"I've got everything under control." — the phrase of a founder who has nearly all his wealth in one asset, one account, one jurisdiction, and has never once checked what happens if that account vanishes on a Tuesday morning.
Here's a scene you'll probably recognize — either because it's you, or because you saw it up close and said nothing. A founder is on a call and spends an hour arguing whether the signup button should be green or orange, because an A/B test on a few thousand users gave a conversion difference of a few percent. All very grown-up: hypothesis, sample, p-value, data-driven decision. And two hours later the same man wires a six-figure sum of personal savings into one account at one bank — and doesn't ask a single question he'd have asked about the green button. What percent of everything is that? What if this bank is put into receivership on Monday? Where's the backup? He tested the button. His own survival — no.
There's a proverb: "the cobbler's children have no shoes." The master who shoes the whole town leaves his own kids barefoot. Only the modern version is harsher: this cobbler isn't just barefoot — he leases a G-Wagon, posts it in stories, and sincerely considers it proof that his finances are fine. The boss walks barefoot, but on display — in expensive untied laces.
Let's name up front where someone's skin is in this game. A text with no personal stake is a lecture. And you've heard enough lectures from people who lost less than you in one bad quarter. For years most people keep the business, the family money, savings, and their own ego in one mental wallet. Not in one account — worse: in one head, with no partitions at all. "How much money do I have" = one blurry warm number in which business revenue, the personal cushion, the partner's share, the taxes not yet paid, and unrealized gains all merge into a single substance called "well, roughly fine." It's like saying "I have a six-pack" and counting the shadow in the mirror together with the Instagram filter. It shows brightest every time a tax bill lands: a person is sincerely surprised why "so little is left" — because in their head it had all long since merged into one warm, unlabeled number.
This habit always existed. But 2023–2026 turned it into a systemic risk of another order. SVB went down in 48 hours in March 2023 — and it turned out a whole layer of the tech business hadn't, infrastructurally, planned for one bank disappearing over a weekend. For Ukrainian founders running businesses across several jurisdictions amid a full-scale war, each of these risks multiplies: a frozen account, a change in currency regulation, a jurisdictional shock — not abstract scenarios but an ordinary quarter. BTC went through another ATH cycle and now occupies, in founders' portfolios, a share they never planned: the asset grew, there was no rebalancing, the concentration quietly reached god-object levels. The system rewards immobility in decisions — until it fines you for it, with interest.
The thesis worth unfolding without anesthesia: whoever runs the product like an engineer and the money like a teenager with an envelope of cash in a drawer will sooner or later learn that the market doesn't distinguish how smart your board is — only how stupidly your balance sheet is built.
 *The same desk, broken into a system. ① capital modules like a server rack — one overflowing, the rest empty; ② a house of cards: one bank's collapse zeroes everything; ③ a "reserve" that only looks like a reserve; ④ you count banknotes by hand while the system-monitor "will I survive a failure" was never even started.*Here's where most founders switch to "I'm bored" and drive past the one formula that could save them from bankruptcy at forty. In finance there's a phrase attributed to Nobel laureate Harry Markowitz: diversification is the only free lunch. The meaning isn't metaphorical but mathematical: if you put into a portfolio assets that don't move in sync, you can lower overall volatility without a proportional loss of expected return. Markowitz won the 1990 Nobel in economics for this model. This isn't YouTube esoterica. It's the basic course a junior analyst learned in year one, and a founder proudly ignores for ten years straight.
And here's the black joke of the whole startup industry: a founder is a person who professionally hunts free lunches. You read threads about LTV/CAC at night, pipe leads through retargeting, dream of a viral clip for zero dollars — your whole craft is built on hunting the asymmetry of "little input, big output." The one free lunch you demonstratively decline is diversifying your own wealth. Why? Because the menu's boring. Because "spread it across baskets" sounds like advice from a retired aunt, not the move of a genius who put everything on one company and won.
Capital should be run like a codebase, not like a wallet: buckets are modules, the cushion is backups, concentration in one asset is a god-object, and buying status on credit is technical debt accruing interest while you sleep. Money has no linter that underlines in red "you've got a god-object here at 90% of net worth." You have to be the linter yourself. Most never switch it on even once.
