Seth Walsh
Iconoclast
Contributor
- Joined
- Jan 12, 2020
- Posts
- 11,045
- Reputation
- 22,604
THE SURVIVAL-FIRST WEALTH PROTOCOL
How to compound for life without ever blowing up
ERGODICITY • POSITION SIZING • ANTIFRAGILITY • OPTIONALITY • COMPOUNDERS
How to compound for life without ever blowing up
ERGODICITY • POSITION SIZING • ANTIFRAGILITY • OPTIONALITY • COMPOUNDERS
Most wealth advice begins in the wrong place.
It asks which stock will rise, which fund has the best return, which asset class will outperform or how to squeeze another percentage point from a spreadsheet.
The first question is more brutal:
WHAT CAN REMOVE YOU FROM THE GAME?
If your answer is margin, one employer, one company, one currency, one customer, one lawsuit, one private key, one adviser or one panicked decision, your portfolio is not optimised. It is merely waiting for the path that exposes it.
A return is not a property you collect in a vacuum. It is something that happens through time to one non-replaceable person.
You do not get the average of one thousand parallel lives.
You get one sequence.
This thread is the complete architecture for making that sequence survive long enough to compound.
- The mathematics of why a positive average can still destroy you.
- Why position size matters more than idea quality.
- How to use Kelly without becoming a Kelly cultist.
- How to build a catastrophe floor, a compounding core and optional upside.
- How to own businesses that can reinvest for decades.
- How to identify hidden short-volatility, bad incentives and fake safety.
- How to spend wealth without turning your life into a terminal-net-worth competition.
This is a decision architecture, not a personal asset allocation. Exact implementation depends on your country, taxes, liabilities, family, human capital and ability to live through losses.
TABLE OF CONTENTS
PART I — THE MATHEMATICS OF NOT DYING
1. Expected wealth can rise while your wealth dies
2. Drawdowns consume time, not merely money
3. Ruin is an absorbing state
4. Position size is more important than idea quality
5. Kelly is a speed limit, not a target
PART II — BUILD THE ARCHITECTURE
6. Your real balance sheet is larger than your brokerage account
7. Build the catastrophe floor
8. The floor–core–opportunity barbell
9. Concentration creates wealth; diversification keeps it
PART III — OWN COMPOUNDING MACHINES
10. Become the tortoise
11. The five tests of a true business compounder
12. Why the index is a brutally strong default
13. Optionality is not the same as gambling
PART IV — REMOVE THE HIDDEN BLOW-UP
14. The inverse-turkey portfolio
15. Debt, leverage and forced selling
16. Audit skin in the game
17. Spend like a mortal
18. Fees, taxes and activity compound backwards
19. Uncertainty requires slack, not fake precision
PART V — THE OPERATING SYSTEM
20. Write the one-page investment constitution
21. Rules for each stage of wealth
22. The 25 non-negotiable laws
23. The 30-minute anti-ruin audit
PART I — THE MATHEMATICS OF NOT DYING
THE DIFFERENCE BETWEEN A GOOD AVERAGE AND A GOOD LIFE
THE DIFFERENCE BETWEEN A GOOD AVERAGE AND A GOOD LIFE
1. EXPECTED WEALTH CAN RISE WHILE YOUR WEALTH DIES
Consider a game.
A fair coin is tossed repeatedly.
- Heads: your wealth rises by 50%.
- Tails: your wealth falls by 40%.
The arithmetic expected return on each toss is positive:
Code:
0.5 × (+50%) + 0.5 × (−40%) = +5%An analyst looking across thousands of parallel players says the game is excellent. Expected wealth multiplies by 1.05 each round.
After 100 rounds:
Code:
Expected wealth = 1.05^100 = 131.50× starting wealthNow stop looking across imaginary parallel people and follow one real person through time.
A typical 100-toss sequence contains roughly 50 heads and 50 tails. The result is:
Code:
1.5^50 × 0.6^50 = 0.9^50 = 0.00515× starting wealthThat person loses 99.485%.
The expected outcome says 131.5×.
The typical time-path says 0.005×.
The gap is more than 25,000-fold.
What happened?
Returns multiply. A 50% gain and a 40% loss do not cancel:
Code:
1.50 × 0.60 = 0.90Every two-round cycle loses 10%.
The relevant long-run quantity is the average logarithmic growth rate:
Code:
g = 0.5 ln(1.5) + 0.5 ln(0.6)
= 0.5 ln(0.9)
= −5.268% per roundThis is the cleanest introduction to ergodicity economics: the average across many simultaneous copies need not describe what happens to one copy across time.
This example proves a narrow and devastating point: expected wealth can be dominated by rare paths and can diverge from typical compound growth.
It does not prove that all expected-utility economics is worthless. Log utility reproduces full-Kelly growth optimisation, while higher risk aversion can represent fractional Kelly, consumption floors and finite lives.
You do not need to win an ideological war to use the result. Keep the mechanism:
Discard the unnecessary claim that one framework has made every other framework obsolete.
It does not prove that all expected-utility economics is worthless. Log utility reproduces full-Kelly growth optimisation, while higher risk aversion can represent fractional Kelly, consumption floors and finite lives.
You do not need to win an ideological war to use the result. Keep the mechanism:
- wealth is multiplicative;
- paths matter;
- volatility creates drag;
- ruin cannot be averaged away.
Discard the unnecessary claim that one framework has made every other framework obsolete.
2. DRAWDOWNS CONSUME TIME, NOT MERELY MONEY
People speak about gains and losses as if percentages were symmetric.
They are not.
