Fund Performance

Royce Small-Cap Opportunity Fund advanced 33.9% for the year-to-date period ended 6/30/26, outperforming its small-cap benchmark, the Russell 2000 Value Index, which was up 23.0% for the same period. The Fund also beat the small-cap Russell 2000 Index, which rose 22.6% for the same period. Longer-term absolute and relative results were equally impressive. The Fund outpaced both indexes for the 1-, 3-, 5-, 10-, 15-, 20-, 25-year, and since inception (11/19/96) periods ended 6/30/26. The Fund’s average annual total return since inception was 12.7%.

What Worked… And What Didn’t

Eight of the Fund’s nine equity sectors contributed positively to performance in the first half of 2026. Information Technology and Industrials made by far the biggest positive impact, followed by Energy. Consumer Staples was the lone detractor. At the industry level, semiconductors & semiconductor equipment (Information Technology), aerospace & defense (Industrials), and energy equipment & services (Energy) contributed most for the year-to-date period, while it services (Information Technology), health care equipment & supplies (Health Care), and health care technology (Health Care) were the largest detractors.

The Fund’s top contributor at the position level was Ultra Clean Holdings, which develops and supplies critical subsystems, components, parts, and related services to the semiconductor industry. While near-term wafer fab equipment shipments remain constrained by cleanroom growth, the capacity cycle underway is providing unprecedented demand visibility in an environment of increased deposition (which adds material onto a chip wafer) and etch (which selectively removes material) share of capital spending and the buildout of new computing loads. We continue to believe Ultra Clean is a differentiated business model with attractive and durable reinvestment opportunities.

Optical semiconductor manufacturer Applied Optoelectronics has been a beneficiary of the buildout of AI Infrastructure thanks to its fast growing demand for optical components. The company has also benefited from its commitment to rapidly raising the technology curve by introducing more advanced high-end optical components, which is allowing it to gain market share with leading AI data center customers. We sold our stake in the portfolio as it reached our valuation target.

Semiconductor test and measurement company Cohu is well positioned to benefit from testing demand associated with High Performance Computing that underlies AI infrastructure while also generating significant revenues from automotive and industrial applications. These last two areas appear to be recovering from something of a cyclical downturn. Growth has been reaccelerating, which we expect to lead to increasing purchases of test and handling equipment. Finally, Cohu has a high recurring revenue mix (test equipment uses disposable contactors in its process), and a pristine balance sheet that we think positions it well for future growth.

Ichor Holdings designs and manufactures gas and chemical delivery systems that are critical components in semiconductor manufacturing. As a result, Ichor has been benefiting from the rapid buildout of AI-related infrastructure. Company management has also been focused on driving vertical integration into its manufacturing processes to improve Ichor’s gross margin structure. While this effort entailed some typical growing pains, the effort now seems to be on track, driving upside to earnings expectations.

Penguin Solutions provides memory and AI infrastructure via infrastructure software, advanced memory, and compute systems. We began purchasing shares in April 2024 when we recognized that the company’s products and solutions appeared to be well suited to the demand related to the buildout of AI initiatives. In the first half of 2026, the market began to re-underwrite the company from a cyclical specialty memory/LED/infrastructure hardware business into a more direct beneficiary of the AI infrastructure buildout. The initial catalyst was not a single dynamic quarter, but a material improvement in visibility for what the company was achieving as Penguin became a “memory wall” and AI factory infrastructure story. Management recently highlighted growing demand from enterprises, governments, and neocloud providers, and in 2Q26 added five AI/HPC (High-Performance Computing) customers, including a Tier One financial institution deploying the company’s MemoryAI CXL-based KV cache server. That positioned Penguin less as a commodity memory supplier and more as a picks-and-shovels enabler of inference-heavy AI workloads, where memory capacity, latency, and cluster management become bottlenecks. External validation helped the re-rating. In June, Penguin was named Dell Technologies’ 2026 Global Alliances Americas AI Partner of the Year, and later became an NVIDIA AI Factory Specialized Partner, validating its role in designing, building, deploying, and managing full-stack AI factory infrastructure. The stock now carries a higher bar—and we sold our stake in the Fund during 2Q26 when it reached our valuation target.

The Fund’s top detractor at the position level was Kyndryl Holdings, which was spun out of IBM in 2021. We started buying shares in September 2022, attracted to the company’s turnaround efforts. Our thesis proved correct at the time as the shares rose from just over $10 to a high of over $43 over two-and-a-half years. Unfortunately, however, the market lost confidence in what had once been a clean turnaround story. The original thesis rested on improving contract discipline, margin expansion, growth in Kyndryl Consult and hyperscaler alliances, and a credible path to more than $1 billion of free cash flow by fiscal 2028. That narrative changed in February, when the company disclosed that its Audit Committee was reviewing cash management practices, related disclosures, and internal controls. The company also said it expected to report material weaknesses across multiple periods and that its prior internal-control assessment, along with the related accounting opinion should no longer be relied upon.

