Worked examples

What sold at Santa Cruz County's June 2026 tax sale — and what nobody wanted

Last verified 2026-06-11 · primary sources

Santa Cruz Worked examples
On this page

Santa Cruz County sold tax-defaulted property at internet auction on Bid4Assets over June 5–8, 2026 — 27 parcels listed, auction IDs 1281234 through 1281260. The storefront banner said “26 parcels”; the auction index has 27 rows, and we counted from the index. At close: 14 parcels sold, 4 drew no bids at all, and 9 were withdrawn before bidding ended. The headline was a Glenwood Drive parcel in the Santa Cruz Mountains that opened at $9,800 and closed at $115,556 — roughly twelve times its minimum bid.

Be clear about what this post is, because with results posts the framing is the honesty. We did not cover this sale live. We published no pre-sale shortlist, ran no pre-sale filters, and made no calls before the close — nothing below should be read as a flag we “would have” raised. This is a post-hoc reconstruction, assembled after the sale from public records: the auction pages we captured on June 11, 2026, Santa Cruz County’s assessor-parcel GIS layer (the county’s own open-data publication), and the county-posted legal descriptions. The county published no 2026 notice PDF and no results sheet — the parcel list existed only on the auction platform. That is also why this post exists at all: we captured the auction pages ourselves, because there is no official record of these outcomes to point you to.

The sale itself is the standard mechanism. California counties may sell tax-defaulted property at public auction once it has been tax-defaulted for five or more years — three or more for nonresidential commercial property (Cal. Rev. & Tax. Code §3691) — under the auction rules at §§3691–3731. The trap mechanics that show up later in this post are covered in our five traps piece; we won’t restate them here. For scale: this county’s June 2024 sale listed exactly 11 parcels, so 27 is more than double the prior outing — and still a different universe from Riverside’s 946. Bidders posted a $3,000 deposit plus a $35 fee by June 1 to participate.

Every parcel that reached the close

The use descriptions below come from the county’s own records — the GIS assessor-parcel layer and the county-posted legal descriptions — not from listing sites. Outcomes are winning bids at close, captured from the auction pages on June 11, 2026. Every parcel that reached the close is also in a sortable results table you can work and download.

APNAreaWhat it is (county records)Min bidBidsWinning bid at close
040-131-20Aptos (Cathedral Dr/Monte Toyon area)Vacant lot, no situs; zoned R-1-15 but only ~3,833 sq ft (0.088 ac) — substandard size, buildability doubtful, not confirmed with planning$1,80028$6,600
040-131-21AptosVacant lot, no situs; same substandard cluster; 0.090 ac$1,8004$2,400
040-131-54AptosVacant lot, no situs; same cluster; 0.099 ac$4,40012$11,700
057-181-04NW of Davenport (Swanton Rd area)Vacant coastal-zone parcel, no situs; 0.43 ac; assessed land value $43,297$7,50068$42,000
058-053-01Davenport Landing155 Davenport Landing Rd — addressed but vacant; $0 improvements; commercial-agriculture zoning; 0.14 ac$4,90019$29,700
058-053-02Davenport Landing145 Davenport Landing Rd — addressed but vacant, same profile; 0.14 ac$4,90085$40,010
075-122-14LompicoCounty use code “011-UNBUILDABLE LOT”; 0.09 ac under 1-acre-minimum zoning$1,5009$2,600
082-155-20Forest Park, west of Boulder Creek”011-UNBUILDABLE LOT”; county-mapped fault zone; 0.39 ac$2,0006$3,100
082-171-14Huckleberry Woods, Boulder Creek”011-UNBUILDABLE LOT”; fault zone; 0.21 ac$1,0009$2,400
082-171-16Boulder Creek”011-UNBUILDABLE LOT”; fault zone; 0.29 ac$1,00011$2,477
082-171-17Boulder Creek”011-UNBUILDABLE LOT”; fault zone; Lot 4, Block 4 of “Riverdale Park,” recorded 1909; 0.093 ac$1,0009$2,400
087-182-17Boulder Creek383 Hillside Dr — tax roll shows $0 improvements; county GIS shows a ~1,031 sq ft building footprint; unresolved; 0.284 ac$8,900102$52,322
093-302-04Glenwood/Summit, Santa Cruz Mountains23980 Glenwood Dr (active county address point) — roll codes a single residence at about $18K total assessed; GIS shows only a ~200 sq ft outline; structure unverified$9,80081$115,556
106-041-09Eureka Canyon Rd at Alexander Dr, hills above CorralitosRoll says vacant; GIS maps a ~1,250 sq ft building ~93% inside the lines; boundary from an unrecorded survey map; corrected legal description posted mid-auction; 0.487 ac$4,80075$25,300

And the four nobody wanted — each offered at a $1,000 minimum, each closing with zero bids:

APNAreaWhat it is (county records)Min bidBidsWinning bid at close
092-032-12Lompico hills”011-UNBUILDABLE LOT”; “Happyland” tract paper; 0.224 ac$1,0000
092-093-07Lompico hillsSame; 0.108 ac$1,0000
092-164-06Lompico hillsSame; 0.455 ac$1,0000
092-172-08Lompico hillsSame; 0.113 ac$1,0000

One note before anyone over-reads the bid counts: bids measure attention, not value. 102 bids on Hillside Drive tells you bidders fought over it; it does not tell you the winner was right.