If you lose a fraction L, the gain required to recover is:
Code:
Recovery gain = 1 / (1 − L) − 1
Code:
LOSS GAIN NEEDED TO RECOVER
10% 11.1%
20% 25.0%
33% 49.3%
50% 100.0%
75% 300.0%
90% 900.0%A 50% drawdown is not “half as bad” as ruin. It demands a doubling merely to return to zero progress.
If your sustainable return is 8% per year, a 50% loss costs roughly nine years of compounding before tax, fees and withdrawals.
A deep drawdown also attacks everything the spreadsheet omits:
- You may need to sell assets to fund life.
- Your lender may change the rules.
- Your income may fall in the same recession.
- Your risk tolerance may collapse after the loss rather than before it.
- The opportunity set may improve exactly when your liquidity disappears.
This is why volatility is not merely emotional discomfort. When it interacts with withdrawals, leverage, job loss, taxes or margin calls, volatility becomes forced action at a bad price.
3. RUIN IS AN ABSORBING STATE
Most financial models quietly assume that after a bad outcome you continue to the next period.
Ruin says otherwise.
If your wealth hits zero, your broker liquidates you, your business loses payroll, your house is repossessed or your health removes your earning capacity, the process does not politely reset.
There is an absorbing barrier.
The probability of at least one catastrophic event over repeated independent exposures is:
Code:
P(at least one catastrophe) = 1 − (1 − p)^nEven small one-period risks accumulate:
Code:
1% annual ruin risk for 40 years = 33.1%
0.1% ruin risk repeated 1,000 times = 63.2%
0.01% repeated 10,000 times = 63.2%Real risks are not perfectly independent. Recessions connect your job, home, stock portfolio, private business and access to credit. Correlation tends to appear exactly when the floor is under pressure.
Your total ruin surface includes:
- market ruin: leverage, concentration, short options, forced liquidation;
- income ruin: one employer, one customer, one industry;
- liability ruin: personal guarantees, lawsuits, uncapped obligations;
- operational ruin: fraud, custody failure, lost keys, tax non-compliance;
- biological ruin: death, disability, addiction, untreated illness;
- behavioural ruin: panic, mania, revenge trading, lifestyle lock-in.
Do not analyse these in separate mental accounts if one of them can destroy the same household.
Ruin is indivisible.
This does not mean taking no risk. Avoiding all volatility can create a slow ruin through inflation, stagnating income and opportunity cost.
It means separating risks that bruise from risks that remove the player.
4. POSITION SIZE IS MORE IMPORTANT THAN IDEA QUALITY
Suppose you have a coin that lands heads 60% of the time.
Every bet pays even money. You have a real edge.
How much should you risk on each toss?
- Risk 5%: cautious growth.
- Risk 10%: strong growth with more survivability.
- Risk 20%: maximum expected log growth under the stated assumptions.
- Risk 50%: the typical path shrinks despite the edge.
- Risk 75%: destruction.
The log-growth rate is:
Code:
g(f) = 0.6 ln(1 + f) + 0.4 ln(1 − f)After 100 typical 60/40 outcomes:
Code:
FRACTION RISKED TYPICAL WEALTH MULTIPLE
5% 2.40×
10% 4.50×
20% 7.49× ← full Kelly maximum
30% 4.37×
40% 0.78×
50% 0.033×
75% ~0.0000000003×Nothing happened to the quality of the coin.
Only the size changed.
At 20%, the edge compounds.
At 50%, the same edge destroys the typical player.
This is why “How confident are you?” is not a sizing system.
Confidence is usually highest when:
- the price has already risen;
- the narrative is socially rewarded;
- your recent bets have worked;
- the downside has not appeared for a long time;
- your estimate error is least visible.
Sizing must be a rule written before the emotional state it is designed to control.
5. KELLY IS A SPEED LIMIT, NOT A TARGET
For a binary bet with probability p, loss probability q and net odds b, full Kelly is:
Code:
f* = (bp − q) / bFor the 60/40 even-money coin, f* = 20%.
Full Kelly maximises long-run expected log growth if:
- you know the true probabilities;
- the payoff distribution is correctly specified;
- the opportunity repeats under stable conditions;
- bets are independent or correlations are modelled correctly;
- you can rebalance continuously without taxes, costs or liquidity problems;
- your objective is log-growth rather than a smoother lifetime spending path.
That describes a clean gambling problem better than it describes your life.
Real investors face finite careers, ageing, dependants, withdrawals, uncertain edges, fat tails, taxes, illiquidity and multiple bets that secretly depend on the same factor.
Full Kelly is therefore best used as an upper-bound diagnostic.
Half Kelly often preserves roughly three quarters of the theoretical growth while materially reducing the violence of the path.
Quarter Kelly may be more rational where the edge is subjective, the distribution changes, exits are illiquid or the position correlates with your income and business.
The correct fraction is not a religious identity.
It is a translation of your total balance sheet, uncertainty and preferences into a survivable exposure.
Use this sizing sequence:
- Write the bear, base and bull cases. Include a probability that the entire model is wrong.
- Calculate maximum permanent loss. Do not substitute recent volatility for downside.
- Map hidden correlation. Salary, employer stock, home, business, country and currency count.
- Calculate full Kelly only if odds are estimable. If they are not, do not fabricate decimals.
- Fraction it aggressively. Model error grows faster than your confidence admits.
- Apply a hard life cap. No formula may threaten the catastrophe floor.
- Size the whole theme. Five “different” AI stocks can be one bet.
A useful approximation for the fraction in a risky asset is:
This teaches three durable things:
It is not a machine for turning a noisy forecast into a precise allocation. Expected return is the least stable input. Use ranges, robust constraints and a smaller position than the most optimistic estimate suggests.
