Everything this site runs on, from scratch.
No prior knowledge assumed. If you have never bought anything more complicated than a savings account, start here and read straight down. Each principle gets a picture, because most of them are easier to see than to describe.
They come from three people — Ray Dalio on how eras turn, Charles Gave on reading the present moment, and Richard Détente on how to decide — plus my own arrangement of them. The order matters: each part depends on the one before it.
The setting
What kind of world you are investing into. Everything downstream depends on this answer, so it comes first.
Empires run an arc, and so does money
Countries rise, peak, decline and reset. It takes centuries, but it is not random — and it repeats.
A country builds strength: it educates people, invents things, sells them abroad, gets rich, and its money becomes the one everyone else wants to hold. That last part is the prize, because it lets the country borrow far more cheaply than anyone else.
Then the same success rots it. People get expensive, so other countries make things cheaper. The generation that built it is replaced by one that inherited it. Borrowing grows because borrowing is easy. Eventually the debt costs more to service than the country can earn, and something breaks.
Ray Dalio traced this through the Dutch, the British and now the Americans. The details change; the shape does not. What matters for you is not predicting the date — it is knowing which part of the arc you are standing on, because the same investment behaves completely differently at different points.
To understand what is coming at you, you need to understand what happened before you.
Money is being made, constantly, and it dilutes yours
Roughly 8% more dollars exist each year. Anything you own that grows more slowly than that is shrinking.
When a government spends more than it collects in tax, it has three options: raise taxes, borrow, or create new money. Raising taxes loses elections. Borrowing has a limit. So it creates money.
Nobody takes anything out of your account when this happens. Your balance is the same number it was yesterday. But there are now more units of that currency chasing the same houses, the same food, the same shares — so each unit buys less. You lost value without a transaction ever appearing.
This is why a savings account paying 2% is not safe. It is a slow, invisible loss. It is also why doing nothing is a decision with a cost, not a neutral position. Once you see this, the whole idea of “playing it safe” has to be re-examined.
The 8% figure is Richard Détente’s estimate of dollar money-supply growth, not an official statistic. Treat the exact number as arguable and the direction as not.
Currencies do not last forever
Of roughly 750 currencies since 1700, fewer than a fifth still exist. Every survivor buys less than it did.
This is the part most people find hardest to believe, because within one lifetime a currency feels permanent. It is not. Money is a promise issued by a government, and governments change, fail, and reissue.
A working currency lasts roughly seventy years before it is replaced or drastically devalued. The current arrangement — the one where the US dollar is the world’s reference — was set up in 1944.
You do not need to conclude the dollar is about to collapse. You only need to stop treating “held in cash” as a position without risk, and to own some things that no government issues.
The compass
Four ideas for deciding what to own. They are not a strategy — they are the things you check a strategy against.
Value moves
What is worth something today was made out of something that did not exist before. It will move again.
When the cart arrived, the porters were finished. When the car arrived, the cart-makers were finished. Every time something new works, value drains out of whatever it replaced — and it drains slowly enough that the people holding the old thing keep believing it will come back.
So the one bet nobody can place for you is this: where is value going in the next ten or twenty years? That is a guess about the future, which makes it speculation. The word has become an insult, but marrying someone, having a child and starting a business are the same act.
Being careful is the dangerous thing
A cautious portfolio is a bet that the next forty years look like the last forty. In a turning era that bet loses.
When someone offers you a prudent portfolio, look at what is in it. It will be the things that worked recently — government bonds, a broad fund, a bit of cash. That is not a judgement about the future. It is the past, copied forward.
It feels safe because it is familiar and because nobody can blame you for it. But if the world is changing, copying the past is precisely the way to be holding carts when the cars arrive. The risk is hidden rather than absent.
There is a very old version of this. In the parable of the talents, the servant who buries his master’s money to keep it safe is the one who is condemned. Not the one who lost some taking a chance — the one who refused to try.
If you accept speculating and enter the fight for survival, you have a 20% chance. If you refuse danger out of prudence, you have a 100% chance of going.
Some things gain from disorder
Most investments quietly need calm. A few do better when things go wrong. Which you want depends on the era.
A normal stock fund performs best when the world is boring. Steady growth, no surprises, no wars. That is a hidden condition attached to it: it needs the world to stay as it is. That property is called fragility.
The opposite exists. Norway sells oil and has almost no debt. When an energy crisis wrecks everyone else, Norway is paid more for the thing causing the wreckage. It does not merely survive the shock — it is a beneficiary of it.
Neither is better in the abstract. If you think the world is settling down, own the first kind. If you think it is turning, you need some of the second.
Two things that do not move together
The only genuinely free improvement in investing: combine holdings that rise and fall for unrelated reasons.
Normally, to earn more you must accept more lurching. There is one exception. If you own two things that move for completely unrelated reasons, they will rarely fall on the same day. Added together, some of the bumps cancel — and you did not give up any return to get that.
