Seven parts, and a banker that runs them for you
A regime model that describes the present, an allocation engine that was decided in advance, alerts that fire on thresholds rather than opinions, and a community that reads the same data. Nothing in here forecasts. And in front of all seven sits the part you actually hire: a private banker you give one instruction to, which then builds the allocation, watches it, prices the rebalance and tells you what your choice costs.
Portfolio construction, handed over
Every view below carries the same affordance in the same place — Ask the banker — and it opens the assistant already holding that screen: which profile you are on, which regime reading, which sleeve is selected, how far the book has drifted, and what your mandate says. It is one assistant entered with context, never a second one.
What it does without being asked
- Takes the mandate. Six standing instructions, each stating who it is for, the horizon and what it will not accept. Picking one sets the weights, the horizon, the drift tolerance and the guard it argues hardest.
- Watches the drift. One check every eleven seconds on the demo clock, and one on every change you make to the profile, horizon, regime or your own sleeves.
- Prices the rebalance. Trades, euro moved, expected slippage, and which legs it skipped on the 20 bps cap. Sells first, so the cash is there before the buys.
- Checks the four house guards. Quarter-Kelly sizing, the −20 % drawdown halt, the 7.5 % single-company ceiling and the 20 bps execution cap — and warns before a breach, not after.
- Triages the week’s notes. It says which ones moved a published weighting and which were flagged and not acted on.
- Audits itself. It publishes its own wrong calls and the stretches where holding it was the worse decision.
What it will not do
- It will not place a trade. “It is priced and it is on the screen; the decision stays yours.”
- It will not size past a published band. Ask for a weight outside the band and it names the nearest weight it can set instead.
- It will not act on a forum post. “The sleeves come from the regime model and the published bands, and nothing else gets a vote.”
- It will not move your book back to the mandate on its own. If you change the profile, it raises the mismatch and leaves the call with you: “you changed it, so you decide.”
- It will not invent a number. Every figure comes from the same functions the panels use. If it cannot compute one, it names the panel that can.
The six standing mandates, and what each one costs
| Say this once | It sets | It refuses | What it costs you |
|---|---|---|---|
| “Protect what I already have” | conservative · 5 y · 1.5 pts · drawdown halt | a deep fall bought with the chance of a better year | Gives up compounding in exchange for a shallower fall. In a rally you will lag, and a tight tolerance also means more trades and more slippage. |
| “Compound this for decades” | growth · 30 y · 3.0 pts · quarter-Kelly | borrowed money, and any sleeve outside its published band | Accepts a noticeably deeper worst fall than the conservative mix — a loss you would have to sit through without selling. |
| “Sleep at night” | conservative · 10 y · 1.5 pts · drawdown halt | a trade whose cost is larger than the drift it corrects | The same compounding given up — “shallowest and best are the same question asked from opposite ends, and I will keep answering the shallow one.” |
| “Stay ahead of prices” | balanced · 20 y · 2.0 pts · execution cap | any single trade priced above the 20 bps cap | Gives up some compounding against growth and falls deeper than conservative; a gap sometimes waits a week rather than closing at a wide spread. |
| “Run this beside a pension” | growth · 20 y · 3.0 pts · 7.5 % ceiling | any single company above 7.5 %, and a satellite that quietly becomes the core | The same deeper worst fall as growth; it holds a volatile sleeve straight through a fall so long as no single company passes 7.5 %. |
| “Understand it before I commit money” | balanced · 10 y · 2.0 pts · 7.5 % ceiling | any change made before you have seen it priced | it acts less: “every proposal waits on you, and a decision you never make is still a decision.” |
Figures are illustrative model output over a 152-month sample with costs taken out. Sizing stays at quarter-Kelly (0.25) inside the 7.5 % per-company ceiling, the −20 % drawdown halt and the 20 bps execution cap in every mandate. Pause, resume, undo the last action and change the mandate are each one click, and changing the mandate never wipes the log.
Three derived indices, one reading
Liquidity (weight 50 %): the 3-month impulse of global broad money and the combined balance sheets of the Fed, ECB, BoJ and PBoC, from FRED and central-bank releases. Inflation momentum (30 %): 3-month annualised core CPI against its 12-month trend, plus 2-year breakevens. Geo-risk (20 %): GDELT conflict tone and conflict-event density over 30 days, scaled 0–100.
