V10 operating economics, definitions and review conclusions

28 Sep 2026 · Analytical clarification with primary-source links

Saved research evidence used by the youtube deck. This is a dated provider snapshot or an EpochX analysis, not a live market feed.

# Hertz operating economics: V10 definitions and conclusions

Research reviewed 28 September 2026. The 70-slide investigation asks whether Palantir-enabled decisions improve Hertz growth and unit economics after the full program cost. The game’s invented encounters provide entertainment. They establish no actual pricing, commercial agreement, surveillance practice or investment recommendation by a cameo character.

## What is established and what remains to measure

Hertz describes deployed operating applications and demonstrated coordination at the vehicle level. A VIN joins records for one physical car. Other systems provide its position, condition and status. Employee task acceptance and completion connect plans to physical work. Hertz separately names Foundry and Navidea in fleet planning. Management describes Palantir-supported labor planning. Public sources do not isolate incremental contribution, program cost or net return attributable to Palantir.

The military comparison concerns readiness, joined data and executable work. It does not transfer an Army performance result or savings estimate to Hertz. Generic Palantir submission criteria establish configured permission checks. They do not establish a separate human approval step or Hertz’s exact production autonomy.

## Economic definitions

- Q2 2026 revenue growth included about 9% higher revenue per day with broadly flat rental days. Better revenue yield alone does not establish AI-created rental capacity.
- Slides 28–29 use total-fleet utilization and average total vehicles. The extra-rental and fixed-demand fleet-reduction illustrations are alternatives for the same released capacity. Do not add their full benefits together.
- The repair example assumes all 3 recovered days sell at $40 incremental contribution, earning $120 against a $100 repair premium. Break-even is $33.33 per recovered rented day. Other transport, warranty and rework differences must also enter a real comparison.
- The labor illustration measures average workload coverage. Variable arrivals and task times can require spare capacity. It does not establish an optimal staffing schedule.
- The queue illustration uses consistent 24-hour days and average elapsed time.50 fewer vehicles in process is a capacity change, not 50 additional daily rentals.
- Q2 scale is 38.646M rental days. At 30% activity coverage, $1 saved on each covered day means $11.5938M per quarter. A fixed $25M recurring cost divided by 4 × Q2 days needs 16.2¢ at 100% coverage, 32.3¢ at 50%, and 53.9¢ at 30%. Actual program cost, adoption and timing are undisclosed.
- The slide 65 scenarios apply cost savings only to baseline rental days. The 60% contribution assumption applies to incremental revenue. The $25M annual cost is an assumption. One-time integration and setup require separate funding.
- Realized 30% of the middle-case modeled gross benefit is approximately $22.54M. With the full assumed $25M recurring cost, net operating benefit is approximately−$2.46M before setup spending. Benefits and actual contract costs may ramp differently.
- The service-only illustration assumes a customer owns and finances 1,000 cars. $4.8M fees less $3M direct cash cost and $0.6M fixed cash cost leaves $1.2M contribution. Program costs, capital spending, working capital, financing and tax must still be considered before net cash and payback. These are invented rates.

## Commercial and funding boundaries

The initial Hertz AV program names Uber, Lucid and Nuro. Counting those participants does not establish three service contracts. Tesla rollout, Avis–Waymo and Hertz/Oro are distinct evidence sets. Contract count, committed volumes, fees and retained margins remain undisclosed in the reviewed sources. Oro’s management revenue expectation includes existing rideshare activity.

The rental network includes company-operated and licensee locations. It is not an inventory of AV-ready depots. Existing facilities have opportunity costs and may require site, power and staffing investment.

A privacy notice describing advertising sale or sharing does not establish a cash data-sales transaction, a Palantir-enhanced data product or a Google trip-history feed. Applicable purpose restrictions, choices and consent depend on jurisdiction and platform. Palantir’s public contract template does not disclose Hertz’s negotiated terms. No separate data-sales premium enters our scenarios.

The corporate-value bridge combines dated equity market value with June nominal net corporate principal. It retains fleet depreciation and vehicle interest in operating economics while excluding fleet debt from that convention. It is not a complete enterprise or liquidation valuation. Operating gains must survive cash needs, financing and dilution before they accrue per share.

## Final verdict

Palantir can plausibly improve a physical fleet business through repair routing, vehicle allocation and staffing. The decisive evidence is an improvement against comparable operations, after demand, price, fleet mix and seasonality, that exceeds the full program cost and produces recurring cash. A disappearance of that advantage, or costs that absorb it, would weaken the thesis. The reviewed public record has yet to isolate that net result.

## Updated slide explanations and source trail

### 04 / Revenue grew 10% with a 1% smaller fleet

Hertz earned more revenue from a smaller fleet. Revenue per day rose nine percent while rental days were broadly flat, so pricing explains much of the growth. Adjusted Corporate EBITDA reached eighty-one million dollars. The AI case needs a second step: identify which deployed workflows reduced waiting, fulfilled bookings or lowered cost, then measure their contribution after software and implementation expense. Fleet renewal, recalls, demand and pricing all belong in that comparison.

- [Hertz Q2 2026 earnings release](https://www.sec.gov/Archives/edgar/data/47129/000165785326000045/q22026earningsrelease.htm) (6 Aug 2026).
- [Q2 2026 management prepared remarks](https://newsroom.hertz.com/articles/article-details/hertz-global-holdings-inc-q2-2026-prepared-remarks/) (6 Aug 2026).
- [EpochX evidence ledger for the Palantir thesis](Research_Sources/palantir_evidence_v9.md) (28 Sep 2026).

