The Optionality Gap: The Options Trader and the Committed Builder

Ireland's largest residential developer holds permissions for over 280 units across four sites in Meath. They are building in Fingal and Kildare instead. Of the ten schemes this developer abandoned across Ireland, eight had An Bord Pleanala involvement. Of the twelve they built, three did.

The same type of one-grant developer commences at 35% in Dún Laoghaire-Rathdown and at 87% in Westmeath. Same scheme sizes. Same regulatory burden. The 50-point spread is what the friction environment costs a developer who has no choice about where to build. Developers who do have a choice — those holding multiple permissions across multiple councils — make that choice visibly in the data. That choice is what this article maps.

Two hundred and sixteen developers hold two or more granted residential permissions for ten or more units (2018-2022, standard permission applications). Between them they hold 581 schemes covering 30,252 units. They commenced 362 and abandoned 219. The profile of what they chose and what they left behind reads as a revealed viability map for Irish residential development.

What do developers choose?

The sites portfolio developers activated and the sites they abandoned differ on one variable more than any other: whether the scheme had an An Bord Pleanala case linked to it.

19% of chosen sites had ABP involvement. 36% of abandoned sites did. Outside Dublin, where building control linkage is more complete (commencement notices match planning records at roughly 55-65% versus 23% in Dublin), the gap widens: 16% chosen, 33% abandoned.

ABP cases come from two directions. Some are applicant appeals against a council refusal. Many more are third-party objections against a council grant — neighbours, environmental groups, competing landowners. The data does not always distinguish, and a council grant overturned by a third-party appeal carries no fault on the council's part. The friction shows up either way: an ABP case adds time, cost, and outcome uncertainty regardless of who triggered it.

ABP involvement is also not a clean variable. ABP sites are more urban on average, with roughly 60% more amenities within a kilometre at the national level (77 vs 49). The appeal process correlates with the broader complexity of urban development. The data cannot fully separate the two. But within every scheme size band we tested outside Dublin, the ABP-linked commencement rate is lower: roughly 30 points lower for small schemes (10-30 units), 15 points for mid-sized (31-60), and 24 points for larger schemes (60+).

What the developers did not use as a selection criterion matters too. RZLT exposure is nearly identical on chosen and abandoned sites (36% vs 41%). Infrastructure metrics (sewered connections, failing wastewater plants, defective blocks risk) wash out completely. Flood zone exposure is directionally higher on abandoned sites (3.3% vs 1.2%) but the numbers are small.

Developers choose the path of least friction, and friction is not about regulation. RZLT, NZEB, Part V, infrastructure are the same on both sides of the choice. The variable that differs is the planning process itself.

What do they walk away from?

The counterintuitive finding: abandoned sites are in better areas than chosen ones.

Census deprivation scores are lower for abandoned sites (-0.69) than for chosen sites (-0.33). More negative means less deprived. The abandoned sites are closer to transport (568m vs 636m), have more amenities (69 vs 50 within a kilometre), and sit in areas with lower unemployment (5.5% vs 5.8%) and lower vacancy (6.9% vs 7.1%).

Developers with options are walking away from the areas with the strongest demand because the friction of building there is highest. They build instead where the process is smoother, even if the demand signal is slightly weaker.

Geographically, this plays out at the council level. Cork County: 73% of portfolio developer permissions activated (n=78). Wicklow: 78% (n=18). Louth: 71% (n=42). At the other end: Meath 33% (n=9), Tipperary 38% (n=13). Sample sizes are small for the low-choice councils, but the direction is consistent.

Meath is a commuter belt county with median new-build prices of €349,000. Cork County: €339,000. The prices are comparable. The choice rates are not. What differs is the planning environment, and Meath is where the consequences of that difference fall on developers without alternatives.

Who is left behind?

When portfolio developers walk away from Meath, the remaining developers are 83% one-grant operators. They hold one permission, on one site, with no fallback. If the numbers don't work, they cannot redirect to a different council.

There are 1,325 of them in the dataset. They hold one permission each, covering 56,817 units. Their aggregate commencement rate is 56%, against 62% for portfolio developers. They face higher ABP rates (35% vs 25%) and longer processing times (159 vs 142 days). They sit in more urban, more complex sites, not by choice but because that is where their single permission happens to be.

ABP friction reaches them too, but it lands differently. Outside Dublin, where building control linkage is more complete, one-grant developers commence at 61% without ABP and 48% with it — a 13-point drop. Portfolio developers commence at 69% without ABP and 46% with it — a 24-point drop.

When a portfolio developer faces an ABP case they walk away, and their commencement drops by 24 points. When a one-grant developer faces an ABP case they push through more often, and their drop is 13 points. Under ABP friction, one-grant developers commence at a slightly higher rate (48%) than portfolio developers (46%).

They push through because they have no alternative.

The friction map varies by council

The same type of developer that builds at 35% in Dun Laoghaire-Rathdown (where 60% of schemes face ABP cases) builds at 75% in Galway County (28% ABP) and at 87% in Westmeath (22% ABP, n=23).

This is not about developer quality. It is about the friction environment.

The pattern is not universal. In Kildare, ABP-linked schemes by one-grant developers commence at a higher rate (73%) than non-ABP schemes (54%, n=22 and n=56 respectively). In some commuter belt counties, the schemes that attract appeals may be the well-located mid-sized developments that get built regardless.