Code:
w* ≈ (expected excess return) / (risk aversion × variance)This teaches three durable things:
- more edge can justify more exposure;
- more variance justifies less exposure;
- more personal risk aversion justifies less exposure.
It is not a machine for turning a noisy forecast into a precise allocation. Expected return is the least stable input. Use ranges, robust constraints and a smaller position than the most optimistic estimate suggests.
PART II — BUILD THE ARCHITECTURE
MAKE THE HOUSE HARD TO KILL BEFORE DECORATING IT
MAKE THE HOUSE HARD TO KILL BEFORE DECORATING IT
6. YOUR REAL BALANCE SHEET IS LARGER THAN YOUR BROKERAGE ACCOUNT
Your portfolio is not the screen your broker shows you.
Your economic balance sheet is closer to:
Code:
Liquid financial assets
+ pensions and long-term claims
+ private business and property equity
+ present value of future after-tax earnings
− debts and guarantees
− present value of essential future spending
= total economic wealthYou do not need a fake-precision present value for every line. A range is enough to expose correlations that conventional portfolios ignore.
Human capital is usually the largest asset of a young person: the after-tax earnings their skills, health, reputation and remaining working years can produce.
The question is not only how large it is.
The question is what it behaves like.
Bond-like human capital
- stable public or tenured employment;
- regulated profession with persistent demand;
- income weakly linked to market cycles;
- portable skills and multiple employers.
Equity-like human capital
- founder or commission-heavy income;
- finance, technology, construction or cyclical industry exposure;
- bonus and employment both depend on asset prices;
- skills tied to one company, geography or boom.
If your salary, bonus, pension, house and stock portfolio all prosper in the same boom, you are not diversified because the tickers differ.
You are one recession trade wearing five outfits.
Examples:
- An employee with company stock should count both the stock and the job as exposure to the same firm.
- A property developer should not pretend a portfolio of banks, builders and local real estate is diversified.
- A founder may rationally keep business concentration, but should isolate personal survival capital outside the company.
- A young person with little financial wealth should spend serious attention on skill acquisition because a 20% increase in lifetime earning power can dwarf a clever fund choice.
7. BUILD THE CATASTROPHE FLOOR
The catastrophe floor is the set of resources and protections that prevent a bad path from becoming an irreversible one.
It is not “dead cash.”
It is stored refusal.
It lets you refuse a fire-sale price, a predatory loan, a fraudulent partner, an abusive employer and a panicked liquidation.
The floor has six layers.
1. Liability-matched liquidity
Ring-fence money for:
- essential living costs;
- taxes already incurred;
- near-term housing, education or medical commitments;
- business payroll and wind-down obligations;
- insurance deductibles and known repairs.
Hold it in assets that match the liability's currency and date. A five-year bond, volatile equity fund or private-credit vehicle is not cash for a bill due next month.
Illustrative runway bands—not universal commandments:
- 6–12 months of essentials for stable, diversified employment and strong support.
- 12–24 months for variable income, dependants or weak support.
- 24–36 months for founders, concentrated wealth, cyclical work or long illiquidity.
2. Catastrophe insurance
Insure losses that could break the floor: health, disability, liability, property and—where others depend on your income—life. Self-insure small, affordable annoyances with higher deductibles where sensible.
3. No callable personal fragility
Avoid structures where another party can force you to transact:
- margin debt;
- cross-collateralised loans;
- uncapped personal guarantees;
- capital calls you cannot meet from safe liquidity;
- short positions with theoretically unlimited loss;
- debt whose rate or maturity mismatches the asset.
4. Operational redundancy
- strong unique passwords and hardware-backed two-factor authentication;
- offline recovery codes and tested backups;
- current beneficiaries, will and powers of attorney;
- clear records of tax basis, account ownership and private keys;
- a second custodian or banking relationship when the sums justify it;
- another competent person who can locate the system if you cannot.
5. Health and earning continuity
Sleep, strength, cardiovascular fitness, preventive care and addiction control are not lifestyle footnotes. They protect the asset that finances every other asset.
6. Behavioural circuit breakers
- no large financial decision during mania, grief, intoxication or sleep deprivation;
- a cooling-off period before new speculative positions;
- no averaging down without a rewritten thesis;
- no changing the leverage rule because the last five trades worked;
- one trusted dissenter for irreversible decisions.
8. THE FLOOR–CORE–OPPORTUNITY BARBELL
Taleb's barbell is often misunderstood as “put 90% in cash and gamble the other 10%.”
That is one example, not a universal portfolio.
The deeper principle is to avoid the dangerous middle where an exposure looks moderate but contains ambiguous, unbounded or correlated downside.
Use three functions:
THE FLOOR — CAPITAL THAT MUST DO ITS JOB
- essential liquidity;
- near-term liabilities;
- catastrophe insurance;
- high-quality safe assets matched to spending currency and horizon.
Its objective is not maximum return. Its objective is to prevent forced action.
THE CORE — CAPITAL THAT COMPOUNDS PRODUCTIVELY
- broad, low-cost ownership of productive businesses;
- tax-efficient long-duration holdings;
- diversified risk sources appropriate to the household;
- an allocation you can hold through a severe drawdown.
Its objective is durable real growth after fees, tax and behaviour.
THE OPPORTUNITY SLEEVE — CAPITAL THAT MAY FAIL SAFELY
- a controlled business;
- concentrated public equities where you have an earned edge;
- start-ups, special situations or digital assets;
- skills, projects and experiments with open-ended upside;
- explicitly convex trades with known maximum loss.
