Global equity portfolio manager Paul Moroz explains why the market trades differently than it did a decade ago and how the team is adapting the way it builds and manages portfolios in response. He describes a market where stocks increasingly move on sentiment, short-term trading, and a handful of data points rather than on how the underlying business is performing, and what that means for sizing positions, deploying capital, and keeping perspective when prices swing. Through it all, the conversation returns to the one question that grounds every decision: what is an investment actually worth? 0:00 Introduction: Market Structure and Outsized Price Moves 0:40 Why the AI Trade Moves Stocks on a Delta, Not on a Beat or a Miss 1:17 Creating Balance as AI Stocks Swing 4:12 Alphabet Earnings: Search Growth Near 12% Against 15% Expected 6:34 The Less Efficient Market Hypothesis 6:59 From 50 Names to Close to 90: More Chess Pieces on the Board 8:54 Is Alphabet Cheaper After the Drop, or De-Rati
A one-sentence AI request might use 10 tokens, while one that needs a live search and a database lookup can use tens of thousands. Mawer Chief Technology Officer Justin Anderson explains what happens between pressing enter and getting an answer, and why that gap drives what AI costs a company. He walks through the value-over-cost equation behind AI spending, describes why Mawer built its own harness, and lays out how to tell an AI dinosaur from a rat racer or an AI native when evaluating management teams. The conversation ends on where human judgment matters: interviewing management, spotting inflection points, and keeping pace with the technology. Key Takeaways: • Every AI request is broken into tokens, and cost follows the token count. A one-sentence summary might use 10 tokens, while a request needing a live search or database lookup can use tens of thousands. • Justin splits the return on AI into value, owned by business teams, and cost, owned by technology. Mawer built a harness t
Mawer equity analyst Irena Petkovic separates what the team knows about the AI build-out from what it doesn't, after several volatile weeks for AI stocks and hyperscaler earnings. She covers the latest hyperscaler results, where growth is now limited by how fast new data centers can be built, and what Anthropic's S-1 filing says about how much room enterprise AI adoption has left. On the unknowns, she lays out the U.S. midterm results, the financing of the data center build-out, and how AI agents, software that uses the internet on a person's behalf without clicking on ads, could change which online businesses capture value. She then explains how the team asks what's in the stock, meaning the assumptions a share price already contains, and what that meant for Alphabet. Key Takeaways: Hyperscaler AI results accelerated again, and supply is the constraint. AWS posted a sixth straight quarter of revenue acceleration, Google Cloud doubled year over year, and all four hyperscalers said grow
Ching Chang has joined Mawer as lead manager of the Global Small Cap strategy. He describes how his focus on quality companies that can grow developed and what three full market cycles in China taught him. He also explains why the small cap premium has gone missing, where small caps still fit in a portfolio, how regional weights will be considered alongside bottom-up stock picking, and what he looks for in a management team. Key Takeaways Over 15 years, beginning at Ontario Teachers', Chang's focus moved toward quality companies that can compound a moat and strong return on capital over time. Three full market cycles in China between 2018 and 2025, including a sell-off of more than 60% from peak to trough and a near doubling in a few months in 2024, built resilience that carries into other markets. Later IPOs such as Uber and Airbnb, and asset-light AI-era leaders such as Google and Amazon, help explain the missing small cap premium. Small caps can still diversify, with more local busi
GLP-1 drugs and the peptide supply chain behind them have grown like no other drug class, reaching 44 million users within five years. Equity analyst Joshua Samuel looks past innovators Eli Lilly and Novo Nordisk to the contract manufacturers and consumables underneath them. He explains why oral GLP-1 pills can need up to 10 times more drug substance than an injection, walks through building a 39-amino-acid peptide like a LEGO set, and lays out why a silica gel supplier with about 70% global market share does not screen as a healthcare company. The conversation lands on a simple idea: growth in this market rests on several independent drivers rather than one. 0:00 - Introduction: GLP-1s Beyond the AI Trade 1:17 - Consumer Biotech: 44 Million GLP-1 Users in Five Years 2:49 - Small Molecules, Biologics, and Peptides Explained 5:01 - GLP-1 Generics, Oral Pills, and Patent Expiry 6:48 - Why Oral GLP-1 Pills Need More API 8:28 - How Peptides Are Made: The LEGO Analogy 10:56 - The Purificati