### 07 / DPU rose 18% to $302 per vehicle per month

Q2 depreciation per vehicle per month rose from 256 to 302 dollars. The scale still makes a small improvement meaningful: 25 dollars less monthly DPU across the Q2 average fleet of 539,118 vehicles would annualize to 161.7 million dollars. That is an expense sensitivity at a fixed fleet size. Cash realization depends on purchase prices, holding periods and eventual sale proceeds. Vehicle mix, used-car prices and disposal channels all matter, so this illustration assigns no benefit to Palantir.

- [Hertz Q2 2026 earnings release](https://www.sec.gov/Archives/edgar/data/47129/000165785326000045/q22026earningsrelease.htm) (6 Aug 2026).
- [EpochX operating sensitivities and assumptions](Research_Sources/operating_models.md) (22 Sep 2026).
- [EpochX Palantir economics model and evidence tests](Research_Sources/palantir_economics_v9.md) (28 Sep 2026).

### 08 / Downtime gives operating intelligence a dollar target

Hertz’s recall experience puts a price on unavailable vehicles. Management estimated a thirty-million-dollar Q2 Corporate EBITDA impact. That is not an addressable saving that software can automatically capture: a car under recall stays unavailable until the legal and physical repair requirements are met. The possible benefit comes from identifying affected cars quickly, arranging parts and repairs, protecting reservations and returning safe vehicles to service sooner. We need measured changes in waiting time and completed repairs before assigning a benefit.

- [Hertz Q2 2026 earnings release](https://www.sec.gov/Archives/edgar/data/47129/000165785326000045/q22026earningsrelease.htm) (6 Aug 2026).
- [Q2 2026 management prepared remarks](https://newsroom.hertz.com/articles/article-details/hertz-global-holdings-inc-q2-2026-prepared-remarks/) (6 Aug 2026).
- [EpochX Palantir economics model and evidence tests](Research_Sources/palantir_economics_v9.md) (28 Sep 2026).

### 09 / The Americas generates 80% of Hertz revenue

Americas RAC produced about eighty percent of second-quarter revenue, so a repeatable improvement there has more initial scale than the same percentage change in a small region. The United States separately represented almost seventy-seven percent of revenue. The segment and geographic views overlap. Neither identifies the actual Palantir deployment footprint or reports city-level profitability. To test causality, we would compare similar locations and workflows rather than assume the entire region adopted the same tools at the same time.

- [Hertz June 2026 Form 10-Q](https://www.sec.gov/Archives/edgar/data/47129/000165785326000046/htzz-20260630.htm) (6 Aug 2026).
- [EpochX evidence ledger for the Palantir thesis](Research_Sources/palantir_evidence_v9.md) (28 Sep 2026).

### 10 / Daytona on the rental lot

Hertz describes Daytona as a way to coordinate the work between a return and the next rental. These company-reported adoption figures establish a large operating footprint. The investment question is what changes in an employee’s working day. Does a person spend less time searching for a car, waiting for instructions, or discovering that a vehicle cannot be rented? A completed task can update the operating picture for the next team. Our analysis follows that chain rather than treating the number of users or events as a financial result. Public materials do not supply a site-by-site deployment or active-use audit.

- [Hertz confirms Daytona deployment](https://www.linkedin.com/posts/hertz_navigating-the-operational-complexity-of-activity-7470127003765985281-u_-A) (2026 / reviewed 28 Sep 2026).
- [Original Hertz and Palantir partnership](https://newsroom.hertz.com/press-releases/press-release-details/hertz-partners-with-palantir-to-drive-operational-excellence-and-enhance-customer-experien/) (19 Oct 2022).
- [Hertz integration evidence and limits](Research_Sources/integration_evidence_v9.md) (Reviewed 28 Sep 2026).

### 13 / Maintenance considers the whole journey

The demonstration exposes three repair paths and checks parts, labor, transport and shop capacity. It also lets the maintenance team consider bundling work into a reliability visit. Our decision rule is to compare the full expected cost of each feasible path, including lost rental contribution. A nearby cheap shop can be the expensive choice when its queue is long. Conversely, an urgent premium repair may waste money when the car would otherwise sit unused. Actual repair outcomes and comparable jobs are needed to test whether recommendations improve decisions.

- [Hertz Daytona demonstration at AIPCon 10](https://www.linkedin.com/posts/palantir-technologies_at-aipcon-10-hertz-reveals-how-palantir-activity-7470120635973799936-nqem) (2026 / reviewed 28 Sep 2026).
- [EpochX logistics analysis and worked examples](Research_Sources/logistics_mechanics_v9.md) (Reviewed 28 Sep 2026).

### 14 / A $100 higher repair bill can save $20 overall

The internal repair costs one hundred dollars and takes four days. The outside repair costs two hundred dollars and takes one day. Assume all three recovered days serve incremental rental demand, each contributing forty dollars after the extra costs of serving that rental. That yields one hundred twenty dollars of contribution for a one hundred dollar repair premium: twenty dollars of net benefit. The break-even contribution is one hundred divided by three, or thirty-three dollars and thirty-three cents per recovered day. At forty dollars a day, the faster repair needs more than two and a half incremental rented days to win. Any extra transport, warranty recovery or rework differences must also enter the comparison. These are illustrative assumptions. Hertz has not disclosed this result for its Palantir program.