The market reaction was severe because the disclosure hit the most sensitive part of the story: Free cash flow quality and management credibility. Kyndryl had already been valued on a forward recovery path rather than current revenue growth, so questions around adjusted free cash flow, working-capital management and disclosure controls undermined the bridge to fiscal 2028 targets. The same release also included leadership changes, with the CFO and General Counsel departing. Operational results were mixed but did not offset the governance shock. Management cut fiscal 2026 guidance to a 2%–3% constant-currency revenue decline and lowered free cash flow guidance to $325–$375 million versus the prior investor-day framework that contemplated a much steeper cash-flow ramp. Knowing it would take several quarters to regain market confidence, we exited the position in February, although Kyndryl still remains a viable candidate for the portfolio once the dust settles.

OptimizeRx Corporation, a healthcare technology company that helps pharmaceutical and biotechnology companies market their products more effectively to both healthcare providers (HCPs) and patients, was another situation where initial success was reversed. We began buying shares in October 2024 as we felt that the company’s growth opportunity was not being recognized. The shares rose from our initial purchase price of just over $7 a share to a high of over $21 before beginning to sell off, which led to OptimizeRx being a meaningful detractor during the first half of 2026 as the market reassessed the durability of the company’s continued growth path. At the end of 2025, the story appeared to be inflecting: full-year 2025 revenue grew 19%, adjusted EBITDA (earnings before interest, taxes, depreciation & amortization) more than doubled to $24.3 million, and operating cash flow was growing. However, in early 2026 management flagged increased end market volatility tied to Most Favored Nation drug-pricing uncertainty, with pharmaceutical customers taking a more cautious approach to discretionary marketing budgets and contract duration. In May, management reduced their outlook.

The issue was not simply a weak quarter, but a loss of visibility. Pharma customers shortened campaign commitments, delayed program timing, and reduced scope as they digested policy and budget uncertainty. That matters for OptimizeRx because the business has meaningful exposure to campaign timing and large enterprise customer behavior. There were, however, positive offsets beneath the headline revenue reset. Adjusted EBITDA increased to $3.3 million in 1Q26 despite lower revenue, GAAP net losses narrowed, management launched cost actions expected to produce $3 million of annualized savings, and the company refinanced its term loan, reducing borrowing costs by roughly 625 basis points and expected annual interest expense by approximately $1.5 million. OptimizeRx is also opening its proprietary EHR (electronic health record) network to demand-side platforms, which management believes can improve the utilization of under-monetized inventory and create a programmatic channel over time. The company remains profitable on an adjusted EBITDA basis and is operating with improved cost discipline. Management continues to point to pharma budget normalization in the second half of 2026 along with DSP (Demand-Side Platform)/programmatic access to restore top-line growth. We added to our position in the first half.

EPAM Systems is a global technology services and consulting company that helps businesses build software, modernize their technology, and undergo digital transformations in industries such as financial services, healthcare, retail, manufacturing, media, and travel. It has a low-debt balance sheet, generates strong free cash flow, and in our view is a healthy company, though it has recently endured slower growth. The shares sold off after management reduced their 2026 revenue outlook, citing greater macro uncertainty and slower client decision-making. The market’s concern appears to be that EPAM is caught in an awkward transition. On the one hand, the company has real AI credibility: AI-native revenue exceeded $125 million in 1Q26, up nearly 20% sequentially, and EPAM announced a multi-year partnership with Anthropic to help enterprises deploy Claude-based applications and AI transformation programs. On the other hand, AI is also pressuring the traditional services model. Management acknowledged that clients are shifting budgets toward AI and automation, away from some legacy digital platforms and e-commerce build-outs. While this creates opportunity in vendor consolidation and AI transformation, it also introduces risk around pricing, utilization, project scope, and how much human labor customers need in future software delivery. Given management’s outlook for slower growth, we exited our position. However, given the company’s healthy financials we continue to monitor how it is executing.

Artivion produces medical equipment targeting cardiovascular procedures. Its most recent quarter was generally mixed. Management guided 2026 earnings lower, which is always challenging for higher multiple med-tech companies. The shortfall was primarily related to slower than expected regulatory approvals of a few new products, as well as the upfront costs associated with the rollout of other new products. Due to the fact that these are typical risks for smaller-cap med-tech companies, we expect them to be overcome in the intermediate term. Meanwhile, we believe Artivion has an exciting pipeline of new products, which we expect to drive growth for the foreseeable future.