Each index is z-scored over a rolling 10-year window and mapped to 0–100. The blend is the risk meter. Above 60 is risk-on, below 40 is risk-off, in between is neutral. A label only flips after an 8-point move that holds for two weeks.
Try the regime explorer in the previewBands, not bets
Every profile has a fixed band per asset. The regime decides where inside the band the weight sits; nothing else does. Balanced, for example: equities 34–55 %, bonds 20–28 %, gold 10–20 %, Bitcoin 6–10 %, cash 5–12 %. The bands are the risk decision. You make it once, when you pick a profile.
- Bitcoin hard cap: 15 % in Growth, 10 % in Balanced, 6 % in Conservative
- Cash floor: 3 / 5 / 7 % — never fully invested
- Weekly change cap: 5 points per asset, so a regime flip takes two to three weeks to fully express
- Gold is never below 8 %: it is the sleeve that pays for the next risk-off regime
A fan, not a number
The builder draws a Monte Carlo fan for your profile and horizon from the historical volatility of the mix. The median path is drawn thicker because people ask for it, not because it is an expectation. We show P5 and P95 next to it so the width of the range is impossible to miss.
Illustrative only. Historical volatility is a weak guide; correlations change in exactly the regimes that hurt. The fan is there to set expectations about dispersion, never about returns.
Change the horizon in the previewThresholds, not opinions
An alert fires when a component crosses a level you can name: CPI above 3 % year-on-year, the liquidity index below −1 %, geo-risk above 70, or the regime label itself flipping. Every alert states what the model did in response — usually "nothing", sometimes "+3 gold, −3 long bonds".
Pro users get alerts in the app and by e-mail, and can write their own rules (for example, "tell me when gold's weight would exceed 18 %"). Alerts are always a description of a change that has already happened in the data.
Watch the alerts feed in the previewLong, calm, charted
Every Thursday: 1,500–2,500 words on what moved the three components, what the model did about it, and what it would take to flip the regime. Written by the Rulebook Wealth editorial desk, with every chart traceable to a public series. No calls to action, no "buy this".
Read the latest briefPeople who hold the same five assets
A moderated forum with three house rules: no leverage talk, no price targets, cite your series. Threads run alongside each brief; the editors answer questions about the model in public. Reading is free; replying is Pro.
Browse the forumTake the weights with you
Rulebook Wealth publishes conclusions; it never places an order and never asks for a trading credential — the optional custody link reads balances only and refuses anything broader. Pro exports the current weights, the band rules and the caps as JSON or CSV, plus a copy-ready order-ticket sheet. Open it in a spreadsheet, paste it into your own tracker, or hand it to whatever actually executes for you. Nothing is pushed from our servers to your machine, and nothing comes back.
Export formats and the file watcherFree vs Pro at a glance
Planned plans. Rulebook Wealth is in early access and not on sale yet.
| Feature | Free | Pro · €19/mo planned |
|---|---|---|
| Ask the banker, in every view header | ✓ Balanced | ✓ all three profiles |
| Standing mandate — one of the six, working on its own | ✓ it reports | ✓ it acts |
| Mandate log: what it did, why, and the number | ✓ | ✓ |
| Pause, resume and change the mandate | ✓ | ✓ |
| Undo the banker's last action | ✗ nothing to undo | ✓ |
| Priced rebalance plan (trades, euro moved, slippage) | summary | ✓ full trade list |
| Its own wrong calls and lagging stretches | ✓ never paywalled | ✓ never paywalled |
| Weekly Macro Brief (web + e-mail) | ✓ | ✓ |
| Risk-On / Risk-Off meter with component inputs | ✓ | ✓ |
| Balanced model allocation, weekly | ✓ read-only | ✓ |
| Conservative & Growth allocations | ✗ | ✓ |
| "Why this weighting" reasoning | summary | ✓ full |
| Projected range fan, any horizon | 10y only | ✓ |
| Macro alerts (app + e-mail) | ✗ | ✓ |
| Custom alert rules | ✗ | ✓ |
| Historical regime archive (since 1986) | ✗ | ✓ |
| Community: read | ✓ | ✓ |
| Community: post & reply | ✗ | ✓ |
| Export weights, bands and caps (JSON / CSV) | ✗ | ✓ |
See it with your own profile
The demo preview runs the full model on simulated data. No account needed.
Educational content. Model allocations are illustrative, not investment advice.