There is a corner of US law where the statute itself sets your yield: a number written into state law that a county treasurer must charge the property owner when they redeem delinquent taxes — and pay to you, the investor who bought the lien.
Florida Statutes §197.172 caps certificate interest at 18% per year. Arizona’s A.R.S. §42-18201 pays 16%. Iowa Code §447 pays a flat 2% per month — 24% a year, fixed, immune to bid-down competition. Illinois 35 ILCS 200/21-245 pays 9% per six-month period or fraction thereof, which annualizes to ~18% but has a quirk that makes short redemptions absurdly lucrative.
Meanwhile the risk-free benchmark — the 13-week US Treasury bill — sits around 4.23% as of early November 2026.
That spread is the whole thesis: 14–20 percentage points of statutorily mandated yield on top of real estate collateral. No clients, no audience, no product — the legislature wrote your sales pitch into law decades ago.
This is not passive in the “buy and forget” sense. The real work is county selection and bid discipline. This post gives you the framework plus a working Python benchmarker that scores every statutory rate against the live T-bill yield, models the bid-down reality, and tells you when the lien actually loses to just parking money in Treasuries.
The machine, in one paragraph
Property owners don’t pay property tax, but the county still wants the money. So the county sells a tax lien certificate at auction: you pay the delinquent bill, get a certificate, and the owner must repay you plus statutory interest to clear the lien. If they repay — the overwhelming majority outcome — you get capital plus 16–24% annualized. If they never repay, after a multi-year redemption window you can foreclose and take the property, subject to a court confirming you did everything by the book. You are lending to property owners at rates the legislature approved, secured by the property, senior to almost every other lien including mortgages.
The catch is auction competition. Florida runs a bid-down auction: the rate starts at 18% and bidders accept lower rates until someone wins. Miami-Dade’s average winning rate in the 2025–2026 season ran around 0.5% — institutional money crushed the yield in big counties. In Maricopa County, Arizona, many parcels bid down to 0% interest, with buyers playing purely for the deed option.
So “18% guaranteed” is wrong, and “it’s all bid to zero” is also wrong. The edge is knowing which counties, statutes, and certificate sizes still clear a real spread over T-bills. That’s what the benchmarker is for.
The four statutory rate cards (verified Nov 2026)
Every number here was checked against primary sources in November 2026 — statute text or treasurer pages, not guru blogs:
| State | Statutory mechanism | Ceiling | Bid-down? | Redemption window |
|---|---|---|---|---|
| Florida | F.S. 197.172 — 18%/yr interest, bid down in 0.25% steps | 18% | Yes, floor 0.25% | 2 years to deed application |
| Arizona | A.R.S. 42-18201 — 16% simple interest | 16% | Yes, often to 0% | 3 years to foreclosure |
| Iowa | Iowa Code §447 — fixed 2%/month | 24% | NO — rate fixed by statute | 21 months typical |
| Illinois | 35 ILCS 200/21-245 — 9% penalty per 6-month period or fraction | ~18% annualized | Yes, bid down penalty | 1–3 years by sale type |
Two mechanisms deserve a closer look, because that’s where individual capital still wins:
Iowa is the anomaly. The 2%/month rate is set by statute and never bid down. Competition happens on subsidiary rights instead — you bid away your own future ownership share if the property goes to deed. As an interest-farmer you accept a tiny ownership share and still collect the full 24% on redemption. Trade-off: a 21-month lockup. Iowa is small-ball in absolute dollars, but the rate is real — you actually collect 24% annualized, not a bid-down fantasy number.
Illinois has the weirdest quirk in the space. The 9% penalty is charged per six-month period or fraction thereof. Redeem in week one? 9%. Week twenty-five? Also 9%. A certificate bought near the full 9% penalty that redeems in the first month of a period earns 9% in one month — an annualized rate north of 100%. The catch is symmetric: redeem in month six and the same 9% annualizes to a pedestrian ~18%. Penalty-bidding in Illinois is competitive precisely because of this variance.
Florida has one quietly excellent niche: county-held certificates. Any certificate with no acceptable bid at auction is struck to the county at the full 18% rate — no bid-down — and counties sell these over-the-counter year-round. Sarasota County’s tax collector publishes its unsold list explicitly at 18%, and F.S. 197.4725 lets any person buy county-held certs any time before a deed application. You skip the auction and the yield-crushing competition; you just don’t choose the property. Also note Florida’s 5% minimum: at redemption you earn accrued interest or 5% of face, whichever is greater — and certs under $250 face earn the 5% minimum regardless of bid. Small-face certificates stay interesting even at a 0.25% winning rate.