- [EpochX Palantir economics model and evidence tests](Research_Sources/palantir_economics_v9.md) (28 Sep 2026).

### 20 / Labor demand changes hour by hour

This worked example shows why demand-based labor planning can matter without assuming layoffs. Ten people provide the day's sixty work-hours if each contributes six productive hours. Half the work arriving in a two-hour peak needs twenty people providing ninety productive minutes each. That is average workload coverage. Variable arrivals and task times may require spare capacity to limit waiting. Shift overlap or reassignment can help, subject to skills, breaks and safety. Hertz says Palantir supports demand-based labor planning, but the staffing numbers here are assumptions. Compare whole-day paid cost, wait times and completed rentals at similar deployed and comparison sites before claiming a saving.

- [Q2 2026 management prepared remarks](https://newsroom.hertz.com/articles/article-details/hertz-global-holdings-inc-q2-2026-prepared-remarks/) (6 Aug 2026).
- [EpochX logistics analysis and worked examples](Research_Sources/logistics_mechanics_v9.md) (Reviewed 28 Sep 2026).

### 21 / Turnaround time ties up fleet capacity

At steady throughput, average work in process equals throughput multiplied by average elapsed time. The fictional depot averages six hundred vehicles per twenty-four-hour day. Reducing average turnaround from six hours to four reduces the cars tied up in the process from one hundred fifty to one hundred on average. The same process boundary and clock must apply to both cases. Find which waiting period shrank: inspection, keys, wash, repair, fueling or movement. With weak demand, the extra ready cars may sit idle. Financial value requires additional profitable rental days or fewer vehicles supporting the same business. Avoid counting both outcomes fully for the same capacity improvement.

- [EpochX logistics analysis and worked examples](Research_Sources/logistics_mechanics_v9.md) (Reviewed 28 Sep 2026).

### 23 / How a recommendation reaches operating systems

Palantir action types can modify operating objects, check submission criteria and trigger notifications or calls to other systems. A configured action can test both the submitting user’s authority and the vehicle’s current status. Those checks do not by themselves establish a separate human approval step. Hertz’s production approval thresholds and autonomy levels remain undisclosed. Our proposed test follows a permitted recommendation into the system staff actually use: confirm the work order, record the actor or approver, check retries for duplicates, then reconcile completion with the reservation record. Failed updates need a visible exception queue. This is the integration work that determines whether a recommendation changes the physical operation.

- [Palantir action types](https://www.palantir.com/docs/foundry/action-types/overview) (Reviewed 28 Sep 2026).
- [Palantir action submission criteria](https://www.palantir.com/docs/foundry/action-types/submission-criteria) (Reviewed 28 Sep 2026).
- [Palantir notifications and external-system writes](https://www.palantir.com/docs/foundry/action-types/side-effects-overview) (Reviewed 28 Sep 2026).
- [EpochX logistics analysis and worked examples](Research_Sources/logistics_mechanics_v9.md) (Reviewed 28 Sep 2026).

### 28 / One total-utilization point represents $30.4M of revenue

At the Q2 fleet size and rental rate, one additional utilization percentage point represents about 490,600 extra rental days and 30.4 million dollars of quarterly revenue. That is the gross opportunity if paying demand exists at the assumed rate. If seventy percent became incremental operating contribution, an explicitly hypothetical assumption, the contribution would be about 21.3 million before program costs. A different use of the capacity is to keep rental days constant and reduce fleet size. Neither calculation demonstrates an achieved Palantir result.

- [Hertz Q2 2026 earnings release](https://www.sec.gov/Archives/edgar/data/47129/000165785326000045/q22026earningsrelease.htm) (6 Aug 2026).
- [EpochX operating sensitivities and assumptions](Research_Sources/operating_models.md) (22 Sep 2026).
- [EpochX Palantir economics model and evidence tests](Research_Sources/palantir_economics_v9.md) (28 Sep 2026).

### 29 / Fleet efficiency can support growth with less capital

Keep the same number of rental days and raise utilization from seventy-nine to eighty percent. The same work then needs about 6,739 fewer vehicles at the Q2 fleet scale. At an assumed twenty-five thousand dollars each, that is about 168 million dollars of gross fleet value. It is not profit or free cash available to shareholders. Vehicle debt, disposal proceeds, taxes and fees determine the actual cash released. This also replaces the extra-rental-day case on slide 28 for the same capacity improvement. A franchise conversion can release capital too, but it changes revenue recognition and must be analyzed separately.

- [Hertz Q2 2026 earnings release](https://www.sec.gov/Archives/edgar/data/47129/000165785326000045/q22026earningsrelease.htm) (6 Aug 2026).
- [EpochX operating sensitivities and assumptions](Research_Sources/operating_models.md) (22 Sep 2026).
- [EpochX Palantir economics model and evidence tests](Research_Sources/palantir_economics_v9.md) (28 Sep 2026).

### 30 / A $1 saving per rental day is $38.6M per quarter

Hertz recorded 38.646 million transaction days in Q2. A dollar of actual cost reduction across every one of those days would save $38.646 million before the program bill. At 30% coverage, that same dollar saves $11.594 million. Paid overtime, outside labor, repair and transport spending must fall against comparable workload to demonstrate the benefit. Annualizing Q2 provides a scale reference. seasonality and the real deployment footprint determine the actual year.

