Opportuna Newsletter #5

This month marks another milestone, as Opportuna SPV-1 went live this week, with plans to secure its first investment by year-end. As always, our newsletter follows the same structured format: 

• highlights of the month - Klarna and ServiceTitan IPO, xAI and Anthropic rounds, and reports worth reading; 

• a chart of the month - Unicorns account for 63% of US VC value;

• insights on current topics - Klarna; and

• analysis of a long-term issue - the rise of “Shadow LLMs'“ within IT organizations. 

This edition remains focused on the convergence of private and public tech investing.

In the “Chart of the Month” and “Current Topics”, we cover secondary proceeds of tech IPOs and provide a deep dive into Klarna’s valuation. Our “Long-Term Theme” examines the rise of large language models and their impact on enterprise cybersecurity.

You can see past editions here. Anybody can sign up following this link.

🚀 Evangelization Work and Unicorn List

Robin was recently featured in a Bloomberg article discussing the growing appeal of venture secondaries, highlighting Opportuna’s approach.

We continue to educate the ecosystem of GPs, BAs, and founders about our solution. Chris and I attended the EPFL Investor Day, where we had meaningful conversations. If you are a founder, executive, BA, or GP looking for liquidity solutions, please do reach out for a confidential chat. For those seeking quality exposure to private markets, with faster liquidity and downside protection, you can book a call with us here.

Finally, we are keen to hear from you about which unicorns you would like us to research! Here is a list to choose from.

🚨 Highlights of the Month

Klarna plans to go public in the US with a $20bn valuation, backed by strong financial results and partnerships with major brands. Utilizing AI technology and innovative products, Klarna has seen significant growth and profitability.  A detailed analysis is available in the Current Topic section of this newsletter.

ServiceTitan, a software company for tradespeople, is preparing for an upcoming IPO with a valuation of $5.2bn. Iconiq Growth, Bessemer and other venture capital firms are expected to benefit from the IPO.

xAI Round. Elon Musk's AI startup, xAI, has reached a valuation of $50bn and attracted significant investment due to his leadership and support for Trump. The company plans to use its $5bn funding to purchase 100,000 Nvidia chips. xAI is set to release a powerful new AI model in December.

Anthropic Round. Amazon invested $4bn in Anthropic, further strengthening the links between the two companies. Anthropic is now naming AWS its primary training partner, in addition to continuing to be its primary cloud provider, and will use AWS Trainium and Inferentia chips to train and deploy its future foundation models.

NVCA-Q324 Report: VC Liquidity Crunch Deepens. The Q3 report highlights a "liquidity drought," with exit values at $10.4Bn—a five-quarter low—and NAV distribution ratios near crisis levels. Fundraising is dominated by established managers, with first-time funds struggling. Deal activity is cautious, returning to pre-pandemic norms, while alternative liquidity options like secondaries gain traction. Investor-friendly terms persist, reflecting LPs' focus on experience amid constrained exits.

According to a Menlo VC report on the state of Generative AI Spending in the Enterprise,  Enterprises are shifting from experimentation to execution, embedding AI at the core of their business strategies. The application layer is heating up, with companies pouring $4.6bn into applications in 2024. The top use cases are code generation, chatbots, enterprise search, data transformation, and meeting summarization. Enterprises are prioritizing value over quick wins when selecting new tools, focusing on measurable ROI and industry-specific customization.

The Information notes that 7 Tech secondaries yielded nearly $20bn in volume in 2023-2024, dwarfing the $12bn raised in Tech IPOs over that period.

📈 Chart of the Month: Unicorns Comprise 63% of US VC Value

Pitchbook’s recent Evolving Economics of 10-Year VC Funds report showcases the underlying reasons for a market in what we call “Blue Chips” to emerge.

The most lucrative VC returns often require extended holding periods, as seen with Uber, Airbnb, and Roblox, which yielded over 50x MOIC but took more than 8.5 years to exit. Stripe, in VC portfolios for 13 years, grew its valuation from $20 million to $70 billion, demonstrating the value of patience. While the median holding period for VC investments is 5.1 years, larger exits ($500M+) require significantly longer timelines, peaking at 9.1 years in 2023. Unicorns, comprising 63.4% of ecosystem value, exacerbate delays by remaining private. With 39.1% of unicorns held for nine years or more, VC funds face challenges in balancing return optimization with distribution timelines as fund lifecycles end.

🌐 Current Topics: Klarna at $20bn…Shopping Time?

