Extreme short-rate declines as equity regime context

A regime check around one equity idea

An equity thesis can look complete on its own price series and still lack a rate context. Intermarket-analysis reads that equity risk through a contemporaneous move in short-term yields rather than through the equity series alone.

The archive treated short-rate moves and equity prices as inversely related. The lagged-equity-response arrived over later days, weeks, or months rather than on the day of the rate change. Two channels motivated that inverse link: rate moves alter financing costs and earnings expectations, and they shift substitution between equities and debt as relative expected returns change.

Place the move on the intermarket

The paired series were daily three-month Treasury bill secondary-market yields and daily S&P 500 closes. When the Treasury market was closed, the prior bill yield was carried forward. Equity changes were measured as non-annualized geometric returns.

A smoothed-yield-path, a short exponential average of two weeks or less, was applied to bill yields before any extreme was classified. The archive argued that this modest smoother made multi-day readings more stable than raw daily spikes.

Grade the move against historical analogs

Historical-analog-comparison sorts past windows by how rare a rate change was, then contrasts later equity outcomes in those windows with the unconditional sample for the same horizon.

Rate declines were grouped on several lookbacks from a few sessions to a few months. Subsequent S&P 500 outcomes after the rarest prior yield declines were tabulated across several forecast horizons and compared with the sample-wide baselines for the same market.

Lengthening the rate-change window was described as reducing the chance of a large later equity decline more reliably than it raised the typical outcome. A medium multi-week change was the default monitor. A longer window was framed as a downside-filter tradeoff.

Ask where the easing sits in the cycle

An extremes-threshold is a one-sided tail rule. It flags a rate change only when the decline sits in the outer 1 percent, 2 percent, or 5 percent of the historical decline distribution. The rule was applied only to falling rates, not to both tails.

Seasonality-analysis places a rate-extreme episode in a multi-week easing cycle: whether it is a first print, a repeat in a cluster, or a late-cycle echo. On the default monitor, the broader tail declines arrived infrequently and in clusters. They tended to show up in the later stage of a declining-rate cycle. That pattern is a rate-cycle-cluster. The first flag was treated as more informative than later repeats.

Extensions that keep the analog design

Suggested extensions kept the same analog design but changed the input: other points on the Treasury curve or credit yields, the speed of the rate change rather than only its size, and equity variability after sharp rate increases as well as after declines.