Get the Major Trend Right

A thousand right moves cannot save a wrong trend

Before the team, before the price, before a single tactic, one thing has to be right: the direction of the trend underneath the bet. Brad Jacobs learned the rule at lunch, at twenty-three, from Ludwig Jesselson, who ran Philipp Brothers, then the largest commodity-trading house in the world. Jesselson traded in a world where a deal worth hundreds of millions closed on a handshake and the written confirmation could trail it by days, with the price of oil or copper moving the whole time. The principle scales far past commodities, and its power is in the converse, which Jacobs states without hedging: "If you get the major trend wrong, you can do a thousand things right, you're still going to lose. You're not going to create alpha. You're not going to create value there."¹ [2:36]

Jacobs runs it at two altitudes

At industry altitude it is his entry screen: the market has to be large, what he calls, borrowing the phrase from Sequoia's Don Valentine, "a big T," and also growing, fragmented, and buyable, and above all the technology trend has to run toward it rather than through it.² That is why he can pick an industry he has never worked in and back it with real money. Building-products distribution, the market he built into QXO, passed because the trend under it is close to unkillable, and he says so plainly, that "it's a safe bet that building products is not going into the metaverse."² At civilization altitude the trend he keeps returning to is technology itself, the long habit of building tools that take over what people used to do by hand, which in his telling runs back millions of years and is now accelerating.¹ In that framing, every business is a technology-adoption story whether it admits it or not.

The proof is in the industries he refused

The discipline shows its worth on the bets he never placed. Screening some 55 industries, he ruled out the online-education company Chegg on the judgment that AI would soon deliver the same tutoring for free, and the stock later fell from around fifty dollars to single digits.¹ [1:33:50] The screen is a decision-tree pruning operation. A correct read of the trend eliminates whole categories of bad bets before anyone argues tactics, because a perfect operator inside a dying market still loses. It is also why his selection funnel starts so wide, roughly 600 companies across those 55 industries, narrowed until a single industry checked every box.²

Easy to state, treacherous to apply

A trend can be rationalized as easily as it can be derived. Ruling out Chegg was clean; most calls are closer, and the same confidence that prunes a bad bet can talk a founder into a bad thesis. The harder limit is that the trend read is a forecast with a clock on it. Jacobs' own screen leans on a filter, that AI will not disrupt his chosen industry soon, and the filter never says how soon. The building-products judgment holds only as long as automation stays out of the trade, and the framework offers no alarm for the day its central assumption expires. Getting the major trend right once is not the same as noticing when it has turned.