What we build
A data stack that turns transaction evidence into live pricing.
Each capability is a production component — not a spreadsheet, not a slide — wired to official data, monitored continuously, and exportable to the format IC actually reads.
Comparable-transactions models
Radius-, similarity-, and date-weighted comps pulled from URA caveats, JPPH, BPN, MLS IDX, and X-Value style private indices — with every weighting exposed so the valuation survives an IC challenge.
Unit-velocity + absorption tracking
Per-project sell-through rates by launch phase, price tier, and unit type. Base, bear, and bull absorption curves projected against demand signals and rate regime.
Supply pipeline by micro-market
Planned, under-construction, and upcoming launches mapped to the postcode or zone. Developer-level exposure, tenure mix, and delivery-date risk surfaced on one view.
Rental yield + capital appreciation
Gross and net yield modelling by asset class, with capital-appreciation paths informed by historical cycles. Ties into DCF and CAP-rate sensitivity for every asset on the book.
Tenant-demand + search signals
DDproperty, PropertyGuru, 99.co, and SRX search volumes plus rental-enquiry velocity — early-warning for shifts in demand before they hit transaction data.
Scenario modelling · counterfactuals
What if GFA +10%, tenure flips to leasehold, or CAP rate widens 40 bps? Every change re-prices the asset and writes an IC-level assumption log against the run.
Who runs on the analytics stack
One engine, every pricing decision.
Same feature store, comps regressor, absorption model, and scenario engine — wrapped differently for each operator class. Developers price launches, REITs underwrite acquisitions, consultancies advise clients, wealth teams place capital. The data stays consistent; the delivery surface changes.
Developer launch pricing
Price a new launch against live comps, floor-premium curves, and tenure-adjusted benchmarks. Pricing workshops run on numbers, not gut feel from the last cycle.
REIT investment arm underwriting
Asset-level DCF, CAP-rate sensitivity, and market-absorption scrutiny for listed REITs underwriting acquisitions, divestments, or development pipelines.
Property consultancies · CBRE, Knight Frank style
Research-house analytics wrapped for advisory teams — radius-weighted comps, micro-market briefs, and institutional-grade pipeline views delivered through a client portal.
Land-banking + site feasibility
Pre-acquisition studies for land-banking outfits — zoning scenarios, GFA uplift economics, absorption-risk modelling, and hold-period yield paths per scenario.
Private-client wealth teams
Investment memos for single-family offices and private-banker wealth teams placing direct real-estate capital. Readable, defensible, and tied to live market data.
Developer-competitive intelligence
See how competing projects are pricing, clearing, and launching — project-level velocity, unit-mix positioning, and launch-phase timing mapped across rivals.
Models inside the stack
Purpose-built models, not a generic regressor.
Each model family handles a distinct job — pricing, absorption, supply, scenario — with latency and accuracy tuned to the decision it supports. Nothing one-size-fits-all.
Gradient-boosted regressor that weights comparable transactions by radius, recency, tenure, and feature-similarity. Calibrated per micro-market rather than on a single global model.
Sequence model that projects unit-velocity curves from launch-phase history, rate expectations, and saved-search signals. Base, bear, and bull paths emitted per project.
LLM-assisted extractor that pulls planned, under-construction, and upcoming launches from URA, BSA, and OSS filings into a queryable graph indexed by micro-market and developer.
Counterfactual agent that re-prices an asset under GFA, tenure, mix, and CAP-rate changes — writing an assumption log per run so IC sees exactly what moved the NPV.
Sources feeding the stack
Every licensed source, every cadence — on the ledger.
Pulled on the cadence the source publishes, normalised into one schema, tagged with period and attribution. What the analyst sees is what an auditor or regulator could reconstruct.
Defensible, not just directional
Every psf number comes with the math behind it.
A single median is easy to produce and easy to attack. Every comparable row in our stack carries its own weighting, radius, date adjustment, and amenity adjustment — so the valuation can be defended line-by-line to an IC, an investor, or a financing banker. When the number gets questioned, you can show exactly which comps moved it and by how much.
- Per-comp weighting surfaced on every derivation
- Radius + similarity + date adjustments transparent
- Amenity and floor-premium curves exposed
- Methodology hash logged per IC-grade run
Frameworks we align to
Licensed sources, defensible methodology.
Why Axccelerate for market analytics
Not a quarterly research PDF.
A live analytics stack.
A PDF report goes stale the day it ships. Our stack refreshes on the cadence each source publishes, ties comparables to the methodology that produced them, and exports what IC, banker, or investor actually reads.
Pricing
Priced to coverage, not to transaction volume.
Analytics deployments are custom — we scope against your markets, asset classes, and delivery surface before quoting.
Glossary
The vocabulary behind every price.
A quick reference for the terms that show up in property analytics — the vocabulary your analysts, IC, and valuers will all use.
- Caveat
- Caveated transaction record
A filed notice of a completed property transaction. In Singapore, URA publishes caveats lodged against land titles — the primary evidence base for comparable-transaction valuation.
- CMA
- Comparative Market Analysis
A valuation built from weighted comparables rather than a full appraisal. Common for pre-launch pricing, broker opinions, and IC-stage underwriting memos.
- psf / psm
- Per square foot · per square metre
The dominant unit for comparing property prices. SG and HK quote psf; most of continental Asia and Europe quote psm. Convert with 1m² = 10.764 ft².
- Absorption rate
- Units sold per period
The rate at which available units in a launch are cleared — usually measured at 3, 6, and 12 months. Core input to developer revenue and cash-flow planning.
- Launch pipeline
- Planned + UC + upcoming
The forward inventory of units being brought to market. Planned, under-construction (UC), and upcoming-launch phases carry different risk and timing profiles.
- Unit velocity
- Rate of unit take-up
Units sold per week or per month across a launch, normalised to launch size. Drives phase-two pricing and the decision to hold inventory vs accelerate.
- Freehold / leasehold
- Tenure classification
Freehold conveys perpetual title; leasehold conveys a finite-term right. In SG, 99-year leasehold is standard; 999-year and freehold command tenure premiums in valuation.
- Gross vs net yield
- Rental return measures
Gross yield divides annual rent by price. Net yield deducts property tax, management, repairs, and vacancy — the number that actually drives investor underwriting.
- Capital appreciation
- Price growth over time
The component of total return that comes from price gains rather than rental income. Modelled against historical cycles, rate regime, and supply-pipeline pressure.
- CAP rate
- Capitalisation rate
NOI divided by property value — the market's implied yield on an income-producing asset. Inverse of the price multiple; every 25 bps move reprices the asset materially.
- DCF
- Discounted Cash Flow
Valuation built by projecting future cash flows and discounting them to present value. The canonical tool for development-feasibility and multi-year hold analysis.
- Comparable selection
- Choosing the comp set
The rules that pick which transactions qualify — same tenure, similar unit size, adjacent floors, recent date, bounded radius. Bad selection is the biggest source of valuation error.
- Radius weighting
- Proximity-based weight
Weighting comparables by distance from the subject — closer transactions count more. Combined with similarity weighting so adjacency alone does not override mismatch.
- Date adjustment
- Time-series correction
Adjusting comparable prices forward (or backward) to the valuation date using a price index — URA Residential PPI in SG, NAPIC in MY. Mandatory for comps more than a quarter old.
Your pricing chain, engineered.
30-minute scoping with a senior engineer and a real-estate analytics operator. You'll leave with a coverage plan, an integration sketch, and a realistic timeline — not a sales pitch.