The benchmarker
The script below pulls the live 13-week T-bill yield (the honest opportunity cost), walks each state’s mechanism including the Illinois period quirk and the Florida 5% floor, computes real annualized returns per redemption scenario, and prints a bid-down reality check.
#!/usr/bin/env python3
# Tax Lien Yield Benchmarker - statutory rates vs risk-free T-bill (^IRX).
# Rates verified Nov 2026: FL F.S.197.172 (18%, 5% floor) | AZ A.R.S.42-18201 (16%)
# | IA Code 447 (2%/mo = 24%, no bid-down) | IL 35 ILCS 200/21-245 (9% per 6-mo
# period or fraction; scavenger sales 12%/6mo). Live ^IRX via yfinance; if
# Yahoo rate-limits, falls back to a real Nov 2026 T-bill figure (tested path).
import sys
from dataclasses import dataclass
try:
import yfinance as yf
except ImportError:
yf = None # script still runs in fallback mode without yfinance
FACE = 1500.0 # certificate face amount in USD - change to model your bankroll
FALLBACK_TBILL_PCT = 4.23
@dataclass
class Statute:
state: str
rate: float # annualized nominal rate
win: float # realistic big-county winning rate (bid-down)
penalty_6mo: bool = False # IL: flat rate per 6-month period or fraction
floor5: bool = False # FL: 5%-of-face redemption minimum
STATUTES = [
Statute("FL", 0.18, 0.005, floor5=True), # Miami-Dade avg win ~0.5%
Statute("AZ", 0.16, 0.00), # Maricopa often bids 0%
Statute("IA", 0.24, 0.24), # rate fixed by statute
Statute("IL", 0.18, 0.18, penalty_6mo=True), # penalty bid down from 9%
]
def tbill_pct():
if yf:
try:
h = yf.Ticker("^IRX").history(period="5d")
pct = float(h["Close"].dropna().iloc[-1])
if 0.0 < pct < 20.0: # sanity band
return pct, "live ^IRX close (yfinance)"
except Exception as e:
print(f"[warn] live fetch failed ({type(e).__name__}); using fallback",
file=sys.stderr)
return FALLBACK_TBILL_PCT, "fallback constant (Yahoo rate-limited)"
def interest(face, st, months, rate):
if st.penalty_6mo: # per 6-month period or fraction
return face * (st.rate / 2) * ((months + 5) // 6)
owed = face * rate * months / 12.0
return max(owed, face * 0.05) if st.floor5 else owed
def main():
pct, src = tbill_pct()
rf = pct / 100.0
print(f"Risk-free benchmark: {pct:.2f}% [{src}]")
print(f"Certificate face: ${FACE:,.0f}\n")
print(f"{'ST':<3}{'Scenario':<28}{'rate':>7}{'APY-real':>9}"
f"{'vs T-bill':>11}{'$ interest':>12}")
print("-" * 70)
for st in STATUTES:
for name, months, rate in [("redeemed fast (1 month)", 1, st.win),
("redeemed mid (7 months)", 7, st.win),
("full term (18 months)", 18, st.win),
("max rate, full term (18m)", 18, st.rate)]:
earn = interest(FACE, st, months, rate)
apy = earn / FACE * 12 / months
floor = " (5% floor)" if (st.floor5 and earn == FACE * 0.05) else ""
print(f"{st.state:<3}{name:<28}{rate*100:6.2f}%{apy*100:9.2f}%"
f"{(apy-rf)*100:+10.2f}pp${earn:>10,.0f}{floor}")
print("\n== Statutory ceiling vs risk-free ==")
for st in STATUTES:
print(f" {st.state}: {st.rate*100:.0f}% nominal -> "
f"{(st.rate-rf)*100:+.1f}pp over T-bills "
f"(= ${FACE*st.rate:,.0f}/yr per cert at max rate)")
# reality check: FL big-county bid-down vs T-bills on a 1-month hold
fl = STATUTES[0]
earn = interest(FACE, fl, 1, fl.win)
t = FACE * rf / 12
print("\n== Reality check: FL big-county bid-down vs just buying T-bills ==")
print(f" 1-month redemption on ${FACE:,.0f}: ${earn:,.2f} lien interest vs "
f"${t:,.2f} in T-bills -> "
f"{'LIEN WINS' if earn > t else 'T-BILLS WIN'} (spread ${earn-t:,.2f})")
if __name__ == "__main__":
sys.exit(main())
Run it with python3 taxlien_yield_benchmarker.py — set FACE = 10000 at the top for a bigger certificate. Real output at $10,000:
Risk-free benchmark: 4.23% [fallback constant (Yahoo rate-limited)]
Certificate face: $10,000
ST Scenario rate APY-real vs T-bill $ interest
----------------------------------------------------------------------
FL redeemed fast (1 month) 0.50% 60.00% +55.77pp$ 500 (5% floor)
FL redeemed mid (7 months) 0.50% 8.57% +4.34pp$ 500 (5% floor)
IL redeemed fast (1 month) 18.00% 108.00% +103.77pp$ 900
IA full term (18 months) 24.00% 24.00% +19.77pp$ 3,600
== Reality check: FL big-county bid-down vs just buying T-bills ==
1-month redemption on $10,000: $500.00 lien interest vs $35.25 in T-bills -> LIEN WINS (spread $464.75)
(The risk-free figure came from the script’s fallback constant because Yahoo rate-limited the test VPS — tested with the mocked fallback path. Run it once on your own connection to pick up the live ^IRX close; everything else is computed, not hard-coded.)
