FDA CDER Q3 2026: twelve novel approvals across the three-month window.
In June Tempora asked a single month to deliver six novel drug approvals. CDER delivered two or three depending on the counting convention, and the reconciliation recorded two defects: the threshold sat near the twentieth percentile of monthly outcomes, and the counting convention had been left loose enough to argue about. This entry runs the same reading across a quarter, with the convention fixed before the window and the run rate quoted.
Falsifier, stated verbatim as the test table carries it: FDA CDER novel drug approvals Q3 2026 (Jul-Aug-Sep combined) at or above 12 NMEs fires MET.
Base rate: CDER approved 46 novel drugs in 2025, 50 in 2024, 55 in 2023, 37 in 2022 and 50 in 2021, which is an average quarter of 11.5, 12.5, 13.75, 9.25 and 12.5 respectively. A 12-per-quarter bar therefore sits close to the middle of the recent five-year distribution and three of those five years averaged at or above it. Base rate taken as 0.50. The 2026 run rate sits just under the line: 22 novel approvals across the first half, about 11 per quarter. Stated probability: 0.52, two points of tilt and no more. Calibration tier: lab, no lift claimed. Reconciliation by 15 October 2026.
What counts as the event, fixed in advance
The qualifying observable is the count of entries on the FDA page titled Novel Drug Approvals for 2026 with an approval date from 1 July 2026 through 30 September 2026 inclusive. That list is CDER's own and it carries new molecular entities together with new therapeutic biologics, which is the convention this test uses. Approvals handled by CBER, including vaccines and most cell and gene therapies, do not count. Supplemental indications, new formulations of already-approved molecules, generics and biosimilars do not count.
Stating the convention before the window is itself the fix for one of June's two defects. That call was gradeable at two or three depending on whether a cell therapy and a biologic counted alongside the strict new molecular entity, and a test that can be graded two ways is a test that has not been written. The line here is the CDER list, as published.
The quarter opens with four on the board through 24 July 2026: gedatolisib (Revtorpyk, 14 July), enlicitide (Lipfendra, 16 July), zidesamtinib (Jideytro, 22 July) and centanafadine (Simtriyo, 24 July). That is a four-per-month pace, which leaves eight needed across August and September. Reference for verification: the FDA Novel Drug Approvals for 2026 page and the CDER new molecular entity and new biologic approvals compilation.
The base rate and the 2026 run rate
CDER's recent annual totals are 46 novel drugs in 2025, 50 in 2024, 55 in 2023, 37 in 2022 and 50 in 2021. Divided evenly those are quarters of 11.5, 12.5, 13.75, 9.25 and 12.5. A 12-per-quarter threshold therefore sits inside the middle of that distribution, with three of the five years averaging at or above it and two below. Taking the base rate at 0.50 is the honest reading of that spread.
The 2026 run rate is slightly under the line. The first half of 2026 produced 22 novel approvals, which is 11 per quarter and 3.67 per month. July's four through the 24th sits a little above that pace. To clear 12 the quarter needs eight more across August and September, which is four per month against a 2026 run rate of 3.67 and a multi-year trend near four. The test is close to the trend line, which is where a rolling-window test belongs, and it is the reason the stated probability of 0.52 carries two points of tilt instead of ten.
Limitation, stated with the signal: approval timing is set by user-fee goal dates and those cluster. A quarter can carry six in one month and two in the next without anything having changed in the pipeline, and Q4 has historically been the heavier end of the year. A quarterly window averages some of that lumpiness and does not remove it.
Method note, and the lesson applied
Lesson applied: a three-month rolling window instead of a single month (M2 in Tempora's miss register). In June Tempora published FDAJUN6NME, which required six or more CDER novel approvals inside June 2026. June delivered two or three depending on the counting convention. The reconciliation recorded the arithmetic: CDER averages close to four novel approvals a month on the multi-year trend, so a six-in-one-month print is around a twentieth-percentile event, and the threshold had been set to that percentile with no seasonal adjustment when June is historically softer than the Q4 cluster.
The fix is window length. A quarter is long enough that user-fee clustering averages out to something near the trend, so the threshold can be set at the trend and the test measures whether the quarter ran above or below it. That is a question with a real answer, and a single soft month no longer settles it.
