LDOS › analyze
LDOS · Analyze
Date: 2026-09-28 | Price: $121.86 (−1.44% today) | Market cap: $15.26B Baseline: analyze-2026-08-04.md — ACCUMULATE [8.0] at $130.60 Also read: analyze-2026-07-14.md — BUY [7.5] at $106.49
The one-line result: the business held completely and the calendar got worse. 85% of Structural and Trend claims survived, none was retracted, superseded or drifted — every status change this pass sits in the Price and Judgment layers. The $115–135 entry zone survives a from-scratch re-derivation, narrowed to $111–131.
1. What this updates
The baseline was written on the Q2 CY2026 print day. No earnings print has happened since, and guidance has not been revised. The event list is therefore short, and it is almost entirely about the government's funding calendar rather than about Leidos.
| Date | Event | Touches |
|---|---|---|
| 2026-08-11 | Director N. Geer sells 10,000 sh at ~$140.6–141.1 | NEW (N6) |
| ~2026-08-20 | Price reaches $146.85 intramonth high (+12.4% above baseline) | J2 |
| 2026-09-01 | Director G. May sells 1,000 sh at $140.85 | NEW (N6) |
| 2026-09-02 | FY2027 Continuing Resolution signed (P.L. 119-103) — funds all 12 bills 10-01 → 12-11 | NEW (N1, N2) |
| 2026-09-16 | FOMC +25bp to 3.75–4.00%; 10yr ~4.96%; ~60% odds of another on 10/28 | P1 |
| 2026-09-16 | Navy NGEN-R SMIT $926M second option year — 425,000 devices, 2,500 sites | S1, N3 |
| 2026-09-23 | Army ARTEMIS extension, up to $127M through summer 2027 + 2 option years | S1, N3 |
| 2026-09-24 | Argus raises PT $185 → $210, Buy | P6 |
| 2026-09-01→28 | Price falls $140.49 → $121.11 low, −13.6% on the month, with no company news | J2, N4 |
Events that touched nothing: the two Sept contract awards moved the stock not at all — it fell through both. That is itself a finding, recorded under N4.
2. The delta ledger
❌ RETRACTED — 1 row
| # | Baseline claim | The error |
|---|---|---|
| P5 | Graham IV $99.00 is "a floor indication"; July wrote it is "understated — penalizes goodwill-heavy book." | The mechanism is backwards. Graham IV = √(22.5 × EPS × BVPS), and BVPS is book value, not tangible book. Leidos carries $6.34B of goodwill against $4.92B of total equity — tangible book per share is negative (≈ −$11/sh before other intangibles). Goodwill inflates the Graham output; it cannot penalise it. Corrected: Graham IV = √(22.5 × 10.70 × 42.11) = $100.69, which is not a conservative floor being held down unfairly — it is an already-flattered number that still sits 17% below spot, and on tangible book the model does not compute at all. Cause: reasoning error carried from the July file into August, not a vendor trap. Consequence: Graham carries zero weight on this name, not the "floor indication" weight the baseline gave it. The verdict does not change — Graham was already the dissenting model — but a false comfort leaves the file. |
🔄 SUPERSEDED — 6 rows
A portfolio-specific passage was removed from the public build.
