DroneShield Limited

The Business

DroneShield sells portable radio-frequency detection systems and AI-based drone classification software to military and government customers across 47 nations, supplemented by a growing commercial security business (roughly 15% of revenue). Its core differentiator is a dismounted, battlefield-portable detection capability that large defence primes have not replicated. Revenue has historically been hardware-led, but a subscription layer tied to its 4,100 installed devices is emerging as the more durable, higher-margin part of the model.

Recent Performance

Revenue grew 74% in the first half of FY26, building on a base that itself had grown strongly the prior year, so the growth rate is decelerating even as the dollar numbers rise. Profitability moved the other way: EBITDA (earnings before interest, tax, depreciation and amortisation) swung from a positive $8.0m to a loss of $12.4m as employee costs grew 166%, well ahead of revenue. The market has largely looked through this, keeping the shares near record highs on the strength of the demand story.

Outlook

We expect revenue growth to decelerate from around 25% toward 15% over the next several years as the business matures, with subscription attach playing a growing role alongside hardware volume. The critical swing factor is cost discipline: our forecast assumes employee costs fall from 45% of revenue toward 30% by the end of the decade, lifting EBITDA margin from negative territory in FY26 toward the mid-teens by FY29. That single assumption does more to determine the investment outcome than any other variable in our model, in either direction.

Key Risks

Employee costs need to fall from 45% to 30% of revenue for the profitability thesis to hold, a trajectory never demonstrated by this management team, with the CEO in the role less than six months. A sustained cost ratio above the high-30s in coming half-year results would be a clear warning sign that the leverage story is not materialising on schedule. Separately, a Ukraine ceasefire or de-escalation in the Middle East could cut the new order pipeline by an estimated 25-40%. DroneShield's footprint across 47 nations and its commercial security segment provide some insulation, but the bulk of recent growth has been conflict-adjacent government spending that could soften quickly if geopolitical conditions ease.

What to Watch

The thesis-defining event is the FY26 full-year result in February 2027, which will show whether the employee cost ratio is genuinely declining or stuck near current levels.

• Feb 2027 FY26 full-year results — the first real test of whether cost leverage is materialising.
• Oct 2026 2H26 trading update — an early read on margin trajectory ahead of the full-year print.
• Ongoing ASIC investigation — resolution would remove an overhang currently sitting on the stock, with the downside risk skewed toward an adverse finding.

Company Description

DroneShield designs and sells counter-drone detection and defeat systems built around radio-frequency sensing and AI-based classification. Hardware sales, mostly ruggedised handheld and fixed-site detection units, still generate the bulk of revenue. A software subscription layer, tied to the installed base of roughly 4,100 devices, is the fastest-growing piece and carries materially higher margins. Government and military customers dominate the business (around 85% of revenue), with commercial applications, airports, events, critical infrastructure, making up the remainder. The company operates from Australia with an expanding European manufacturing footprint aimed at meeting EU defence procurement preferences.

Where the Growth Is

Subscription revenue grew 229% year-on-year to $11.5m in the first half of FY26, still only 9% of total revenue but scaling fast off a small base. Each detection device sold creates a recurring software relationship, and if that mix reaches 15-20% of revenue by FY28, it would materially lift the margin expansion built into our forecasts.

This is the part of the business the market should be watching most closely, because it is what would convert DroneShield from a hardware cycle into a more durable, recurring-revenue franchise.

Competitive Position

DroneShield's advantage rests on a portable, dismounted detection capability combined with an AI classification engine that no large defence prime currently matches in the field. The installed base of 4,100 devices creates switching costs: once a military customer standardises on DroneShield's detection network, replacing it means re-training and re-certifying an entirely new system. Regulatory approvals, including inclusion on US and NATO procurement panels, add a further 12-18 month barrier for new entrants. That said, this is a narrow advantage rather than a deep one. Well-funded competitors, including Anduril and the major US primes, are investing heavily in counter-drone capability, and we estimate the current competitive window at roughly three to five years before that lead narrows materially.

Management & Capital Discipline

Management has framed the 47% headcount increase as deliberate investment ahead of a larger revenue base, but employee costs grew 166% against 74% revenue growth in the first half of FY26, a gap that has yet to reverse. A $3m settlement with a former key management person and $8.7m of "business disruption" costs add to the sense of an organisation still finding its footing. CEO Angus Bean has been in the role less than six months, following a governance incident that triggered an ongoing regulatory investigation. Revenue delivery has been strong and largely credible; cost discipline remains entirely unproven.

Financial Position

DroneShield holds $180m in cash with zero debt, giving it 24-36 months of runway even if losses persist at recent rates. This balance sheet strength is the single biggest reason the business can absorb a slower-than-hoped margin recovery without needing to raise capital in our base case. It does not, however, change the underlying economics: the company still needs to convert that cash into a profitable operating model within the next two to three years.