In programming there's the god-object — a class that knows and does everything, that half the system is wired to, that's terrifying to touch because you don't know what falls off. Every engineer hates it. Every founder is one, in the financial sense.
There's research (Moskowitz & Vissing-Jørgensen, 2002) where the typical picture looks like this: most private entrepreneurs hold the lion's share of their wealth — often 60–80% and up — in one private firm. This isn't an anomaly, it's the genre's norm: you build one thing, pour years, cash, and nerves into it, and your balance sheet mirrors the obsession. The puzzle researchers named "the private equity premium puzzle": entrepreneurs systematically put it all on one card and systematically get underpaid for the risk they take blind. Holding everything in one company is like storing your whole ammunition in one depot near the front line: convenient for logistics, perfect for one accurate strike.
Now — why it's a god-object, not the "focus" people love to call it at conferences. Focus is when attention concentrates on one thing. A god-object is when survival in all other domains at once depends on one thing: the mortgage, the kid's school, parents' treatment, self-esteem, and pension are all plugged into one API called "my only business." When that API falls — and any one does, sooner or later — a person's production and personal life collapse at once, because the system never assumed these were separate circuits.
And here's the finest place where concentration becomes a personal mirror. It's easier to skip than to admit. When 90% of the money and 100% of the identity is one company, a person is physically incapable of making sober decisions about it. Sell a stake at the peak? That's self-betrayal. Take a few chips off the table? That's disbelief in your own dream. Capital concentration quietly becomes ego concentration — and then you're defending not the money but the image of the self who bet everything and didn't blink.
That's why "I never sell" stories are so popular: they let you pass off fear as principle. It's not conviction. It's dependence on one mirror in which you like yourself. When 90% of capital and 100% of self-esteem hang on one company, that's no longer focus — it's a gallows with a pretty logo. The most expensive position in a founder's portfolio isn't Bitcoin or his company. It's his certainty that this won't happen to him.
But here's what's worth saying plainly before moving to tables. Not all concentration is a mistake. Truly outstanding companies — from Apple to Monobank — were built by people who went all-in and were right. For an early-stage founder, full concentration is sometimes the only possible decision: without it there's neither bandwidth nor resource. Concentration at the start isn't an anti-pattern, it's a survival strategy. But here's what that doesn't explain: why a founder who has already moved past survival and built a stable cash flow keeps holding 80% of his wealth in the same god-object — not because it's mathematically optimal, but because selling even a stake is psychologically impossible. The system rewards the founder's inseparable identification with the asset, because it creates founders willing to work for the minimum market order: a person whose ego = company won't leave the board even when he should. The VC industry knows this mechanic well. So do banks: the one-account founder is a better client than the distributed one.
| Anti-pattern in code | The same anti-pattern in capital |
|---|---|
| God-object: one class holds everything, scary to touch | 60–80%+ of net worth in one illiquid company |
| Single point of failure: one service brings the system down | All money in one bank / one jurisdiction |
| No backups: disk fails — data gone | No cushion: income falls — no month of life left |
| Hardcoding secrets into the repo | Business cash and personal money on one account |
| Technical debt disguised as "we'll rewrite it later" | Leasing status disguised as "I earned this" |
| No tests: "it works for me anyway" | No stress test: "I've got everything under control" |
Here's a concrete scene, then the theory — because that way it cuts sharper. Imagine: a "unique opportunity" came to join a friend's round. You pulled money out of your head — and grabbed the wrong stack. The one that should have read "this is the state's, I'm the courier." April. Tax authority. You're sincerely surprised the till is empty. It isn't. This is what a mental wallet looks like in action. Not a tragedy. A comedy with one actor and one act, repeating every year.
The solution is so simple it's embarrassing to even put in a text for adults: buckets. The concept of laying capital out by purpose and horizon has been tied to financial planning since the 1980s. The essence: every dollar should know which module it belongs to and what it mustn't touch. Business operating cash — a separate circuit that doesn't know your vacation exists. The tax reserve — a concreted module from which you can't call the function "buy a watch," because that money isn't yours anymore — you're just its temporary custodian on behalf of the state. The personal cushion — read-only for any "profitable opportunity." Investment / risk capital — the only bucket allowed to hurt.
