Spectra Systems Corporation (SPSY.L) - Deep Dive Research Report

Prepared 15 June 2026. Listing venue: London Stock Exchange AIM (unrestricted line SPSY; Regulation S line SPSC). Reporting currency: US dollars. Functional reporting cadence: half-yearly. All segment figures expressed as mix percentages; absolute revenue, margin and earnings figures are deliberately excluded per mandate.

Section 1: What the company does

Spectra Systems makes the invisible ink that central banks use to tell a real banknote from a fake one, and it makes the machines that read that ink at high speed. That is the core of the business. When a used banknote is fed into a sorting machine at a central bank's cash centre, the machine has a fraction of a second to decide whether the note is genuine before it is recirculated or destroyed. Spectra supplies two halves of that decision: a covert chemical "taggant" that is built into the note during manufacture, and a sensor that detects the unique optical signature of that taggant as the note flies past. This is called Level III authentication, meaning it is machine-only and secret, readable by neither the public (Level I, the features you can see) nor a shop clerk with a UV lamp (Level II), but only by the issuing authority itself.

The company was founded by Dr. Nabil Lawandy, a former tenured professor of engineering and physics at Brown University (1981 to 1999), who incorporated the present entity in July 1996 (originally Spectra Science Corporation, renamed Spectra Systems in 2001). Lawandy has run it ever since and still owns roughly 8% of the equity. The business grew out of his academic work in spectroscopy and optical materials, and that scientific origin still defines it: the moat is chemistry and physics that took decades to develop and that, by design, the customer is contractually and operationally committed to keeping secret. Spectra says its features are embedded in roughly 45 billion banknotes in circulation worldwide, and its Level III solutions are used by around 20 central banks, including two G7 central banks (spsy.com, businessabc.net).

The value proposition is precision plus secrecy plus speed. A central bank cannot simply switch authentication chemistry the way a factory switches a supplier of bolts. Once a covert feature is designed into a banknote series, it is locked in for the life of that series, often seven to ten years, and the sensors that read it are calibrated to that exact chemistry. That creates a relationship with very long tails and very high switching costs, which is why a company this small can sit inside the cash infrastructure of G7 economies.

Over the last few years Spectra has broadened from "the taggant and sensor company" into a small group with four activities: banknote authentication (the engine), security printing (postage and tax stamps, acquired through Cartor in December 2023), gaming/lottery integrity software, and brand protection. The single most important recent event is a sensor manufacturing contract worth about $39.6 million signed in July 2024 with an existing central bank customer, which is the reason 2025 was the most profitable year in the company's history and the reason the share moved on results day (proactiveinvestors.com, accesswire.com).

"The combination of the sensor revenues expected with the execution of the manufacturing contract, the expected sensor maintenance contract in 2025, the increased opportunities for optical materials with smartphone verification, and downstream polymer substrate sales, give the board confidence for significant increases in revenues, cash generation, and continued long-term growth." - Dr. Nabil Lawandy, FY2024 audited results, 31 March 2025

Section 2: Business segments

Spectra reports three operating activities. Below is each in turn, with the FY2024 revenue mix used to convey relative scale (FY2024 is the most recent year with a clean, fully disclosed segment split; in FY2025 the sensor surge pushed the authentication share materially higher).

2.1 Physical & Software Authentication Systems (~63% of FY2024 revenue)

This is the original Spectra and the profit engine. It sells covert taggant materials to central banks, the high-speed and offline quality-control sensors that read those materials, and increasingly the funded development and manufacture of sensors themselves. It also houses the smartphone authentication technology (TruNote for banknotes, TruBrand/TruStamp for brands and tax stamps) and the optical/covert materials used in passports and laminates.

The core capability is materials science that is hard to reverse-engineer and is, by contract, never published. Spectra invents a luminescent or otherwise optically active compound, formulates it so it survives the brutal physical life of a banknote (folding, laundering, ageing), embeds it at parts-per-million concentrations, and then builds a detector sensitive enough to find that signal at sorting speeds of tens of notes per second with effectively zero false positives. Getting a feature qualified into a live currency series can take years of central-bank testing. That is the barrier, and it is why this segment carries the group's margin.