On November 12^th, Klarna filed a confidential draft of its Registration Statement for a proposed IPO. The Swedish fintech company, which popularized Buy Now, Pay Later (BNPL), is now valued at $15-$20bn. As it expands in the US and introduces new services like a wallet, Klarna is transitioning into a fully-fledged neobank.

Klarna’s core offerings include:

• Pay in 3 or 4: Splits the cost of a purchase into interest-free installments.

• Pay in 30: Allows consumers to pay within 30 days with no interest or fees if paid on time.

• Long-term loans: Finance larger purchases with longer-term loans.

As a consumer, Klarna seems magical. How can I get interest-free loans during times of high interest rates? The answer is that the merchant, not the consumer pays for the interest. Klarna charges merchants a fixed fee and percentage of the purchase,ranging from 0.99% to 4.99%, depending on the country. Merchants accept this because, as Klarna argues, BNPL users spend more, and if they don’t offer it, they risk losing the sale to competitors. Merchants receive the full purchase value, while Klarna assumes the credit risk and finances what is essentially a consumer loan. BNPL providers may do this on their balance sheet if they hold a banking license, or partner with a bank that originates the loans which may later be resold to the provider. Over time, Klarna has expanded into adjacent offerings, like card issuing, advertising for merchants, and a consumer wallet.

Over the past two years, Klarna has seen dramatic changes. While it still relies heavily on Europe (Germany, the UK, and Sweden account for 51% of net operating income), these markets are plateauing. This is why its US expansion, which began in 2015, is critical. The US represents a $230bn GMV opportunity, growing 20% annually, and has become Klarna’s largest market, contributing 31% of its net operating income.

Klarna is also diversifying its revenue streams. The Klarna card brings in card issuer fees, while the new wallet aims to compete with PayPal and Cash App by creating a two-sided network to sell banking products.

Peers like Affirm, Zip, and Sezzle serve as benchmarks for Klarna’s BNPL model. They get a higher value per $ of GMV processed, reflecting the lower Transaction Gross Profit per $ of GMV earns. Its rapidly growing US market and expanding business model might change that…

If you would like a detailed report on Klarna, fill this form.

Table 1: Klarna dwarfs Public Peers in Volume

🧭LT: “Shadow LLMs” Is The New “Shadow IT”

This section was prepared with substantial input from Mark Campbell, a long-time friend of Opportuna.

The rise of large language models (LLMs) in the workplace is creating cybersecurity risks reminiscent of shadow IT. Employees are increasingly turning to tools like ChatGPT for efficiency, often bypassing official approvals or oversight. This unchecked adoption exposes enterprises to significant vulnerabilities.

The most pressing threat is data leakage. Unlike traditional data loss prevention (DLP) solutions, which focus on personally identifiable information (PII), LLMs create novel pathways for exfiltration. Employees may inadvertently feed sensitive corporate data into public LLMs, exposing proprietary information. More concerning, attackers can exploit vulnerabilities in these models, using crafted prompts to extract confidential training data.

LLMs also expand the attack surface. Prompt injection attacks can manipulate models to perform unauthorized actions or disclose sensitive data. Another critical issue is insecure code generation; LLMs trained on outdated codebases may produce software laden with vulnerabilities, exposing organizations to breaches, lawsuits, and reputational damage.

The burgeoning LLM ecosystem presents opportunities for innovators. Tools to monitor and secure unauthorized AI use are becoming increasingly vital. For instance, adapting cloud access security broker (CASB) solutions to track LLM activity are already gaining traction.

AI-specific security seems very promising. Conventional methods are poorly suited to safeguarding LLMs, necessitating tailored approaches. Startups like Patronus AI are emerging to secure model outputs, prevent data leaks, and ensure compliance with company policies. The sector is ripe for innovation, with venture capitalists likening its potential to the explosion of CASB and CSPM platforms during the SaaS and cloud revolutions.

Startups will capitalize on these trends by addressing four key areas: 

• LLM-aware DLP solutions to prevent unauthorized data exposure.

• AI firewalls capable of analyzing and filtering prompts to counter injection attacks.

• Secure code generation platforms to vet LLM-produced code for flaws.

• AI risk awareness programs to train employees on responsible model usage.

The adoption of LLMs marks an inflection point in enterprise cybersecurity. While these tools promise efficiency gains, their unregulated use is a Pandora’s box of risks. Companies will act swiftly to safeguard their data, leveraging the emergence of specialized AI-security tools. As with shadow IT, a lack of oversight today could lead to costly lessons tomorrow.