📉 DRIFTED — 1 row
| # | Baseline | Now | Where it breaks |
|---|---|---|---|
| St4 | Q2 underlying EPS grew +1.6% ($3.21 → $3.26); the $0.35 "beat" is against consensus, not against the business | TTM EPS $10.70 vs FY25 $11.14 — earnings growth is now −7% y/y outright. The VA Health drag has moved from "flattens a quarter" to "bends the trailing line down." | It breaks if FY26 lands below the $12.20 guide floor. The full-year claim (+9.5% to +12.2% over FY25's $11.14) requires a materially stronger H2 than H1 delivered. The Nov print is the test. Until then the trend line and the guide point in opposite directions, and the guide is the one with management's name on it. |
🔁 REFRESHED — 5 rows
| # | Old → New |
|---|---|
| T4 | Shares out 126.38M (FY25) → 125.49M. Buyback still running, −0.7% since year-end. Diluted 3yr CAGR −2.0% intact. |
| T6 | Dividend yield 1.49% → 1.39–1.41% at $121.86 on a $1.72 annualised rate; payout ratio 15% → 16%. ⚠️ The baseline's 1.49% does not reconcile: $1.72 ÷ $130.60 = 1.32%. Small vendor/arithmetic discrepancy, not thesis-bearing, but recorded rather than quietly replaced. |
| P4 | Forward P/E 10.6× → 9.87× (on guide mid $12.35 at $121.86) · P/FCF 10.4× → 9.75× · EV/EBITDA 8.9× → 8.98×. |
| P6 | Consensus target $150.67 → $157.27 (Yahoo mean). Argus $185 → $210 on 9/24. ⚠️ A secondary aggregator reported a $188.31 mean across 16 analysts, which does not reconcile with $157.27. I used $157.27 and did not rely on the $188.31 figure — single-source and unreconciled. |
| P7 | 23% below central $161 → 21% below central $154. The discount is essentially unchanged: price fell 6.7%, fair value fell 4.3%. |
✅ CARRIED — 15 rows
| # | Claim | Check |
|---|---|---|
| S1 | Moat = incumbency inside classified and mission-critical programmes; barriers are procedural (cleared staff, past-performance qualification, facility accreditation, program-of-record positions) | Strongest row of the pass — and it moved from asserted to evidenced. Two incumbent retentions in eight days: NGEN-R SMIT $926M second option year and ARTEMIS up to $127M through summer 2027. The July file asserted high recompete win-rates with no data behind it; there is now data. |
| S2 | The real threat is procurement reform, not a rival prime; this is why services trade 10–12× | No procurement-reform event fired. Tension noted: the CR's new-start freeze is an adjacent, milder form of the same pressure — see N1. |
| S4 | Capex 0.7% of revenue, so OCF ≈ FCF; no capex cliff | $125M on $17.17B. Unchanged; no new statements filed. |
| S5 | Evergreen 7/10; backlog gives 2–3 years of visibility in a political sector | Backlog $48.7B unchanged pending the Nov print. The CR is exactly the political risk this rating already priced. |
| S6 | LDOS preferred over BAH; do not need both | Carried and strengthened. BAH: revenue −4% y/y, earnings −24% y/y, "hold", 10.49× fwd. LDOS: revenue +7%, 9.87× fwd. LDOS is now both cheaper and growing while BAH shrinks. The divergence widened. |
| T1 | Revenue 3yr CAGR +6.1% ($14.40B → $17.17B) | Vendor y/y revenue growth still +7%. |
| T2 | Operating income 3yr CAGR +23.0%; margin 7.8% → 12.2% | Operating margin 12% current. |
| T3 | FCF 3yr CAGR +23.5% ($863M → $1.62B) | Unchanged; no new statements. |
| T5 | Debt flat-to-down over 4 years ($5.49B → $5.24B) while $2.64B went to buybacks — the buyback is cash-funded | Unchanged. |
| St1 | Q2 CY26: revenue $4.6B +7%, adj EBITDA 13.8%, EPS $3.26, backlog $48.7B, leverage 2.5× | Historical fact, undisturbed. |
| St2 | FY26 guide raised: revenue $18.2–18.4B, EPS $12.20–12.50, OCF raised | Not revised. Still the live guide as of late September. |
| St5 | VA suspended MDE incentive payments for the rest of 2026, absorbed into the raised guide; ex-Health the rest of Leidos is guided ~+7% revenue / ~+19% adj EBITDA | Unchanged. The suspension carries; the recompete is J3 and it moved. |
| St6 | Debt/assets 38.8%; $300M commercial paper repaid | D/E 123; debt/assets basis unchanged. |
| St7 | Net margin 8% < operating margin 12% → no equity-mark distortion | Still 8% vs 12%. [[pitfall-unrealized-equity-marks-break-headline-pe]] does not fire. |