This segment exists as the strategic centre because everything else is downstream of it. The 2024 sensor manufacturing contract (about $39.6m, recognised on a cost-accounting basis from Q1 2025 through Q4 2027 with a trailing tail to roughly 2029) and the associated five-year sensor maintenance contract (about $6.7m, 2026 to 2030) both live here, and both convert a historically lumpy materials-order business into something with multi-year visibility (accesswire.com, marketscreener.com). Within the segment, gaming/lottery software (the Secure Transactions Group) is sometimes carved out separately, see 2.3.

Competitively, this is where Spectra meets Authentix (its Jewel Level III covert system and 550-plus central-bank sorter sensors) and the authentication unit formerly owned by De La Rue and now inside Crane NXT. Spectra wins on installed-base lock-in with specific central banks and on the breadth of pairing materials with proprietary sensors; it is exposed by being far smaller than Crane NXT and by depending heavily on a handful of central-bank relationships.

2.2 Security Printing - Cartor (~33% of FY2024 revenue)

Cartor Security Printers was acquired in December 2023 for up to £10.5 million (£5.5m cash plus 947,494 shares at £2.11, with a contingent £2.0m tied to qualifying the Fusion polymer substrate) (investegate.co.uk, cartor.com). Cartor prints postage stamps for more than 180 postal administrations worldwide and produces hybrid stamps, tax/excise stamps, vouchers and coupons, with manufacturing in the UK and (historically) France.

The strategic logic was vertical integration. Spectra's Fusion product is a polymer banknote substrate with a Level III covert feature embedded in its core. To sell substrate to central banks, you need a printing/converting capability; Cartor provides the industrial base to produce substrate "for over 2.5 billion notes per annum" and to consolidate margin in-house rather than paying a third-party printer (proactiveinvestors.com). It also gives Spectra a credible bid vehicle for government tax-stamp contracts.

This segment is the lowest-margin part of the group (FY2024 adjusted EBITDA of roughly $1.6m on a third of revenue) and it is the one being actively restructured. Management has initiated workforce reductions and a (repeatedly delayed) closure of the French operation, targeting the security-printing reshaping by Q2 2026, with the stated aim of "repeatable and significant profitability even in the absence of Fusion sales." It is best understood as a strategic option: a low-return printing business today that becomes valuable if and when Fusion qualifies into a live currency. The most important recent win here is the HMRC UK vaping duty stamps contract (see Section 3).

2.3 Gaming Software - Secure Transactions Group (~4% of FY2024 revenue)

The smallest activity. It supplies the Premier ICS internal control system, software that monitors lottery and sports-wagering operations for fraud and reconciles the integrity of transactions for gaming operators and state lotteries. It was roughly breakeven in FY2024 (a small EBITDA loss, with management noting staffing costs ran ahead of plan) and then delivered record revenue and profitability in FY2025 (investegate.co.uk, uk.advfn.com). It is a high-recurring, niche software annuity that gives the group a third leg with different cyclicality from currency and printing.

Section 3: Products and business detail

Banknote authentication core

• Covert taggant materials. The foundational product: parts-per-million optical compounds embedded in banknote paper or polymer, readable only by the central bank's own sorters. Sold as recurring materials orders to the installed base of central banks.
• High-speed and QC sensors. The detectors that read the taggants on sorting machines, plus offline quality-control vision systems used during banknote production. The 2024 manufacturing contract is for building sensors for an existing central-bank customer; first units shipped in 2025, with full delivery targeted for H2 2026.
• Fusion. A transparent, machine-readable polymer banknote substrate with the Level III feature embedded in its core, designed so that even a counterfeiter with access to commercial polymer film cannot reproduce it. As of FY2025, Fusion was selected by a major banknote printer as part of a response to a central-bank tender expected to be awarded in H2 2026, and a Middle Eastern central bank received a further 10,000-sheet sample for testing (proactiveinvestors.com / FY2025 reporting).
• TruNote. A smartphone-based banknote authentication system pairing a physical covert material with server-side algorithms, so authentication can happen outside the central-bank sorter.
• Aeris / BDS-1000. Banknote decontamination and SARS-2 disinfection hardware. These are legacy/opportunistic products from the COVID period and are not the growth story.