Its objective is asymmetric upside. Its maximum total failure must leave the floor and core intact.
No universal percentages exist.
A founder's business may dominate total wealth. A retiree's floor may be large. A young employee with stable human capital may hold more productive risk. Someone whose income and home already load on one country may need more global diversification.
The architecture is universal.
The allocation is personal.
Rebalance with bands, not feelings.
- Use new contributions and cash flows first.
- Set tax-aware percentage or risk bands.
- Refill the floor after using it.
- Trim an opportunity position when it threatens the household, not merely because it won.
- Do not “rebalance” a broken thesis back to its original size.
9. CONCENTRATION CREATES WEALTH; DIVERSIFICATION KEEPS IT
This slogan is useful only when you understand both halves.
Extreme wealth is often created through concentration:
- founding a company;
- owning a meaningful stake;
- developing a rare skill in one field;
- holding a great business through years of growth.
But concentration is rational only when at least one of these is true:
- you have control;
- you possess a real informational or analytical edge;
- the concentration is the unavoidable price of creating the asset;
- the maximum failure does not destroy the household.
Concentration without control, edge or a protected floor is not conviction.
It is undiversified faith.
Public equities have extreme positive skew. A small minority of companies create a disproportionate share of total market wealth, while many individual stocks underperform safe bills over their full lives.
That creates two problems for the stock picker:
- You must avoid a large population of permanent losers.
- You must also own and not prematurely sell the rare extreme winners.
A diversified market-cap-weighted index solves this with a simple mechanism:
- any one loser can only fall by the amount invested;
- a winner can become hundreds of times larger;
- the index keeps increasing exposure to proven winners without forecasting their identity in advance;
- turnover and fees can remain low.
This is not magic and it does not remove market risk. It explains why broad ownership is a powerful default.
The correct transition is usually:
CREATE WITH CONCENTRATION → ISOLATE THE FLOOR → DIVERSIFY THE SURPLUS → KEEP OPTIONAL UPSIDE
Do not diversify a great controlled business so early that you never create wealth.
Do not remain all-in so late that one idiosyncratic failure erases a life-changing outcome.
The line is not determined by ego. It is determined by how much permanent loss would change the rest of your life.
PART III — OWN COMPOUNDING MACHINES
LET TIME DO THE HEAVY LIFTING
LET TIME DO THE HEAVY LIFTING
10. BECOME THE TORTOISE
There are three archetypes in capital markets.
THE SHARK wants a spectacular internal rate of return on a rapid deal. It needs motion, exits and another meal.
THE ELEPHANT wants more assets under management. It gets paid on scale whether the client's life improves or not.
THE TORTOISE wants survival, reinvestment runway and duration.
The tortoise understands the arithmetic everybody quotes but almost nobody obeys:
Code:
10% compounded for 10 years = 2.59×
10% compounded for 20 years = 6.73×
10% compounded for 30 years = 17.45×
10% compounded for 40 years = 45.26×
10% compounded for 50 years = 117.39×This is an illustration, not a promised return.
The point is that duration eventually dominates theatrical intensity.
A sustainable 10% for half a century beats a sequence of dazzling gains interrupted by one extinction event.
The tortoise's enemies are not merely bad stock picks:
- deep permanent loss;
- excess leverage;
- high recurring fees;
- tax-creating turnover;
- prematurely selling extreme winners;
- lifestyle withdrawals that rise with every good year;
- changing systems after every period of underperformance;
- losing attention to a constantly moving scoreboard.
That requires a portfolio and a psychology designed for boredom.
The tortoise protocol:
- Own assets that can produce and reinvest cash.
- Pay a price that leaves room for error.
- Keep costs and turnover low.
- Protect the floor so volatility cannot force a sale.
- Judge business progress over years, not price noise over days.
- Interrupt only when the underlying compounding engine changes.
11. THE FIVE TESTS OF A TRUE BUSINESS COMPOUNDER
A company with a high historical return on capital is not automatically a compounder.
The critical question is what it can do with the next unit of capital.
A rough operating identity is:
Code:
Long-run operating growth ≈ reinvestment rate × incremental return on capitalIf a business earns 30% on its old capital but can reinvest almost nothing, it may be a wonderful cash cow but a limited internal compounder.
If it can reinvest heavily but new projects earn mediocre returns, growth can destroy value.
The exceptional case combines both.
TEST 1 — INCREMENTAL RETURNS
Ask:
- For each additional unit retained, how much additional sustainable after-tax operating profit appears?
- Is growth organic or purchased through acquisitions?
- Are reported returns flattered by underinvestment, capitalised costs, stock compensation or an old low-cost asset base?
- Does cash conversion confirm accounting profit?
Use the numerator and denominator consistently. A ratio is not insight if the accounting perimeter moves whenever management needs a better slide.
TEST 2 — REINVESTMENT RUNWAY
High returns matter most when they can persist across a large opportunity set.
Look for:
- a large or expanding addressable market;
- new products that use the same distribution advantage;
- geographic expansion without collapsing unit economics;
- customer retention and repeat purchasing;
- low marginal capital needs;
- room to deploy capital without attracting ruinous competition.
Time increases the value of an option only if the company survives and keeps the option alive.
TEST 3 — COMPETITIVE DURABILITY
“Moat” is not a magic word. Identify the mechanism.
- switching costs;
- network effects;
- cost advantage;
- brand embedded in repeated behaviour;
- regulatory licence;
- unique data or distribution;
- scale economies shared with customers;
- a culture that improves the product faster than rivals.
Then ask how the mechanism breaks.