| J2 | "Why not higher than [8.0]" — it ran +10.3% in three weeks; [[pattern-shortlist-runs-when-unexecuted]]; "cheap, not free" | Carried and vindicated. The stock ran a further +12.4% to $146.85, then gave the entire overshoot back. It is still +5.4% above the pre-print $115.23 close of 31 July — the print gain held, the overshoot did not. The baseline's most cautious sentence was its most accurate one. |
⏳ UNTESTED — 3 rows, named
| # | Claim | Why it could not be tested |
|---|---|---|
| S3 | The mix is shifting to the durable end — munitions, unmanned surface vessels, space payloads | No new bookings disclosure since Q2. The two September awards are ISR and Navy IT — services, not product. That is neutral-to-mildly-against the mix-shift reading, but two awards are not a mix. Next data point: the Nov print. |
| S7 | The MDE platform (15,000 providers, 90 clinics, 2.8M exams/yr) is not trivially replicable; the VA's quality pivot favours the incumbent | No new evidence either way. This is now the highest-stakes untested row in the file, because J3 put a date on the event it governs. |
| St3 | Book-to-bill 1.1× total, Defense 2.2× in-quarter / 1.9× ttm; backlog 2.66× FY26 revenue | Quarterly disclosure; the Nov print is the only source. A search result purporting to give "Q3 2026" bookings of $5.9B / 1.3× book-to-bill describes an earlier fiscal quarter and was discarded — Q3 CY2026 has not been reported. |
🆕 NEW — 6 rows
N1 — The CR passed, and the specific risk this run was commissioned to test did not fire. The FY2027 Continuing Resolution (P.L. 119-103) was signed 2026-09-02 and funds all twelve appropriations bills from 2026-10-01 through 2026-12-11. There is no shutdown at the fiscal year start. Defense programs continue at FY26 levels; no furloughs, no unpaid troops. But a CR is not neutral. It funds at prior-year levels and blocks new program starts — which throttles exactly the defense-tech ramp (munitions, USVs, space) that carries the S3 mix-shift thesis. And none of the twelve FY27 regular bills has been enacted: the House has passed three, the Senate Appropriations Committee has reported none.
N2 — Two December binaries stack, and the November print resolves neither. The CR expires 2026-12-11. The MDE IDIQ completion date is 2026-12-31. They are 20 days apart, and both sit after the early-November Q3 print. A buyer today is buying into a quarter with two unresolved dated events at the end of it and one earnings report in the middle that will speak to neither with finality. This is the single largest change to the risk picture since the baseline, and it is a calendar change, not a business change.
N3 — The moat claim is now evidenced rather than asserted. NGEN-R SMIT second option year, $926M, announced 2026-09-16 — Leidos manages more than 425,000 Navy and Marine Corps devices at more than 2,500 sites, on a vehicle originally worth up to $7.7B. ARTEMIS, up to $127M, announced 2026-09-23, extending the modified Challenger 650 ISR platform through summer 2027 with two option years, under a $7.9B ceiling vehicle. Two incumbent positions defended in eight days. The July report claimed "recompete win-rates in this industry run high" with nothing behind it; that sentence now has evidence under it.
N4 — The peer set has split, and the "sector de-rating" reading does not survive. | | LDOS | BAH | SAIC | CACI | |---|---|---|---|---| | Off 52w high | −41% | −35% | −8% | −11% | | Revenue y/y | +7% | −4% | +6% | +18% | | Earnings y/y | −7% | −24% | −12% | −1% | | Operating margin | 12% | 10% | 8% | 10% | | Forward P/E | 9.87× | 10.49× | 11.41× | 16.16× |
This is not a uniform government-services de-rating — CACI and SAIC sit near their highs. BAH's fall is explained by its own numbers. LDOS's is not explained by its numbers. And neither prior report noted that LDOS fell from $205.77 (Nov 2025) to $98.86 (June 2026), a 52% drawdown, before recovering. This is a broken chart in recovery, not a steady compounder, and that changes how an entry zone should be read: the zone is not catching a dip in an uptrend, it is sizing into an unfinished repair.
