Recession Probability Rises as Inflation Expectations Increase

US Treasury yields increased last week after Federal Reserve (Fed) Chair Powell gave a hawkish speech. He stated that the FOMC would hike again if inflation was not on a sustainable path back to the 2% target and that December was a live meeting. Additionally, The University of Michigan’s preliminary reading for its Consumer Sentiment Index dropped to 60.4 (the lowest level since May) and consumers’ outlook for inflation rose MoM. Dominque continues to expect a December hike based on PCE remaining close to the SEP end-2023 forecast, no tightening of financial conditions from current levels, and continued above-trend growth. 

Turning to market moves, US 10Y yields closed the week at 4.61% (+4bps WoW, +3bps MoM). Meanwhile, the yield on the policy-sensitive US 2Y closed the week at 5.04% (+21bps WoW, +5bps MoM). In terms of yield curve inversion, the magnitude of the 2s10s inversion sat at -43bps on Friday, up from year lows of -109bps seen in early July. The probability of recession increases with yield curve inversion. 

The probability of recession within the next twelve months, assigned by the 2Y10Y part of the yield curve, increased to 78% (previously 68%, Chart 1). Meanwhile, the Fed’s recession model, which uses the 3M10Y part of the yield curve, produced a 45% chance of recession (Chart 2). 

Background to Models 

We introduced two models for predicting US recessions using the slope of the US yield curve. When long-term yields start to fall towards or below short-term yields, the curve flattens or inverts. This has often predicted a recession in subsequent months. Our model is based on the 2s10s curve compared to a model from the Fed that is based on 3M10Y curve. We believe that the 2Y better captures expectations for Fed hikes in coming years and is therefore more forward-looking. 

Dalvir Mandara is a Quantitative Researcher at Macro Hive. Dalvir has a BSc Mathematics and Computer Science and an MSc Mathematical Finance both from the University of Birmingham. His areas of interest are in the applications of machine learning, deep learning and alternative data for predictive modelling of financial markets.

(The commentary contained in the above article does not constitute an offer or a solicitation, or a recommendation to implement or liquidate an investment or to carry out any other transaction. It should not be used as a basis for any investment decision or other decision. Any investment decision should be based on appropriate professional advice specific to your needs.)