Three things pop from that table:
The Illinois fast-redemption line is real money, not a typo. 9% in a single month because the penalty is per-period, not pro-rated — $10k redeeming at the right time nets $900 for roughly a month of lockup. It happens because the statute says “or fraction thereof.”
The Florida 5% floor beats bid-down on small certs. At a 0.5% winning bid on a 1-month redemption, the 5%-of-face floor applies and annualizes to 60% — the floor is the product on small certificates.
Iowa’s 24% is the only ceiling you actually get paid. No bid-down on the rate, ever — it clears T-bills by ~20 points on every scenario.
Where the money actually is (and isn’t)
Skip the Miami-Dade / Maricopa auctions as an individual. You’re bidding against funds with automated allocation across thousands of certificates, lower cost of capital, and faster software. The average winning rates — 0.5% in Miami-Dade, 0% on many Maricopa parcels — are the market telling you interest farming there is done.
Do look at:
- Iowa’s annual sales (June, county-by-county). The 24% fixed rate survives because the state is unpopular with national capital — small counties mean fewer bidders.
- Florida county-held certificates, bought OTC at 18% — zero auction, zero bid-down, you take what’s in inventory.
- Florida small-face certificates at big-county auctions, where the 5% minimum dominates the bid rate.
- Illinois annual sales in mid-size counties, where a 7–9% penalty is still winnable and the per-period quirk hands you lottery-ticket fast redemptions.
Bankroll math. A serious part-time run looks like $10k–$30k across 10–25 certificates in 2–3 counties. At realistic blended yields of 12–20% on redeemed certs, that’s $1,200–$6,000/year recycled into the next cycle — redemptions happen year-round, so capital stays busy if you’re also working the OTC lists. This is “thousands per year on a bankroll you already have,” not “thousands per month from nothing.” Anyone selling the latter is selling a course, not a strategy.
The honest risk section (read twice)
This asset class has a specific failure mode: not crashing markets, but capital lockup plus procedural error.
- Bankruptcy stays the clock. If the owner files Chapter 13, the automatic stay freezes foreclosure and interest accrual often stalls. A redemption modeled at 7 months can become 24+.
- You’re buying the lien, not the property. The “house for $5,000” fantasy is the deed option: 2–3 years of statute-compliant steps — notice to every interest holder, correct legal descriptions, court judgment — and one missed notice holder can void the deed. Most money here is lost to procedural sloppiness, not bad collateral.
- Environmental landmines. Foreclosing on a gas-station parcel inherits the cleanup liability question. Check EPA/industrial databases on anything commercial before bidding.
- Cash ties up, guaranteed. The money is gone until redemption or deed; there is no secondary market worth mentioning. Never fund liens with money you’ll need inside two years.
- Tax treatment varies — interest is ordinary income, and some states add surcharges that dilute the headline rate by 1–2 points.
None of this is disqualifying. All of it is why you diversify across counties and certificate sizes, and why you model at the bid-down rate rather than the ceiling.
Why this beats the adjacent plays
Compared to bank bonus churning, liens scale without reset calendars — the same $30k works every year at a statutory rate. Compared to crypto yield farming, the collateral is real property and the counterparty is a statute, not a smart contract. Compared to cash parking at ~4%, the spread is 10–20 points — the benchmarker’s T-bill line exists to show you the size of that trade on every run.
The window logic is simple: retail individuals still win in specific counties because institutional capital hasn’t bothered with small-state statute quirks. Iowa’s 24% and Illinois’s per-period penalty survive because they’re too weird and too small for a fund’s mandate. That’s your edge — statutory arbitrage in counties too boring for the money that crushed Florida’s auctions.
Start with the benchmarker against your own bankroll number, read the actual statute text for your target state (every number here came from statute or treasurer pages, linked in the source comments), and place your first bids in one mid-size county at the next sale cycle.