Current balances
update monthly
Investing waterfall — every dollar knows its next job
Fill each step before feeding the next. Hard rules: no stock-picking, no crypto, no margin. Plan of record, not financial advice.
- 1 · Starter emergency fund — $1,000
One car repair or laptop death without touching the grad fund or a credit card.
- 2 · Tax reserve — skim 27% of every 1099 dollar on arrival
Handshake pays gross. Self-employment + income tax is not optional; the reserve makes April boring.
- 3 · Grad fund — CASH instruments only (HYSA / 26-week T-bill ladder / I-bond decision)
2-3 year horizon: equities can sit 30% down exactly when Part III tuition is due. This bucket exists to be THERE, not to grow.
- 4 · Roth IRA — up to min(earned income, annual limit), ONE broad index fund
Student-bracket Roth dollars are the cheapest you will ever buy; contributions stay retrievable in a true emergency.
- 5 · Taxable brokerage — only after 1-4 are full
No rush to reach this step as a sophomore. Money arriving here means the plan is working.
Projection desk — $0 waterfall mix, to grad school
Scenario B — one alternative, same engine
Fill any Scenario-B field — blank fields inherit your current assumptions.
New contributions only — current balances live in the hero above, not in these curves. Includes the two REU stipends (summers '27, '28) and a senior-internship lump. The 80% equity line exists to show WHY it is excluded before grad school. Deterministic illustration, not advice.
The account stack — 0/12 done
Monthly review — 0/5
Key numbers (2026)
| Roth IRA contribution limit (2026) | $7,500 |
| Roth phase-out (single, MAGI) | $153,000–$168,000 |
| Top HYSA APY | ~4.00–4.50% |
| Self-employment tax rate | ~15.3% |
| Suggested gig tax reserve | ~27% |
Income streams — annual estimate $0
Savings trajectory — running total $0
Should I buy this? — the reality check
Name a purchase and a price — you get the percent of monthly savings it eats, months-to-save, the opportunity cost, and how far it pushes your target.
Look-through allocation — what your hours actually buy
The Metron look-through table, transposed: the portfolio is 733 training hours; here is its true exposure.
| Phase | Weeks | Gates | Hours | Share | What it buys |
|---|---|---|---|---|---|
| Boot | 2 | 15 | 52 | 7% | proof language + the analysis/NT foundations everything compounds on |
| Toolkit | 11 | 193 | 308 | 42% | the eleven technique families — the A1/A2/B1/B2 conversion engine |
| Mastery | 15 | 252 | 373 | 51% | timed conversion, spiral retention, mock calibration — the score itself |
SaaS bench — scored, parked, one in validation
| # | Idea | Moat · Effort · Market | Line |
|---|---|---|---|
| 1 | REU + grad-deadline tracker · 90-day validation | 4 · 4 · 3 = 11 | The digest productized. Mostly already built; every STEM undergrad has this problem every winter. |
| 2 | Quant-interview trainer | 3 · 3 · 4 = 10 | Timed Green-Book-style drills + SRS. Deep-pocketed users with proven willingness to pay. |
| 3 | KeYmaera X spec-visualizer | 5 · 2 · 2 = 9 | URAP moat nobody else has; small research/defense market, hard tech. |
| 4 | Putnam prep platform | 4 · 2 · 2 = 8 | This app productized. Living-lab moat, tiny seasonal niche. |
| 5 | UWB immersive-audio tooling | 4 · 2 · 2 = 8 | Ramsay-lab moat; hardware-adjacent effort, niche venue market. |