N5 — LDOS is the cheapest name in its own peer set, on the group's best margin. 9.87× forward on the guide midpoint, against SAIC 11.41×, BAH 10.49×, CACI 16.16×. Only CACI grows faster, and it costs 64% more per unit of forward earnings. The relative-value case is stronger today than it was at the baseline, because the price fell and the peers did not.
N6 — Insider selling into the August run; no open-market buying anywhere in 2026. Director Noel Geer sold 10,000 shares at ~$140.6–141.1 on 2026-08-11; director Gary May sold 1,000 shares at $140.85 on 2026-09-01. Both sold within days of the $146.85 high. The last open-market purchase by any insider was November 2024 at $161.13. Mild negative, and internally consistent with a board that did not regard $141 as cheap — which is also what this file now says, since $141 sits above the $131 zone top. RSU grants were separated out and not counted as buys, per [[pitfall-yahoo-insider-purchases-counts-rsu-grants]].
3. How the close calls were decided
J3 — does a dated binary supersede, or merely refresh? The baseline's sentence contained two assertions: the MDE recompete is the thesis-breaker, and it is not scheduled. The first carried. The second did not. A refresh would have been the wrong status because the change is not a number moving — it is an unknown becoming a known, and the known is worse than the assumption the baseline reasoned from. Opposing force: the date rests on a single contract-action family, and no 2026 confirmation was obtained. That opposing force is why the row is SUPERSEDED with an explicit caveat rather than driving a larger conviction cut.
P1 — is a fair-value cut on rates alone legitimate? The instruction in [[pitfall-stale-fair-value-is-most-costly-on-winners]] is to change the inputs before changing the assumptions, and report both separately. So: inputs unchanged — guidance $12.20–12.50 is the same number, FCF/share $12.50 is the same number, share count is marginally better. The only thing that moved is the required return: 10yr 4.96%, a Fed that raised in September with another likely in October. That is a real, external, non-discretionary input to any multiple, and it argues for the same treatment on every name in the book, not special pleading here. The band moved 12–14× → 11.5–13.5×, a half-turn at each end. I did not let the September price fall influence the fair value — that would be circular, and it is the error the pitfall note warns about in mirror image.
St4 vs St2 — the Trend row and the State row contradict each other. Trailing EPS is −7% y/y; the guide implies +9.5% to +12.2% for the full year. One of them is wrong. Resolved in favour of the guide, but held open: the guide has management's name on it, was raised twice this year, and the gap is fully explained by a dated, quantified, already-disclosed VA Health headwind that management has absorbed. But that is the baseline's reasoning, not new evidence, so the row is DRIFTED with a named break level rather than CARRIED. If the Nov print does not close the gap, the next pass should treat St4 as the leading indicator and St2 as the lagging one.
N4 vs P2 — does a −41%-off-high chart argue against buying in the zone? Weighed against: the fall is not explained by the company's own numbers, the peer set did not fall with it, the business claims all carried, and the discount to a freshly derived fair value is unchanged. A price move with no fundamental counterpart is a Price force and belongs in the Price rows — it is not permitted to retroactively break a Structural claim it never touched. It does, however, justify buying the lower half of the zone rather than the top, which is where §6 lands.
J1 — why 7.5 and not 7.0 or 8.0. Not 8.0, because the single identified thesis-breaker acquired a date inside 90 days and a second dated binary landed 20 days before it. Not 7.0, because 85% thesis persistence with nothing retracted, superseded or drifted below the Price layer is a strong result, and because the discount to fair value is unchanged after a 7% price fall. A half-point is what the calendar is worth; the business earned none of the cut.
4. ⭐ Did the $115–135 entry zone survive? — Yes, substantially intact.
This is the question the run was commissioned to answer, so it gets its own verdict.
| Baseline 2026-08-04 | Re-derived 2026-09-28 | Move | |
|---|---|---|---|
| Central fair value | $161 | $154 | −4.3% |
| Zone floor | $115 | $111 | −3.5% |
| Zone ceiling | $135 | $131 | −3.0% |
| Spot | $130.60 (96th pctile of band) | $121.86 (54th pctile) | — |
The zone survives, and the buy signal survives with it. The re-derivation moved both ends down about 3%, and $121.86 sits mid-band rather than pressed against the ceiling as it was in August. Derivation: zone ceiling = central FV × 0.85, floor = central FV × 0.72 — the same margin-of-safety discipline, re-applied to a fresh fair value, not inherited.