A moat that depends on customers being trapped and angry may invite regulation, technological substitution or a focused competitor.
TEST 4 — MANAGEMENT AS CAPITAL ALLOCATOR
Management can operate the business brilliantly and allocate the resulting cash terribly.
Examine:
- insider ownership bought with real money, not merely granted options;
- compensation tied to per-share value rather than empire size;
- honest treatment of dilution and stock compensation;
- willingness to shrink, divest or stop failed projects;
- acquisition discipline;
- buybacks only below reasonable value;
- clear distinction between organic progress and purchased revenue;
- language that names mistakes without redefining the metric.
There are owner-like managers and managers who rent the company for salary, status and acquisition activity.
Do not confuse polished communication with aligned behaviour.
TEST 5 — BALANCE SHEET AND PRICE
A great company can be a terrible investment at the wrong price.
Your return depends on:
- business growth;
- cash distributed or reinvested;
- change in valuation multiple;
- dilution;
- debt and interest burden;
- tax;
- the price you paid.
Compare every new idea with the expected after-tax return from simply keeping your existing portfolio.
Your opportunity cost is not cash yielding zero. It is the best realistic alternative already available to you.
Code:
INCREMENTAL RETURN ON CAPITAL /10
REINVESTMENT RUNWAY /10
CUSTOMER RETENTION / PRICING POWER /10
BALANCE-SHEET SURVIVABILITY /10
MANAGEMENT ALIGNMENT /10
ORGANIC CASH GROWTH /10
VALUATION / MARGIN OF ERROR /10
THESIS FALSIFIABILITY /10
Subtract:
CUSTOMER CONCENTRATION /10
CYCLICAL / REGULATORY FRAGILITY /10
DILUTION / ACQUISITION DEPENDENCE /10
ACCOUNTING COMPLEXITY /10Do not mechanically add the numbers into a fake scientific target price. The scorecard exists to force complete questions and expose what the narrative omitted.
12. WHY THE INDEX IS A BRUTALLY STRONG DEFAULT
A broad, low-cost, market-cap-weighted equity index is not intellectually impressive.
That is part of its power.
It provides:
- ownership of productive businesses;
- extreme diversification across individual failure;
- automatic retention of rare giant winners;
- automatic reduction of companies that shrink;
- low decision load;
- low fees and usually low turnover;
- no requirement to identify tomorrow's champions today.
It also has real weaknesses:
- it can suffer severe market drawdowns;
- market weights can embed expensive sectors or countries;
- tax and fund rules differ by jurisdiction;
- currency may mismatch near-term spending;
- an investor can still destroy the result by buying high and panic-selling low.
The index is not “safe” in the sense of stable price.
It is robust in the sense that it does not require repeated acts of genius.
Active concentration must clear a high hurdle:
- What do you know or understand that the marginal owner does not?
- Why is that edge durable rather than a story you read publicly?
- Why does the expected excess return survive tax, fees, error and opportunity cost?
- What evidence would falsify the thesis?
- What position size leaves you alive if you are completely wrong?
If you cannot answer, broad ownership is not surrender.
It is the rational refusal to pay tuition to the market for the pleasure of feeling exceptional.
13. OPTIONALITY IS NOT THE SAME AS GAMBLING
An option is valuable because it creates asymmetry.
You can choose to continue after good information and stop after bad information.
Real optionality has four properties:
- bounded cost: the maximum loss is known and survivable;
- open upside: gains can be many times the cost;
- choice: you are not obligated to keep funding failure;
- learning: small exposures produce information before large commitment.
Positive examples
- building a prototype before funding a company;
- learning a scarce skill with applications across industries;
- publishing work that can create unlimited distribution;
- a small investment in a business whose downside is the stake and whose upside is large;
- running many cheap experiments, then concentrating resources in the few that show evidence.
Counterfeit optionality
- weekly out-of-the-money calls bought without an informational edge;
- a start-up that repeatedly demands rescue capital;
- an illiquid fund with surprise capital calls;
- a “small” leveraged position whose loss can exceed the premium;
- a side project with low financial cost but unlimited attention drain;
- averaging down because stopping would force you to admit error.
An option contract is not automatically good optionality. It is often a fairly or expensively priced zero-sum transfer after spreads and fees.
Likewise, an opportunity with extraordinary upside may deserve a smaller position, not a larger one.
Why?
If a 1% position can transform your life in the success state, increasing it to 20% may add little useful upside while making the failure state twenty times worse.
This is one of the least intuitive rules in wealth management.
Short selling belongs here as a shield, not a sword.
A short has at most 100% gross upside if the asset goes to zero, potentially unbounded loss if it rises, borrowing cost, recall risk and hostile timing. It can hedge a specific exposure when bounded carefully. It is structurally poor as the main engine of a decades-long compounding programme.
PART IV — REMOVE THE HIDDEN BLOW-UP
THE BEST ADDITION IS OFTEN A SUBTRACTION
THE BEST ADDITION IS OFTEN A SUBTRACTION
14. THE INVERSE-TURKEY PORTFOLIO
A turkey is fed every day and updates its model: humans are safe, food arrives reliably, volatility is low.
Then one day invalidates the entire track record.
The inverse turkey appears in investing.
Some strategies produce frequent small gains and conceal a rare catastrophic loss:
- naked option selling;
- leveraged carry trades;
- selling insurance without enough capital;
- martingale averaging-down systems;
- illiquid credit marked by models rather than transactions;
- strategies that borrow short and lend long;
- businesses whose profit depends on never seeing a simultaneous claim.
Their record looks safest immediately before it matters least.