Almost nobody does it properly, and the reason is psychological rather than technical. Our opinions hang together. A portfolio of genuinely unrelated holdings does not feel coherent — it feels like a mistake. Put a flat in Singapore, a data centre in Texas and some Bitcoin in the same list and you will irritate everybody who sees it.
That irritation is the evidence it is working. If every holding you own makes sense next to every other one, you have not spread your bets. You have made the same bet several times.
There is one free lunch in finance, and it is the decorrelation of assets inside a portfolio.
Danger is not the same as risk
One word is used for two different things. Confusing them is how careful people lose everything.
Risk, as the industry uses the word, means how much the price jumps around. If something is worth 100 today, 60 next month and 140 the month after, that is high risk.
Danger is something else entirely: the chance the money never comes back at all.
These are separate, and they often point opposite ways. An Argentine government bond barely moved in price — and went to zero, nine times. A salary has no volatility whatsoever until the Monday morning it stops. Bitcoin swings a few percent before breakfast and has never gone to zero.
If you only measure the first one, you will systematically buy quiet things that can kill you and refuse violent things that cannot. That single confusion explains most of what gets sold as conservative.
Illustrative positions, not recommendations. The point is the shape of the plane, not the exact spot of any one dot.
Reading the moment
The compass tells you which way to face. This part tells you where you are standing right now.
An economy is energy, transformed
Everything an economy makes is energy turned into something else. So the price of energy is the master variable.
A factory takes electricity and raw material and produces goods. A lorry takes diesel and produces delivery. Even an office takes power and produces decisions. Strip it back and every economy is a machine for converting energy into things people want.
That gives you a test. If a country’s companies are growing faster than the cost of the energy they consume, the machine is working. If energy costs are climbing faster than company values, it is not — and no amount of official growth statistics will save you, because those get revised and massaged.
Charles Gave’s version is a simple division: the country’s stock index divided by the oil price, averaged over seven years. It uses market prices, which no committee can quietly adjust.
Four weathers, four answers
Growth and inflation make four combinations. Each one rewards a different thing to own.
Two questions describe the whole environment. Is the economy growing or shrinking? Are prices rising quickly or not? Answer both and you land in one of four boxes.
- Growing, prices calmThe best weather. Own companies — especially the efficient, inventive ones.
- Growing, prices risingOwn gold and metals alongside shares. Money is losing value even as business does well.
- Shrinking, prices risingThe worst of the four. Hold cash in a currency run by serious people, keep everything short-dated, own energy.
- Shrinking, prices calmLong government bonds, and essentially nothing else.
The second question — are prices rising? — has its own market test: gold divided by that country’s long government bond, again over seven years. Gold is the store of value nobody issues; the bond is the one the state issues. When savers move from the second to the first, they are telling you they have stopped trusting the state’s money.
See all 32 countries placed →Remove the loser instead of picking the winner
You do not need to know what will do best. You need to know what this weather destroys, and not own it.
Picking winners is genuinely hard and almost nobody is reliably good at it. Spotting what is about to be punished is much easier — and it is enough.
Of the four things you can own — company shares, bonds, gold, cash — the current weather usually makes one or two of them obviously wrong. Take those out. Hold what remains. You have improved the outcome without having predicted anything.
It is how you would bet on a horse race if you were serious. You do not need to know which horse wins. You need to know which three cannot.
You are building a balance sheet, not trading
The question is not what you made this year. It is what you own, and whether it survives the decade.
Most people treat investing as a scoreboard: what did I make, what did I lose, how am I doing this year. That framing pushes you into activity, and activity mostly costs money.
The alternative is to think about what you own. Is it genuinely yours? Can somebody freeze it, print more of it, or default on it? Will it still be worth something in ten years? Those are balance sheet questions, and they are duller and far more useful.
This is also why bad years matter less than they feel like they do. As Warren Buffett put it, a serious downturn is when assets return to their rightful owners — the people who can hold through it.
What it builds
Put the principles together and they produce a specific shape. This is that shape.
Separate pockets, not one pot
A few holdings, each bold on its own terms, none of them moving for the same reason.
Follow the principles through and you get somewhere quite unlike a standard portfolio. Value moves, so you have to take positions. The era is turning, so those positions should gain from disorder rather than merely endure it. That means accepting things that jump around.
And the only way to hold jumpy things without risking everything is to make sure they are not connected. So: a small number of separate pockets. If one fails completely, it costs roughly a year of what the others produce — not the structure.
This is the opposite of the usual advice, which is to own forty things. Forty things that all fall on the same Tuesday is not spreading your bets. It is one bet with extra paperwork.
The liquid pocket
Expresses the current quadrant. The only pocket that changes when the macro changes.
This is the pocket that reads the cadran and acts on it. Listed instruments, daily liquidity, sized by conviction, with every position carrying a written reason to exist and a written condition that ends it.
