Gundlach Sees Yields Climbing. Sort Your Stocks by Duration, Not by Story.

The 10-year Treasury pushed past 4.8% this week, its highest reading since late 2023. That alone deserves your attention even if you never trade a bond, because the 10-year is not really a bond thing. It is the market's discount rate — the number used to mark down dollars you expect to collect far in the future. When it climbs, the present value of distant earnings shrinks, and that re-pricing runs through every stock you own.

A climbing 10-year is a re-pricing lever, not a bond story

That is the transmission belt behind the call from Jeffrey Gundlach, founder and CIO of DoubleLine Capital, who argues the long stretch of falling interest rates is finished and that long-term yields have further to climb. In late July, when the Federal Reserve left its benchmark rate at 3.5% to 3.75%, the 30-year Treasury jumped to 5.213% — the highest since 2007 — and Gundlach read the move as the bond market telling the Fed it would have to actually hike to prove it is serious about 2% inflation. "If you really want to get to 2%, I think you have to raise interest rates," he said.

Whether his forecast is right is not the most useful question for an equity investor, because the mechanism works regardless of the exact endpoint. Yields are the "r" in the present-value math every stock runs through. The productive question is which stocks the rising discount rate hits and which it passes over. The answer sorts by something called duration: how much of a company's value sits in earnings far in the future versus cash it produces in the next year or two. The further out the promised dollars, the more they are discounted by a higher yield, and the harder they fall.

Duration is the factor that sorts winners from losers

Run the current market through that lens and the spectrum is easy to see. On the long-duration end sits a high-multiple growth name like Palantir, trading at roughly 135 times trailing earnings and 66 times sales with no dividend — a price that assumes enormous growth years from now, so nearly all of its value lives in the discount rate's crosshairs. On the short-duration end sits a bank like JPMorgan: about 14.8 times trailing earnings, an 18% return on equity, and a dividend raised for 14 consecutive years. The whole big-bank group trades at modest multiples — 12 to 15 times earnings across JPMorgan, Bank of America, Wells Fargo, and Citigroup — because banks convert rates into earnings now rather than decades from now.

Here is where the "safe income" instinct leads you astray. When yields rise, the default move for a dividend investor is to reach for what feels defensive, and utilities are the classic stand-in. But NextEra Energy trades at about 30 times forward earnings with a roughly 2.9% dividend, and its steady regulated cash flows are priced off long-term rates. It behaves more like a long bond than a banknote. What actually looks rate-sheltered under a climbing 10-year is near-term cash flow: the financials and earning machines that deliver value now instead of promising it later. High yields compress the growth sleeve and flatter the cash-flow sleeve — that is the re-ranking the market quietly does for you.

Build a barbell, not a forecast

Which is why the disciplined response to Gundlach's call is structure, not more conviction. I cannot out-forecast the long end, and I treat anyone who claims they can with some skepticism. But I do not need the exact endpoint to act on the direction: higher yields are a regime for the discount rate, and a regime change calls for a barbell, not a louder prediction. On one side, keep the rate-sheltered near-term cash-flow names as the hedge. On the other, hold quality growth — sized by how much duration you can genuinely stomach. And notice that even the low-duration sleeve wobbled on the day the yield spiked: the banks pulled back with the broader market. Nobody escapes the day-to-day noise; a barbell is a position that absorbs it, not a forecast that avoids it.

The point of a systematic screen in this regime is to keep the comparison set honest. Do not reach for a name because it pays a dividend, and do not avoid a growth stock simply because the yield is high. Rank by duration first, then by the factor stack — valuation, growth, profitability, and the earnings-revision trend that confirms the timing — and update the rank as the yield moves. If yields keep climbing, keep re-ranking rather than defending a prior belief. In this process a rating is always today's rating, and the yield curve is simply the largest input not standing still.