Why LDOS survived where three of four in-zone held names failed the same test on 2026-09-22. Those names failed because they ran — the price moved away from a zone nobody re-derived, and the refreshed fair value could not catch up. LDOS did the opposite: the price came toward the zone. Between 2026-08-04 and today, price fell 6.7% and fair value fell 4.3%. The gap to central fair value went 23% → 21% — statistically unchanged. A zone goes stale when the price and the value diverge. Here they moved together, in the same direction, by almost the same amount. That is the clean counter-case to [[pitfall-stale-entry-zone-suppresses-a-name]]: a zone that was still approximately correct after a real move, demonstrated rather than assumed.
The honest caveat: the zone survived on arithmetic. The reason to act inside it is weaker than it was in August, because of N2's stacked December binaries. Survival of the zone and enthusiasm about the entry are two different findings, and only the first is unqualified.
5. Thesis persistence and conviction delta
Thesis persistence: 11 of 13 Structural + Trend claims carried or refreshed = 85%. The other two are UNTESTED, not broken. Zero Structural or Trend claims were retracted, superseded or drifted. Every single status change this pass sits in the Price and Judgment layers.
That is the classic profile the command was built to detect, and it points the opposite way from the price: the business did not move and the multiple did. 85% persistence against a −6.7% price move, in a name already −41% from its high, is a re-rating setup, not a deterioration.
Conviction [8.0] → [7.5] ⬇️
| Direction | Rows | Weight |
|---|---|---|
| ⬇️ J3 — the one identified thesis-breaker moved from undated to ~90 days out | J3 | Largest single mover |
| ⬇️ N1/N2 — CR new-start freeze; two binaries stacked 12-11 and 12-31, neither resolved by the Nov print | N1, N2 | Large |
| ⬇️ St4 — trailing EPS now −7% y/y, contradicting the guide until the Nov print settles it | St4 | Moderate |
| ⬇️ P1 — fair value cut 4.3% on a higher required return | P1 | Small |
| ⬆️ S1/N3 — the moat claim moved from asserted to evidenced on two incumbent retentions | S1, N3 | Large |
| ⬆️ N5/S6 — cheapest in the peer set on the best margin; the BAH divergence widened | N5, S6 | Moderate |
| ⬆️ P7 — discount to fair value unchanged at 21% after a 7% price fall | P7 | Moderate |
| ⬆️ Persistence 85% with nothing broken below the Price layer | — | Moderate |
Net −0.5. The upward forces are real and roughly cancel the fundamental downward ones. What does not cancel is the calendar: two dated binaries inside 95 days that were not on the board in August. The half-point is the calendar's, and the business earned none of it.
6. Updated verdict — 🟢 ACCUMULATE · conviction [8.0] → [7.5] ⬇️
Valuation
| Method | Input | Result | Weight |
|---|---|---|---|
| Forward P/E | $121.86 ÷ $12.35 (FY26 guide mid) | 9.87× | High |
| P/FCF | $121.86 ÷ $12.50 FCF/share | 9.75× | High |
| EV/EBITDA | 8.98× | Medium | |
| Peer relative | vs BAH 10.49× · SAIC 11.41× · CACI 16.16× | cheapest, best margin | High |
| Bogle expected return | 1.41% yield + ~10% EPS growth | ~11.4% + multiple change | Medium |
| Dividend Yield Theory | 1.41% vs 5yr average 1.25% | 13% above its own average — but the yield is high because the price fell, not because the dividend grew | Near zero |
| DDM | 16% payout ratio | N/A — the dividend is not the thesis | — |
| Graham IV | √(22.5 × 10.70 × 42.11) = $100.69 | Zero weight — see the RETRACTED row. The $42.11 book is $6.34B goodwill against $4.92B equity; tangible book is negative. The model is flattered by intangibles and still prints 17% below spot. | Zero |
Fair value: $142–167, central $154. That is 11.5–13.5× the FY26 guide midpoint — a half-turn below the baseline's band, on rates rather than on earnings. The top of the band is deliberately unheroic: services businesses do not earn a market multiple, and 13.5× is roughly where the better-run peers top out, not a re-rating fantasy.