- [Hertz Q2 2026 earnings release](https://www.sec.gov/Archives/edgar/data/47129/000165785326000045/q22026earningsrelease.htm) (6 Aug 2026).
- [EpochX operating sensitivities and assumptions](Research_Sources/operating_models.md) (22 Sep 2026).
- [EpochX Palantir economics model and evidence tests](Research_Sources/palantir_economics_v9.md) (28 Sep 2026).

### 31 / 30% rollout needs 54¢ saved per covered day

Hold the assumed recurring program bill at $25 million. Across four times Q2 rental days, the hurdle is 16.2 cents per day. If the tools affect half of those days, it rises to 32.3 cents. at 30% coverage, 53.9 cents. Coverage here means rental activity that can realize the modeled savings, rather than a percentage of locations. At full coverage, the earlier $10 million and $50 million cost cases require 6.5 and 32.3 cents. None is Hertz guidance or disclosed Palantir pricing. Include licenses, cloud usage and ongoing operations in recurring cost, and fund integration, data cleanup and initial training separately. The rollout pace, employee use and cash savings must justify both bills.

- [Hertz Q2 2026 earnings release](https://www.sec.gov/Archives/edgar/data/47129/000165785326000045/q22026earningsrelease.htm) (6 Aug 2026).
- [EpochX Palantir economics model and evidence tests](Research_Sources/palantir_economics_v9.md) (28 Sep 2026).

### 32 / The reported results give a mixed first answer

Revenue per day rose, which can reflect pricing and product mix. Higher utilization means a larger share of the fleet was on rent. Neither measure isolates the effect of Palantir. Depreciation per vehicle and reported operating expense per day also rose, while the daily revenue-to-operating-expense spread expanded. A younger fleet, market prices, recalls and commercial changes all affect the comparison. To attribute the gain, we need comparable operations with known deployment dates and full program costs. These results give the operating thesis a starting point. they do not provide the missing counterfactual.

- [Hertz Q2 2026 earnings release](https://www.sec.gov/Archives/edgar/data/47129/000165785326000045/q22026earningsrelease.htm) (6 Aug 2026).
- [Hertz June 2026 Form 10-Q](https://www.sec.gov/Archives/edgar/data/47129/000165785326000046/htzz-20260630.htm) (6 Aug 2026).
- [Q2 2026 management prepared remarks](https://newsroom.hertz.com/articles/article-details/hertz-global-holdings-inc-q2-2026-prepared-remarks/) (6 Aug 2026).
- [EpochX evidence ledger for the Palantir thesis](Research_Sources/palantir_evidence_v9.md) (28 Sep 2026).

### 33 / The labor thesis still needs a clean cost comparison

Management says Palantir-supported labor planning better matches staffing to demand and reduces overtime and outside labor. The published figures still show higher cost per day, alongside a roughly 2% improvement after management normalizations. The proof would be payroll and agency invoices for comparable work at deployed locations. Track paid hours per 100 vehicle turns, overtime, outside labor, customer waits and rework against matched locations and seasonality. Separate labor gains from revenue-linked expenses, recalls and sale-leaseback changes. Count the program cost once when calculating total net benefit. This is a proposed measurement plan. Hertz has not published that isolated comparison.

- [Hertz June 2026 Form 10-Q](https://www.sec.gov/Archives/edgar/data/47129/000165785326000046/htzz-20260630.htm) (6 Aug 2026).
- [Hertz Q2 2026 investor presentation](https://s204.q4cdn.com/384814028/files/doc_financials/2026/q2/Q2-2026-Investor-Presentation-8-5-2026-Final.pdf) (6 Aug 2026).
- [Q2 2026 management prepared remarks](https://newsroom.hertz.com/articles/article-details/hertz-global-holdings-inc-q2-2026-prepared-remarks/) (6 Aug 2026).
- [EpochX evidence ledger for the Palantir thesis](Research_Sources/palantir_evidence_v9.md) (28 Sep 2026).

### 34 / Higher utilization can earn less revenue

Intelligent allocation should maximize economic value. A system that fills more rental days through deep discounts can reduce revenue per available car, and relocation or variable costs can make the profit result worse. The chart shows that trade-off with hypothetical prices and utilization. Conversely, a slight fall in utilization can accompany a stronger contribution if the price and mix improve enough. The software needs a business objective with actual costs, service commitments and demand constraints. An isolated utilization headline is insufficient.

- [Integrating pricing and capacity decisions in car rental](https://www.sciencedirect.com/science/article/pii/S2214716018302707) (2018 / accessed 22 Sep 2026).
- [EpochX operating sensitivities and assumptions](Research_Sources/operating_models.md) (22 Sep 2026).
- [EpochX Palantir economics model and evidence tests](Research_Sources/palantir_economics_v9.md) (28 Sep 2026).

### 39 / Non-operating gains lifted Q2 GAAP income

Q2 GAAP net income of 64 million dollars includes 98 million of public-warrant fair-value gains, 51 million of net derivative gains and 64 million from selling non-vehicle capital assets. The full company reconciliation, including taxes and other adjustments, produces an adjusted net loss of 47 million dollars. Ordinary rental-car disposal economics remain in fleet depreciation. To judge Palantir, follow deployment-level operating improvements through recurring earnings and cash.

