BofA's 2% S&P 500 Forecast Isn't a Market Call — It's a Stock-Picking Call

On Monday, Bank of America raised its year-end 2026 target for the S&P 500 to 7,400 from 7,100 — and the move barely registered, because the bank's 12-month view is hardly more generous. BofA sees the index at 7,800 twelve months from now, roughly 2% above where it trades today. When the market's been up big for two years running, a call for another 2% is Wall Street's polite way of saying "the fun part is over."

That forecast reads as one cautious bank being timid. It's actually more interesting than that, because of what sits right next to it. BofA also expects S&P 500 earnings per share to grow 33% in 2026 and 12% in 2027. Earnings growth like that is the sort that normally underwrites a double-digit market return. So how do you get 33% earnings growth and only 2% of price appreciation out of the same index?

Through the one number that connects them: the price-to-earnings multiple.

Earnings and price are not the same thing. Price is what the next buyer pays for a share. The P/E ratio is what you pay for each dollar of expected earnings. If earnings climb a third and the price climbs 2%, then the multiple investors are willing to pay has to fall — sharply — to make the arithmetic work. BofA's 2% forecast is not a bet against corporate profits. It is a forecast that the market will pay less for those profits. The simple version: nobody doubts the earnings; the question is whether the valuation can hold.

That distinction matters, because the compression would not land evenly. The cap-weighted index is dominated by a handful of mega-cap names whose multiples tower above the rest of the market, and those are exactly the shares most exposed when investors decide to pay less per dollar of earnings. Take Nvidia: 83% year-over-year revenue growth and a 74% gross margin — a genuinely strong factor stack — yet it trades around 27 times earnings and 17 times sales. Apple sits near 38 times earnings. These are the names with the most multiple to lose. When the P/E has to compress, the discount lands hardest on the segment that started out richest.

Across the index, the starting valuations are not close. JPMorgan trades around 15 times earnings and yields roughly 1.7%; Chevron and Exxon carry enterprise values of about 8 to 10 times EBITDA with dividends near 2.5% to 3.3%. The same 2% index forecast contains stocks with wildly different valuation hair and, with it, wildly different amounts of room to absorb multiple compression.

That is not my framing alone — it is BofA's own positioning. The bank says it prefers large-cap value and the equal-weighted S&P 500 over the cap-weighted index, a structural way to reduce the overhang of the highest-multiple mega-caps, and it names energy and financials as favored segments. Behind the call is Savita Subramanian, head of U.S. equity and quant strategy, who has held the lowest year-end S&P 500 target in Bloomberg's survey of more than 20 strategists while the group average sat closer to 7,900. Her caution comes less from a bearish view of earnings than from flags on the way they're being paid for: rising leverage tied to the AI build-out, unusually tight credit spreads, and a tech basket she'd rather underweight than chase until she sees real monetization from the spending.

For an investor who works off factor data rather than macro feels, that's the productive read. A 2% index forecast is not a reason to leave the market — it's a reason to be selective, because when index beta is expected to do almost nothing, the whole game becomes which companies you own and what you pay for them. The 2% is an average, and the average is not your destiny unless you buy the whole index at its most expensive point. Structure handles the uncertainty better than conviction does: quality growth where the multiple is defensible, paired with lower-multiple cash-generative names whose starting price already builds in the compression BofA is warning about.

The judgment stands — not on whether 7,800 proves right, but on what the forecast presumes. If real AI monetization arrives and multiples hold, the cautious calls get clawed back as fast as they were issued. If the compression shows up, it lands in the places that started most expensive. Either way, the factor question was already the better one to ask.