Bond spreads as an equity regime lens

An outside window on a broad equity ETF

A broad US equity ETF can be treated as a market-wide proxy whose level depends on many macro and sentiment drivers. An outside asset class can therefore supply an independent valuation window.

Intermarket analysis uses another asset class, especially credit and Treasury markets, as that independent window on equity valuation and regime. The credit market is framed as a complementary gauge of what bonds are saying about equities. The source workflow cited a US bond market that was more than double nominal GDP.

What a usable equity-level model needs

A usable equity-level model is described as needing a growth factor, a defined equity risk premium, an inflation measure, and both stock and bond valuations. The equity risk premium is the extra expected return demanded for holding equities instead of a safer government bond.

A simple earnings-yield versus 10-year yield comparison, often called the Fed model, treats stocks and bonds as competing income assets. That comparison is presented as breaking down when inflation and rates are low. It also faces a high-growth case that can produce an unusable negative price.

Growth plus three credit-Treasury spreads

The proposed specification treats the equity ETF price as a function of a daily growth term plus high-yield-minus-Treasury spreads across short, intermediate, and long government maturities.

Each credit-Treasury spread is the price gap between a high-yield corporate bond fund and a Treasury fund of a chosen maturity, used as a predictor of broad equity level. Linear regression then fits the equity price to that small set of bond-spread and growth predictors.

Drop overlapping predictors

Backward elimination starts with all candidate predictors, drops the term with the smallest absolute t-statistic, and repeats until every remaining predictor is statistically significant. The loop is meant to leave a set that is significant and less collinear.

Correlation analysis measures how closely credit-Treasury spread pairs move together so redundant predictors can be removed from a joint market model. On the out-of-sample fit, overlapping information appeared as a very high negative correlation between two intermediate and long credit-Treasury spreads and a similarly high negative correlation between another pair. The common weaker term was removed.

Fit through a chosen starting price

The regressions were run through the origin by setting the equity ETF price on the first sample day to zero, then recovering the plotted level by adding back that first-day price. Through the origin is a regression constraint that forces the fitted series through a chosen starting price so later levels can be recovered in the same way.

Editorial reading of the residual

Editorial: once the reduced fit is in hand, TradersWeek reads large, persistent gaps between the ETF and the credit-plus-growth level as a weekly regime overlay. Those residual bands are cross-asset context for the broader tape, not another equity-only oscillator.