If you still need a term in English to lay money out in stacks, we have here not a financial problem but an aesthetic one. Label the stacks. Imagine the cash on the table is labeled. On one, in marker — "this is the state's, I'm the courier." On the second — "this is a month of life if it all burns." On the third — "this can be lost without dying." That's the whole architecture. The rest is implementation detail. The tax authority comes not as an enemy but as someone you already owe. It's just the first to read your financial code without the comment "I'll figure it out later."
The quietest benefit of buckets isn't even safety. It's that you finally stop making financial decisions by mood. When everything is merged into one number, every purchase is a negotiation with your own anxiety. When money is laid out into circuits, the decision becomes mechanical: this bucket isn't for that — full stop. Discipline isn't willpower. Discipline is an architecture in which a bad decision takes more effort than a good one.
No engineer in their right mind keeps production without a backup. Ask anyone who survived a disk crash with no copy — their eye still twitches. And now the same person in their personal life lives revenue to revenue, with a zero cushion, and calls it "I reinvest everything into growth." In code it's called something else: "we have no backups because we spent the budget on new features." Everyone knows how that ends. The only question is whether the data center burns down before you set up replication.
The safety cushion is your life's off-site backup. Financial planners usually recommend having on the order of 3–6 months of expenses in instantly accessible cash; for a founder with irregular income and dependents, the reasonable ceiling shifts up. But the number isn't what matters — what matters is the humiliation it buys you out of. A person without a cushion sells the company's stake, and himself, left and right — as long as it's by Friday. A person with a cushion can at least haggle over the price of his own humiliation. The cushion is a month's supply of dignity; without it you sell not just your time but your right to choose whom to bow to.
And here's the genre's favorite self-deception. The founder says: "my money isn't lying dead in cash, it's working in the business." Sounds smart. Translation: "I took the backups off prod because the disks they were on can also be used for load." Yes, they can. Right up until the incident. A cushion that "works in the business" is a backup stored on the very disk you're afraid to lose. That's not a cushion. It's a decorative pillowcase on the same god-object. At the funeral of such a business, everyone notes not the losses but the brand of the hearse.
Hot memory — RAM, warm — SSD, cold — the archive in Glacier you reach in hours, but cheap and reliable. Money has exactly the same hierarchy — and it's almost never built consciously, because "liquidity" sounds boring until it suddenly becomes the only thing that matters.
To Bitcoiners this metaphor is even literally native: "hot wallet" versus "cold storage" — a wallet connected to the network for speed, versus keys taken offline for security. The same principle, just applied to all money, not only crypto. Liquidity tiers are the question "in how many hours can I turn this into bread without killing its value."
| Tier | System analog | What's here | Access time | Function |
|---|---|---|---|---|
| Hot | RAM / hot wallet | Current account, operating cash | Seconds | Pay today |
| Warm | SSD | Cushion, short deposits | Days | Survive a shock |
| Cool | HDD archive | Liquid investments, ETFs, bonds | Weeks | Growth + reserve |
| Cold | Glacier / cold storage | Stake in the business, real estate, BTC "forever" | Months–years | The long game |
Here's the real caricature — and it's most likely about you or someone you know. A founder with almost all his capital in the coldest tier (an illiquid company plus "bitcoin forever"), and exactly zero in hot and warm. He's a millionaire on paper and bankrupt on Tuesday, when the rent bill is due. On the balance sheet — a fortune. In the wallet — empty. The illusion of wealth that doesn't convert into bread without a discount and a three-week wait. Millions in the cap table don't keep you warm when, on Friday, your card declines at the supermarket and you stand at the register counting in your head how much honor you've got left to call a friend and ask to borrow for groceries.
Thousands go through this same trap in a milder form: formally, everything is fine — but the free, instantly accessible cash isn't enough for peace of mind. The discovery, banal to the point of shame: peace isn't a function of total wealth. Peace is a function of the share of wealth in the warm tier. You can have a lot and not sleep. You can have less and sleep, because you know three months are covered, whatever happens to the cold tier.
Now the most technical part — and the most beautiful, because it has an exact formula founders ignore most demonstratively. The question sounds simple: if you have an edge — how much capital to bet on one move? The founder's intuitive answer: "if I'm sure — then everything." The mathematical answer: almost never everything, and here's exactly how much.
