Arron Jones/iStock via Getty Images
In the first quarter of 2026, Deep Sail Capital Partners (the "Fund") returned -17.8% net of fees while averaging 84% net long exposure. Since the end of the first quarter the fund has returned 20.4% net of fees and now sits flat year to date. Please consult your individual capital account statements for your individual net returns.
Performance Summary Net Returns Deep Sail Capital Vs Benchmark Q1 - 26 1 Yr 3 Yr Annualized Since Inception Annualized Q1 - 26 1 Yr 3 Yr Annualized Since Inception Annualized Deep Sail Capital Partners LP -17.8% 11.0% 12.9% 9.5% Russell 2000 0.9% 25.7% 14.8% 8.9% -18.7% -14.7% -1.9% 0.7% Russell Mid Cap Growth Index -1.4% 15.4% 19.4% 11.4% -16.3% -4.4% -6.5% -1.9%
In the first quarter, the fund underperformed both of its benchmarks, the Russell 2000 Mid Cap Growth Index and the Russell 2000 Index. The result was somewhat atypical, as performance tracked broadly in line with benchmarks for most of the quarter before several positions experienced material drawdowns late in March. This follows a long stretch of seven consecutive quarters of positive net returns for the fund.
Performance in the quarter was driven by weakness in the long portfolio, partially offset by gains in the short portfolio. The drawdown in long positions was concentrated toward quarter-end and coincided with broader market volatility surrounding the Iran war selloff. Notably, these losses have completely reversed in the second quarter as of the market close on April 23rd.
While the short portfolio generated positive returns, as would be expected in a declining market, the magnitude of the contribution was below my expectations. Following strong performance in the prior quarter and given the significant declines of many lower-quality, long-duration equities (a source of the fund's short exposure) in Q4 2025, the short book entered the quarter with significant gains from Q4. As a result, the short return did not materially offset the losses generated on the long side as has historically been the case.
My quarterly commentaries have been very skewed toward the negatives over the last 12 months, but this time I will shift a bit to cover why AI and the associated data center ramps might just be the only thing that matters to the market trajectory for the next few years.
The global economy has entered a period of structural transformation that transcends the typical ebbs and flows of the business cycle with the adoption of LLMs, AI agents, and soon physical AI. The unprecedented capex cycle required to build the physical infrastructure of the machine-intelligence age is beyond massive. For investors, the data center ramp is no longer merely a "story" but rather a fundamental economic driver upon which all future economic activity is being built.
The scale of the current investment in AI infrastructure is perhaps the most visible indicator of its economic dominance. The Mag7 have shifted from being capital-light software providers to some of the most capital-intensive private enterprises in history. By 2026, the combined annual capex of the five largest U.S. hyperscalers—Amazon, Microsoft, Alphabet, Meta, and Oracle—is projected to reach between $660 billion and $690 billion. This represents a near-doubling of spending from 2025 levels and is more than four times the annual capital investment of the entire publicly traded U.S. energy sector.
Source 2025 Est. 2026 Forecast Long-Term Projection (through 2030) Dell'Oro Group $489.5 Billion ~$1 Trillion (Global) $1.7 Trillion (Annual) by 2030 Bloomberg Intelligence $450 Billion $750 Billion (Top 14 firms) $3 Trillion (Aggregate) by 2029 Goldman Sachs / Analysts $410 Billion $700 Billion (Hyperscalers) $5–$7 Trillion (Cumulative) by 2030 JPMorgan $342 Billion (Hyperscalers) ~$500 Billion (Reinvestment) $5 Trillion (Cumulative) by 2030
This surge is not a temporary spike but a reflection of a fundamental shift in the lifecycle of computing assets, or as Jensen likes to say, " accelerated computing." This level of spending is fundamentally rewriting the risk-reward profile of the equity market. While these companies are technically categorized as "Technology," their balance sheets and operational hurdles now mirror those of heavy industry like the railroads and oil refineries before them. The sheer magnitude of the capex can't be funded from the Mag 7 cash flows alone. Significant debt will be required as the need for AI compute continues to expand. That means that the stability of the credit markets is now inextricably linked to the continued demand for AI compute; if the demand or use cases falter, we might hit a brick wall economically.
On top of the availability of the capital to do all this is the question of "Are there enough materials and energy?" This lack of components and availability of energy (both from an energy production and energy grid capacity standpoint) has become the limiting factor in the AI buildout.