Other strategies produce frequent small losses and conceal rare giant gains:
- many bounded experiments;
- venture portfolios with disciplined sizing;
- trend or crisis protection bought at a tolerable cost;
- creative work where most attempts fail but one can scale globally;
- a diversified equity portfolio that retains extreme winners.
Do not judge them with the same superficial statistic.
The hidden-tail audit
- Where is the leverage?
- Who can demand cash and when?
- Are losses marked by a liquid market or by the manager?
- What happens when every exit is crowded?
- Does the strategy add risk after losses?
- Can one bad period erase ten good years?
- Does reported “income” include compensation for selling catastrophe insurance?
- Who keeps past fees after the blow-up?
15. DEBT, LEVERAGE AND FORCED SELLING
Debt is not automatically evil.
It is a contract that moves future flexibility into the present.
The danger depends on the mismatch.
More defensible debt
- funds a durable productive asset;
- has a fixed or controllable cost;
- maturity matches the asset's cash flow;
- cannot be called because market price falls;
- leaves a large coverage margin;
- does not expose unrelated personal assets through guarantees.
Fragile debt
- funds consumption or status;
- has variable rates against fixed income;
- is short-term against a long illiquid asset;
- is secured by volatile collateral;
- cross-defaults across otherwise separate assets;
- depends on refinancing under friendly market conditions;
- creates personal liability for a speculative business.
Leverage does more than multiply gains and losses.
It transfers control of time to the lender.
An unlevered owner can wait through a 60% price decline if the asset survives. A margined owner may be forced to sell at the exact point expected returns improve.
This is negative optionality: the market chooses when you stop.
Stress every leveraged structure against a combined scenario:
- asset price falls 50%;
- income disappears for a year;
- credit cost rises;
- refinancing closes;
- the asset becomes temporarily illiquid;
- a tax or legal payment arrives on schedule anyway.
If the plan survives only because these events “would never happen together,” you have built a crisis correlation trade.
16. AUDIT SKIN IN THE GAME
Advice is not independent of the adviser's payoff.
The classic agency trade is simple:
- the decision-maker receives salary, bonuses or fees during the calm years;
- the hidden tail grows somewhere else;
- clients, shareholders or taxpayers absorb the failure;
- past compensation is not returned.
The person had upside in your game without symmetric downside.
Before trusting a manager, adviser, founder or product seller, ask:
- How are they paid? Fixed fee, commission, percentage of assets, spread, performance allocation?
- What do they personally own? Real net worth at risk or promotional token exposure?
- What happens if the advice fails? Clawback, reputation, lost capital—or simply a new sales job?
- Who controls the valuation? Public market, independent administrator or the same manager charging on the number?
- Can you exit? At what price, delay and tax cost?
- What is omitted from the headline return? Fees, financing, dilution, tax, illiquidity, survivorship?
- What alternative pays them less? Did they show it to you?
Skin in the game is not proof of competence. A reckless person can lose beside you.
No skin is not proof of fraud. A good fixed-fee lawyer need not invest in your company.
It is evidence about incentives, and incentives belong inside the analysis.
Prefer:
- transparent, comprehensible fees;
- written advice with assumptions and conflicts;
- independent custody;
- long records that include hostile regimes;
- managers whose own capital is invested on similar terms;
- clear capacity limits and willingness to return capital;
- decision-makers who name what would make them wrong.
17. SPEND LIKE A MORTAL
The purpose of wealth is not to die with the highest score.
Wealth funds:
- safety;
- time;
- health;
- family and relationships;
- autonomy;
- experiences;
- creative work;
- gifts and institutions that outlive you.
An investment system that maximises terminal capital while wasting the only years you can use it has optimised the wrong objective.
But rigid spending can also create ruin.
Use a floor-and-flex system.
Essential spending
Housing, food, healthcare, dependants, insurance and core obligations should be protected by safe assets, reliable income and insurance appropriate to the horizon.
Discretionary spending
Travel, luxury, gifts and optional projects should be allowed to move with wealth and opportunity. A percentage of current resources is safer than pretending a fixed real amount can never change.
Capitalise recurring lifestyle commitments.
A one-time €10,000 purchase costs €10,000.
A permanent €10,000 annual lifestyle increase requires a large additional capital base to support indefinitely. The subscription is usually more dangerous than the splurge.
Good spending converts money into durable capability:
- health and energy;
- time bought back from low-value friction;
- education with a credible skill or access payoff;
- a stable home that supports work and relationships;
- experiences with people you value;
- tools that increase earning or creative capacity.
Bad spending converts volatile income into rigid overhead:
- housing at the maximum a lender permits;
- cars, clubs and subscriptions that require continued peak earnings;
- status purchases financed by debt;
- a lifestyle calibrated to bonuses;
- dependants or relatives promised an undefined permanent bailout.
18. FEES, TAXES AND ACTIVITY COMPOUND BACKWARDS
A fee is not “only 1%.”
It is 1% this year, plus the future return on that 1%, plus the return on the return, repeated for decades.
Illustration:
Code:
€100,000 compounded at 7% for 40 years = €1,497,446
€100,000 compounded at 6% for 40 years = €1,028,572
One percentage point of annual drag = 31.3% less terminal wealthThat does not mean every adviser or fund charging 1% is worthless. It means the service must create enough after-tax, after-behaviour value to overcome an enormous lifetime hurdle.
Audit all friction:
- fund expense ratios;
- adviser fees;
- performance fees;
- platform and custody charges;
- bid–offer spreads;
- financing and borrow costs;
- foreign-exchange conversion;
- tax triggered by turnover;
- your time and attention.