In 1956, Bell Labs engineer John Kelly Jr. derived the formula for optimal bet size — the Kelly criterion. The essence in human terms: bet size should be proportional to your edge and inversely proportional to the odds of losing. Even with a real edge, Kelly almost always says to bet a fraction, not everything — because total ruin has an infinite cost: from zero there's no multiplier that returns you to the game. Edward Thorp, the mathematician who beat both blackjack and Wall Street, describes applying this principle in his memoir "A Man for All Markets" — and there he's very concrete about the bet sizes he allowed even with an obvious edge.
And here's why it's destructively important for you personally. The business is already the maximum Kelly bet on yourself: you put in time, reputation, and most of your money. When on top of that you keep the whole personal cushion in the same asset, that's exceeding Kelly — an over-bet, a bet larger than optimal. And beyond Kelly begins a paradox few feel in their gut: by increasing the bet past the optimum, you raise not the expected result but the probability of ruin. An over-bet is like revving the engine past the red zone: on the first lap you feel like a Formula 1 driver, on the third you collect the wreckage. How this works together with ergodicity and survivorship bias is unfolded in "Leap into the Night," the next text in this series. Here one phrase is enough.
Let's translate it to daily life. Imagine two players at the same table with the same honest edge. The first bets a sensible fraction of his stack on a hand, loses, stays in the game, recovers on later hands where the edge finally materializes. The second bets the whole stack "because I know I have an edge," loses one hand — and walks home on foot, forever, never living to see his rightness come true. Both were right about the edge. One was right about size. The wallet answers the question "how much do I have." The stack answers "what falls first and whom it crushes on the way down."
 *Two with the same edge. One bets a fraction and stays at the table. The other bets it all — his chair already pushed back. A crowned duck has settled on the doomed stack: concentration dressed up as ego again.*In code, technical debt is when you take a quick dirty decision today to hit a deadline, and pay later, in interest, as bugs and hours of refactoring. In capital, its exact twin is buying status on credit or lease: a quick dirty decision "to look like I'm already there," for which you pay interest every month while it quietly gnaws at liquidity.
A leased G-Wagon is the canonical example, which is why it sits in the first section's heading. It's not a purchase. It's a subscription to a self-image, debited monthly from the same account the cushion should be growing in. A founder who can't afford to lose a month of income, but "can afford" a premium crossover at thousands of euros a month, did exactly what an engineer who wrote no tests but spent a week on a beautiful logo animation during loading did. The product crashes — but crashes elegantly. Leasing status is when you pay monthly for the chance to seem like the person you lacked the courage not to be.
This is not to say status spending is evil. Sometimes an expensive thing is a tool: it opens doors, signals to partners, saves your time. Technical debt in code can also be rational — consciously taken, with a repayment plan. The difference is one question: did you take this debt consciously, with a plan, or did you just convert anxiety about your own weight into a monthly payment that now demands the business never fail? Conscious debt is a lever. Unconscious debt is a noose you bought yourself on installments — and thanked the seller for.
Every "leased G-Wagon" is a god-object that quietly wired self-esteem to itself too. Now you can't sell the car, because that means admitting you're not where you pretended to be. The founder can no longer slim down financially without gaining the weight of shame: selling the car is like going out in stories with no filter and no caption "this was planned." Every status payment is a hidden margin call on self-esteem: the moment income sags, the bank takes first not the car, but the feeling that you're still someone. Ego debt is the only creditor paid not in money but in an ever-narrowing corridor of decisions.
The most elementary principle of reliable systems is removing the single point of failure. Any senior will design a service so one node's fall doesn't bring it all down. And now the same senior, as a private person, keeps all his money in one bank, all his keys in one place, all his jurisdiction in one country — and considers it not a risk but order. Two banks and two jurisdictions isn't paranoia. It's a simple healthcheck: if there's nowhere to switch, there's no system, only hope.
March 2023: Silicon Valley Bank, the pillar of a large part of the tech community, falls in 48 hours — one of the fastest bank runs in US history. On Friday the regulator closes the doors. And thousands of startups suddenly find they kept all operating cash in one account at one bank, because "everyone sat there." Some couldn't make payroll on Monday not because they were insolvent, but because their money froze at one point they never duplicated. Picture the scene: you're "worth" several million, and that same Friday you stare into the bank's app, which says "temporarily unavailable," and learn for the first time in your life what your capital looks like without the UI shell of a banking app. Over those two days you maybe understood for the first time the difference between "having money" and "having access to money."
