Cathie Wood Goes Bargain Hunting: 3 Stocks She Just Bought

Cathie Wood added to Palantir, Cerebras Systems, and Coinbase on Thursday, but the article highlights sharp drawdowns across all three names: Palantir is down 48% from its November peak, Cerebras has fallen more than half from its first-day high, and Coinbase has been cut in half from its 52-week high. Palantir's revenue growth is accelerating and earnings are improving, while Cerebras posted 94% revenue growth but with margin pressure and customer concentration risk. Coinbase remains under pressure as crypto demand weakens, with revenue down 22% and 31% in its last two quarters.

Analysis

The common thread here is not “Wood bought three names,” but that the market is re-pricing duration risk across the most crowded growth narratives. When a favored allocator adds on weakness, it can stabilize sentiment for a few sessions, but it does little to change the underlying dispersion: the market is rewarding revenue quality and penalizing any hint of balance-sheet or customer-concentration fragility. That makes the next leg less about headline growth and more about whether growth can be funded without future dilution or margin compression.

PLTR is the cleanest expression of this setup. The business is entering a phase where operating leverage can compound for multiple quarters, so the multiple compression is more a function of positioning reset than deteriorating fundamentals. That creates a favorable asymmetry for the next 3-6 months: if earnings revisions continue higher, the stock can rerate quickly because it has already de-grossed from its peak-owned status. The risk is less execution and more expectation management; any deceleration in government or commercial deal conversion would reopen the de-rating.

COIN is the opposite: it is a high-beta macro proxy with a business model that gets punished twice when crypto volumes weaken—first on transaction revenue, then on operating leverage. The key second-order effect is that, in a risk-off crypto tape, investors often rotate into lower-friction “picks and shovels” exposure elsewhere rather than back into COIN, so the stock can lag even if BTC merely stabilizes. Until crypto spot prices and retail activity inflect, each rally is vulnerable to fading.

Cerebras is the most fragile of the three because the market will not underwrite backlog headlines if customer concentration and margin quality stay poor. The open question is whether this becomes a strategic infrastructure winner or another capital-intensive AI supplier with lumpy demand. In the near term, that tension argues for treating any strength as a fade unless there is evidence of broader customer diversification and repeatable gross-margin expansion.

Key Decisions for Investors

• Long PLTR on 1-3 month horizon via call spreads or stock on pullbacks; risk/reward favors upside if earnings revisions continue to outpace multiple compression. Use a trailing stop on a close below the post-earnings breakout zone.
• Avoid or short COIN on any crypto-led relief rally over the next 2-6 weeks; pair long BTC-related asset exposure with short COIN to isolate the operating leverage discount if trading volumes stay soft.
• Treat CBRS as a tactical trading vehicle only: sell strength, avoid core long exposure until the next quarter shows customer diversification and margin stabilization. Best expressed as a short against a basket of higher-quality AI infrastructure names.
• Pair trade: long PLTR / short COIN for the next earnings cycle. This isolates “durable software compounding” versus “macro-dependent monetization,” with PLTR the higher-quality duration asset if growth momentum persists.
• If BTC reclaims and holds trend support for multiple weeks, reconsider COIN for a short-covering trade, but only with tight risk control; the upside is fast, but the thesis depends on a sustained crypto activity inflection.