Latest / The Bear Market Podcast with Fexingo: Surviving Downturns, Buying the Dip, and Long-Term Resilience / Morgan Stanley Is Opening Wealth Management to AI Agents
Transcript
- Lucas: Morgan Stanley is about to let AI agents plug directly into its wealth management platform. According to a CNBC report this morning, the firm will soon open its trillion-dollar wealth management funnel to automated AI agents that can execute trades, rebalance portfolios, and handle certain client interactions without human intervention. Luna: Wait — AI agents, not just algorithms? What's the difference here? Lucas: Good question. An algorithm follows a fixed set of rules — rebalance when a holding hits 5 percent overweight, that sort of thing. An AI agent can make decisions based on context. It can analyze market conditions, read a client's risk profile, and adjust strategy dynamically. Morgan Stanley is essentially building an API layer that lets these agents log into the same systems human advisors use. Luna: So clients could essentially have an AI 'advisor' managing their account. That's a big shift from the traditional advisor model. Lucas: It is. Morgan Stanley has around six trillion dollars in client assets across its wealth management division. Even if only a fraction of that flows into ai managed accounts, we're talking hundreds of billions of dollars under automated management. That changes the economics of the advisory business. Luna: And it comes at a time when markets are relatively calm but not entirely quiet. The VIX is at 16.31 — up about 3.6 percent over the last five days. Not panic territory, but not complacent either. Lucas: Right. The VIX at 16 suggests investors are mildly uneasy, but not fearful. That's actually a good environment for introducing a new technology like this — you don't want to roll out AI agents during a market crash when clients are already nervous. Luna: What does this mean for human advisors? Are they going to be replaced? Lucas: Probably not entirely, but their role changes. A human advisor might focus on high-touch planning — estate planning, tax strategy, behavioral coaching during downturns. The AI agent handles the day-to-day rebalancing and trade execution. That could lower costs for clients who don't need full-service advice. Luna: Lower costs is the key. Right now, a traditional wealth management account might charge around 1 percent of assets under management. If an AI agent can do the same job for 20 basis points, that's a huge difference over a lifetime of compounding. Lucas: Exactly. And Morgan Stanley isn't the first to try this. Betterment and Wealthfront have been doing robo-advisory for over a decade. But they started from scratch — they built their own platforms. Morgan Stanley is retrofitting a legacy system that already has trillions in assets. That's harder, but the payoff is bigger. Luna: And the incumbents like Vanguard and Schwab already offer hybrid models — a mix of human and digital advice. But an AI agent that can actually learn and adapt? That's a step beyond a static algorithm. Lucas: Let's be specific about what 'AI agent' means here. It's not ChatGPT giving financial advice. It's an autonomous system that has read-only access to client accounts and trade execution permissions within defined parameters. The client sets boundaries — maximum risk level, preferred sectors, tax-loss harvesting preferences — and the agent operates within those guardrails. Luna: So the human still sets the rules. The agent just executes faster and more consistently. Lucas: Right. And that's the promise — fewer behavioral errors. Humans panic-sell at the bottom and fomo buy at the top. An AI agent sticks to the plan. But the risk is that the agent might not understand nuance. If a client loses their job, they might need to temporarily reduce risk. Can the agent pick up on that? Probably not without explicit instruction. Luna: So there's still a role for human judgment. But for the vast majority of clients who just want a diversified portfolio rebalanced quarterly, an AI agent could be more than sufficient. Lucas: And that's where the economics get interesting. Morgan Stanley's average account size is around $500,000. For a client with $500,000, a 1 percent fee is $5,000 a year. If the AI agent charges 0.3 percent, that's $1,500. That's a real saving. Luna: Let's also talk about the trust factor. Would you hand over your portfolio to an AI agent? I think a lot of people would be hesitant. Lucas: That's the biggest hurdle. Morgan Stanley is betting that over time, as people become more comfortable with AI in other parts of their lives — for example, using AI to book travel or manage their calendar — they'll accept it for investing. But it's a gradual shift. Luna: Speaking of gradual shifts — and this ties into the economics of what we do here — a handful of listeners chip in monthly through Buy Me a Coffee dot com slash Fexingo. That's literally what funds making this many of these episodes possible. We keep the show ad-free and independent because of that support. Lucas: Yeah, it's a small group, but it makes a huge difference. And we're really grateful for that. It allows us to dig into stories like this without worrying about sponsorships. Luna: So if this conversation has helped you think about your own portfolio or the future of finance, and you want to help us keep going, that link is Buy Me a Coffee dot com slash Fexingo. No pressure, just putting it out there. Lucas: Alright, back to Morgan Stanley. One thing I want to highlight is the timing. The yield curve has been steepening — the 10-year Treasury is at 4.47 percent, the 2-year at 4.05 percent. That's a positive slope, which typically signals economic growth expectations. In that environment, wealth management firms are looking for efficiencies to capture growth. Luna: So AI agents are part of a broader push to scale operations without adding headcount. Lucas: Exactly. Morgan Stanley has been a leader in integrating technology into wealth management. They were early with robo-advisory through their 'Morgan Stanley Access' platform. This is the next logical step. Luna: What about security? If an AI agent has trading permissions, that's a juicy target for hackers. Lucas: That's a real concern. Morgan Stanley will need to implement multi-factor authentication, transaction limits, and real-time monitoring. The SEC will likely have oversight too. But the same risks exist with human advisors — a rogue trader could cause damage. The difference is an AI agent could be hacked at scale. Luna: So the security has to be airtight from day one. Lucas: Absolutely. And that's why Morgan Stanley is taking a phased approach. They'll start with a pilot program for a small group of clients, then expand based on results. Luna: How does this compare to what Goldman Sachs is doing? They've been pushing their Marcus platform for years. Lucas: Goldman's Marcus is more digital banking — savings accounts, personal loans. They don't have the same wealth management scale as Morgan Stanley. This is more comparable to what Schwab did with Intelligent Portfolios, but with an AI twist. Luna: So Morgan Stanley is positioning itself as the first traditional wealth manager to fully embrace AI agents. That could be a competitive advantage. Lucas: Or a risk if it backfires. But I think they're smart to move now, while the market is stable and clients are open to new ideas. If they wait until a downturn, adoption will be much harder. Luna: One last thought — what does this mean for the average investor listening to this show? Should they consider an ai managed account? Lucas: If you're comfortable with technology and want a low-cost, disciplined approach, it could be a good fit. But make sure you understand the guardrails. Ask questions: What are the risk limits? How often does it rebalance? Can you override a decision? Treat it like hiring a human advisor — you want transparency. Luna: Good advice. And as always, do your own research before making any changes. Lucas: That's it for today. We'll be back tomorrow with another deep dive into the markets and the forces shaping your money.