- [Hertz Q2 2026 earnings release](https://www.sec.gov/Archives/edgar/data/47129/000165785326000045/q22026earningsrelease.htm) (6 Aug 2026).
- [Hertz Q2 2026 investor presentation](https://s204.q4cdn.com/384814028/files/doc_financials/2026/q2/Q2-2026-Investor-Presentation-8-5-2026-Final.pdf) (6 Aug 2026).
- [EpochX evidence ledger for the Palantir thesis](Research_Sources/palantir_evidence_v9.md) (28 Sep 2026).

### 42 / Oro creates a possible new customer for fleet expertise

Hertz’s initial AV program supports Uber robotaxis using Lucid vehicles and Nuro technology. That can be a growth avenue for expertise in maintenance, charging and vehicle readiness. It is one announced program with several participants. The service contract count, committed fleet volume and pricing are undisclosed in the reviewed material. The next step in the thesis is operating proof: vehicles serviced, paid revenue and retained contribution. We have no basis to attribute that potential return to a Palantir deployment within Oro.

- [Oro and Uber fleet partnerships](https://newsroom.hertz.com/press-releases/press-release-details/hertz-and-uber-partner-to-power-autonomous-robotaxi-and-driver-led-fleet-operations/) (30 Apr 2026).
- [Q2 2026 management prepared remarks](https://newsroom.hertz.com/articles/article-details/hertz-global-holdings-inc-q2-2026-prepared-remarks/) (6 Aug 2026).
- [EpochX evidence ledger for the Palantir thesis](Research_Sources/palantir_evidence_v9.md) (28 Sep 2026).

### 43 / Autonomous cars still need physical turnaround

Removing the driver does not remove cleaning, charging, damage checks, tires or maintenance. Oro’s announced service scope includes depot staffing as well as the physical work. A shared schedule could reduce waiting between tasks, but output still depends on usable chargers, available power, cleaning bays, technicians and shifts. Measure ready vehicles per shift, queue time, paid labor and cost per serviced vehicle. More available road time produces value only when demand and contract terms let the operator earn from it. This is a potential coordination benefit. the reviewed sources do not measure a Palantir return within Oro.

- [Oro and Uber fleet partnerships](https://newsroom.hertz.com/press-releases/press-release-details/hertz-and-uber-partner-to-power-autonomous-robotaxi-and-driver-led-fleet-operations/) (30 Apr 2026).
- [Avis and Waymo fleet operations agreement](https://avisbudgetgroup.gcs-web.com/news-releases/news-release-details/avis-budget-group-announces-multi-year-strategic-partnership) (29 Jul 2025).
- [EpochX Palantir economics model and evidence tests](Research_Sources/palantir_economics_v9.md) (28 Sep 2026).

### 44 / Depot productivity can improve service margins

These invented service rates assume a customer owns and funds one thousand vehicles for a full year. A four-hundred-dollar monthly fee produces 4.8 million dollars of revenue. Direct cash costs of three million and fixed cash costs of six hundred thousand leave 1.2 million: a twenty-five-percent operating contribution, before program costs, depot capital spending, working-capital needs, interest and tax. A twenty-five-dollar monthly cost saving adds three hundred thousand before any fee reduction or new software cost. The contract decides who keeps it. Net cash and payback need the missing investment and contract inputs. this is neither a Hertz forecast nor a measured Palantir return.

- [EpochX autonomous servicing illustration](Research_Sources/autonomous_service_model.md) (22 Sep 2026).
- [EpochX Palantir economics model and evidence tests](Research_Sources/palantir_economics_v9.md) (28 Sep 2026).

### 45 / AV growth needs contracts with attractive economics

The number of technology companies named in an announcement is a poor measure of commercial value. Hertz has identified one initial AV program. contract count, committed volumes and fees remain undisclosed. A service-only contract differs from owning and financing the cars. Faster cleaning, charging and repairs can support more paid volume or lower labor cost, but the fee basis and reimbursement rules determine who keeps those gains. We need minimum volumes, depot funding, residual-value exposure, uptime penalties and termination terms before estimating a return. These are contract-evaluation questions, not disclosed Oro terms. We assign no value to a hypothetical Tesla contract.

- [Oro and Uber fleet partnerships](https://newsroom.hertz.com/press-releases/press-release-details/hertz-and-uber-partner-to-power-autonomous-robotaxi-and-driver-led-fleet-operations/) (30 Apr 2026).
- [Q2 2026 management prepared remarks](https://newsroom.hertz.com/articles/article-details/hertz-global-holdings-inc-q2-2026-prepared-remarks/) (6 Aug 2026).
- [Avis Budget Q2 results](https://www.avisbudgetgroup.com/home/news-and-media/press-release/avis-budget-group-reports-second-quarter-results-2026) (28 Jul 2026).
- [Tesla Cybercab availability and commercial inquiries](https://www.tesla.com/support/robotaxi/cybercab) (Accessed 24 Sep 2026).
- [EpochX Palantir economics model and evidence tests](Research_Sources/palantir_economics_v9.md) (28 Sep 2026).

### 46 / Geographic mix can change margins without an AI gain

The International segment earned a higher Q2 margin on a smaller revenue base. A shift in geographic mix could therefore change consolidated margin even if each operation performed identically. Seasonality, franchise mix, fleet economics and accounting definitions also matter. This is another reason the Palantir test needs comparable operations, not a single consolidated margin before and after implementation.

