Skip to content

A fixed synthetic scenario · 195 kernel-checked theorems

A 6.49% mortgage against the market. 195 theorems, one dial.

One fixed synthetic household: $500K at 6.49%, $5K of slack each month, a 24% marginal tax rate. Prospect theory prices the gut. Kelly prices the edge. Tax theorems set the break-even at 5.803%. The scenario is fictional, and nothing here is financial advice.

195
kernel-checked theorems
23
Lean 4 source files
5.803%
break-even pre-tax return
2.25
loss-aversion ratio λ
?6.49% vs μf* = μ_excess / σ²file : line

Act I · Prospect theory

The gut is a parameter, not a flaw.

Kahneman and Tversky's 1992 cumulative parameter fit puts loss aversion at λ = 2.25 with curvature α = β = 0.88. The corpus proves the algebra behind the feeling: for any positive x, |v(−x)| / v(x) = λ exactly. A $1,000 loss weighs 2.25 times the same gain. The theorem does not tell you to feel less. It tells you how much you feel.

Act II · The guaranteed side

Paying early earns 4.9324%. Guaranteed.

At a 24% marginal rate, itemizing drops the 6.49% nominal mortgage to 6.49% · (1 − 0.24) = 4.9324% after tax. Every extra principal dollar saves that rate with zero variance. The corpus proves the after-tax rate is weakly below nominal, and that the scenario's first-year interest clears the standard deduction, so the itemizing is real, not assumed.

Act III · The growth side

Kelly prices the edge at 156%. Take half.

The Kelly fraction is f* = μ_excess / σ². At the scenario's 10% expected return and 15% volatility, the excess over the 6.49% mortgage is 3.51%, so f* ≈ 1.56: full Kelly says borrow and hold 156% in equities. The corpus proves f* beats every other fraction on log-growth, and it proves the discipline too: half-Kelly keeps 3/4 of the growth at 1/4 of the variance, and betting past 2f* is strictly worse than optimal.

Act IV · The synthesis

Moderate aversion splits the $5K. High aversion pays the house.

For risk aversion γ between 2 and 4, the recommended split's realized 6.8% dominates the 4.9324% after-tax mortgage rate; at γ = 3 the Merton fraction puts 52.86% of the flow in equities, about $2,643 of the $5K. The corpus keeps the other answer honest too: for loss aversion λ ≥ 2 the market's certainty equivalent falls below the guaranteed mortgage return, and the key-insight theorem records a Sharpe under 0.25 with over 1% volatility drag. The defensible default leans mortgage. The split is the compromise.

Proof receipts

Each receipt is a kernel-checked Lean 4 theorem with its file and line. The theorems prove the algebra of the scenario; they do not validate the market inputs, and they are not financial advice.

THEOREM

Becky.Finance.KellyCriterion.kelly_maximizes_log_growth

Substituting f* = μ_excess / variance makes the log-growth derivative zero.

Becky/Finance/KellyCriterion.lean:59
THEOREM

Becky.Finance.KellyCriterion.kelly_growth_dominates

Log-growth at f* strictly exceeds log-growth at every other fraction.

Becky/Finance/KellyCriterion.lean:68
THEOREM

Becky.Psychology.ProspectTheory.loss_gain_ratio

For x > 0 and α = β, |v(−x)| / v(x) = λ exactly.

Becky/Psychology/ProspectTheory.lean:75
THEOREM

Becky.Psychology.ProspectTheory.loss_aversion

For x > 0 and α = β, v(x) < |v(−x)|. Losses outweigh equal gains.

Becky/Psychology/ProspectTheory.lean:84
THEOREM

Becky.Finance.TaxEfficiency.after_tax_mortgage_rate_le

Itemizing weakly lowers the mortgage rate: r · (1 − τ) ≤ r.

Becky/Finance/TaxEfficiency.lean:41
THEOREM

Becky.Application.BeckyScenario.becky_breakeven_above_mortgage

The break-even pre-tax market return, 5.803%, sits strictly above the 4.9324% after-tax mortgage rate.

Becky/Application/BeckyScenario.lean:90
THEOREM

Becky.Application.BeckyRecommendation.becky_recommendation_dominates_pure_mortgage_for_moderate_aversion

For γ ∈ [2, 4] the recommended split's realized return, 6.8%, dominates the after-tax mortgage rate.

Becky/Application/BeckyRecommendation.lean:103

Formal statements

theorem
Becky.Finance.KellyCriterion.half_kelly_three_quarter_growth
logGrowthRate (f*/2) μ σ² = 3 / 4 * logGrowthRate f* μ σ²
Becky/Finance/KellyCriterion.lean:83
theorem
Becky.Finance.KellyCriterion.half_kelly_quarter_variance
(f* / 2)² * variance = f*² * variance / 4
Becky/Finance/KellyCriterion.lean:90
theorem
Becky.Finance.KellyCriterion.becky_kelly_over_one
becky_kelly > 1
Becky/Finance/KellyCriterion.lean:124
theorem
Becky.Quant.MertonPortfolio.becky_merton_at_gamma_3
|beckyMertonFraction 3 − 0.5286| < 0.001
Becky/Quant/MertonPortfolio.lean:105
theorem
Becky.Application.BeckyRecommendation.becky_ce_above_mortgage_with_discipline
certaintyEquivalent beckyStrategy.realizedReturn becky_effective_sigma 1 ≥ becky_r
Becky/Application/BeckyRecommendation.lean:119
theorem
Becky.Application.BeckyScenario.becky_key_insight
annual_rate > 0.06 ∧ becky_sharpe < 0.25 ∧ becky_volatility_drag > 0.01
Becky/Application/BeckyScenario.lean:316

The receipts are open.

The archive is free to read. The explorer carries every theorem with its file and line. The scenario on this page is fixed and synthetic; nothing here is financial advice.