Limitation, stated with the fix: a wider window also dilutes any chart-side signal that is genuinely dated. If the reading being tested is about a specific week, a quarterly count will not see it. This entry accepts that trade because the June instance showed the monthly version was not gradeable against a sensible threshold, and the tracker records the tier accordingly.
Chart-side reading, and what it cannot do
The period state on the USA 1776 chart is engine-verified: MD Venus with AD Sun through 10 August 2026, then AD Moon from 10 August 2026 to 9 April 2028. Venus is the classical karaka of pharmaceutical and medical substances, and the antardasha boundary falls inside the measured quarter, six weeks before it closes. A correction belongs here: the June instance of this signature cited AD Rahu for the same chart. The engine returns AD Sun and then AD Moon across this window, so the June article's period claim was wrong and the state above is the corrected one. That is recorded here rather than quietly fixed.
What the reading does and does not support: Tempora's calibrated table carries no regulatory-approval-count signature at quarterly resolution, so no lift is claimed and the chart-side layer is not carrying this call. The prior is CDER's run rate. The chart-side content is the antardasha boundary date, entered as a signature under test, and a MET on a 12-approval quarter that arrives evenly across the three months would give that boundary no support at all. The reconciliation will report the monthly split so the distinction stays visible.
One further constraint, declared: the transit fact sheet for this publication batch was unavailable at drafting, so this entry claims no transit contact dates and quotes only dasha states the engine reproduces on demand.
What would make this fail
A thin goal-date calendar. August and September carry fewer user-fee goal dates than the quarter needs and the count finishes at nine to eleven. Given the 2026 run rate of 11 per quarter, this is the central failure branch.
Agency capacity or a funding interruption. Staffing losses, a shutdown or a review-resource squeeze push goal dates into Q4. The approvals still happen and they land outside the window, which grades FAILED without saying anything about the pipeline.
Complete response letters at the margin. Two or three refusals in a quarter is ordinary variance and it is also the whole margin of this call.
Late posting. An approval dated inside September can appear on the FDA list after 30 September. The count uses the list as it stands at reconciliation on 15 October 2026 and any later additions are recorded as a caveat without changing the grade.
How this grades
Reconciliation publishes by 15 October 2026 on the public Lab surface. The verdict quotes the full list of qualifying approvals with brand name, active moiety and approval date, the monthly split across July, August and September, the quarter total against the 12 line, the 2026 year-to-date count, and a note on whether the second-half-of-quarter timing pattern held or the count arrived evenly.
Frequently asked
What is the FDA CDER Q3 2026 call?
FDA CDER novel drug approvals for July, August and September 2026 combined total at or above 12. Twelve or more fires MET. Eleven or fewer fires FAILED. Stated probability 0.52 against a base rate of 0.50 drawn from CDER's 2021 to 2025 annual totals.
What counts toward the twelve?
Entries on the FDA page titled Novel Drug Approvals for 2026 with an approval date from 1 July through 30 September 2026 inclusive, which covers new molecular entities and new therapeutic biologics on that list. CBER approvals including vaccines and most cell and gene therapies do not count, and neither do supplements, reformulations, generics or biosimilars.
What is the base rate?
CDER approved 46 novel drugs in 2025, 50 in 2024, 55 in 2023, 37 in 2022 and 50 in 2021, which is an average quarter of 11.5 to 13.75 in three of those years and below 12 in two. That puts a 12-per-quarter bar at the middle of the distribution, so the base rate is taken as 0.50. The 2026 run rate is 22 novel approvals in the first half, about 11 per quarter.
What lesson from an earlier miss does this apply?
A three-month rolling window instead of a single month. The June call FDAJUN6NME required six or more novel approvals inside June 2026 and CDER delivered two or three depending on the counting convention. CDER averages close to four a month, so six in one month is around a twentieth-percentile event. This entry sets the threshold at the trend across a quarter and fixes the counting convention before the window opens.
When does Tempora reconcile?
By 15 October 2026, with the full list of qualifying approvals, the monthly split across the quarter and the quarter total against the 12 line. Approvals posted to the FDA list after reconciliation are recorded as caveats and do not change the grade.
Read next
Lab-tier forward call published by Tempora Research. Lab entries are experimental signatures under test and carry no calibrated lift. Methodology reproducible against the public engine using Swiss Ephemeris with True Pushya Paksha ayanamsa (PVRN Rao). Internal audit log maintained. This article does not constitute investment, financial, legal, medical or professional advice. First published 30 July 2026 by Tempora Research.