While the capex story is clear for at least the next few years, there are still massive questions that are not answered around post-AI employment, productivity gains, the reshoring of high-value production (like silicon, rare earths, drones, and nuclear materials), and how/when physical AI will have its “ChatGPT” moment.
As we look across the next few years, the reality is that we are witnessing the birth of a new kind of civilization, one built on a substrate of silicon with massive intelligence limited by compute. Will the AI Supercycle eventually crash and cause a recession? At this point I would say it is hard to tell exactly if or when that will happen, as we are still barely getting into the real AI Supercycle.
The impact of AI is just so large relative to any other market forces at this time, that it may truly be all that matters to drive the market up or down.
Q1 Average Long Exposure 128.0% Contribution (gross) -21.6%
Best Performing Q1 Contribution Worst Performing Q1 Contribution Kraken Robotics (KRKNF) 4.1% Hydreight Technologies (HYDTF) -5.0% Vertiv Holdings (VRT) 2.6% SANUWAVE Health -4.2% Lumentum Holdings (LITE) 1.0% ClearPoint Neuro (CLPT) -2.6%
In the first quarter, the long portfolio underperformed both of our benchmarks, the Russell 2000 Index and the Russell Midcap Growth Index, by a significant margin. There were a few large detractors at the end of the quarter, including On Holding (ONON), Credo Technology (CRDO), Amplitech Group, and Sanuwave Holdings. Amplitech and Sanuwave were both down on negative reactions to their Q4 earnings reports, which I believe were both unwarranted. Credo Technologies was down over 50% from its ATH by the end of March on the narrative that “Cooper is dead” in data centers. That narrative has since changed in Q2 with the market beginning to realize that Credo’s ZeroFlap optics platform is a real second leg for the company. Credo went on to make a transformative acquisition of Dust Photonics just last week, which not only makes Credo a viable optics supplier but also creates a vertically integrated platform for connectivity in the data center space.
In Q1, the fund closed one position and opened five new positions. The new positions in Q4 included AmpliTech, MDA Space, and Micron (MU).
After several years of watching Constellation Software and following the company, in Q1 we finally had the opportunity to buy shares at a reasonable valuation. We have long owned Topicus Software (which is a spinout from Constellation Software), but only recently did the valuation and forward IRR of Constellation Software land it in a viable range for the fund to purchase shares. I believe the AI narrative around VMS software is largely wrong, and AI will provide a nice tailwind for the industry over the next five years.
AmpliTech Group (AMPG) operates a high-quality, diversified business model centered on designing, developing, and manufacturing advanced RF and microwave signal-processing components and systems. The company serves multiple end markets, which include satellite communications, 5G/6G telecom, quantum computing, defense, aerospace, and space. It operates through five divisions: AmpliTech Inc., Specialty Microwave, Spectrum Semiconductor Materials, AmpliTech Group Microwave Design Center, and the fast-growing AmpliTech Group 5G Division.
AmpliTech's 5G ORAN product is the material driver of the current revenue growth that the company has seen. ORAN (or O-RAN) stands for "Open Radio Access Network." ORAN, led by the O-RAN Alliance, is an agreed-upon framework that has been adopted by telecom operators for 5G networks, mainly in non-US markets. 5G ORAN is an open interface for building a 5G network. This allows telecom operators to mix and match vendors, reducing their dependence on proprietary, single-vendor systems while improving the supply chain and driving down costs for the telecom operators.
A key point about true 5G networks vs. 4G or 4G LTE is that due to the coverage efficiency, bandwidth used, and latency requirements, 5G networks will have materially more endpoints (RAN) devices required than 4G networks. This means that ORAN makers will be in high demand as the replacement cycle from 4G to 5G continues, which is a massive tailwind for AmpliTech for the next 3 years.
The company is currently forecasting at least $50m in 2026 revenue (100% growth over 2025), which will be driven by two large LOI's that they have secured. The company announced in 2025 they have received an LOI for $78 million in ORAN radio systems with an Asian telco and a $40m LOI for a North American MNO. These LOIs are both just getting ramped up and will support the 2026 revenue growth.
