Six years on 3.5 acres beside an interchange nobody funded, sold at a loss

I do small multifamily. In 2019 I bought 3.5 acres of scrub about a mile from where a new highway interchange was going to go, $28,000, cash, because a broker showed me the county's long range transportation plan with the interchange drawn on it and I thought I was being clever.

Here's what I got wrong at the very first step. The interchange was in the long range plan, which in that county meant a 25 year horizon document listing projects with no money attached. There was a separate short range program, four years, with actual funding sources named. The interchange was never in that one. Not delayed, never in it. Nobody lied to me. I read the wrong document and never asked which document I was reading.

Six years of carry on a parcel that produced nothing:

Taxes started at $310 and finished at $890, because the property never qualified for the agricultural classification, it was scrub with no agricultural use, and I didn't understand that classification was a thing you apply for based on use. That's about $3,600 total.

A weed abatement citation in 2022, $200, plus $350 to have it cleared.

A road improvement special assessment in 2023 when the county paved and widened the county road my frontage was on. $1,400, spread over three years. Ironically the only infrastructure that ever arrived, and I paid for it.

Sold in July for $26,500 with a 6% commission, so $24,910 net. Against $28,000 in and about $5,550 of carry, I'm down roughly $8,600 over six years, and the $28,000 didn't earn anything for six years while my duplexes did.

What I'd do differently, plainly. I'd ask for the short range, funded program and read that instead of the long range wish list, and I'd treat a project with no identified funding source as not existing. I'd check whether the parcel can qualify for a reduced tax classification before buying and price the full undiscounted tax bill if it can't, and those classification rules differ by state so I'd ask the assessor rather than the broker. And I'd write out twelve years of carry rather than five, because five was a number I picked because it sounded like a long time.

Answer

2sonia_alvarez Get yoursPRO6

The two-document thing is the most useful part of this post and I'd put it plainer for anyone reading. Transportation planning at the regional level generally has a long range plan covering twenty or more years, which is a list of projects the region would like, and a short range program covering roughly four years, which is the list with money identified. Names and lengths vary by state and by the regional planning organization, so ask what each document is called locally. The test is whether a funding source and a fiscal year are named next to the project. If they aren't, the project is a drawing.

The special assessment is the other thing worth pulling out. When a county or a district builds an improvement that benefits specific parcels, it can often charge those parcels directly, and raw land with frontage gets charged the same as improved land does. Whether that's allowed, how it's calculated, and how you can protest it all depend on state law and the local ordinance. You had no income and a bill arrived anyway. That's the carrying cost people leave out of their model, because taxes are predictable and assessments are not.

I'd disagree with you on one point. You describe the cause as reading the wrong document, and that was the trigger. The reason a $28,000 mistake became an $8,600 loss is that the carry ran at a rate you hadn't priced and you had no classification relief. If the interchange had been real but ten years late, you'd have been in the same squeeze with a better story. Price the carry first, then form an opinion about growth.

3 replies

1Rena Expert

Adding the mechanism behind @tender's last point, because it generalizes.

Land banking has no internal source of funds. Every dollar of property tax, nuisance compliance, special assessment, and insurance if carried comes from outside the asset for the entire hold. That means the hold length and the carry rate multiply together into the real cost basis, and both are uncertain. @cairn's parcel carried at roughly $925 a year against a $28,000 basis, about 3.3% a year, so six years added close to 20% to what the land had to appreciate before breaking even. At a $310 tax bill it would have been closer to 7% over the same period. The classification question is worth as much as the growth question over a long hold.

On the growth side, the underlying thesis for land banking rests on development and population actually reaching the parcel. National housing shortage and movement toward less dense areas support the direction in aggregate. Neither supports any particular parcel, and the timing and direction of growth at a specific location remain genuinely uncertain. A funded capital program is evidence. An unfunded long range plan is not evidence of timing at all.

Two things to confirm in writing before buying raw land for a long hold, both of which vary by state and often by county: the current property tax classification of the parcel and what use is required to obtain or keep any reduced classification, from the assessor, and whether any improvement district, road district, or assessment district covers or is proposed to cover the parcel, from the county. Then write the carry out to a hold length longer than you expect, because the constraint that ends most land banks is the owner running out of patience or cash before growth arrives.

2ehernandez74PO2BPRO5

I nearly did the same thing in 2021, different county, same shape. Broker had a printed map with a proposed arterial extension on it and I asked what stage it was at, and he said it's in the plan. I asked which plan and he didn't know, which is what saved me, because it made me go find out. It was in a corridor study, which as far as I can tell means somebody was paid to draw options.

The part of your post I'll actually use is the twelve years instead of five. I keep writing five year models because five feels patient and it isn't. Sorry about the eight and a half thousand.