Brand protection

• SpectraGuard, TruBrand, TruStamp. Covert physical-plus-software authentication for labels, holograms and product packaging, verifiable by smartphone. The notable live deployment is a major Chinese tobacco manufacturer running TruBrand taggant smartphone authentication on roughly 6 to 10 million units per year, with additional customer trials primarily in China (spsy.com).

Security printing (Cartor)

• Conventional and hybrid postage stamps, supplied to 180-plus postal administrations; tax/excise stamps; vouchers and coupons; and passport laminates that can carry Spectra's covert optical materials. Recent wins: a $4m four-year hybrid-stamp contract, and the headline UK HMRC vaping duty stamps contract.

The HMRC vaping duty stamps contract (Jan/Feb 2026). Cartor, partnering with tax-stamp specialist SICPA, was selected as the supplier for the new UK vaping duty stamps scheme. The total contract is around £32m excluding VAT (£38.4m including VAT) over five years with a one-year option, and Spectra expects to receive roughly 15% of the total five-year revenue (tipranks.com, investegate.co.uk). This is strategically important well beyond its size: it validates Cartor as a government-grade excise-stamp supplier and creates a reference for further postal/tax-stamp tenders.

Manufacturing and geography. Materials and sensors are made under secure conditions (a requirement for central-bank work). Printing is in the UK, with the French facility being wound down by Q2 2026. The customer base is global: central banks across multiple regions including two G7 banks, postal authorities worldwide, a Chinese brand-protection deployment, and UK government (HMRC).

Section 4: Customers

Central banks are the defining customer. The buying decision sits with currency/technology directorates inside the bank, and the criteria are reliability (false-accept and false-reject rates that must be near-zero at sorting speed), absolute secrecy, and proven longevity across a currency series. Sales cycles are measured in years because a feature must be tested, qualified, designed into a new note series, and then procured. Once embedded, the relationship is extremely sticky: the central bank's own sorters are calibrated to Spectra's chemistry, and changing it means re-tooling the cash infrastructure and waiting for the next series. This is the source of the installed-base lock-in and the recurring materials orders, and it is why a roughly 20-central-bank customer list, including two G7 banks, is a durable asset rather than a marketing line.

Postal and government authorities buy through Cartor. Postal administrations (180-plus) buy stamps on recurring annual programmes; tax authorities such as HMRC buy excise-stamp schemes through competitive tenders, typically multi-year. These are more contestable than central-bank work but still relationship- and reference-driven.

Brand owners buy brand protection. The anchor is a major Chinese tobacco manufacturer running TruBrand at scale; the rest is trials. The decision-maker is the brand's anti-counterfeit/supply-chain function, and the criterion is whether covert smartphone verification meaningfully reduces grey-market and counterfeit losses.

State lotteries and gaming operators buy Premier ICS as a fraud/integrity control layer, a recurring software relationship with operational lock-in once integrated into a lottery's control environment.

Concentration and contract structure. Revenue predictability has historically been the weak point: central-bank materials orders are lumpy, which is exactly why management has prized the move to multi-year structures. The 2024 sensor manufacturing contract (~$39.6m, recognised 2025 to 2027 with a tail to ~2029) and the linked five-year maintenance contract (~$6.7m to 2030) convert a single large customer relationship into visible recurring cash. The flip side is concentration: a meaningful slug of near-term growth depends on one central-bank customer executing the sensor programme on schedule.

Section 5: Competitive landscape

This is a small, secretive oligopoly. Only a handful of firms in the world are trusted by central banks to supply covert Level III features and the sensors to read them, and the customer's own secrecy requirements make the market opaque and very hard to enter. The barriers are not capital; they are accreditation, decades of materials know-how, a track record of zero-failure performance in live currency, and the security clearances and facilities to do classified work.