Tax is part of return, not the only objective.
Optimise after-tax wealth, but do not accept catastrophic concentration merely to defer a tax bill. Do not buy a bad asset for a deduction. Do not make a legally or operationally fragile structure to save a visible percentage while creating an invisible tail.
Tax rules are jurisdiction-specific and change. Use written professional advice for pensions, trusts, companies, residence, inheritance and large realised gains.
Activity has a burden of proof.
Every trade must overcome:
SPREAD + FEE + TAX + ERROR RISK + LOST ATTENTION + OPPORTUNITY COST
If the thesis is “the price moved,” the burden has not been met.
19. UNCERTAINTY REQUIRES SLACK, NOT FAKE PRECISION
Risk is often described as a known distribution.
Investing gives you something worse: you do not know the true distribution, and it changes while you estimate it.
Expected return is especially noisy. Volatility, correlation, tail shape, liquidity and your own future behaviour are also uncertain.
An optimiser can turn tiny input differences into enormous allocation differences. That is not sophistication. It is a precision amplifier attached to weak evidence.
Use robust decision-making instead:
- Ranges, not point forecasts. Write bear, base and bull assumptions.
- Base rates before stories. Start with how similar assets usually fail.
- Parameter haircuts. Reduce the edge; widen the tails; raise assumed correlation.
- Fractional sizing. The less measurable the edge, the less capital it deserves.
- Constraints. Position, sector, liquidity and leverage caps prevent the optimiser from becoming insane.
- Slack. Spare liquidity, time and borrowing capacity protect against model error.
- Reversibility. Prefer decisions that can be stopped cheaply after new information.
- Pre-mortems. Imagine the position failed and ask which omitted mechanism caused it.
Uncertainty is not permission to do nothing.
It is a reason to choose structures that do not require perfect prediction.
The phrase “ergodicity economics” contains two separable claims.
Claim one: Multiplicative dynamics can make ensemble averages radically different from typical time paths.
That is demonstrable and central to this thread.
Claim two: This observation invalidates expected-utility economics and uniquely determines rational behaviour.
That is disputed. Expected utility can represent log-growth and more risk-averse objectives; real people may rationally prefer smoother spending, lower drawdowns or bequests over asymptotically maximum log wealth.
The intelligent move is not to join a tribe. It is to retain the mechanism that improves decisions:
Then choose an objective that reflects the life the capital is meant to serve.
Claim one: Multiplicative dynamics can make ensemble averages radically different from typical time paths.
That is demonstrable and central to this thread.
Claim two: This observation invalidates expected-utility economics and uniquely determines rational behaviour.
That is disputed. Expected utility can represent log-growth and more risk-averse objectives; real people may rationally prefer smoother spending, lower drawdowns or bequests over asymptotically maximum log wealth.
The intelligent move is not to join a tribe. It is to retain the mechanism that improves decisions:
PATH DEPENDENCE + MULTIPLICATION + RUIN + FINITE LIFE
Then choose an objective that reflects the life the capital is meant to serve.
PART V — THE OPERATING SYSTEM
TURN THE PHILOSOPHY INTO RULES THAT SURVIVE YOUR MOODS
TURN THE PHILOSOPHY INTO RULES THAT SURVIVE YOUR MOODS
20. WRITE THE ONE-PAGE INVESTMENT CONSTITUTION
A good investment policy is written when you are calm and consulted when you are not.
It should fit on one page.
Code:
PURPOSE
What is this capital meant to fund, for whom and when?
TOTAL BALANCE SHEET
Financial assets:
Human capital character:
Business / property concentration:
Debt / guarantees:
Essential annual spending:
CATASTROPHE FLOOR
Runway target:
Near-term liabilities:
Insurance:
Custody / operational backup:
CORE ALLOCATION
Target ranges, not fake-precise points:
Rebalancing bands:
Liability currency:
OPPORTUNITY SLEEVE
Maximum total allocation:
Maximum single-position permanent loss:
Required thesis fields:
Kelly fraction / uncertainty haircut:
ABSOLUTE PROHIBITIONS
Margin:
Naked options:
Personal guarantees:
Unfunded capital calls:
Unknown maximum loss:
BUY RULE
Edge, base rate, valuation, size, correlation, falsifier.
SELL RULE
Thesis invalidated; fraud/incentive break; better after-tax
opportunity; concentration threatens floor; liability arrives.
GOVERNANCE
Who can trade, approve, access and inherit?
REVIEW
Quarterly operational check; annual policy review;
event-driven review after major life change.A price decline is not automatically a sell signal.
Sell because:
- the business economics or thesis changed;
- the balance sheet became fragile;
- management integrity broke;
- your original evidence was wrong;
- a better opportunity clears tax and friction;
- position size now threatens lifetime objectives;
- the capital has reached the liability it was meant to fund.
Do not sell merely because:
- a winner looks large relative to its original cost;
- a loser would make you feel stupid if realised;
- the market produced a scary headline;
- a forecaster changed a twelve-month target;
- you are bored.
Review the system on a calendar, not on an adrenaline spike.
21. RULES FOR EACH STAGE OF WEALTH
STAGE 0 — EXPOSED
One modest shock threatens food, housing, health or employability.
Priority: stop ruin.
- stabilise housing, health and essential bills;
- remove predatory/high-cost debt;
- build the first month of runway, then three;
- insure catastrophic exposures;
- increase reliable earning capacity;
- do not speculate with the floor.
STAGE 1 — BUILDER
Income is stable but financial capital is small.
Priority: human capital, savings rate and habit.
