[Week 52 of 2025] Sweep, Sales, and Savants

Welcome back to Price and Prejudice with a few musings from Week 52 of 2025. I had to skip week 51 due to my honeymoon, but hopefully no more breaks henceforth!

Robo Revolution Ends in a Cash Sweep

This article discusses how the robo-advice revolution never really happened. Automated portfolio construction became cheap, ubiquitous, and ultimately indistinguishable, leaving little reason for investors to care who provided it. Wealthfront’s low-key IPO is treated as the epilogue — not a failure exactly, just the end of a story that once promised to replace human advisors and instead turned into a cash-management business with an investing sidecar.

Why did this play out so quietly? One way to read this is that robo advisors successfully unbundled the most scalable slice of advice — asset allocation, rebalancing, basic tax mechanics — and competed it down to a commodity. What remained was the hard-to-automate bundle: planning, coordination, and behavioral coaching, which is basically the part where someone absorbs responsibility when decisions are stressful and irreversible.

There’s also a credence-good / platform story. Advice quality is hard to verify even after the fact, so trust, reputation, and a relationship act like a substitute for due diligence; “someone to call” is a form of insurance against panic and regret. Meanwhile once the product is commoditized, distribution matters more than product, and the platform’s incentives start to shape outcomes. If revenue comes from cash spread, cash will mysteriously become a prominent “feature.”

The implication for AI is that it probably lands in the same equilibrium, just further along the curve. AI will compress the cost of standardized tasks and then disappear into infrastructure, raising the relative value of whatever remains scarce: judgment, accountability, and trust under pressure. At the same time it will advantage whoever owns the rails — data, default settings, and distribution — so the world gets “better tools” and “more platform power” in the same breath.

Taxing the Dollar Trade

South Korea is trying to lean against a weakening won by incentivizing retail investors to come home. The policy is a temporary capital-gains tax break: sell overseas stocks and reinvest the proceeds in Korean equities, and the sale proceeds (up to a per-person cap) get favorable tax treatment. The backdrop is a won that slid to around ₩1,484 per dollar and official anxiety about a psychologically loud ₩1,500 level, alongside a surge in Koreans’ US equity holdings to roughly $160bn.

Why might this be happening in the first place? One story is fundamentals and risk premia: if growth expectations soften or the economy looks more exposed to global shocks, the currency can cheapen even without a melodramatic “crisis,” because investors demand a bigger discount to hold won assets. A big example in the same news cycle is the $350bn US investment pledge and related constraints: large, politically salient dollar outflow plans narrow the government’s perceived room to maneuver, and markets tend to price that sort of commitment faster than the cash actually leaves the country. But it’s also easy to see why policymakers latch onto the mechanical flow story: retail US equity ownership is observable, record-high, and framed as a behavior that can be nudged with tax rules, defaults, and jawboning, whereas “worse fundamentals” is not something you can fix with a Wednesday press release.

What can the government do? The near-term toolkit is basically: (1) incentives to reduce the marginal attractiveness of outbound flows (this tax break), (2) make hedging easier/cheaper for households and institutions so the “US assets” trade doesn’t automatically mean “short won,” and (3) enlist large public balance sheets to smooth spikes — the National Pension Service’s more flexible hedging has already been a meaningful signal. The harder challenge is making domestic assets easier to own with confidence: governance, payout policy, credible macro policy, and fewer reasons for Koreans to treat USD assets as the default store of value. The tax break can shift flows at the margin; it can’t substitute for fundamentals.

Historical Genius

Historically, we have enough records to see that brilliance in science, politics, or art doesn’t reliably translate into investment success. This article is a fun, seasonally appropriate exercise building on that theme: imagine hedge funds time-travelling through history to recruit the “best minds,” then check whether those minds actually would have made money.

It raises the bigger question of what skills investing actually selects for, and the common threads aren’t especially romantic. First, constraints matter: having surplus capital and a long horizon is a skill-adjacent advantage, because it lets you survive drawdowns and avoid forced selling (Churchill is a case study in how being “right eventually” is useless if your broker calls you today.) Second, temperament is a real edge: Newton’s problem wasn’t calculus, it was re-entry. The market punishes the urge to be involved, especially when involvement is really just emotion dressed as strategy. And probably the third, process beats raw intelligence: Keynes looks good not because he’s the smartest person in the room (he is), but because he learns to concentrate, lower turnover, and bet on things he understands, which is basically “institutional memory” applied to capital allocation.

There’s also a quietly uncomfortable implication in the premise that hedge funds “steal talent”: it assumes money is a clean measure of social contribution. The more mundane economics story is that finance is where scalable decision-making gets bid up, and markets pay a lot for the ability to move large sums with a small perceived edge, even if that edge is mostly discipline and structure rather than genius. If anything, the article hints that the most recruitable historical genius is the one who can do boring things consistently, which is not how we usually define genius at all.