The competitive map has just shifted materially. De La Rue, the historic British currency company, sold its Authentication division to Crane NXT in May 2025, and De La Rue itself was taken private by Atlas Holdings in July 2025. The net effect is that Crane NXT (which already owned Crane Currency) is now the consolidated heavyweight in central-bank authentication, while the De La Rue banknote-printing/substrate business is now a private entity (securingindustry.com, printweek.com).

Where Spectra wins: it is the incumbent inside specific central banks, including G7 banks, with chemistry and sensors that those banks have already qualified and locked in. That incumbency is nearly impossible to dislodge mid-series. Where Spectra is exposed: it is a fraction of the size of Crane NXT, it depends on a handful of relationships, and on the substrate/printing side (Fusion, Cartor) it competes against far larger and better-capitalised players (De La Rue's Safeguard substrate and ASSURE covert, Giesecke+Devrient, SICPA's track-and-trace). On tax stamps, SICPA is simultaneously a competitor and Spectra's partner on the HMRC vaping scheme.

Market-cap figures are a peer-size reference only (Crane NXT ~$2.2bn as of June 2026, per companiesmarketcap.com); the privately held competitors do not disclose comparable figures.

Section 6: Industry

Demand for Spectra's core product is driven by the simple fact that physical cash is still everywhere and still being counterfeited. There are roughly 180 currency-issuing authorities globally, and each one needs to authenticate notes at scale as they recirculate. Even as digital payments grow, the absolute volume of banknotes in circulation has continued to rise in most major economies, which keeps demand for authentication features and sorter sensors steady. The Crane NXT currency business pointing to an upbeat 2026 currency outlook is a useful read-through on overall central-bank demand (finviz.com).

The second industry driver is excise and tax-stamp enforcement. Governments are expanding duty-stamp regimes to capture tax on tobacco, alcohol and now vaping products, supported by frameworks like the WHO FCTC track-and-trace protocol and the EU Tobacco Products Directive. The UK vaping duty stamps scheme that Cartor won is a concrete example of new government-mandated demand. The third driver is brand anti-counterfeiting, a large but fragmented market where Spectra plays a niche covert-smartphone role. The fourth is gaming/lottery integrity, a small regulated software niche.

Spectra sits at the high-value, low-volume end of the supply chain: it does not print most of the world's money, it supplies the secret ingredient and the reader. The market for Level III covert features is tiny in unit terms but high-margin and defended by secrecy and qualification cycles. Regulation is central to the whole industry: central-bank procurement, security clearances, and government excise mandates all gate who can play. Cyclicality is low for the authentication core (cash and tax enforcement are non-discretionary) and somewhat higher for printing, which depends on government and postal budgets and on the lumpy timing of central-bank tenders.

Section 7: Growth triggers

All items below are drawn from the five reporting periods (FY2023, H1 2024, FY2024, H1 2025, FY2025). Spectra reports half-yearly, so the "concalls" are the results announcements.

• Sensor manufacturing contract delivery ramping through 2026. First sensors delivered in 2025 ($5.7m payment); cash from the sensor contract is expected to rise by a further ~$26.6m in 2026 as remaining sensors are built and delivered (H1 2025 results, 29 Sep 2025; FY2025 results, 30 Mar 2026). Repeated across H1 2025 and FY2025.

  "Unrestricted cash was expected to rapidly increase beginning in H2 2025 with the first payment for sensors, and in 2026 ... cash from the sensor contract is expected to increase by an additional $26,600k." - H1 2025 results, 29 Sep 2025