AmpliTech's business segmentation structure creates natural cross-selling opportunities and a somewhat resilient revenue base, although their 5G ORAN business is lumpy. The legacy components business (low-noise amplifiers (LNAs) and low-noise blocks (LNBs)) provides stable, high-margin recurring demand from satellite, defense, and test/measurement customers. Their 5G systems business, which was materially scaled up in 2025, delivers a scalable, high-volume platform revenue opportunity within the 5G Open RAN (ORAN) space.
Fiscal 2025 revenue reached a record $25.2 million, up approximately 165% year-over-year from $9.5 million in 2024. Growth was driven by strong demand for LNA/LNB products, expanded 5G product lines, and a rebound in the Spectrum division's Asian markets. Gross profit rose to $6.0 million (from $3.5 million), but gross margin compressed to 23.9% (from 36.7%).
This compression in gross margin was largely a one-off event tied to three factors: (1) elevated legal and professional fees associated with capital-raising activities and compliance during the year's rapid scaling; (2) one-time costs from the registered direct and rights offerings that strengthened the balance sheet; and (3) the deliberate early-stage ramp-up of carrier-grade 5G ORAN radio production, including competitive pricing to secure Tier-1 operator relationships and initial manufacturing scale-up inefficiencies. Management has explicitly guided that these headwinds are transitory and gross margins should return to the ~40% range in the back half of 2026 and be trending toward the 30–50% telecom industry benchmark as fixed costs are absorbed and pricing normalizes on established contracts. This margin improvement should be steadily driven as the company shifts from market-penetration pricing and low-volume pilot deployments to recurring, high-volume orders with better gross margins.
The model is further strengthened by vertical integration with design and manufacturing in the U.S., patented front-end and LNA technologies, and a transition from pure component supplier to full systems provider. AmpliTech's products are 100% American-made, which may be a key selling point with US telcom providers due to onshoring efforts in the future.
Amplitech is led by its founder, chairman, president, and CEO, Fawad Maqbool. Fawad has 30 years of microwave engineering experience with dual BSEE degrees in microwave and biomedical engineering. He has scaled the business from nothing to $50m in revenue forecasted for 2026. Maqbool founded the company in 2002 and has steered it from a bootstrapped startup to a NASDAQ-listed growth platform while maintaining hands-on oversight of RF innovation.
Outside of Fawad, AmpliTech's leadership team combines deep technical expertise with proven operational discipline. Complementing him is COO Jorge Flores, who joined in 2021 with over 30 years of operations and program management experience across complex manufacturing environments.
Management's execution has been strong within the last few years. In 2025 they delivered 165% revenue growth while simultaneously certifying the industry's first U.S.-manufactured 64T64R CAT-B massive MIMO ORAN radio. This 5g ORAN product allowed them to secure the $78m LOI with a Tier-1 operator. While their 4Q results showed a gross margin decline due to the 5G ramp, they continue to be transparent and unwavering about the long-term vision for the company. As they approach a positive operating profit (likely later in 2026), they have proactively raised capital on favorable terms to fund the 5G ramp once the business inflects.
Guidance has been consistently met or raised, and the team's willingness to accept short-term gross-margin dilution for strategic design wins demonstrates a sophisticated understanding of carrier procurement cycles. AmpliTech's management is well positioned to convert early ORAN leadership into multi-year, high-margin revenue.
My biggest issue with management is their IR has not been great. They had not scheduled an earnings call for Q4 results and released the results, and the stock sold off 30% after the results. Then, in what seemed kind of reactionary, they scheduled a call to discuss the results. They also do not do a very good job at getting back to investors who have questions. This is not an idea killer to me if the business decisions seem correct and they are focusing on execution, but I do think they could improve IR.
AmpliTech sits at the epicenter of the secular shift to Open RAN in 5G infrastructure. Traditional 4G RAN was monolithic and proprietary, locking operators into single-vendor ecosystems with high costs and limited flexibility. ORAN disaggregates the radio, baseband, and software layers, enabling multi-vendor interoperability, which is why Amplitech has succeeded in selling into much larger MNO telecom companies. ORAN is an interesting product within 5G that is more flexible and adoptable than its 4G counterpart.
Specifically, AmpliTech's 5G ORAN product offers ~30% cost savings, faster innovation, and domestic supply chain security, which is a critical sales advantage over Chinese competition due to geopolitical concerns.
