[Week 28 of 2026] Fragrance and Flows

Welcome back to Price and Prejudice with a few musings from Week 28 of 2026.

Scent and Sensibility

This FT article reports that big companies are commissioning their own bespoke aromas. Apparently, JPMorgan has a "signature scent" wafting through its new Park Avenue headquarters, and same goes for Goldman's London offices reportedly carrying a whiff of Byredo's Mojave Ghost. Apparently there is some research showing that a good smell lengthens visits and lifts spending – which makes sense, since a nicer-smelling store keeps you browsing, and even with a fixed probability of browsing turning into buying, this means higher sales.

In offices, though, surveys keep finding that workers rank air odor well below privacy, adjustable temperature, and a view of a window. An office has no customer to convert, so scent is probably less of a productivity booster but more of a signal. A nice lobby fragrance is the olfactory version of a marble atrium: the point isn't the marble, it's that you could afford the marble. The audience for the scent is never the analysts but the clients and the new recruits walking through, deciding whether these people look serious enough for delegating their future.

Reshuffling Rents

This Bloomberg piece reports that two trading pods at Millennium made about $3.7 billion between them in June, more than half of the roughly $6.6 billion the whole firm generated before fees. That's a lot of money! Both desks, Glen Scheinberg's SRBL group in New York and Pratik Madhvani's team in Dubai, specialize in index rebalancing, i.e. highly leveraged bets on which securities are about to enter or exit an index. June was an unusually rich month for the trade because five events landed almost at once: the S&P 500 quarterly rebalance, the Nasdaq 100 changes, the Russell annual reconstitution, the fast-track inclusion of SpaceX, and quarter-end multi-asset rebalancing.

Just a recap of the broad trade: when a stock joins an index, every index fund tracking that index has to buy it, on roughly the same day, in size, more or less regardless of price. This forced, price-insensitive demand is the "index effect." The strange part is that recent research has shown the measured index effect has shrunk close to nothing. So the real trick is now to rent out liquidity: the pods forecast the adds and deletes weeks ahead, take the other side of the index funds' forced flow, and get paid for absorbing demand that has to trade against a deadline.

This is a nice example of a "demand-based" view of markets. When a large, price-insensitive buyer must trade, the price is set by whoever is willing to accommodate them, and that accommodation is a service that carries a fee. Of course, this is not free money since you have to forecast the adds, the resulting flows, how much liquidity your competitors will supply, and carry the risks for weeks. But it sounds like it's pretty worth it.