• Sensor maintenance contract running 2026 to 2030 (~$6.7m). A five-year annuity attached to the sensor programme, providing recurring revenue visibility (flagged FY2024, confirmed FY2025).
• Fusion polymer substrate central-bank tender, award expected H2 2026. Fusion was selected by a major banknote printer as part of its response to a central-bank tender; a Middle Eastern central bank received a further 10,000-sheet test sample (FY2025 results, 30 Mar 2026). Repeated theme since FY2024, now advanced to a live tender.
• HMRC vaping duty stamps scheme (Cartor + SICPA), five years from 2026. ~£32m total scheme value, Spectra share ~15%, validating Cartor as a government excise-stamp supplier (announced Jan/Feb 2026, executed and referenced into FY2025 reporting).
• Security-printing restructuring completing Q2 2026. French facility closure and workforce reductions intended to deliver "repeatable and significant profitability even in the absence of Fusion sales" (FY2025 results, 30 Mar 2026).
• Smartphone tax-stamp adoption (TruBrand/TruStamp). Successful smartphone authentication trials with a major tax-stamp partner and Middle East governments, with potential multi-billion-unit adoption (FY2024 and FY2025 results). Repeated.
• Hybrid stamp expansion to additional postal authorities, building on the $4m four-year hybrid-stamp win (FY2025 results, 30 Mar 2026).

Section 8: Key risks

Single-customer / single-contract concentration on the sensor programme. The step-change in 2025 profitability and the 2026 cash inflow both hinge on one central-bank customer executing the ~$39.6m sensor manufacturing contract on schedule. If delivery slips, if the customer changes scope, or if the trailing maintenance contract is not renewed, the near-term cash and earnings trajectory weakens materially. Management itself frames 2026 cash around this contract completing, which is exactly the dependency.

Fusion has been "nearly qualified" for years without a commercial currency win. The polymer substrate has been described as in central-bank qualification since the Cartor acquisition rationale in 2023, and the contingent £2m of Cartor consideration was tied to qualifying Fusion. As of FY2025 it is "selected as part of a response to a tender" with an award only "expected" in H2 2026. This is a high-value option that has repeatedly been close but not converted; if the tender is lost or delayed again, a central pillar of the printing-integration thesis is undermined.

Security printing is structurally low-margin and mid-restructuring. Cartor earned thin EBITDA in FY2024, and management is closing the French operation and cutting headcount, a delayed and execution-sensitive process targeted for Q2 2026. Restructurings of acquired industrial assets routinely cost more and take longer than planned, and Cartor only becomes attractive if Fusion and tax-stamp wins materialise.

"Modeling predicts that restructuring will generate repeatable and significant profitability even in the absence of Fusion sales." - FY2025 results, 30 Mar 2026

That statement is a forecast, not a result; the risk is that the printing business stays a drag if both the restructuring and the Fusion/tax-stamp pipeline underwhelm.

Lumpiness and visibility. Outside the sensor contract, central-bank materials orders and printing tenders are episodic. A clean year can be followed by a gap year if no large order lands, which makes the half-yearly numbers volatile and the share sensitive to single announcements.

Key-person and small-company governance. The founder-CEO has run the company for nearly 30 years, embodies the scientific IP, and holds ~8% of equity; the company is small, AIM-listed, and US-incorporated with a complex dual-line share structure. Concentration of knowledge and authority in one individual is a genuine continuity risk.

Scale disadvantage post-consolidation. With Crane NXT now combining Crane Currency and the former De La Rue Authentication unit, Spectra competes for central-bank attention against a far larger, integrated rival. In a procurement environment that values supplier durability, being sub-scale is a standing exposure.

Section 9: Walk the talk

Five reporting periods used: FY2023 audited (Apr 2024), H1 2024 (Sep 2024), FY2024 audited (31 Mar 2025), H1 2025 (29 Sep 2025), FY2025 audited (30 Mar 2026). The most recent is within 90 days of today.

The throughline is a management team that has been directionally honest and, on the biggest call, delivered. The story really starts with FY2023, the first set of numbers carrying the newly acquired Cartor, where management set up the integration as the route to consolidating Fusion substrate production and lifting margins. Through H1 2024 and especially the July 2024 announcement of the ~$39.6m sensor manufacturing contract, the narrative shifted decisively to "a large, multi-year, visible revenue event is coming." That was a specific, datable, checkable promise.

At FY2024 (31 Mar 2025) management put a stake in the ground:

"The Board therefore believes that the Company is on track to achieve record earnings in 2025." - FY2024 results, 31 March 2025

That was delivered. FY2025 was reported as the most profitable year in the company's history, with adjusted EBITDA up roughly 83% and the dividend raised. The sensor contract recognition that management described at H1 2025 (recognising contract revenue on a cost-accounting basis, first batch in H2 2025, completion H2 2026) tracked the guidance through the year. On the single largest commitment, the company said what it would do and then did it.