In the end the depositors were saved — but the lesson cost each of them one heart-attack weekend. The moral isn't "banks are bad." The moral is the system was designed so that one node's fall took down all of production, and that was called convenience. Convenience and fault-tolerance are antagonists. A founder who keeps everything in one bank deserves not sympathy but a boot-camp course in internet banking.
 *One node with everything plugged into it. A founder in pajamas at dawn watches his single point of failure burn; payroll is Monday. A crowned duck sits on the flaming server, slightly singed, crown still on.*| Single point of failure | Failure scenario | Replication (no SPOF) |
|---|---|---|
| One bank | Receivership / frozen account | 2–3 banks, operating cash spread out |
| One currency | Devaluation eats the cushion | A currency basket matched to your spending |
| One asset (BTC / your own company) | Drawdown on the god-object | Concentration limit, rebalance |
| One key / one medium | Loss = loss of everything | Backup keys, distributed storage |
| One country | Jurisdictional / political shock | A backup jurisdiction, mobility |
The last engineering habit, without which a system rots, is versioning and regular commits. A codebase untouched for months isn't "stable" — it goes stale: dependencies aged, security holes opened, no one remembers anymore why it's like this. Capital behaves the same. A portfolio you haven't looked at for a year isn't disciplined. It just quietly skewed: one asset grew and now takes 70% instead of the 30% you signed up for.
A rebalance is a commit with a description of what changed and why. Not a heroic gesture, not "I'm taking profits because I got scared," but mundane hygiene: once a quarter, sit for an hour with a spreadsheet — "it was 30% in the business, now it's 55%; it was 10% in BTC, now 35%" — and physically press "sell" where a module overran its limit. Founders hate rebalancing for the same reason they hate deleting their own favorite but dead code: selling part of an asset that grew feels like betraying a winning bet. But it's not a betrayal. It's a refusal to let one module quietly become a god-object just because it got lucky.
If you're reading this with your phone open on a BTC chart and a zero cushion — this paragraph is about you, not Saylor. You bought BTC at one price, it grew several times over, but you still have a zero warm tier — that's not a strategy. It's a refusal to look at the dashboard of your own financial state. "HODL, never sell" is a beautiful mantra while the asset rises; it also becomes the mechanism that forbids rebalancing exactly when it's most needed. What happens to a machine that refuses to look at the meter is covered in a neighboring text of the series — about Saylor's stalling flywheel. A dream you can't trim slightly is no longer a position, it's a cult. A version with no commits isn't immortal. It just hasn't fallen in front of everyone yet.
Look at your balance sheet through this prism — and honestly note which layer you're stuck on. Any financial structure runs through these levels. The first four make it work. The rest keep it from quietly degenerating into a god-object holding 90% of your life.
| Layer | Engineering principle | Question to capital |
|---|---|---|
| 1. Modules | Separation of concerns | Does every dollar know its bucket? Or is it all one warm number? |
| 2. Backups | Off-site replica | How many months of life in the warm tier, untouchable for "opportunities"? |
| 3. Tiers | Hot / warm / cold storage | What can you turn into bread in hours, and what in years? |
| 4. Sizing | Position sizing (Kelly) | Does no bet kill the system entirely, even if you're sure? |
| 5. Concentration | Anti-god-object | What's the max share in one asset before the bell rings? |
| 6. Fault tolerance | No single point of failure | Does one bank / key / country falling bring it all down? |
| 7. Debt | Conscious vs hidden tech debt | Is this payment a lever with a plan, or a lease on anxiety? |
| 8. Versioning | Regular commits | When was the last rebalance — or did the system quietly skew? |
| 9. Monitoring | Alerts and dashboards | Do you see the structure monthly — or once a year, from a heart attack? |
| 10. Auditor | External code review | Who, that you can't talk around, sees your balance without your illusions? |
Most founders live on layer zero — "how much money do I have," one warm number — and quietly go bankrupt on it, never learning that under the proud "I'm all-in" no architecture ever assembled. Serious money — funds, treasuries, family offices — have long lived on layers three through ten: they don't trust the feeling of abundance, only a structure with limits, tiers, and an audit. A god-object in a portfolio is a black hole: at first it looks like a star that warms everyone, then it bends space so that any decision only flies one way. The difference between "just earning" and "running capital" is the difference between a wallet and a stack.
