Stage 2
51 questions
Why can't a high return make up for a total loss?
Because returns act multiplicatively. A factor of zero makes the whole product zero.
What's an absorbing state?
A state with no way back, here meaning total loss.
Can a strategy with positive expected value still lead to certain ruin?
Yes, if it reaches the absorbing state with sufficient probability over time.
What exactly does the one-percent rule limit?
The maximum loss of a position, not the capital deployed.
Why is only a fraction of the Kelly stake used?
Because win rate and win-loss ratio are estimated and usually overestimated.
Why isn't the rule enough applied per position alone?
Because correlated positions move together and their risk nearly adds up.
What's the basic formula for share count?
The allowed loss in currency divided by the distance between entry and stop per share.
What happens if you set the stop tight just to make the position bigger?
The position is highly likely to get stopped out by normal fluctuation.
What is risk measured against with leveraged products?
The notional value of the position moved, not the collateral posted.
What win rate does a three-to-one ratio need to break even?
25 percent, since 1 divided by 1 plus 3 gives 0.25.
Why doesn't a higher price target automatically improve a strategy?
Because more distant targets get hit less often. The win rate falls to match.
Why does the realized ratio fall below the planned one?
Because gains often get taken early, and losses can run larger through slippage and gaps.
How is a trade's expected value calculated?
Win rate times average gain, minus loss rate times average loss, each after costs.
What happens with negative expected value if you trade more often?
The loss arrives faster. Frequency amplifies the sign.
Why aren't thirty trades a reliable basis?
Because the standard error of the estimate only falls with the square root of the number of observations.
Why does the needed recovery grow disproportionately?
Because it's applied to the shrunken base. The expression 1/(1−d) − 1 grows toward infinity.
How should a historical maximum drawdown be read?
As a floor on what's possible in the future, not a ceiling.
Which risk can be diversified away, and which can't?
Holding-specific risk vanishes as the count rises; shared market risk remains.
Why does ten stocks in the same industry provide little diversification?
Because their covariance sits close to their individual variance. They move together.
What is correlation breakdown?
During stress periods, correlations between risk assets rise, which is exactly when diversification works worst.
At what price move is a 25x-leveraged stake wiped out?
Four percent, since 100 divided by 25 is 4.
Why does growth rate fall past a certain leverage?
Because variance grows with the square of leverage, while expected value only grows linearly.
What happens to daily leveraged products in sideways markets?
They systematically lose value through path dependence.
Why does the actual liquidation threshold sit closer to entry than 1 divided by leverage?
Because financing costs, fees, and the maintenance margin get subtracted as well.
What does margin-call liability mean?
That claims can arise beyond the amount you originally staked.
How does a liquidation cascade form?
Forced selling pushes the price down, which pushes further positions to their own threshold and gets them closed too.
What do you own when you buy a CFD?
Nothing. You have a contract to pay a price difference with the provider.
Where does the loss-rate figure in the advertising come from?
From the provider itself. It's a mandatory regulatory disclosure about its own retail customers.
What conflict of interest can exist?
If the provider itself holds the opposite side, customer losses become its revenue.
Why do good resolutions fail during a crisis?
Because under uncertainty, time pressure, and arousal, the fast, affect-driven system dominates.
What's the return gap for fund investors?
Investors earn less than their own fund, because they buy and sell procyclically.
Why do automated savings plans work?
Because they remove the option to act, instead of requiring you to master the emotion.
Why does FOMO structurally get you in late?
Because a move has to be big first before it draws any attention at all.
Why is your impression from social media skewed?
Gains get shared, losses don't. The perceived success rate sits above the real one.
What's a simple countermeasure?
A fixed waiting period of about a week between wanting to buy and actually buying.
What does Prospect Theory say about losses?
They get evaluated relative to a reference point and weigh roughly twice as heavily as equally sized gains.
Why does someone become more risk-seeking after a loss?
Because the value function is convex in the loss domain. Getting back to the reference point gets overweighted.
What question exposes a held losing position?
Whether you'd buy it today, at this price, from scratch.
On which levels does confirmation bias operate?
Choice of sources, weighting of evidence, and memory.
What is a falsification criterion?
An event defined in advance that would tell you your thesis was wrong.
What is a premortem analysis?
Assuming the decision has already failed, and gathering the most plausible reasons for it in advance.
Why is the revenge trade the opposite of growth-optimal behavior?
Because the optimal stake is proportional to capital and should fall after a loss, not rise.
How strongly do trading costs act at high frequency?
Linearly with frequency. They can eat up a large share of a typical expected risk premium.
What rule protects on bad days?
A fixed daily loss limit or a maximum number of trades, regardless of how you feel in the moment.
Why does the thesis need to be recorded before the outcome?
Because of hindsight bias. Once the outcome is known, the remembered reasoning adapts to fit.
What metric matters more than win rate?
Expected value, together with the share of rule-compliant trades.
Why aren't past losses an argument for continuing?
Because sunk costs are irrelevant to a future decision. Only expected future return matters.
Why isn't a losing streak alone proof of a bad strategy?
Because it can also arise from normal variance around a positive expectation. Rule compliance also needs to be considered.
When does this stop being a math question?
When trading starts interfering with sleep, relationships, or work. Then talking to someone and getting professional support is what's called for.