- build 6–12+ months of essentials according to risk;
- capture genuine employer/state contribution advantages where available;
- own low-cost diversified productive assets;
- avoid lifestyle inflation after every pay rise;
- take career risks with bounded downside and skill upside;
- keep speculation small enough to become tuition, not trauma.
STAGE 2 — ALLOCATOR
Financial capital can now materially change life outcomes.
Priority: total-balance-sheet construction.
- measure employer, home, country and sector correlation;
- reduce fee and tax drag;
- formalise rebalancing and position caps;
- improve custody, records and legal documents;
- fund known liabilities separately;
- demand an earned edge before concentration.
STAGE 3 — CONCENTRATED CREATOR
A business, carried interest, employer stock or property dominates wealth.
Priority: preserve the engine without letting it own the household.
- move taxes and personal runway outside operating risk;
- remove unnecessary guarantees and cross-collateralisation;
- avoid making the home, investments and lifestyle depend on the same boom;
- pre-plan liquidity, succession and diversification;
- convert enough success into permanent autonomy;
- keep meaningful upside if control and economics remain exceptional.
STAGE 4 — PERMANENT CAPITAL
Capital exceeds reasonable lifetime consumption needs.
Priority: stop playing a game you have already won.
- lower household ruin probability before chasing marginal return;
- define spending, gifts, bequests and institutional purpose;
- diversify custody, managers and decision authority;
- write governance for incapacity, death and conflict;
- train heirs in judgment before transferring control;
- measure success by durable freedom and useful output, not rank.
22. THE 25 NON-NEGOTIABLE LAWS
- Survival comes before optimisation.
- Never confuse expected wealth with the wealth one person typically experiences through time.
- Multiplication punishes volatility and deep loss asymmetrically.
- Ruin is one state across all causes; do not hide it in separate accounts.
- A positive edge does not rescue a lethal position size.
- Treat full Kelly as a theoretical ceiling, not a masculinity test.
- The less measurable the edge, the smaller the bet.
- Count salary, business, house, debt and future spending in the portfolio.
- Match near-term liabilities with safe assets in the right currency and duration.
- Cash is not the growth engine; it is the option not to sell.
- Insure catastrophes; self-insure inconveniences you can afford.
- Avoid any structure whose maximum loss you cannot explain.
- Avoid callable leverage and personally guaranteed speculation.
- Concentration requires control, edge or a protected household.
- Diversification protects access to rare extreme winners.
- Do not cut every winner merely because it became a winner.
- A compounder needs high incremental returns and reinvestment runway.
- Management integrity is an asset; dilution and empire-building are costs.
- Price matters even for the world's best business.
- Optionality means bounded loss, choice, learning and open upside.
- Smooth returns may conceal a short-volatility catastrophe.
- Audit how every adviser gets paid and what happens when they are wrong.
- Fees, taxes, turnover and attention compound negatively.
- Use ranges, haircuts, constraints and slack when inputs are uncertain.
- Convert wealth into a life before time converts you into an estate.
23. THE 30-MINUTE ANTI-RUIN AUDIT
Do this tonight.
Minute 0–5: calculate the floor
- What is one year of truly essential spending?
- How many months are liquid and safe today?
- Which known payments are not included?
Minute 5–10: calculate real concentration
- Largest single position as a percentage of total economic wealth?
- What percentage depends on your employer, industry, country and currency?
- Does your home rise and fall with the same local economy as your income?
Minute 10–15: run the combined crash
- Markets −50%.
- Income gone for twelve months.
- Credit tight.
- One major unexpected bill.
What are you forced to sell, cancel or borrow against?
Minute 15–20: find hidden convexity
- Any margin, short options, personal guarantees or capital calls?
- Any asset whose quoted value is not a real exit price?
- Any “yield” you cannot explain without leverage, illiquidity or insurance selling?
Minute 20–25: total the drag
- Weighted fund and adviser fee?
- Financing and platform cost?
- Tax created by last year's turnover?
- Hours spent on decisions that did not beat the simple alternative?
Minute 25–30: write one rule
Complete this sentence:
Examples:
- one stock / my family's housing;
- my business / money already owed in tax;
- a manager / custody of all assets;
- a bonus year / permanent lifestyle overhead;
- fear of tax / necessary diversification;
- a drawdown / a panicked rule change.
Then move one piece of the system tomorrow.
Not after the next crash.
Not after the next raise.
Not after the concentrated position doubles.
Tomorrow.
THE ENTIRE THREAD IN ONE PARAGRAPH
THE FIRST RULE OF COMPOUNDING IS TO REMAIN COMPOUNDABLE.
Find the one line in your balance sheet that can end the game.
Fix that before chasing another return.
Find the one line in your balance sheet that can end the game.
Fix that before chasing another return.
SOURCES & FURTHER READING
- An Introduction to Ergodicity Economics — official book page
- Ole Peters & Murray Gell-Mann — Evaluating Gambles Using Dynamics
- J. L. Kelly Jr. — A New Interpretation of Information Rate
- The Missing Billionaires — official publisher page
- Nassim Nicholas Taleb — Antifragile
- Nassim Nicholas Taleb — Skin in the Game
- Hendrik Bessembinder — Do Stocks Outperform Treasury Bills?
- Alexis Akira Toda — a critical challenge to the strongest claims of ergodicity economics
All numerical examples are illustrations using stated assumptions. They are not return forecasts. Personal tax, pension, insurance, estate and regulated-investment decisions require jurisdiction-specific advice.
If this changed how you see risk, post the single exposure most capable of removing you from the game. That answer matters more than your favourite ticker.
@KeepCopingLads @AverageCurryEnjoyer @Macan
