How a deal becomes a pick.

Every week we pull every multi-unit and house-hack-viable listing in the Bay Area ZIPs we cover. Coverage isn't accidental. Each ZIP has to clear three criteria before it's eligible:

Fail any of the three, the ZIP doesn't get scanned. That's where most of the "great Bay Area deals" you see on Instagram die.

Every listing runs through the same financial model. The number we care about is S&P breakeven: the S&P 500 return required, per year over 10 years, for the same down payment in an index fund to come out ahead of the deal.

Inputs are sourced, not invented:

• Rents come from documented neighborhood comps for the same unit type. Not Zillow rent estimates, which we've found wildly off in multi-unit Bay Area inventory.
• Vacancy, CapEx reserves, property-management drag, all at realistic levels. Not optimistic.
• Property taxes use the actual jurisdiction's effective rate, not a national average.
• Cautious appreciation pulled from each ZIP's pricing framework, with named catalysts adjusted in or out.

Every input is something you could pull yourself in an afternoon. We pull it for ~14,000 listings a week. That's the whole trick.

One rent number is a guess. So we model each deal three times:

Each scenario produces its own S&P breakeven and 10-year wealth margin in dollars. Not percentages. Percentages on $1M+ assets are how people lose track of real money, and the difference between a 4% spread and a 7% spread is six figures of net worth at the end of the horizon. We want that visible as real money.

The benchmark is the same starting capital going into an S&P 500 index fund for the same horizon, compounded at the same investment-return rate. We report the difference in dollars at the end of the horizon, not in CAGR or ROI.

On the property side, the model credits rental-portion depreciation (27.5-year straight-line on the non-land basis of rented units) and rental-portion mortgage interest at your marginal tax rate. The slider defaults to 0% — opt-in, conservative. Set it to your bracket to enable the deduction.

What we don't model on either side, we name. The calculator's disclosure block lists every tax event we don't compute — Schedule E income tax on positive net rental, §469 phaseout above $150k MAGI, §1250 recapture at exit, §121 owner-occupied exclusion, state income tax, S&P long-term capital gains on realization — with magnitude or rule-of-thumb where we can quantify it. The realistic after-tax benefit captured on the property side is typically 40–70% of the headline "Tax Savings" line.

Most house-hack calculators ignore the index alternative entirely. This one names what it models and what it doesn't.

Modeling alone isn't enough. Every issue passes two review gates.

Picks are ranked by conservative S&P breakeven ascending, ties broken on conservative wealth margin descending.

If the cautious case loses to the S&P, it's not ranked. Some weeks the list is short. Some weeks it's: don't buy. We'd rather lose a subscriber to a missing pick than keep one we've talked into a $1.5M mistake.

Every issue ships with that week's screening funnel right in the body of the email:

You see the kill rate up front. Most weeks under 5% of active listings clear the cautious-case S&P benchmark. That's the entire reason this newsletter exists.