- [Hertz June 2026 Form 10-Q](https://www.sec.gov/Archives/edgar/data/47129/000165785326000046/htzz-20260630.htm) (6 Aug 2026).
- [EpochX evidence ledger for the Palantir thesis](Research_Sources/palantir_evidence_v9.md) (28 Sep 2026).

### 47 / Existing infrastructure is a starting advantage

Management says existing facilities and charging networks give Oro a useful starting point. That does not make every rental location an AV-ready depot: the published network includes licensees, and each operation needs suitable sites, partner acceptance, charging power and staffed service capacity. The reviewed sources do not provide an AV-ready depot count or quantify the incremental investment. Existing space also has an opportunity cost if it displaces rental work. The bull case is more profitable throughput from the infrastructure. the test is incremental cash after upgrades, working capital and program costs. Palantir-specific AV returns remain unmeasured in these disclosures.

- [Hertz Q2 2026 earnings call, Q&A](https://s204.q4cdn.com/384814028/files/doc_financials/2026/q2/Hertz-Global-Holdings-Inc-Q2-2026-Earnings-Call_Transcript_EditedCopy_20260806155605.pdf) (6 Aug 2026).
- [Oro and Uber fleet partnerships](https://newsroom.hertz.com/press-releases/press-release-details/hertz-and-uber-partner-to-power-autonomous-robotaxi-and-driver-led-fleet-operations/) (30 Apr 2026).
- [EpochX Palantir economics model and evidence tests](Research_Sources/palantir_economics_v9.md) (28 Sep 2026).

### 50 / Permission to share data does not establish revenue

Hertz’s notice includes vehicle data and advertising-related sale or sharing, subject to applicable rights and choices. California uses sharing specifically for cross-context behavioral advertising. that label alone does not establish cash revenue. Hertz’s disclosure does not identify a Palantir-enriched product, its buyer, price or net contribution. It also does not establish that Google receives identifiable vehicle trip histories. A commercial data service would need permitted uses, the necessary rights and a paying customer. Those questions are separate from documented operating workflows.

- [Hertz current rental privacy policy](https://www.hertz.com/rentacar/privacypolicy/index.jsp?targetPage=privacyPolicyView.jsp) (Updated 9 Mar 2026).
- [California AG: sale, sharing and sensitive information](https://oag.ca.gov/privacy/ccpa) (Updated 28 Aug 2026).
- [EpochX evidence ledger for the Palantir thesis](Research_Sources/palantir_evidence_v9.md) (28 Sep 2026).

### 51 / Palantir supplies software while data rights stay bounded

Palantir says it licenses software and does not sell personal data. Its published GCP template leaves content rights with the customer and limits Palantir’s use to providing and supporting services. That template is not Hertz’s disclosed agreement. Hertz’s rights against its software provider do not establish permission for every use of renter-linked information. Owning a fleet likewise does not settle the permitted uses of its telematics. A proposed advertising product needs its own rights and commercial case. operational returns can come from labor, allocation and turnaround.

- [Palantir proxy: customer data and business model](https://www.sec.gov/Archives/edgar/data/1321655/000132165526000019/pltr-20260423.htm) (23 Apr 2026).
- [Palantir published GCP terms, customer content](https://www.palantir.com/assets/xrfr7uokpv1b/2ajxcAAipDS2EAkFxZZvDa/ed556b05d8ef5d80b91a0e7bf8aff559/Palantir_Terms_and_Conditions_for_GCP.pdf) (Accessed 22 Sep 2026).
- [EpochX evidence ledger for the Palantir thesis](Research_Sources/palantir_evidence_v9.md) (28 Sep 2026).

### 52 / Travel offers are an optional revenue experiment

A contextual café offer could produce commissions or sponsorship revenue. The economics depend on a partner, customer acceptance, conversion, revenue sharing and delivery costs. For publishers subject to Google’s location policy, precise-location use requires express opt-in consent. That platform rule is not evidence of a Hertz–Palantir–Google agreement. The example is a possible product design. Measure incremental revenue per customer after costs and any effect on rental retention. It remains outside the base valuation.

- [Google publisher policy on precise location](https://support.google.com/publisherpolicies/answer/10437073?hl=en-GB) (Accessed 22 Sep 2026).
- [California AG: sale, sharing and sensitive information](https://oag.ca.gov/privacy/ccpa) (Updated 28 Aug 2026).
- [Hertz current rental privacy policy](https://www.hertz.com/rentacar/privacypolicy/index.jsp?targetPage=privacyPolicyView.jsp) (Updated 9 Mar 2026).
- [EpochX Palantir economics model and evidence tests](Research_Sources/palantir_economics_v9.md) (28 Sep 2026).

### 53 / The core AI thesis needs no data-sale premium

The thesis is stronger when it does not depend on an undisclosed data buyer. We can test whether intelligence improves labor, turnaround and fleet decisions using the existing business. A future data product would be additive only after Hertz establishes lawful rights, a paying customer and net contribution. We give it no separate premium in the valuation. This keeps optional opportunities from substituting for proof in the core operation.

- [Hertz current rental privacy policy](https://www.hertz.com/rentacar/privacypolicy/index.jsp?targetPage=privacyPolicyView.jsp) (Updated 9 Mar 2026).
- [Palantir proxy: customer data and business model](https://www.sec.gov/Archives/edgar/data/1321655/000132165526000019/pltr-20260423.htm) (23 Apr 2026).
- [EpochX evidence ledger for the Palantir thesis](Research_Sources/palantir_evidence_v9.md) (28 Sep 2026).