Where the record is weaker is on Fusion qualification, which has been a recurring "almost there" since the Cartor deal. At FY2024 management listed full central-bank qualification and substrate tender invitations as near-term opportunities; a year later, at FY2025, the status is "selected as part of a response to a tender" with an award only expected in H2 2026. This is not a broken promise (they never claimed it was won), but it is a target that has been carried forward repeatedly, and a skeptical reader should treat the H2 2026 award as hoped-for, not banked. Similarly, the Cartor restructuring and French closure have been "delayed", management has been candid about the delay rather than hiding it, but it is a commitment still in progress.

On capital returns the team has been measured and consistent: dividend held flat through the lean years (FY2023, FY2024) and raised only once the sensor cash was visible (FY2025), plus a modest first buyback in late 2025. That is conservative and matches the conservative tone of the guidance.

Net assessment: this is management that does what it says on the financially material items and is honest about slippage on the optionality (Fusion, restructuring). They lean conservative rather than promotional. The one watch-item is the multi-year habit of carrying the Fusion qualification forward.

Section 10: Shareholder friendliness index

Dividends. Spectra paid $0.116 per share for FY2023 and held it flat at $0.116 for FY2024, then raised it to $0.136 for FY2025 (paid mid-2026), an increase of roughly 17% (investegate.co.uk FY2024, proactiveinvestors.com FY2025). The pattern is deliberate: the board held the payout flat through the years before the sensor cash arrived and lifted it only once delivery was visible, which reads as disciplined rather than stingy. The dividend has been a consistent feature for years, signalling a genuine capital-return culture rather than a one-off.

Buybacks and dilution. For the recent (last ~90 day) window I cross-checked the company's own RNS: Spectra launched its first share buyback programme on 24 November 2025, an irrevocable arrangement with Zeus Capital to repurchase up to US$350,000 of shares at a maximum price of £1.35, with transactions in own shares executed late November and early December 2025 (investegate.co.uk buyback, investing.com). That is a small, almost symbolic programme (around 20,000 shares ended up in treasury). For the longer three-year window, web searches surface no prior buyback programmes; capital return historically came entirely through the dividend. Against that, the share count has been slowly growing from continuous employee/director option exercises at a 25p strike (issued share capital rose from roughly 47.8m unrestricted shares in late 2025 to roughly 48.4m by mid-2026), only marginally offset by the tiny buyback. So shares are being modestly created, not retired.

Verdict: Returns Capital (mild) - a long-standing and now-growing dividend backed by real cash, but with persistent low-level option dilution and only a token buyback, so it returns capital chiefly through the dividend rather than by shrinking the share count.

Section 11: Insider activities

Listing venue is UK AIM, so the primary source is RNS "Director/PDMR Shareholding", "Holding(s) in Company", "Exercise of Options" and "Transaction in Own Shares" announcements via the London Stock Exchange / Investegate. No MoatMap database block was injected for this venue; the data below is drawn from RNS filings located via Investegate.

Recent transactions (most recent first):

Buys - read the signal. I could not identify any open-market purchase by a director or officer in the last 12 months. The insider-side activity is option exercises at a 25p strike, with the shares well in the money against a market price around £1.20, so these are economic gains being realised by employees and option holders, not conviction open-market buying. There is therefore no bullish cluster-buy signal to flag.

Sells - work out the why. There are no large disclosed open-market director sales in the window either. The recurring "Holding(s) in Company" notices are substantial-shareholder threshold crossings by institutions (the register is held mostly by institutions and individuals; named holders include Raymond James ~11%, Close Asset Management ~9.8%, Charles Stanley ~9.8%, Nicholas Slater ~8.9%, and founder-CEO Lawandy ~8.1% as of late 2025, per Yahoo/Simply Wall St ownership data), reflecting normal institutional position changes rather than insider conviction signals.

