Risk · how safe the book is today
On a normal bad day the book might lose about 1.8% — roughly $3,341 — it would take an unusually rough day to lose more than that. If a crash like 2008 financial crisis hit, it would drop about 28.2%.
Normal bad day
1.8%
In dollars
$3,341
A 1-in-100 day
3.0%
If it's worse than that
$4,789
10,000 simulations, fat-tailed · cov 120d daily · eod 2026-07-31 · src massive
The book is within its risk limits — no single name dominates. Nothing to do today.
The book →What lenders charge risky companies to borrow — High-yield option-adjusted spread: the extra yield junk-rated borrowers pay over Treasuries. When lenders get nervous, this widens — usually 2 to 4 weeks before stocks feel it. tend to move before stocks do.
HY spread
284bp
Engine word
Complacent
Past month
+10bp
De-risk line
66bp away
Spreads at 284bp mean lenders are charging little for default risk — how late cycles look, not how safe ones feel. The engine watches the 350bp line: crossing it has led equity weakness by 2 to 4 weeks, and that line is 66bp away.
ICE BofA HY OAS · FRED BAMLH0A0HYM2 · as of 2026-07-30 · fetched by the nightly refresh
How much of the book’s swing is just the Risk that comes from the whole market rising and falling — it hits everything at once and can't be diversified away. versus risk Risk unique to the individual companies held. Unlike market risk, spreading across more names reduces it. — and which positions carry the most of it.
Yearly swing (vol)
18.3%
From the market
38.3%
Name-specific
61.7%
The book run through the actual moves of past crises — approximate analogues, not predictions. Worst first.
S&P 500 −35.7%
S&P 500 −27.0%
S&P 500 −15.1%
S&P 500 −6.3%
| Scenario | Book | In dollars | What drives it |
|---|
Trading days to sell each position without pushing its price around, at 20.0% of its recent daily volume. The whole book takes about 0 days; the slowest, ETOR, takes 0.
| Name | Shares | Daily volume | Days to exit |
|---|---|---|---|
| ETOR | 347 | 1,030,000 | 0 |
| RDDT | 76 | 5,340,000 | 0 |
| NTES | 126 | 690,000 | 0 |
| HALO | 333 | 3,650,000 | 0 |
| WM |
The loss figure up top comes from three models. The fat-tailed one is the honest one: A model whose extreme outcomes are more likely than a normal bell curve predicts — because real market crashes happen more often than 'normal' math expects. than a bell curve, and that’s where the money is actually lost.
Method and limits: the covariance is the stored correlation and swings over the lookback, which is regime-dependent and breaks in a true tail. The simulation is seeded for reproducibility; the fat-tailed version adds realism but is still a model. Stress moves are approximate historical analogues run through first-order (marginal) betas. Liquidity uses recent daily volume as a proxy. Educational, not investment advice.
| Name | Weight | Daily swing | Share of risk |
|---|---|---|---|
| KGC | 9.7% | 3.4% | 16.1% |
| RDDT | 5.7% | 4.7% | 14.7% |
| HALO | 14.7% | 2.1% | 13.6% |
| SOUN | 5.1% | 4.5% | 12.6% |
| PYPL | 9.8% | 2.6% | 12.0% |
| ETOR | 6.6% | 3.3% | 10.1% |
| PLMR | 10.4% | 2.4% | 7.4% |
| NTES | 9.0% | 2.0% | 7.2% |
| TSM | 6.9% | 2.9% | 6.9% |
| MCD | 8.0% | 1.3% | 1.9% |
| Factor | How strongly the book moves when this factor moves. 1.0 means one-for-one; 0.5 means half as much. | Standalone swing (yr) |
|---|---|---|
| S&P 500 | 0.79 | 11.3% |
| Gold | 0.31 | 8.7% |
| Oil | -0.11 | 6.7% |
| US Dollar | -1.22 | 7.0% |
Standalone swings overlap and are not additive.
cov 120d daily · eod 2026-07-31 · src massive
| 2008 financial crisisSep–Nov 2008: equities collapse, oil craters, gold holds. (Rate-driven flight embedded in the equity move.) | −28.2% | −$52,762 | S&P 500 −35.7% |
| 2020 COVID crashFeb–Mar 2020: fastest bear market on record, oil to negative, a brief gold liquidation. | −21.4% | −$40,115 | S&P 500 −27.0% |
| 2022 rate shockFull year 2022: the Fed hikes hard, equities derate, energy outperforms. | −17.0% | −$31,840 | S&P 500 −15.1% |
| Oil / geopolitical shockAn oil spike on a Middle East escalation: crude jumps, gold bid, equities wobble. | −6.9% | −$12,923 | S&P 500 −6.3% |
marginal betas · cov 120d daily · eod 2026-07-31 · src massive
| 58 |
| 3,620,000 |
| 0 |
| PLMR | 145.98 | 468,000 | 0 |
| XLE | 225 | 39,700,000 | 0 |
| TSM | 32 | 25,800,000 | 0 |
| KGC | 782 | 8,080,000 | 0 |
| MCD | 55 | 8,110,000 | 0 |
| PYPL | 322 | 27,200,000 | 0 |
| SOUN | 1,560 | 23,400,000 | 0 |
daily volume as the liquidity proxy · eod 2026-07-31
| Model | The most the book is likely to lose on a normal bad day — 19 days out of 20 it stays better than this. | VaR 99 | If the loss does blow past the VaR line, this is the average size of that worse loss. | VaR 95 USD |
|---|---|---|---|---|
| Normal curve (parametric) | 1.9% | 2.7% | $4,447 | $3,548 |
| Simulation (normal) | 1.9% | 2.7% | $4,446 | $3,554 |
| Fat-tailed simulation ( A model whose tails are fatter than a bell curve, so it expects big crashes more often — closer to how markets really behave., df 5) | 1.8% | 3.0% | $4,789 | $3,341 |
10,000 sims · cov 120d daily · eod 2026-07-31 · src massive
