While Donaldson Cuts Guidance, Atmus Just Walked Into Data Centers: Buy ATMU, Sell DCI

Atmus appears to be outperforming: its Koch Filter acquisition is already showing 21.9% EBITDA margins in the first quarter of ownership, while Donaldson's industrial margins fell 420 bps YoY and management cut guidance. Donaldson also paid 20x EBITDA for Facet, with no earnings accretion expected until FY2027. The article argues Atmus deserves a relative valuation premium given its 16.64x forward P/E versus Donaldson's 21.03x.

Analysis

The market is still underpricing execution dispersion in filtration, and that creates an obvious relative-value setup: the company with the cleaner integration story and faster end-market mix shift should trade at a premium, not a discount. The key second-order effect is that data-center filtration is not just another vertical; it is a structurally higher-growth, less cyclical pocket with better pricing power and a more favorable mix than legacy industrial exposure. That makes margin durability more important than headline scale, and it should force a rerating as investors model a longer runway for accretion rather than one-off revenue contribution.

The loser is not only the acquirer’s near-term P&L, but also its multiple: paying up for an asset with delayed earnings impact while showing margin compression creates a credibility gap that can persist for several quarters. In a slower industrial tape, investors typically punish deals that promise synergy far out on the horizon, because the market will discount the benefits heavily and capitalize the margin slippage immediately. That also pressures peers and adjacent suppliers: if one consolidator is proving that bolt-ons can still be immediately accretive in a specialty niche, capital will migrate toward businesses with tighter integration discipline and better end-market exposure.

Near term, the main risk to the relative-value call is a broad industrial de-rating that drags both names lower, but the spread should still be less exposed on the better-executing operator. The catalyst path is likely 1-2 quarters: another margin print confirming accretion, plus any evidence that the acquired asset is expanding into adjacent hyperscale customers, should force estimate revisions and multiple expansion. Conversely, if the weaker operator stabilizes margins faster than expected or if data-center demand normalizes, the trade could stall; that would likely take months, not days.

The contrarian view is that the premium/discount gap may be less about quality and more about expectations already embedded in the stock prices. If the better operator is already being valued for flawless integration, upside could be capped unless management raises guidance again. Still, the asymmetry favors owning the cleaner compounder and shorting the more expensive, lower-conviction execution story until the market either closes the valuation gap or the laggard proves it can deliver sustained accretion.

Key Decisions for Investors

• Long the better-executing filtration name / short the weaker acquirer as a 3-6 month pair trade; target spread compression from current mispricing, with downside limited if industrials weaken broadly because the long should hold up better on quality.
• If using listed exposure, buy near-term call spreads on the higher-quality operator into the next earnings cycle; this offers defined risk and levered upside if management confirms margin accretion and guidance momentum.
• Short the acquirer on rallies over the next 1-2 quarters, using a tighter stop if margins stabilize; the market is likely to punish delayed EPS contribution and integration risk until there is visible earnings traction.
• For conservative portfolios, stay long the higher-quality name and hedge with a sector ETF short rather than a full single-name short; this captures idiosyncratic outperformance while reducing macro-industrial beta.
• Reassess after the next quarterly print: if the premium name fails to expand margins or the laggard re-accelerates, cover the pair; otherwise, expect the relative-value spread to work over a multi-quarter horizon.