As I mentioned above, the key investment insight about the 5G opportunity ahead of AmpliTech is understanding the 5G infrastructure rollout versus the previous 4G infrastructure. A key differentiator is the dramatically higher density of endpoints required for 5G versus 4G. 5G demands gigabit-per-second speeds, ultra-low latency, and higher bandwidths. These requirements force denser network topologies as higher-frequency bands, which have shorter propagation distances. The shorter distances of propagation force telecom companies to deploy many more small cells and massive MIMO radios per square kilometer (5-20x depending on the buildings and structures nearby). This density is required for applications that require low latency and fast two-way signal strength, like autonomous vehicles, AR/VR, drones, and other new technologies.
Industry forecasts project the global 5G infrastructure market growing at a 22%+ CAGR by 2035. ORAN will be a major subset of that growth. AmpliTech's U.S.-made, OTIC-tested, O-RAN-compliant, and interoperability-proven device positions the company as a rare domestic pure-play beneficiary.
Current analyst consensus (one covering firm) forecasts align with management's reaffirmed FY2026 revenue guidance of at least $50 million (roughly 2x FY2025), which to me is an easy forecast to hit with just the existing two LOIs. But the company recently discussed that they are in the process of RFPs and other discussions for additional LOIs on their 5G ORAN products. I expect we'll hear more about further LOIs by mid-summer, which could materially bolster the 2025 revenue and support another sizable revenue growth in 2026.
The Legacy RF and Spectrum divisions should continue growing mid-teens, providing a stable base while 5G systems become the dominant, higher-margin contributor. Gross margins normalize to ~40%+ by H2 2026 and sustain or expand thereafter, driving robust free-cash-flow generation and potential for margin expansion through operating leverage.
AmpliTech reported what was viewed by investors as a bad quarter in Q4 2025. Although revenue grows 111% y-o-y, it sequentially declined for the last two quarters. On top of that, the gross margins went down to 24%. Management did a fairly good job at explaining the Q4 numbers, although the conference call was a week after the earnings announcement.
The 2026 revenue guidance of “at least $50m” in revenue is still intact, and management confirmed that on the call. There will likely be some lumpy quarters as they ship orders against both of their LOIs. Investors never like a lumpy business, so I suspect that is why its current enterprise value is just $46m.
Below is my base case forecast for the company based on the existing 5g ORAN LOIs and the other business lines.
Quarter Actuals Q2 2023 Actuals Q3 2023 Actuals Q4 2023 Actuals Q1 2024 Actuals Q2 2024 Actuals Q3 2024 Actuals Q4 2024 Actuals Q1 2025 Actuals Q2 2025 Actuals Q3 2025 Actuals Q4 2025 Forecast 2026 Forecast 2027 Forecast 2028 Total Revenue ($m) 4.1 3.4 4 2.3 2.5 2.8 1.9 3.6 11 6.1 4.5 50.0 65.0 84.5 Revenue Growth (YoY) -39% -18% -53% 57% 340% 118% 137% 98% 30% 30% Gross Profit ($m) 1.9 1.5 2 0.9 1.1 1.3 0.2 1.2 0.9 3 1 18 26 33.8 Gross Margin (%) 46% 44% 50% 39% 44% 46% 11% 33% 8% 49% 22% 36% 40% 40% SG&A 1.7 2 1.8 2 2.2 1.9 1.8 2.3 2.1 2.5 3.7 11 12.5 14.5 R&D 0.7 0.4 0.7 0.3 0.5 0.6 2.3 0.7 0.7 0.7 0.6 2.5 3 4 EBITDA ($m) -0.4 -0.8 -0.4 -1.3 -1.4 -1 -3.7 -1.8 -1.7 0.2 -2.8 4.5 10.5 15.3 EBITDA Margin -10% -24% -10% -57% -56% 25% -195% -50% -15% 3% -62% 9% 16% 18% EV/EBITDA Multiple 10.3 4.4 3.0
Note that the company completed a rights offering in January 2026, in which they sold 2.2m shares at $4 which came with two rights: Series A right with $5 strike and expiry in July 2026, Series B right with $6 strike and expiry in November 2026. This would be about a 18% share dilution if the rights were executed, which suggests some resistance around $5-6/share. What is interesting about this rights offering is, why did management execute this with such short expiry times? $5 in July 2026 and then $6 in November 2026. It is a bit of foreshadowing from management, about what they see on the horizon for 2026.
There is a bull case where they receive new LOIs for 5G ORAN, as management has suggested on the recent call. If that is the case, their revenue growth might be significantly higher than 30% in 2027 and 2028.