At $121.86 the stock is ~21% below central fair value — essentially the same discount the baseline found, reached by a different route.
Zones
- Entry: $111–131. Spot $121.86 is mid-band.
- Accumulate in the lower half ($111–121), and hold size back for December. The name is cheap and in-zone, but it is not at the floor, and two dated binaries sit at the end of the quarter. Buying the top of the band ahead of them is paying full price for an unresolved calendar.
- Trim: 13.5× fwd.
EPS basis: FY26 guide midpoint $12.35 → derived trim level ≈ $167.
⚠️ Vendor-staleness check per [[pitfall-multiple-trim-inherits-the-broken-vendor-field]]:
Yahoo carries
PE(fwd) 9.54at $121.59, implying a vendor EPS of $12.75 — above the guide midpoint, consistent with an NTM blend reaching into FY27. The site will therefore render 13.5 × $12.75 ≈ $172, about 3.1% above the $167 this analysis derives. Within tolerance and recorded, not silently accepted.
Break triggers
- The VA MDE franchise is lost, or re-awarded at materially reduced scope or economics. The IDIQ completion date is 2026-12-31. This remains the single identifiable thesis-breaker, and it is now dated.
- FY26 guidance is cut at the Q3 print. The guide has been raised twice this year; a cut would be a regime change, not a wobble.
- TTM book-to-bill falls below 1.0× for two consecutive quarters.
- An FY27 appropriation that cuts the defense topline, or a CR extended past ~Q1 2027 that keeps the new-start freeze in place through the ramp the mix-shift thesis depends on.
Upgrade conditions (back toward [8.0]+)
- MDE re-awarded on comparable economics — this alone is worth most of the half-point back.
- Q3 print holds or raises the $12.20–12.50 guide with ttm book-to-bill ≥ 1.1×, closing the St4/St2 contradiction in the guide's favour.
- A full-year FY27 defense appropriation enacted, ending the new-start freeze.
7. What this pass did NOT test
- S3 — the mix shift toward munitions, USVs and space payloads. No bookings disclosure since Q2. The two September awards are services-side, which is mildly against the reading. Next test: the Nov print.
- S7 — whether the MDE platform is genuinely hard to replicate, and whether the VA's quality pivot favours the incumbent. No new evidence. This is the highest-stakes untested row in the file, because J3 gave a date to the event it governs. It should be the first thing tested next pass.
- St3 — current book-to-bill and backlog. Quarterly disclosure only. A search result offering "Q3 2026" figures of $5.9B bookings / 1.3× book-to-bill described an earlier fiscal quarter and was discarded. Q3 CY2026 has not been reported.
Single-source conclusions flagged, so the next run knows which of today's calls are thin:
- J3's 2026-12-31 MDE date rests on one contract-action family, reported by three outlets describing the same January-2025 award. No 2026 confirmation of a follow-on, extension or re-solicitation was obtained. This is the thinnest load-bearing input in the file, and it is the row that drove most of the conviction cut.
- The September price fall has no confirmed cause. The best-supported reading — that the post-print overshoot to $146.85 simply unwound, leaving the stock still +5.4% above its pre-print level — is arithmetic and holds up, but it is an inference, not a reported fact. N4 shows the peer set did not fall alongside, which rules out a sector explanation without establishing what the real one is.
- Analyst consensus is unreconciled between $157.27 (used) and $188.31 (not used).