### 54 / Cash flow is the test after the operating improvement

The cash reconciliation prevents an operating story from becoming a financing illusion. Hertz reported positive adjusted free cash flow in Q2, with fleet investment and vehicle financing included in the calculation. Vehicle financing is legitimate to the business model, but it is not software productivity. A Palantir return should improve recurring cash generation after license, implementation, working-capital and capital costs. Separate asset-sale proceeds and financing movements before claiming the turnaround can fund itself.

- [Hertz Q2 2026 investor presentation](https://s204.q4cdn.com/384814028/files/doc_financials/2026/q2/Q2-2026-Investor-Presentation-8-5-2026-Final.pdf) (6 Aug 2026).
- [Hertz Q2 2026 earnings release](https://www.sec.gov/Archives/edgar/data/47129/000165785326000045/q22026earningsrelease.htm) (6 Aug 2026).
- [EpochX Palantir economics model and evidence tests](Research_Sources/palantir_economics_v9.md) (28 Sep 2026).

### 55 / Liquidity determines how long improvements have to work

At June 30, corporate liquidity was $984 million, combining cash and revolver availability. The separate $673 million of restricted cash cannot be treated as general spending money. Unused fleet commitments require eligible collateral. Hertz reported covenant compliance, including minimum liquidity at month end of $400 million in the second and third quarters and $500 million in the first and fourth. The rollout must fund its costs while preserving that cushion.

- [Hertz June 2026 Form 10-Q](https://www.sec.gov/Archives/edgar/data/47129/000165785326000046/htzz-20260630.htm) (6 Aug 2026).
- [Hertz Q2 2026 earnings release](https://www.sec.gov/Archives/edgar/data/47129/000165785326000045/q22026earningsrelease.htm) (6 Aug 2026).
- [EpochX Palantir economics model and evidence tests](Research_Sources/palantir_economics_v9.md) (28 Sep 2026).

### 58 / Better operations still face dilution risk

The $380 million covers the new first-lien issue. Other exchangeable notes remain outstanding. PIK interest grows its principal. Hertz may settle exchanges in cash, shares or both, subject to the terms. A shareholder model must update debt and any cash paid alongside the share count. The borrowed shares carry return obligations and are excluded from reported basic and diluted EPS. ATM issuance can add permanent shares. Measure the operating gain per share after the relevant financing claims.

- [Hertz June 2026 Form 10-Q](https://www.sec.gov/Archives/edgar/data/47129/000165785326000046/htzz-20260630.htm) (6 Aug 2026).
- [June first-lien exchangeable financing](https://www.sec.gov/Archives/edgar/data/1657853/000110465926078857/tm2619276d1_8k.htm) (29 Jun 2026).
- [EpochX Palantir economics model and evidence tests](Research_Sources/palantir_economics_v9.md) (28 Sep 2026).

### 59 / A rollout comparison would make the case stronger

A practical case study needs a credible comparison. Match locations with similar demand, vehicle mix and operating constraints. Compare their labor cost, turnaround and rental contribution around actual rollout dates, ideally using a phased rollout with a comparison group. Record adoption and exceptions so a nominal software installation does not count as effective use. This would help separate the deployment from market pricing or a younger fleet. It is our proposed test, not an experiment Hertz has publicly reported.

- [EpochX Palantir economics model and evidence tests](Research_Sources/palantir_economics_v9.md) (28 Sep 2026).
- [EpochX evidence ledger for the Palantir thesis](Research_Sources/palantir_evidence_v9.md) (28 Sep 2026).

### 60 / Q3 is the next public checkpoint

The next earnings report can test the direction of the business. Management guided Q3 Corporate EBITDA of 275 to 325 million dollars, transaction days up about one percent and monthly net DPU of 285 to 295 dollars. We will compare the reported numbers with those expectations and ask what caused the difference. Strong results support the broader turnaround. A quantified labor, turnaround or allocation contribution would strengthen the specific Palantir case.

- [Q2 2026 management prepared remarks](https://newsroom.hertz.com/articles/article-details/hertz-global-holdings-inc-q2-2026-prepared-remarks/) (6 Aug 2026).
- [EpochX evidence ledger for the Palantir thesis](Research_Sources/palantir_evidence_v9.md) (28 Sep 2026).

### 61 / The $1B ambition requires a large earnings bridge

The gap between the 2026 guidance midpoint and a one-billion-dollar earnings level is about 750 million dollars. Management’s comments distinguish the 2027 ambition and run-rate potential from earnings actually booked in a full year. A credible bridge would specify rental pricing, volume, operating productivity, depreciation, service growth and overhead, without overlap. Palantir may enable part of that bridge. Crediting it with the entire gap would outrun the disclosures.

- [Q2 2026 management prepared remarks](https://newsroom.hertz.com/articles/article-details/hertz-global-holdings-inc-q2-2026-prepared-remarks/) (6 Aug 2026).
- [EpochX Palantir economics model and evidence tests](Research_Sources/palantir_economics_v9.md) (28 Sep 2026).

### 63 / The equity sits below $5.67B of net corporate debt

The 28 September equity snapshot plus June net corporate principal gives a reference corporate value of about 6.21 billion dollars. This convention pairs Adjusted Corporate EBITDA with nominal corporate debt while retaining fleet depreciation and vehicle interest in operating economics. It is a reference bridge, with mixed dates and other claims still to consider. A recurring operating gain can matter to the small equity residual, but debt, valuation multiples and dilution determine the share outcome.