And here's where it's worth pausing for a second. Most of the tools described above — buckets, rebalancing, jurisdictional diversification — assume the existence of capital there's something to distribute. A pre-seed founder with no runway can't even think about diversification: there's a fight for survival, and concentration is the only available tool. This architecture is for those already past survival. For those still going through it, the rules are different. But here too there's a distributive asymmetry not to ignore: access to the VC network that advises how not to diversify and stay all-in belongs mostly to founders who already have something to lose. A leased G-Wagon at €1,500/month is an option for people with the corresponding income. BTC cold storage is for those who already have something to take cold. These decisions aren't neutral to your starting position. The system rewards those who can afford pretty mistakes.
 *Four buckets, one overflowing. The crowned duck moved into the fullest box — concentration always grows an ego. A hand froze over a decision that should have been mechanical.*Because all of the above reads as "optimize every cent, turn life into a dashboard" — and that's a straight road to the person who counts the cost of every hour with their child and proudly reports the ROI of dinner with their wife. The most expensive mistake of the engineering approach to money is applying it where it bleeds you dry: generosity, risk for the sake of loved ones, money given without counting, help that doesn't get logged in a table.
The boundary is precise, as it should be. Every dollar with which you provision the system must know its bucket. But not every dollar must pass through a spreadsheet, because part of the money exists exactly to step outside accounting. The cushion, the tax reserve, concentration limits — that's infrastructure, and it should be cold and counted. But a gift, a risk for a friend, a contribution to something you simply believe in — no: the moment you attach a KPI to them, they stop doing what they existed for. A miser with perfect capital architecture and an empty life is also a failed design, just failed elegantly, with pretty charts and evenings where no one calls, because you once counted everyone.
| Money | Bad spreadsheet | Honest function |
|---|---|---|
| Business operating cash | — | Cold calculation, separate circuit, rebalance |
| Tax reserve | — | Concrete it over, you're only the custodian |
| Cushion | "Let it work in the business" | Warm tier, untouchable, read-only |
| Gift to a loved one | Count the ROI of gratitude | Just give it, no table |
| Help you believe in | Optimize "efficiency" | A separate "no-accounting" bucket you don't report on |
Go back to the founder from the first section — the one who argued for an hour about the button color and two hours later wired all his savings into one god-object without blinking. That's not an abstraction. That's most people who right now run a smart business and a child's wallet in the same head. Adulthood here isn't "learn to earn more." Earning a lot and keeping it all in one unlabeled number is just scaling fragility: the bigger the god-object, the louder it falls.
So don't fear concentration as an idea — at the start you must be all-in on one thing, or you'll build nothing. Just never stop at that version. Lay buckets under it that know their purpose. Move the backup into a warm tier, untouchable for the next "unique opportunity." Set a concentration limit while your inner "hodl forever" hasn't yet forbidden you any sober rebalance. And, hardest of all — have the courage to switch on, for your own balance sheet, the same linter you long ago switched on for your code: underline in red "you've got 90% of your life in one file with no backup."
Not everyone needs a complex portfolio. But almost every founder carries in his head one warm god-object called "my money," which was once an honest simplification and now quietly holds hostage both his liquidity and his sleep and his ego. Maturity is revoking one number's root access to your whole life: leave the business as the main bet, but take away its keys to the cushion, the taxes, the family, and your ability to sleep calmly on a Tuesday when the next bank "everyone sat in" falls. Because a founder's real status is measured not by the number of zeroes in a bank screenshot, but by how many of them you can lose overnight and still sleep.
Adulthood is not believing you have "everything under control," but keeping a record of what architecture your money has. Part is a god-object, and you see it. Part is spread across buckets. Part is moved into the cold. That's more honest than any warm number: not to feel rich, but to run capital like a system — and find the courage to rewrite it before one bank's collapse rewrites it for you.
An analytical breakdown of an engineering habit carried onto personal finance. Not financial, investment, or tax advice. No sentence here is a recommendation to buy, sell, or hold a specific asset. Make decisions about your own money with a qualified advisor and at your own responsibility.