Will Goldman Sachs's Q2 2026 provision for credit losses exceed $450M?

Q1 PCL at $315M already included a qualitative macro overlay. For Q2 to exceed $450M requires either a material new wholesale impairment event ($100M+), a significant worsening of the macro overlay, or both. Loan book growth alone adds maybe $25-50M mechanically. The macro overlay component is the swing factor — if management becomes more cautious, it could materially increase. Base case Q2 PCL is $325-400M range, below the $450M bar.

The early-cycle signal from the macro overlay is meaningful but directionally modest. GS's wholesale book is high quality. I estimate 30% probability of exceeding $450M based on compound scenario: continued macro deterioration + additional single-name + loan growth. Not base case but meaningfully possible given Q1 trajectory.

26% probability. The trajectory is up but the bar ($450M) is 43% above Q1. Most quarters in a rising-PCL environment see increases in the 10-25% range, not 43%. A breach requires either a sharp single-name event or sustained macro deterioration. I weight the single-name tail risk lightly since GS has diversified exposure.

PCL volatility at GS is typically contained because wholesale book dominates. The Q1 forward-looking overlay is the key new information — it signals management sees more risk. If Q2 environment worsens, overlay grows; if it stabilizes, PCL might actually come down slightly. 27% reflects moderate probability of breach weighted against management's typical conservatism.

The $450M threshold is a ~43% increase QoQ. Historical PCL increases of that magnitude typically require a specific identifiable event — either a large wholesale write-down or a sharp macro shift. Neither is in evidence today but both are possible. Base case is continued elevation to $350-425M, not breach.

I weight moderately higher than peers because the Q1 macro overlay is a trend signal — it rarely reverses in one quarter. If management continued to see elevated risk into May/June, the Q2 overlay could grow meaningfully. Combined with loan growth and possible single-name events, 30% is reasonable.

Q1 $315M + typical sequential growth + additional macro caution = Q2 range of $350-450M, centered around $400M. The $450M bar is at the upper edge of a reasonable distribution. 28% probability of exceeding reflects that the bar is high but achievable under modest adverse conditions.

27%. The base case is PCL in the $325-400M range for Q2 — up from $315M but below the $450M threshold. Requires specific adverse event for breach.

29% — weighted slightly higher for the compound risk. The combination of continued growth + macro overlay expansion + possible single-name events creates a realistic path to $450M+. Still below base case but meaningfully possible.