It runs as a barbell rather than a balanced fund: a protective layer that holds the line when the regime is unclear, and an offensive layer that expresses the specific thing the current quadrant is rewarding. Gold and a serious currency on one side; the growth engine on the other.
It is also the only pocket here with published positions, because it is the only one where a position is a public, checkable claim rather than a private arrangement.
Moves with the global macro cycle — which is exactly why it needs the other pockets, not more of itself.
Property
Real collateral. The thing that still exists when the financial layer is having a bad decade.
Property is the clearest example of the danger-versus-risk distinction working in your favour. It moves slowly and it is a nuisance to sell, so conventional risk measures treat it as safe for the wrong reason. What actually makes it safe is that it is collateral: something physical stands behind the claim, and it does not go to zero because a government changed its mind.
The location logic follows from the era rather than from taste. As the world regionalises into blocs, a small number of hinge jurisdictions — Switzerland, Singapore, the Gulf, a few others — become the places where capital from both sides can still meet. They are also where globally mobile wealth chooses to live, and that population is growing more concentrated, not less.
The cost is real: capital is frozen, entry costs are high, and you cannot rebalance out of it in an afternoon. That is the trade you are making for collateral.
Driven by local supply, local law and where mobile capital chooses to sit — almost none of which is the global rate cycle.
Compute and energy
The asset class being born right now. Roughly one appears every thirty years.
Compute is becoming an asset class in the formal sense — priced, financed, borrowed against, with futures and options forming around it. When the largest asset manager in the world says so publicly, the positioning is already done; the announcement is the receipt, not the news.
What makes it interesting is that the stack inverts the usual depreciation logic. Hardware that used to be worthless after four years now clears at a premium on secondary markets, because the constraint is not chips but the power to run them. That pushes the real value down the stack into energy — which is why the serious version of this pocket is energy first, hardware second, software third, rather than buying whichever accelerator is fashionable.
It is also the most dangerous pocket here, and should be sized accordingly. New asset classes are where the returns are and where the failures are, in the same place, for the same reason.
Driven by AI capex and electricity supply — a physical build-out that keeps running through monetary conditions that would stop a normal cycle.
Bitcoin
A bet that the thing everyone converges on when they cannot coordinate is changing.
Only if you accept the premiseOnly if you accept the premise. If you do not think the monetary Schelling point is moving, this pocket is a zero, not a small position.
For thousands of years the answer to "what will a stranger accept?" was gold. It is the monetary Schelling point — the thing everyone independently converges on precisely because everyone expects everyone else to. Bitcoin is a bid to take that position, and either it does or it does not.
That makes the sizing question tractable in a way conviction arguments are not. Take gold's market capitalisation, take Bitcoin's, and hold the ratio. You end up with a low single-digit percentage. If it goes to zero it does not change your life; if it takes the position it can double the book. You do not need to be certain — you need to be sized so that being wrong is survivable and being right is material.
Its real portfolio function is decorrelation. Over long windows it is the least correlated liquid thing available, which means it can be volatile without making the whole book volatile. It also carries something no custodied asset has: no counterparty, and nothing to freeze — a property the 2022 reserve seizures made concrete rather than theoretical.
Amplification is a separate decision and a harder one. Volatility has compressed as institutions have absorbed the supply, and there are ways to restore convexity — leveraged treasury-company equity, or a small allocation to long-dated calls where the loss is capped at the premium. Both raise danger, not just risk. Neither belongs anywhere near a core position.
Over long horizons the least correlated liquid asset available. Correlated in short panics, like everything is.
Then check that it actually worked
Decorrelation is easy to believe you have and easy not to have. So measure it.
The whole structure rests on the pockets being unrelated. That is a claim, and claims should be checked rather than assumed. The table below compares every pocket against every other. Low numbers mean they genuinely move independently. A grid of high numbers would mean the same bet, taken four times, wearing four hats.
| Liquid pocket | Property | Compute and energy | Bitcoin | |
|---|---|---|---|---|
| Liquid pocket | 1.00 | 0.15 | 0.35 | 0.25 |
| Property | 0.15 | 1.00 | 0.05 | 0.05 |
| Compute and energy | 0.35 | 0.05 | 1.00 | 0.30 |
| Bitcoin | 0.25 | 0.05 | 0.30 | 1.00 |
1.00 means “moves identically”. 0.00 means “no relationship at all”. These are long-horizon estimates, not computed from a return series — the shape is the point, not the decimals.
Only one pocket has published positions.
The liquid pocket carries a real, checkable book — every holding with a reason to exist and a condition that ends it. The other three are research on how the pocket should be built, not a disclosure of what I own. Where that line sits matters, so I would rather state it than let the layout imply otherwise.
None of this is advice, and I am not a financial adviser. It is a structure I find defensible, published so it can be argued with.