The analysis puts the S&P 500's Shiller cyclically adjusted price-to-earnings ratio at 40.3 at the close on Tuesday, Sept. 15, describing it as the index's second-highest valuation since the dot-com bubble peak in 2000. It identifies Middle East tensions, rising oil prices and pressure on inflation and interest rates as potential sources of market stress.
It also cites November's congressional midterm elections and calls by AI developers for slower development as possible risks to high-growth sectors. Against that backdrop, the author's interest in the two companies is conditional: a broader market decline could offer lower entry prices. The analysis does not establish that such a decline will occur.
DigitalOcean focuses on small and mid-sized enterprise customers, offering cloud services with straightforward pricing, personalized support and deployment tools designed to be accessible to businesses without their own technical teams. It is extending that approach to its AI-Native Cloud platform, which has five layers for deploying AI software.
At the infrastructure level, the company operates 20 data centers with specialized chips from suppliers including Nvidia and Advanced Micro Devices. Customers lease computing capacity and can use an inference engine offering models from developers such as OpenAI and Anthropic, alongside more than 70 open-source models. Those services support the creation of agents, chatbots and other applications.
DigitalOcean reported $1.1 billion in annual recurring revenue as of June 30. AI customers accounted for $234 million, up 212% from a year earlier. Its backlog of orders from customers awaiting additional data-center capacity reached $894 million, a twelvefold increase over the same period.
The valuation is a restraint on the analysis's enthusiasm. DigitalOcean's price-to-sales ratio of 13.2 exceeds its average of 8.6 since its 2021 stock-market debut. The author would prefer a valuation nearer that historical average rather than buying solely on the strength of AI growth.
Lemonade, meanwhile, provides homeowners, renters, life, pet and car insurance to more than 3.3 million customers in the United States and Europe. It uses AI for customer service, pricing and operations. Its Maya chatbot can provide a quote in under 90 seconds, while a separate assistant, Jim, can handle claims in seconds without human intervention.
In the second quarter of 2026, Lemonade's loss adjustment expense ratio fell to a record 5%, compared with an industry average of 9% cited in the analysis. The ratio measures the share of premiums spent handling claims, making it a different measure from the share paid out to policyholders.
Lemonade ended the quarter with more than $1.4 billion in in-force premiums, representing premiums from active policies, up 32% year over year. Its gross loss ratio, the percentage of premiums paid as claims, was 60%. Quarterly revenue rose 79% to $294.4 million.
Management believes in-force premiums could reach $10 billion by 2034, a projected increase of 600%. That is a management expectation rather than an achieved result. Lemonade's price-to-sales ratio of 4.3 is below its three-year average of 5.3, in contrast to DigitalOcean's premium to its own historical average.
The comparison therefore rests on two separate questions: whether the companies can sustain their operating growth and what price investors pay for it. DigitalOcean's expanding AI revenue and backlog draw attention to demand for computing capacity, while Lemonade's figures emphasize insurance growth and the cost of processing claims.