Sources
- FY2027 Continuing Resolution, P.L. 119-103 — Congress.gov CRS R49353
- House passes funding stopgap, averting government shutdown in October — Breaking Defense
- Leidos secures $926M Navy, $127M Army awards (NGEN-R SMIT, ARTEMIS) — Virginia Business
- Leidos wins Army ARTEMIS contract worth up to $127 million — Seeking Alpha
- VA awards 4 spots on Medical Disability Exam Support IDIQ — GovConWire
- VBA awards medical disability examination services contract to Leidos QTC — Leidos IR
- Leidos Holdings stock gains on analyst targets and fresh guidance — ad-hoc-news
- Statements, snapshot and peer quotes: Yahoo Finance via
.mcp/fin.py LDOS --newsand.mcp/fin.py BAH CACI SAIC --quote - Price history and insider transactions: Yahoo Finance MCP (
get_historical_stock_prices1y/1mo,get_holder_info insider_transactions) - Prior: analyze-2026-08-04.md · analyze-2026-07-14.md · Knowledge/Themes/consulting-it-services.md
ADDENDUM — 2026-09-28, same day: J3's untested date, resolved
Why this exists. §"What this pass did NOT test" listed as item 1: "J3's 2026-12-31 MDE date rests on one contract-action family." A targeted primary-source check was run hours later, prompted by a sizing question. It resolves the row, and it materially softens the risk the conviction cut was built on. Recording it here rather than in a new report, because it tests one claim rather than re-deriving the name.
What 2026-12-31 actually is
Not a competition Leidos might lose. It is the expiry of a bridge extension Leidos already holds a seat on.
- In January 2025 the VA extended the four incumbents rather than running a full recompete — Washington Technology: the VA "has essentially extended the contracts of four companies". The holders are Maximus/VES, Leidos QTC, Optumserve (UnitedHealth) and Loyal Source.
- The "MDE 2025 Regions 1-4 Recompete" already happened, and Leidos won its seat — the follow-on multiple-award IDIQ was awarded to those same four.
- Leidos is the LARGEST holder on the vehicle, not a marginal one: $5.1B obligated, against Maximus $3.7B, Optumserve $3.4B, Loyal Source $500M+.
- The vehicle's ceiling was raised from $6.8B to $13.2B because demand exceeded the original early. It covers 9M+ exams. This is a program the VA keeps enlarging.
What this does to the ledger
| Row | Was | Now |
|---|---|---|
| J3 | "The recompete is scheduled ~90 days out" — read as a franchise binary | 🔁 REFRESHED. The date is real, but it is a period-of-performance end on a vehicle Leidos already won, under an agency that bridged rather than competed at the last decision point. Realistic downside is task-order share at the next vehicle, not franchise loss. |
| S7 (MDE platform replicability) | "Highest-stakes untested row, because J3 put a date on it" | Still untested, but lower stakes — the incumbent-displacement scenario it guards against is not what 12/31 presents. |
The genuinely new flag, which the report did not have
No published follow-on solicitation for the post-2026 period could be found, with expiry ~3 months out. Given the VA bridged rather than competed in Jan-2025, and demand is running above the original ceiling, another bridge is the base case. But "no visible procurement path 90 days out" is a real unknown — it is simply a different unknown from the one §J3 described, and it resolves by a solicitation appearing (or not) rather than by an award being lost.
Conviction
Held at [7.5]. Not raised, and the reasoning matters more than the number.
J3 was "the biggest mover" in the cut, and roughly half of what it was carrying has now evaporated — the franchise-loss reading was wrong. That argues for giving some back. Against that: the cut's other legs are untouched (TTM EPS −7% y/y, the CR new-start freeze through 12-11, the required-return move), and the replacement risk — an unscheduled procurement path — is less severe but less legible than the one it replaces. Re-rating on the same facts re-weighted is what [[pitfall-stale-fair-value-is-most-costly-on-winners]]'s sibling failure looks like, so the honest move is to fix the description of the risk and leave the number until a solicitation appears.
If a follow-on solicitation or a second bridge is published naming Leidos QTC → [7.5] → [8.0], and that is a cleaner upgrade gate than the one the main report wrote.
Break trigger, rewritten
The original — "VA MDE franchise lost or re-awarded at materially reduced scope/economics" — described an event that is not on the calendar. Replace with:
No MDE follow-on or bridge published by ~2026-12-01 (procurement path invisible <30 days from expiry) · a published follow-on that excludes Leidos QTC or cuts its regions · FY27 MDE delivery order materially below the FY26 $392.7M on comparable exam volume