The nine that never reached the close

Nine of the 27 were withdrawn before bidding ended. We are reporting these in aggregate only — no APNs, no addresses — on purpose. Withdrawal before close is typically the owner paying the taxes, though the county doesn’t publish per-parcel reasons, and people who most likely just kept their homes don’t belong in a results table.

The pattern is what matters, and it can be stated without identifying anyone: all nine were improved properties. Single-family homes in Santa Cruz, Watsonville, Ben Lomond, Boulder Creek and the Corralitos hills; a coastal lodging property; a home assessed at over $1.1 million.

Now hold that against the table above. Everything withdrawn was improved. Everything that actually went to auction was land, or near-land — the contested-record parcels are the “near.” That is the plainest lesson this sale offers a beginner: the houses get redeemed; the auction is what’s left. If you read a pre-sale list and start mentally pricing the million-dollar home on it, this is the base rate you’re fighting. The split here was absolute — improved, withdrawn; land or contested-record, auctioned — though one sale doesn’t make a law, and we’ll say so again below.

Happyland, recorded 1911, scored 2026

The four no-bid lots deserve their own section, because the market just graded two of our five traps at once.

All four sit in the Lompico hills. The county-posted legal descriptions trace them to “Happyland” Tracts 7, 8 and 10 — promotional subdivisions recorded on October 4, 1911. The county’s own GIS codes every one of them “011-UNBUILDABLE LOT.” No water service and high mapped fire severity, per the county’s own layers. Three of the four sit in a county-mapped fault zone. That is the paper-lot trap in its purest form: lots that exist on a recorded map and nowhere else, 115 years from a promoter’s subdivision to a $1,000 auction nobody bid on.

Stacked on top is the second trap. The deeds conveying these four lots to the defaulting owner were recorded sequentially in 2024 — one batch purchase by a single owner, which went into tax default almost immediately. We don’t name owners; the shape is the point. Somebody bought a bundle of 1911 paper in one go, and within about two years the county was auctioning it. When a cluster of cheap lots on a tax-sale list all changed hands the same way at the same time, you are usually looking at the tail end of someone else’s mistake, or someone else’s exit. The market’s verdict was unambiguous: at $1,000, with the county itself stamping “unbuildable” on the record, zero bids on all four.

One contrast worth holding next to that: 082-171-17, in Boulder Creek, is the same vintage of paper — a 1909 “Riverdale Park” lot, also county-coded unbuildable, also in a fault zone. It sold, $2,400 on 9 bids. Old promotional paper isn’t automatically worthless; it’s automatically suspect. The difference is the subject of the next section.

Signals point; verification decides

If you’ve read our Riverside case study, you know our highest-leverage filter is brutally simple: does the parcel have a street address? In Riverside it did 89% of the cull. Santa Cruz is a useful corrective on three fronts, because here the cheap signals — address, use code, assessed value — pointed the wrong way more than once.

“No address” did not mean worthless. APN 057-181-04 has no situs at all — an unaddressed 0.43-acre coastal-zone parcel off Swanton Road northwest of Davenport. It drew 68 bids and closed at $42,000, essentially its assessed land value of $43,297. An address filter would have binned it with the paper lots; the market treated it as real land that happens to lack a mailbox. The five county-coded unbuildable lots that sold tell the same story at smaller scale, clearing $2,400 to $3,100. Somebody values these — adjoining neighbors, specialists, buyers with a use the county’s use code doesn’t capture. We made exactly this point in five-traps; the market here agreed.

“Has an address” did not mean a house. 155 and 145 Davenport Landing Road are county-confirmed addresses in a coastal hamlet, and both parcels are bare — $0 improvements on the roll, commercial-agriculture zoning, 0.14 acres each. They sold for $29,700 and $40,010, the latter on 85 bids. An address is a proxy for “probably a structure.” Proxies miss.

Two parcels carried contradictory county records — the kind of thing only a per-parcel records check catches. At 383 Hillside Drive in Boulder Creek (087-182-17), the tax roll shows $0 improvements while the county’s GIS shows a roughly 1,031-square-foot building footprint. What stands there today, we don’t know — we didn’t inspect, and we’re saying so. The market bid as if the roll were the wrong half: 102 bids, the most in the sale, $52,322 at close. On Eureka Canyon Road above Corralitos (106-041-09), the roll says vacant while GIS maps a ~1,250-square-foot building with roughly 93% of it inside the parcel’s lines — a boundary that derives from an unrecorded survey map, on a parcel where the county posted a corrected legal description mid-auction. That is a boundary-ambiguity wildcard, and it still closed at $25,300 on 75 bids.