Tandem Diabetes Care is a medical device company that develops and commercializes insulin delivery systems for diabetics, primarily those with Type 1 diabetes and insulin-dependent Type 2 diabetes. Its core business is making insulin pumps and integrating them with continuous glucose monitors (CGMs) to automate insulin dosing. We began buying shares just under a year ago as the stock was just above $10, which proved to be well timed as the shares rose to over $28 in February of 2026 before selling off, despite several pieces of good fundamental news. The stock initially rallied as the company crossed $1 billion of annual sales, reported record quarterly revenue of $290 million, posted a record 58% gross margin, and generated positive free cash flow. The problem was that the 2026 outlook embedded a transition year: management guided to only $1.065–$1.085 billion of sales, despite expecting U.S. pump shipments to grow 10%–11%, because the shift to a new U.S. pay-as-you-go pharmacy model is expected to create a $70-$80 million revenue headwind. International direct-distribution changes add another roughly $15 million headwind.

The selloff reflected a classic “better long term, worse near term” problem. We were aware that the new PayGo model could ultimately improve affordability, expand access, and make revenue more recurring, but it also reduces upfront revenue. Investors therefore began to discount the reported revenue air pocket, the timing of pharmacy adoption, and the risk that gross-margin improvement would not translate quickly enough into durable profitability. Results for 1Q26 were not bad but also did not fully resolve those concerns. Tandem reported record first-quarter pump shipments, sales of $247 million, U.S. sales growth of 7%, gross margin of 55%, positive adjusted EBITDA, and positive free cash flow, while reaffirming full-year guidance. The stock briefly reacted positively, but that move faded quickly; the shares rose more than 10% after the release and then fell -14.5% the next day. We continue to see value in Tandem’s product franchise, particularly Control-IQ+, Mobi, the move toward pharmacy reimbursement, and future Mobi Tubeless optionality. The company also received FDA clearance for Control-IQ+ use in pregnancy complicated by Type 1 diabetes, a differentiated clinical label expansion. But the burden of proof has shifted. For the stock to recover, Tandem needs to show that PayGo is expanding the market rather than merely deferring revenue, that pump shipment growth is converting into recurring economics, and that Mobi Tubeless can narrow the competitive gap with Insulet. Until then, the market is likely to treat it less like a diabetes-tech growth asset and more like a small-cap med-tech transition story with reimbursement, competitive, and timing risk. Believing that the market will come around to the former view, we built our stake in the first half at what we thought were attractive prices.

The Fund’s advantage over the Russell 2000 Value was mostly due to sector allocation decisions, though stock selection also contributed positively to first-half results. At the sector level, the Fund’s much larger weighting in Information Technology had the biggest positive effect by a substantial margin, followed by stock selection and, to a lesser extent, a higher weighting in Industrials and stock selection in Materials. Conversely, stock selection in Health Care and Consumer Staples detracted most from relative performance.

Current Positioning and Outlook

The ongoing physical infrastructure buildout to power the AI revolution is creating significant winners among semiconductor stocks, optical component makers, and certain industrial power related stocks that we own in the portfolio. Several other portfolio companies are benefiting from employing AI tools in their businesses and are realizing productivity gains as a result. We believe the physical buildout of the AI infrastructure is a multi-year structural phenomenon and continue to own positions in many of the aforementioned areas. And while we believe we are in the early innings of this phenomenon, we have also begun reallocating some of the outsized gains to areas of the market that we see as offering significant value.

With energy prices retreating as the supply shock created by the Iran war abates, we foresee lower inflationary pressures and increasing strength in the U.S. economy. This should improve growth and margin prospects for the cyclical parts of the market, such as Consumer Discretionary companies, agriculture-oriented industries, transportation, and industrials with oil-derived inputs. We have been allocating capital to several of these sectors to take advantage of the broadening out of the market. In addition, the longer-term impacts from the closure of the Strait of Hormuz should result in a diversification of energy supply sources that will benefit the North American oil and gas industry, both onshore and offshore. We have been adding capital in this area as well. Another area of the market where valuations have been highly dislocated and disjointed from long-term fundamentals is software, driven by a perceived threat to business models from AI. We believe there are pockets of the software industry that will actually be beneficiaries of AI, with the possibility of expansion in their addressable market and an increased need for their services. As a result, we have begun to take positions in cybersecurity and niche vertical software stocks whose valuations do not appear to reflect the durability of their growth.

Average Annual Total Returns Through 06/30/26 (%)

Annual Operating Expenses: 1.24

¹ Not annualized.

Important Performance, Expense and Disclosure Information