Second-quarter S&P EPS results point to $85 of quarterly EPS or $340 annualized

Half of second quarter S&P earnings have been reported so far and the results are terrific.  The final result is likely $85, up nearly 30% y/y, ex. the big securities gain at Alphabet. The season kicked off with impressive results from big banks and Financials appear to deliver 25% earnings growth year-over-year (y/y) on strong capital markets activity and bank loan growth. Such Financial sector profit growth in the middle of an economic cycle would normally grab most investor attention, but even this boom pales in comparison to the 70% y/y growth likely from Tech. Within Tech, the mother of all profit booms is unfolding with Semiconductor profits up 200% y/y. We see this boom as part of a cycle, but with an uncertain peak and duration. This is now a crucial debate for chip stocks and key to forecasting aggregate S&P earnings per share (EPS)  the next few years as analysts expect Semiconductors to be 20% of estimated 2027 S&P EPS.

We raise 2026E S&P EPS to $345 and 2027E EPS to $375 with upside

Given current annualized earnings and the relentless climb in second half quarterly estimates to over $90, we raise our 2026 S&P EPS estimate to $345 from $335 and 2027 estimate to $375 from $365 with upside risk. The bottom-up aggregate of analyst estimates for 2027E S&P EPS is now $405-410. Analysts forecast that Semiconductor profits climb to $685bn in 2027 from about $450bn this year, which is up from $225bn in 2025 and $150bn in 2024. We aren’t saying the analysts are wrong, but if they are right we think it means that semiconductors reach peak profitability earlier than we forecast and we think semiconductor profits likely plateau at that level for 2 or more years. In other words, if 2027 S&P EPS is $400, we think 2028 S&P EPS would be about $400 too and then likely 8-10% growth thereafter. Our S&P 500 intrinsic valuation model includes adjustment for normalized vs. current earnings, but given the record share of S&P profits from a single industry (GICS 3, (Global Industry Classification Standard)) that semiconductors is expected to become, we address this mother of all profit cycles in our headline estimates.

EPS isn’t the only factor in flux in our valuation model: Can’t ignore yields

While we raise our S&P EPS outlook for next year, we also raise our S&P 500 real cost of equity (CoE) estimate. Observed 10-20-year Treasury Inflation-Protected Securities (TIPS) yields have climbed to the highest levels in 25 years reaching 2.4-2.7%. We raise our long-term real risk-free interest rate assumption to 2.0% from 1.75% and maintain a fair S&P 500 Equity Risk Premium (ERP) estimate of 3.25%, thus raising our real cost of equity estimate to 5.25% from 5.0%. Every 25bp change in the real CoE estimate is 5% to intrinsic value all else the same. We estimate fair real CoE with a focus on the sum of the real Rf (risk free rate) and ERP as these two parts tend to be somewhat offsetting with stickiness observed in the whole. However, we might be observing a significant secular shift in normal real interest rates, which appear to be ending the 2009-2020 regime. This is also an important new cycle to evaluate of uncertain peak and duration with the usual challenges in deciphering cyclical vs. secular.

Higher real CoE estimates across the S&P 500 had a more adverse impact to the value of bond substitute equities and slower earnings growers in our framework. Industries with stronger earnings trends and Financials may help mitigate higher interest rates’ impact.

The quality of non-GAAP S&P EPS remains good with a slightly improving trend. Albeit fourth quarter and full cycle results are important to assessing this factor, but it should not be confused with the difference between earnings and free cash flow. Non-GAAP accounting quality issues aside, the difference between earnings and free cash flow is investment activity and the quality of this investment requires consideration of the expected return on such.  Employee stock option expense remains a significant accounting non-GAAP (Generally accepted accounting principles) EPS quality issue, many Software companies still exclude this expense and some Semiconductors exclude it yet some big ones don’t; making for substantial inconsistency in oft quoted Price-to-earnings (P/E) ratios. DWS Americas generally votes against Say on Pay and certain executive compensation proposals when S&P 500 companies exclude employee stock option expense from non-GAAP EPS measures presented in investor communications.

Warsh can’t ignore yields either: Will the Fed repress market rates?

Kevin Warsh revealed quite a bit of his philosophy and convictions about sound monetary policy at Wednesday’s press conference. He tipped his hat to the importance of market signals and observed market Treasury yields across the curve without excessive influence from Fed views. He welcomed markets to “play the ball and not the referee.”  We’re encouraged by how he thinks and this transparency in philosophy, but a difficult courage of convictions test might lie ahead. We note 9 of 12 decided to hold rates, not just the Chair, but unless inflation falls sharply and/or TIPS yields fall sharply by September, we think this Fed runs the risk of being seen as repressing real interest rates from what markets deem appropriate for this cycle. This could lead to wider norms in term premiums that could push long-term yields higher and steady-state PEs lower.