The sale’s headline lives in the same fog. 23980 Glenwood Drive (093-302-04) is roll-coded as a single residence with a decades-old assessed basis — about $18K total assessed — while GIS shows only a small ~200-square-foot outline, and the consumer listing sites we’d normally cross-check were blocked. Structure detail: unverified. The market didn’t wait for the fog to clear: 81 bids, $115,556, roughly twelve times the minimum. The assessed value told you nothing about price. The bid count told you about attention. Neither told you what’s standing on the lot — only records work plus an inspection answers that, and we have only done the first half.

The calibration confession

Here is the part we’d rather not write, which is why it’s here. Our published Riverside filter stack uses a minimum-bid band of $10,000 to $300,000. Every one of the 18 parcels that went to auction in Santa Cruz opened below $10,000 — the highest minimum in the sale was Glenwood’s $9,800. Apply our Riverside-tuned floor to this list naively and the output is an empty page. The filter would have dropped the entire sale, including the parcel that closed at $115,556.

The floor isn’t a bug in Riverside. A 946-row list needs a tractability cut, and on a list that size a sub-$10,000 minimum overwhelmingly marks junk-priced paper lots — the floor exists partly to steer a capital-constrained beginner away from exactly the Happyland-shaped junk above. But the floor encodes an assumption about what “cheap” means on a big list, and Santa Cruz breaks that assumption from the other side: when the improved properties get withdrawn and what reaches auction is land, the whole sale lives under $10,000 — including the one parcel the market took to six figures.

A 27-row list doesn’t need a tractability filter at all. You can read 27 rows in one sitting. The work shifts entirely to per-parcel depth — the GIS cross-checks, the legal descriptions, the contradiction-flagging you just watched. The division of labor doesn’t change, only its center of gravity: deterministic rules reduce, research explains, a human decides. This is what “the filters are per-sale calibrated” means in practice, and we’d rather show you the failure mode on a closed sale than quietly retune the band and pretend it was always right.

What a $1,000 bid actually costs

One pricing note for the small end, because beginners get this wrong. A $1,000 winning bid on this sale did not cost $1,000. Bid4Assets charged a 5% buyer’s premium with a $100 minimum, the county added a $35 per-parcel administrative fee, and California documentary transfer tax applies on top. So the cheapest possible win here was $1,000 + $100 + $35 = $1,135 before transfer tax — about 13.5% over the hammer price. On a $50,000 close the $100 minimum and the $35 fee disappear into the flat 5%; on a $1,000 lot the fixed pieces are a meaningful slice of the deal, worth pricing before the bid rather than after.

What this does not establish

  • These are winning bids at close, not completed transfers. Full payment was due by 4:00 PM Eastern on June 11, 2026, and a winning bidder who fails to settle forfeits the deposit. Final recorded deeds are what confirm a transfer, and we have not pulled deeds.
  • It’s one sale. 27 parcels, one county, one June. The improved-withdrawn, land-auctioned split was absolute here, but this post proves the pattern for exactly one sale, nothing more.
  • Signals are not verdicts. We pulled no title on any of these parcels, inspected nothing in person, and put no buildability question to county planning. The record contradictions above are flagged, not resolved. Nothing here says any winning bidder overpaid or got a bargain — “the market agreed” is a statement about prices, not about wisdom.
  • There is no official tally to reconcile against. The county published no results. Our counts come from our own capture of the auction pages, and the per-parcel numbers are all shown so you can check them against anything that surfaces later.

If you want this kind of read before a sale instead of after one, that’s the work we publish — free. For each upcoming California sale we do the records work you just watched done post-hoc — county GIS against the roll, legal descriptions, contradiction flags, plain-English reasoning per parcel — and publish it in full before bidding opens. Judge the work first: the free sample brief is our complete Riverside research, nothing held back. If you want an email when the next county list drops, the newsletter does that and nothing else.


The Santa Cruz County sale closed June 8, 2026, and is presented here as post-hoc market reporting and a methodology case study, not a recommendation — we made no pre-sale calls on this sale and claim none. Parcel descriptions come from Santa Cruz County’s published records as captured June 11, 2026, and may contain county-side errors we cannot rule out. Nothing here is legal, tax, or investment advice. We surface signals; you make the decision. We are not licensed real estate brokers, attorneys, or financial advisors in any jurisdiction, and the single statutory point above — confirmed against the California Revenue & Taxation Code as published by the state legislature — states a general rule whose application to any specific parcel is a question for a licensed professional.

Want the next California sale like this one — the full read, before bidding opens?

Or see the free sample brief · browse the rest of the research.

Worked examples

We read all 1,614 parcels in San Bernardino's July 2026 tax sale. Most of them will lose you money.

San Bernardino County is auctioning 1,614 tax-defaulted parcels July 11–17, 2026 — mostly raw desert and mountain land, plus 364 timeshares. We cross-checked every non-timeshare parcel against the county assessor/GIS: about 72% open at or above the county's own assessed value, and the genuine discounts are a minority of 148. A free, pre-sale desk analysis — and a worked example of how we read a sale.

San Bernardino Worked examples