Builder Cancellation Rate: Which Denominator?

A builder cancellation rate divides canceled home contracts by a disclosed denominator, such as gross orders generated during the period. Cancellations can include contracts signed earlier, so the percentage does not automatically measure the share of that quarter's new buyers who will fail to close. This distinction follows from issuer definitions of orders and backlog.

Which denominator does the builder use?

Start with the earnings release's label and the operating table's footnotes. These disclosures, available by September 29, 2026, illustrate the gross-order approach:

Meritage explicitly defines its numerator as units canceled during the period. These two disclosures use the same denominator concept, but their quarters cover different months. The percentages alone cannot rank buyer reliability.

An older filing shows how much the denominator can matter. In its October 25, 2006 release, Meritage reported third-quarter cancellations equivalent to 37% of gross orders or 19% of beginning backlog. Both percentages described the period's cancellations. This is a historical reporting example, not a current benchmark or evidence that Meritage uses a backlog denominator today.

Gross orders count contracts generated during an interval. Beginning backlog counts homes already under contract but unclosed at the interval's start. The two bases produce ratios that answer different questions.

Reconcile gross orders, cancellations and net orders

Great Southern Homes' historical homebuilding disclosure supplies the convention for the example below: net new orders equal gross sales minus customer cancellations received during the same period. United Homes Group filed the disclosure as Exhibit 99.3 on April 5, 2023.

Net orders = gross orders − cancellations during the period

Consider a hypothetical builder. All figures below count home contracts or homes, not dollars. Assume consistent reporting scope, no acquisitions or other backlog adjustments, and the following activity during one quarter:

The same 20 cancellations produce 20% of gross orders: 20 ÷ 100. Dividing them by beginning backlog instead produces 6.7%, rounded: 20 ÷ 300. Neither calculation changes the underlying activity.

The backlog reconciliation is:

Ending backlog = beginning backlog + net orders − closings

Net orders already subtract cancellations. Subtracting the 20 again when calculating ending backlog would count them twice. This volume reconciliation also helps explain how backlog relates to future deliveries.

Reconstructing counts from a reported rate

If the builder reports net orders and an exact gross-order cancellation rate below 100%, you can rearrange the equations:

• Gross orders = net orders ÷ (1 − cancellation rate)
• Cancellations = gross orders − net orders

Use the rate as a decimal. In the example, 80 ÷ (1 − 0.20) yields 100 gross orders and therefore 20 cancellations. These equations require a gross-order denominator and the same net-order convention used above. A rounded published percentage produces only estimated counts; do not present them as exact issuer disclosures.

Why the rate is not a buyer failure probability

In the hypothetical quarter, 14 of the 20 cancellations came from the 300 contracts already in backlog. Those buyers did not belong to the group that signed this quarter.

Six of the 100 new contracts canceled by quarter-end. For that signing group—a contract cohort—the observed cancellation share so far is 6%, even though the period's headline rate is 20%. The other 94 contracts have not canceled by that date; some may have closed, so they are not necessarily all still in backlog.

The 6% is not the cohort's eventual cancellation rate. Contracts still open at quarter-end can cancel later. Measuring eventual outcomes requires following the same group through closing or cancellation. A comparison over a shorter window needs a consistent observation period.

A backlog denominator requires the same care. Total cancellations divided by beginning backlog were 6.7%, but cancellations drawn from that opening group were 14 ÷ 300 = 4.7%, rounded. Before interpreting the ratio as that cohort's cancellation share, restrict the numerator to cancellations from that cohort.

Why the rate can rise without more cancellations

Keep the hypothetical cancellation count at 20, but reduce gross orders from 100 to 50. The gross-order cancellation rate rises from 20% to 40%, while net orders fall from 80 to 30.

Fewer new contracts are replacing canceled ones, but the count of canceled contracts has not increased. Attributing the rate increase to more buyers backing out would misdescribe the change.

Likewise, with 20 cancellations, increasing beginning backlog from 300 to 600 reduces the total-cancellations-to-backlog ratio from 6.7% to 3.3%, rounded. The lower rate comes entirely from the larger denominator.

When comparisons between builders are defensible

When comparing homebuilder operating results, keep the definition beside each percentage and check four things:

• Reporting window: Match the actual start and end dates, not just labels such as “third quarter.”
• Measurement: Use the same cancellation numerator, denominator and unit of measurement. Keep contract counts separate from order values.
• Scope: Check whether both figures cover comparable businesses and whether acquisitions or reporting changes affect the series.
• Contract exposure: Consider the mix of homes sold before construction versus near completion, and the time contracts remain open. A shared formula does not establish comparable cohorts.

Where disclosed counts permit it, calculate a common ratio and label it as your calculation. Where they do not, preserve the issuer's definition and explain the limit.

For land or credit analysis, request gross orders, cancellation counts, opening backlog and the signing dates of canceled contracts. Those inputs help distinguish weaker new demand from losses among earlier buyers—something a standalone percentage cannot establish.