Month-End US Pension Rebalancing: $14bn Equity Sell / Bond Buy

The month-end pension rebalance adds a modest but real supply risk into an otherwise constructive equity tape. The desk’s model estimates $14bn of US equities to sell from US pensions into month-end, with an equivalent amount of bonds to buy, reflecting the fact that equities have materially outperformed fixed income in May.

Month-to-date, the relative performance gap is large:

- **S&P 500 total return:** +3.77%

- **10-year Treasury total return:** -1.03%

- **Equity outperformance versus fixed income:** +4.80%

That equity outperformance mechanically pushes balanced pension portfolios above target equity weights, requiring sales of equities and purchases of bonds to rebalance.

## How the estimate evolved

The rebalance estimate started May as insignificant. During the first week of the month, it remained small until the 1.46% equity move higher on May 6, which pushed the model to a modest $3bn equity sell estimate.

From there, the estimate grew in a relatively linear fashion as equities continued to outperform bonds. By May 18, the model entered double-digit territory, with roughly $12bn of equities to sell.

More recently, the estimate has fluctuated in a $3-4bn range and currently stands at $14bn of US equities to sell, with a matching $14bn of bonds to buy.

## Historical context

The $14bn sell estimate is meaningful, but not extreme.

On an absolute dollar basis, across both buy and sell estimates:

- It ranks in the 69th percentile over the past three years

- It ranks in the 80th percentile going back to January 2000

On a net basis, however, the sell pressure is less extreme:

- It ranks in the 25th percentile over the past three years

- It ranks in the 13th percentile going back to January 2000

The reason for the difference is that $14bn is a relatively large absolute rebalance, but as a directional equity sell estimate it is not especially large compared with the biggest historical sell events.

The desk also notes that this is the 12th largest non-quarterly sell estimate on record since 2000. Historical returns around the 11 larger non-quarterly sell estimates are mixed, though the three-day pre-rebalance return was positive nearly three-quarters of the time.

## Market implications

This flow is a headwind into month-end, but it does not appear large enough by itself to overwhelm the broader tape. It should be understood as incremental supply, not a stand-alone bearish catalyst.

The context matters:

### Supportive forces

- Geopolitical headlines have improved, supporting risk appetite.

- Buybacks remain active, with 99% of S&P 500 corporates in the open window.

- Hedge funds have been buying again, especially macro products, discretionary, and tech.

- Short exposure in US index and ETF products is at a 10-year high, creating squeeze risk.

- Software, most-shorted names, and unprofitable tech are catching a bid.

- VIX has fallen toward 15.7, and front-end implied vol remains soft.

### Offsetting headwind

- Month-end pension rebalancing implies $14bn of equity supply.

- That equity sale is paired with $14bn of bond demand, which may modestly support Treasuries.

- If bonds rally further and yields fall, that could partially cushion the equity impact by supporting duration-sensitive growth and software.

## Link to current tape

The rebalance estimate lands at an interesting moment. Equities have just rallied on hopes for a US/Iran de-escalation and a broadening software/short-covering squeeze. The S&P 500 closed at 7,564, and the near-term implied move through tomorrow was roughly 0.47%, or about 36 points, implying a range around 7,528 to 7,600.

The pension rebalance could make it harder for the index to break meaningfully above the top of that implied range into month-end, especially if discretionary sellers and asset managers remain net sellers. But the associated bond buying could help keep the 10-year yield contained, which would be supportive for the same parts of the market that have led the recent broadening: software, consumer discretionary, and higher-duration growth.

The model points to $14bn of US equity selling and $14bn of bond buying from month-end US pension rebalancing. This is a meaningful flow, ranking in the 80th percentile in absolute terms since 2000 and the 12th largest non-quarterly sell estimate on record, but it is not extreme on a net historical basis.

For trading, it argues for expecting some month-end equity supply and potential bond support, rather than a major de-risking event. The bigger directional driver remains geopolitics and positioning. If US/Iran headlines continue to improve, the market can likely absorb the rebalance. If headlines deteriorate, the rebalance could amplify near-term downside pressure.