South Dublin shows the same picture. One-grant ABP schemes there commence at 74% versus 71% for non-ABP, essentially no penalty (n=33 across the two cohorts). The council also processes applications faster than any other in the country (32 days on average). ABP is the strongest single correlate we can measure in the aggregate, but it works differently in different places. In some councils an appeal is the friction that pushes a marginal scheme below viability. In others it is noise around schemes that get built anyway.

The cross-council pattern is the central finding of this article. A 50-point spread between top and bottom councils, for the same type of developer, building the same size scheme. The system treats one-grant developers very differently depending on where their single site happens to be.

The friction that doesn't matter

Every article in this series has tested whether the Residential Zoned Land Tax correlates with worse outcomes. In every analysis, it does not.

The same pattern holds at the developer-cohort level. RZLT-liable sites held by one-grant developers outside Dublin commence at a higher rate than non-RZLT sites: 66% vs 60% without ABP involvement, 54% vs 45% with ABP. The likely explanation is the underlying zoning. Land that is RZLT-liable is zoned residential, which signals development-ready infrastructure and council support. The tax is a cost (as Part 3 documented). The zoning is a viability signal.

The RZLT is a cost on land that the planning system has already identified as suitable for development. Friction lives in the planning process itself, not in the tax.

What this means before you submit

If you are a one-grant developer, the data says your commencement probability ranges from 35% to 87% depending on council and ABP exposure. The system-wide 56% average is the wrong number to plan against; your council's number and your ABP status are the ones that matter.

If you are advising on site selection, the band with the highest commencement rate for one-grant developers is 61-to-100 units (70% outside Dublin). Below 60 units, the rate settles around 55%. Above 100 units, 38% commence nationally and 24% outside Dublin. ABP involvement correlates with a commencement penalty in every size band, though the magnitude varies by council and year.

If you are a portfolio developer, the data says your revealed choices map to where friction is lowest. The places multi-grant developers consistently abandon (Meath, Tipperary, Westmeath) are where one-grant developers hold 80-95% of remaining permissions, face higher ABP rates, and build without fallback.

Methodology

Population. 1,906 residential permission grants of 10 or more units, received between January 2018 and December 2022, across all 31 local authorities. This excludes extensions of duration, compliance submissions, retention, and other ancillary application types to ensure like-for-like comparison of genuine development permissions. Developer classification is based on applicant name: developers with two or more grants in the population are classified as "multi-grant" (216 developers, 581 schemes), and developers with exactly one grant are "one-grant" (1,325 developers). Developer counts can shift by a few percentage points when variant spellings of the same applicant are reconciled.

SPV caveat. Portfolio developers routinely use Special Purpose Vehicles (separate legal entities per site), which means some developers classified as "one-grant" may be subsidiaries of larger operators. Approximately 14% of one-grant developers share a planning agent with multi-grant developers, suggesting possible SPV contamination. The true one-grant population may be 10-15% smaller than reported.

Commencement. Classified via linked National Building Control Office records. Building control data matching is incomplete, particularly in Dublin where the linkage baseline is approximately 23% versus 54-66% elsewhere (varies by year). All headline comparisons use outside-Dublin figures except where stated. Dublin council figures (DCC, DLR) are lower bounds on true commencement.

ABP involvement. A boolean flag on the planning register indicating the application has a linked An Bord Pleanala case. This includes third-party appeals against council grants, applicant appeals against refusals, and direct ABP applications (SHD, strategic infrastructure). For Dublin councils, the appeal type is not classified in the data, meaning we cannot distinguish third-party opposition from developer appeals. The "friction" interpretation assumes third-party opposition is the dominant pathway, but this is unverified for the largest urban councils.

ABP penalty stability. The aggregate commencement penalty associated with ABP involvement is approximately 11-12 percentage points (outside Dublin, one-grant developers), but this varies substantially by year: roughly 25pp in 2018, under 10pp in 2019 (SHD era), near zero in 2020 (Covid disruption), approximately 9pp in 2021, and 19-22pp in 2022. Two of five cohort years show near-zero or reversed penalties. All ABP claims should be read as "on average across the period" rather than as a stable annual effect.

Council-level claims. Sample sizes vary widely. Councils cited individually in the body text: Cork County (n=78 multi-grant, n=117 one-grant), DCC (n=42 multi-grant, n=124 one-grant), Fingal (n=48, n=66), Wicklow (n=18, n=58), DLR (n=7, n=57), Meath (n=9, n=44), Westmeath (n=9, n=23), Carlow (n=4, n=13). Councils with fewer than 20 schemes are flagged as indicative in the text. The cross-council pattern across all 29 testable councils is more robust than any individual comparison.

The Kildare exception. In Kildare (n=78 one-grant), ABP-linked schemes commence at 73% (n=22) compared with 54% for non-ABP schemes (n=56). This directly contradicts the aggregate ABP friction finding and is addressed in the body text. The reversal may reflect the nature of ABP cases in commuter-belt counties: well-located schemes that attract appeals but are commercially viable regardless.

RZLT. RZLT liability defaults to false for sites outside the RZLT map. The finding that RZLT-liable sites have higher commencement rates likely reflects the underlying zoning (zoned residential = infrastructure-ready, council-supported) rather than a benefit of the tax.

Figures deliberately excluded. The 100+ unit scheme activation rate (21%, n=28) is noted as indicative only. Council-level choice rates for councils with fewer than 10 multi-grant schemes (Meath n=9, Tipperary n=13) are flagged with sample sizes in the body text. The multi-grant Westmeath commencement rate (33%, n=9) is not used as a headline figure.

Sources