- [FMP quote, estimates and filing inventory](Research_Sources/provider_snapshot_2026-09-28.json) (28 Sep 2026).
- [Hertz June 2026 Form 10-Q](https://www.sec.gov/Archives/edgar/data/47129/000165785326000046/htzz-20260630.htm) (6 Aug 2026).
- [EpochX Palantir economics model and evidence tests](Research_Sources/palantir_economics_v9.md) (28 Sep 2026).

### 65 / A net-return model makes the assumptions visible

Here is a transparent sensitivity rather than a promised result. We annualize Q2 operating scale, assume sixty percent of extra rental revenue becomes contribution, and assume a twenty-five-million-dollar recurring program cost. Cost savings apply only to baseline rental days so that the extra-day contribution does not receive the same saving again. The middle illustration combines half a utilization point with twenty-five cents saved per baseline rental day and produces about fifty million dollars of net annual operating benefit. The small case loses money after program cost. Hertz has not disclosed that these effects, margins or costs are achieved. One-time implementation spending is additional.

- [Hertz Q2 2026 earnings release](https://www.sec.gov/Archives/edgar/data/47129/000165785326000045/q22026earningsrelease.htm) (6 Aug 2026).
- [EpochX Palantir economics model and evidence tests](Research_Sources/palantir_economics_v9.md) (28 Sep 2026).

### 66 / Adoption and ramp time can change the outcome

The middle scenario is worth roughly seventy-five million dollars before recurring program cost at full assumed coverage. If only thirty percent of that benefit is realized while the full twenty-five-million-dollar annual cost is due, the net is a small loss. Real adoption may differ from installed coverage, and benefits may arrive slowly. Integration costs, employee training and data problems can postpone payback. This is why deployment announcements and employee counts are early indicators rather than the final proof. The actual contract could have a different cost ramp.

- [EpochX Palantir economics model and evidence tests](Research_Sources/palantir_economics_v9.md) (28 Sep 2026).

### 68 / The evidence that would change our verdict

The next proof would compare similar operations using the new workflows with a credible baseline, accounting for vehicle age, demand, seasonality and recalls. It should show actual paid labor, turnaround and contribution changes, then subtract software, integration and ongoing costs. Customer service must hold up, and cash must remain after the fleet’s needs. Persistent gains would strengthen the thesis. If reported improvements disappear after these controls, or if costs absorb the benefit, we would weaken it. This makes the conclusion testable.

- [EpochX Palantir economics model and evidence tests](Research_Sources/palantir_economics_v9.md) (28 Sep 2026).
- [EpochX evidence ledger for the Palantir thesis](Research_Sources/palantir_evidence_v9.md) (28 Sep 2026).

### 69 / Consensus expects modest growth before full recovery

Consensus implies about 366 million dollars of additional revenue in 2027, roughly four percent growth. Average EPS remains slightly negative. Management expects full-year net income profitability and positive free cash flow in 2027. Those expectations differ in direction, but the provider snapshot does not identify the EPS adjustment basis, so this is not a like-for-like earnings reconciliation. For the AI thesis, track whether better availability produces more profitable rental days without proportional fleet growth. The forecast itself supplies no Palantir attribution.

- [FMP quote, estimates and filing inventory](Research_Sources/provider_snapshot_2026-09-28.json) (28 Sep 2026).
- [Q2 2026 management prepared remarks](https://newsroom.hertz.com/articles/article-details/hertz-global-holdings-inc-q2-2026-prepared-remarks/) (6 Aug 2026).
- [EpochX Palantir economics model and evidence tests](Research_Sources/palantir_economics_v9.md) (28 Sep 2026).

### 70 / Can Palantir improve Hertz’s economics?

Yes, Palantir can plausibly improve Hertz’s growth and unit economics. Better repair routing and vehicle placement can turn downtime into profitable rental days. Staffing decisions can reduce paid overtime and outside labor. Live workflows support that mechanism. The public record still does not isolate the added contribution or match it to the full program cost. Our test is sustained improvement versus comparable operations, after demand, pricing, fleet mix and seasonality. If that advantage disappears or program costs absorb it, the investment thesis weakens. If it persists and produces cash after fleet needs and financing, the case strengthens. Shareholders also need the benefit to survive dilution.

- [Hertz Q1 2026 earnings call, company transcript](https://s204.q4cdn.com/384814028/files/doc_financials/2026/q1/Q1-Earnings-Call-Transcript.pdf) (7 May 2026).
- [Q2 2026 management prepared remarks](https://newsroom.hertz.com/articles/article-details/hertz-global-holdings-inc-q2-2026-prepared-remarks/) (6 Aug 2026).
- [Hertz Q2 2026 earnings release](https://www.sec.gov/Archives/edgar/data/47129/000165785326000045/q22026earningsrelease.htm) (6 Aug 2026).
- [Hertz June 2026 Form 10-Q](https://www.sec.gov/Archives/edgar/data/47129/000165785326000046/htzz-20260630.htm) (6 Aug 2026).
- [EpochX Palantir economics model and evidence tests](Research_Sources/palantir_economics_v9.md) (28 Sep 2026).
- [EpochX evidence ledger for the Palantir thesis](Research_Sources/palantir_evidence_v9.md) (28 Sep 2026).