The only analyst covering AmpliTech (Maxim Group) revised down their price target from $9 to $7 a share after the Q4 2025 earnings report on margin concerns. Currently Maxim Group is forecasting 2026 EBITDA at $11m, which would put the valuation at 3.5x 2026 EBITDA. I am a bit suspicious of that figure, as I think it assumes full-year gross profit at 40%, instead of a blended full-year number closer to 35%. Even so, my 2026 EV/EBITDA multiple is just 10.3x, and then it drops to 4.4 for 2027. This is more than reasonable if the company can continue its path to EBITDA and cash flow breakeven while growing the topline at +30%. I believe shares are worth north of $6 (210% upside) once the company can prove out gross margins and scale their 5G ORAN business.
The worst seems to be behind AmpliTech in terms of the valuation drop after the recent earnings. Q1 has historically not been a great quarter for the business, but they expect to begin to start executing against the LOIs they have in hand, so I would expect to see a solid Q1 report, which will put investors at ease with the story.
Q1 Average Short Exposure 37.6% Contribution (gross) 4.5%
Best Performing Q1 Contribution Worst Performing Q1 Contribution Robinhood (HOOD) 1.0% New Era Energy & Digital -1.3% Oklo (OKLO) 0.9% Fastly (FSLY) -0.6% D-Wave Quantum (QBTS) 0.9% Clear Secure (YOU) -0.4%
The short portfolio underperformed my expectations in the first quarter. The short portfolio had a stellar return in Q4 2025, which generated a 7.4% contribution in the face of the Russell 2000 being up 2.2%. Coming off the back of that, the mini-bubble in low-quality, long-duration stocks had largely already subsided. The core drivers to the short book in 2025 were quantum-related stocks, which by early 2026 I had either closed or they saw the downside volatility decline significantly. While the short portfolio contributed positively to the overall returns of the fund in Q1, there was less opportunity than we saw in much of 2025 for “easy” short setups both fundamentally and technically.
I expect this opportunity set to materially improve in the second quarter, as we have already seen quantum stocks take off and several short squeezes occur, suggesting ample opportunities in the short portfolio.
Long Portfolio Industry/Segment % of NAV Short Portfolio Industry/Segment % of NAV Kraken Robotics Defense Technology 14.3% Cava (CAVA) Quick Serve Restaurants -3.8% Mercado Libre (MELI) Ecommerce 11.8% Carvana (CVNA) Ecommerce -3.1% Shelly Group IoT Devices 10.0% ARK Blockchain & Fintech Innovation ETF ETF -3.1% Hydreight Technologies Healthcare 7.8% ARK Next Generation Internet ETF ETF -3.0% Arista Networks (ANET) Networking 7.6% Clear Secure Software -2.7%
At the end of the fourth quarter the fund held 24 long positions and 20 short positions. The fund ended the year with an exposure of 129% long and 41% short or an 88% net long exposure.
Sincerely,
Sean
Disclaimer Deep Sail Capital LLC (“Deep Sail Capital”) is an investment adviser to funds that are in the business of buying and selling securities and other financial instruments. This information is provided for informational purposes only and does not constitute investment advice or an offer or solicitation to buy or sell an interest in a private fund or any other security. An offer or solicitation of an investment in a private fund will only be made to accredited investors pursuant to a private placement memorandum and associate subscription documents. Past performance is no guarantee of future results. “Deep Sail Capital Partners” returns in this document are shown as net returns or gross returns where stated. Historical net returns assume a 1.5% and 15% management and performance fee, respectively. For Net Returns of fees and expenditures figures please reach out to the fund manager at the email info@deepsailcapital.com. “Deep Sail Capital LLC” name was changed on April 7 th 2022 from the previous name “Organon Capital LLC”. “Deep Sail Capital Partners LP” name was changed on April 6 th 2022 from the previous name “Westropp Funds LP”. * - “Strategy Since Inception” refers to the Strategy inception date of July 2016. Deep Sail Capital Partners LP’s predecessor incubator fund, “Westropp Funds LP” pivoted from a Value Investment style to a Growth at a Reasonable Price (GARP) style fund on that date. For more details on this transition or the calculation behind the “Strategy Since Inception” returns please reach out to the fund manager at info@deepsailcapital.com.
Editor's Note: The summary bullets for this article were chosen by Seeking Alpha editors.
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。