SpaceX’s underwriters published their price targets this week, and the range is a punchline. Goldman Sachs (GS) says $205. Morgan Stanley (MS) says $300. UBS says $210. And Raymond James says $800 — a number that would value a money-losing rocket company at roughly $10 trillion. The stock closed Wednesday at $148.30.
A price target is an analyst’s published guess at where a stock is headed. That is all it is — a guess with a logo on it. When four banks that took the same company public cannot agree within a factor of four, that is not a forecast. That is a food fight with a research header, and the confetti is other people’s money.
The data has an opinion
Here is the part that turns a joke into a lesson. Trivariate, a research firm, finds that wide target dispersion historically underperforms. In plain English: when analysts disagree badly about a stock, that stock tends to disappoint. Confusion at the top is not a buying signal. It is a warning label.
Why would that be? Because dispersion measures how little anyone actually knows. When a company’s future is countable, the estimates cluster. When the estimates are scattered from $205 to $800, the honest translation is: nobody can count anything yet, and everybody is guessing in public.
Targets chase prices
The deeper tell is which way targets move. Oracle (ORCL) would need to rise 80% just to reach its own average target — a target the stock shows no interest in visiting. Meanwhile, Nvidia’s (NVDA) targets have politely chased its price higher all year, the analysts adjusting their math after each move like commentators updating the score.
Put those two together and you get the whole problem. Targets follow prices. They are rearview mirrors dressed up as telescopes. A portfolio that follows targets is therefore following the past — at a delay, with extra confidence.
That does not make analysts useless. Their work on businesses — the products, the contracts, the costs — is often excellent. It is the single glowing number stapled to the front page that deserves your suspicion, especially when it comes from the same bank that earned fees taking the company public.
What we actually do
So SpaceX stays on the watch bench. A $205-to-$800 range is wide enough to be useless, and we do not buy uselessness at any price. Nvidia stays a core hold — not because of anyone’s target, but because of its own signed demand: contracts, not commentary.
Notice the difference between those two decisions. One rests on published guesses that disagree by a factor of four. The other rests on orders a customer already signed. Only one of those can be checked. We hold the checkable one.
The discipline generalizes to every holding we own. When a price target makes you feel something — excitement, envy, fear of missing out — that is marketing, not math. We do math here. The math is slower, quieter, and much better company in a downturn.