Two of the largest US advisers booked a fifth more in merger fees this quarter, and the queue of deals behind them got shorter.

What did two of the largest US investment banks report for the third quarter?

Higher merger fees and lower underwriting. Goldman Sachs earned $1.654 billion in advisory fees in the third quarter of 2026, up 20 percent from the second quarter and 18 percent from a year earlier; JPMorgan earned $1.222 billion, up 21 percent and 32 percent (Goldman Sachs, Form 8-K, 13 October 2026; JPMorgan Chase, earnings release and financial supplement, 13 October 2026).

Every other fee line went the other way on the quarter. Equity underwriting fell 32 percent at both banks, to $666 million at Goldman and $564 million at JPMorgan. Debt underwriting fell 12 percent at Goldman, to $904 million, and 15 percent at JPMorgan, to $1.219 billion. Total investment banking fees were $3.224 billion at Goldman, down 5 percent on the quarter, and $3.005 billion at JPMorgan, down 8 percent (same filings).

Against a year earlier the picture is stronger. Goldman's investment banking fees were up 21 percent and JPMorgan's up 14 percent, and across the first nine months advisory fees were up 34 percent at Goldman and 42 percent at JPMorgan (same filings). Goldman's filing calls the advisory gain a reflection of "an increase in completed mergers and acquisitions volumes".

The two prints are not quite the same story. For Goldman the third quarter was its strongest advisory quarter of 2026: our arithmetic from the nine-month total puts the first quarter near $1.494 billion, above the second quarter and below the third. For JPMorgan it was not: the bank earned $1.266 billion in advisory fees in the first quarter of 2026, more than the $1.222 billion it booked in the third (JPMorgan Chase, financial supplement, 13 October 2026). Our arithmetic: advisory was 51 percent of Goldman's banking fees in the quarter against 41 percent in the second, and 41 percent of JPMorgan's against 31 percent.

Why does a strong advisory quarter describe the spring rather than the autumn?

Because merger fees are paid when deals close, and deals close months after they are signed. Goldman's own explanation for its advisory gain is completed volume, not announced volume (Goldman Sachs, Form 8-K, 13 October 2026). A transaction that pays a success fee in September was typically agreed in the spring and spent the summer in regulatory review, financing and closing conditions.

Underwriting works on a shorter clock. An equity or bond issue is priced and paid in days, so the underwriting lines are close to a live reading of the capital markets in the quarter itself. Both banks' equity underwriting fell by about a third from the second quarter, and Goldman attributes its debt underwriting to investment-grade activity, which is not the market most private companies borrow in (same filings).

Read together, the lines describe a market that closed a strong spring and then slowed. JPMorgan's chairman said as much in the release: client activity was strong, "although activity did moderate somewhat during the quarter" (JPMorgan Chase, earnings release, 13 October 2026). The fee jump and the moderation are the same quarter seen at two points on the deal clock. That is our reading of the filings, not the banks' framing.

For an owner, the practical consequence is that a headline about surging merger fees is not evidence that buyers are hungrier this month than last. It is evidence that deals agreed earlier in the year got done. The forward measure is elsewhere.

What does a shrinking backlog tell a seller planning for 2027?

That fewer mandates are queued at the largest adviser than three months ago. Goldman reported that its investment banking fees backlog "decreased compared with the end of the second quarter of 2026 and decreased slightly compared with the end of 2025" (Goldman Sachs, Form 8-K, 13 October 2026). The backlog is the bank's estimate of future fees from transactions it is already working on.

The filing does not give the size of the backlog or of the decline, and it labels the measure a forward-looking estimate: transactions in it may be modified or may not complete. So the direction is the signal, not the magnitude. A backlog that shrinks in a quarter of strong completions means more deals left the queue by closing than entered it by signing.

For a seller that cuts both ways, in our judgement. A thinner queue means less competition for buyer and lender attention when the next wave of processes launches, and the large sponsor-owned assets that crowd a market are not yet lined up behind it. It also means boards and sponsors chose not to launch in late summer, which is a statement about the price they expected to get.

Neither reading argues for waiting on the banks. The backlog moves on the decisions of large companies and large funds; a privately held business is sold on its own preparation clock, which runs from clean accounts and a credible forecast to buyer outreach. An owner who starts that preparation now arrives in the market when the queue refills, rather than joining it at the back.

What should an owner take from a quarter that moderated in September?

Treat it as a timing signal, not a valuation signal. Neither filing says anything about the price paid for private companies, and the year-on-year figures, banking fees up 21 percent at Goldman and 14 percent at JPMorgan, describe an active market rather than a closing one (same filings).

Three things follow, in our judgement. First, a process that launches in the first quarter of 2027 meets buyers whose own pipelines were thinned by autumn closings and who have capital to put to work; that is a better audience than one distracted by a full calendar. Second, the softer underwriting lines say financing markets were less open in the quarter than merger fees suggest, so the debt package behind any bid deserves early attention. Third, the moderation was described as somewhat; the banks did not report a stop.

The figures that would change this reading arrive in January, when both banks report the fourth quarter and Goldman states whether its backlog grew again. Until then, the September pause is a reason to prepare quietly, not a reason to rush or to wait.