Net assessment. Insider activity over the last 12 months is dominated by employee option exercises (mildly dilutive) and a small company buyback, with no open-market director buys and no material open-market director sells. The founder-CEO retains a meaningful ~8% stake, which aligns him with shareholders, but there is no fresh open-market buying to read as a conviction signal. The honest read is neutral: no red flag, but also no bullish insider tell. Detailed share counts and values for several PDMR/option lines were not fully retrievable within the search budget; the directional picture (option exercises plus a token buyback, no open-market director purchases) is reliable.

Section 12: Scenarios

Bull case. The sensor manufacturing contract delivers on schedule through 2026, the ~$26.6m of contract cash lands, restricted balances unwind, and the five-year maintenance annuity locks in recurring revenue to 2030. The H2 2026 central-bank tender that selected Fusion is won, finally converting a decade of substrate development into a live currency programme and proving the Cartor vertical-integration thesis, with Fusion capable of supplying substrate for billions of notes a year. The restructured, French-closed Cartor turns durably profitable, the HMRC vaping stamps scheme runs cleanly and becomes a reference that wins further excise-stamp tenders across postal authorities, and the China-anchored smartphone brand/tax-stamp technology scales to multi-billion-unit adoption. Gaming software keeps compounding off its newly profitable base. In this world Spectra has transformed from a lumpy materials supplier into a diversified security group with multiple multi-year contracts, visible cash, and a rising dividend, with the founder still aligned at ~8%.

Base case. Management delivers roughly what it has guided. The sensor contract completes on time and drives the 2026 cash inflow, the dividend continues its measured rise, and the maintenance annuity provides a steady recurring layer. Cartor's restructuring lands close to plan and the printing business stops being a drag, but Fusion qualification converts slowly or only partially, so the substrate upside is a real option rather than a banked win. Brand protection and gaming stay small but contribute. Results remain somewhat lumpy half to half, the share trades on contract announcements, and the company looks like a cash-generative, dividend-paying niche security business with embedded optionality that the market prices cautiously until Fusion or the next big central-bank order converts.

Bear case. The sensor contract slips or its scope shrinks, deferring the 2026 cash and exposing how concentrated the recent step-change was on one customer. Fusion loses or fails to convert the H2 2026 tender, after years of being "almost qualified", and the substrate thesis that justified the Cartor acquisition deflates. The Cartor restructuring costs more and takes longer than planned, leaving security printing a persistent low-margin weight rather than the integrated margin engine management modelled. With Crane NXT now combining Crane Currency and the former De La Rue Authentication unit, a far larger rival crowds Spectra in central-bank procurement and the next large materials order does not arrive on time, so revenue reverts to its historically episodic pattern. Layer on key-person risk around a founder-CEO who is the scientific core of the company, and a small AIM stock with thin liquidity re-rates down on a single disappointing half.

A note on completeness: Spectra reports half-yearly, so the five "reporting periods" used throughout are FY2023, H1 2024, FY2024, H1 2025 and FY2025 results, the most recent (FY2025 audited, 30 March 2026) falling within 90 days of today. SemiAnalysis, Stratechery and MBI Deep Dives were searched and none has covered Spectra Systems, so the Further Reading section is correctly omitted.

Sources:

• Spectra Systems FY2025 results - Proactive Investors
• Spectra Systems FY2025 results - ADVFN
• Spectra Systems FY2024 audited results - Investegate
• Sensor manufacturing contract - Accesswire
• Sensor maintenance contract - MarketScreener
• Cartor acquisition - Cartor
• HMRC vaping duty stamps - TipRanks
• HMRC vaping stamps successful bid - Investegate
• Share buyback programme - Investegate
• Buyback launch - Investing.com
• RNS announcements list - Investegate
• Company products/about - Spectra Systems
• Brand protection - Spectra Systems
• Company history - businessabc
• De La Rue / Atlas Holdings takeover - SecuringIndustry
• De La Rue acquisition completes - Printweek
• Crane NXT market cap - CompaniesMarketCap
• Authentix banknote sensors
• Spectra Systems ownership - Yahoo Finance / Simply Wall St

Would you like me to save this as a .md file in the working directory?