regime_Allocator.xlsx — READ ME, a Returns sheet for five series and up to 240 months, a Risk Parity sheet (volatility, correlation, covariance, dollar vs. inverse-vol vs. equal-risk weights, risk shares, target-volatility scale factor), a 30-step ERC Solver, a Quadrants sheet (the four environments, your current reading, your history per quadrant) and a Believability sheet (weighted vote + mistake log), all live formulas with amber input cells
- dalio_regime_allocator.py — stdlib-only CLI: the same statistics, weight sets and risk shares for any number of series, the quadrant table when growth and inflation labels are present, and an optional target-volatility scale factor
- returns_sample.csv — 60 EXAMPLE_ months of made-up returns for five asset types with growth and inflation labels
- README.txt — where the model comes from (Bridgewater's public material, not the book), inputs, the math, how to run, limits and the educational-use disclaimer
Pack security
Macro-free Excel · plain-text Python you can read before you run it · no installers, no network access — the code works only on files you provide. Served only from closelook.net; we never distribute through download portals or email attachments. How to verify in 30 seconds →
SHA-256 aca756d8214a709d5259801e03af04dde49aba29fdb6e8ae11239c48d5d9f888
The pack is a free-membership download — no card, free forever. Membership also delivers the Daily Pulse and Weekly Signal to your inbox.
Join the Look — free → Already a member? Sign in →
Educational templates — a research diary companion, not investment advice.
Closelooknet publishes a market diary, not investment advice. This condensed read restates the book's ideas in our own words for education — for the author's full argument, go to the source.