FactSet Trades at a Third of MSCI's Multiple — the Catch Is the Growth It Still Has to Prove

FactSet's shares slid again on Friday — down about 6% on the day and roughly 28% below their 52-week high — the morning after OpenAI unveiled a ChatGPT built specifically for financial services. That is the same AI fear that already cut this stock from roughly $450 to about $185 in the span of a year, and the fact that FactSet is listed as a partner in the launch, rather than a casualty, hasn't stopped investors from selling first and asking questions later.

For a retail investor watching this from the outside, the natural question is whether FactSet is a falling knife or a mispriced cash machine. The valuation data says it's a genuine cash machine priced like it isn't. That gap is the whole story, and it's worth unpacking before the company reports fiscal fourth-quarter results on September 30.

The cheapness is real — run the comparison

No stock is "cheap" on its own terms. It's cheap only relative to a credible peer set, and FactSet's natural comparison is the growthier data-analytics names it competes with. Against them, the discount is not a rounding error.

FactSet trades around 16.5 times trailing earnings, against roughly 29 for MSCI and about 25 for S&P Global. On price-to-sales it's 3.8 times versus 11.9 and 7.5. On EV/EBITDA, 11.2 versus 23.4 and 15.8. Put plainly, FactSet carries about a third of MSCI's earnings multiple and a little more than half of S&P Global's — with a $9 billion market cap against their $40 billion and $121 billion.

The profitability sitting under that multiple is not the usual value-trap signature. Operating margin is near 30%, return on equity is 27%, and the company converts about 29% of revenue into free cash flow. That cash funds a 1.8% dividend yield with 24 consecutive years of payouts, 17 straight years of increases, and a payout ratio around 37% — meaning roughly two-thirds of cash flow stays in the company. Free cash flow itself grew about 23% over the past year. On valuation and profitability together, this scores like the sort of name a systematic screen flags, not buries.

The catch is growth — the factor the market is waiting on

Here is where the comparison cuts the other way, and it's the whole reason the cheapness exists. Revenue rose only about 6.6% over the past year, and gross profit just 2.7% — a tighter spread that reflects money being spent on AI. MSCI and S&P Global compound revenue considerably faster, and the market pays up for that growth. A stock selling at 16 times earnings with mid-single-digit growth can sit at that multiple for a long time; a low absolute P/E does not re-rate until growth accelerates. On a PEG basis — the price-to-earnings ratio divided by the growth rate — FactSet lands near 1.9, which tells you the low multiple is really just a fair price for a slow grower, not an obvious bargain.

That growth story is the specific thing management has pointed to as improving. The wealth-management segment had already returned to double-digit growth on the strength of a contract to power UBS adviser desktops, and newly monetized AI products were expected to add a few tenths of a point to growth. But the banking segment has stayed weak, with long buying cycles, and organic annual-subscription-value growth has run in the mid-single digits. The market's skepticism is not baseless; it's just not the whole story either.

AI is both partner and predator — and the street says Hold

The AI story is where narrative and data collide. OpenAI named FactSet a partner for its new financial-services ChatGPT, alongside Morgan Stanley, Evercore, and S&P Global. On its face that reads as an endorsement — FactSet's data feeding OpenAI's model. The announcement still pushed the shares of data providers lower, because the same capability looks like a threat to the terminal business that generates the revenue. Both readings are live, and that unresolved tension is exactly what keeps the multiple low.

The analyst committee is describing the same standoff. Deutsche Bank raised its price target while holding its rating, and the wider street has settled into the same camp: analysts polled by FactSet carry a consensus Hold with a mean target near $315 — roughly 20% above today's price — and the distribution is dominated by Holds rather than conviction in either direction. AInvest's aggregate signal also labels FactSet a Hold. In this process, Hold is not Sell: nobody is telling owners to exit, but nobody is convinced enough to call the re-rating either.

The timing factors aren't cooperating yet, either. After the run from the low, momentum has rolled over — the relative-strength reading sits below 40 and the stock has fallen back under its 50-day average — so this reads as a value-plus-profitability setup whose entry timing is still being worked out.

The portfolio role: the defensive side of the barbell

So where does this belong? This is the income-and-durability half of a barbell, not a momentum speculation. FactSet is a durable cash-flow business that pays you a growing dividend while you wait to see whether AI is partner or predator. It's the sleeve you reach for when you want yield and quality over frothy AI growth, and it happens to hedge against the very AI-growth side of the market that's been driving the indexes.

What would change the read is one specific thing: organic revenue growth re-accelerating. The wealth-management strength and the AI-product contributions are the evidence stream to watch in that September 30 report. If growth breaks back into the double digits, the Hold turns into a Buy and the multiple re-rates with it. If it stays in the 5% to 6% range, the cheapness is a fair price — and you keep the dividend and the discipline either way.