(Bloomberg Intelligence) — This autumn’s weakness in stocks may be quickly forgotten in 2022, as a steady hand at the Fed and faster-than-expected economic recovery could lead to the S&P 500 stretching above 5,000 within 12 months in the most bullish scenario of our macroeconomic model. A 4,628 base case also assumes recent weakness will pass, but with the Fed set to detract from multiples, earnings are the source of optimism. Higher-for-longer inflation remains a risk.

The economy’s continued improvement should spur better-than-average EPS growth in the year ahead, pushing the S&P 500 higher despite multiples that are topping out. Our base-case scenario assumes the consensus view of a modestly higher two-year Treasury yield, some yield-curve steepening and falling corporate credit spreads to arrive at 22.6x P/E in the next 12 months. Earnings growth of about 14% seems achievable, given strong new orders and falling unemployment.

We believe the primary risks to our model are a hawkish shift in monetary policy and stronger-for-longer rate pressure. If the Fed tapers quickly and hikes rates during 2022, higher bond yields could sink equity multiples. At the same time, persistently high costs may pressure bottom-line results and limit market upside.

The Fair Value Model sees solid EPS gains in the year ahead in both our base and bullish scenarios, while even a bearish scenario of hardly any economic improvement suggests moderate earnings growth. The base case assumes 5% new-orders growth, a 28-bp change in the two-year Treasury yield, a 100-bp decrease in the 12-month moving average of the unemployment rate and 5% commodity growth, arriving at 13.8% earnings growth in the year ahead. Our bullish case pushes new-orders growth to 7% and bumps up yield forecasts while suggesting stronger-than-consensus employment gains and 10% commodity growth could lift EPS 24%.

A weak recovery scenario based on only 2% new-orders growth, no change in unemployment, a drop in commodity prices and no change in rates suggests a 4.8% EPS gain.

The three scenarios for S&P 500 multiples suggest the market’s P/E has little room to grow over the next year, but it could be aided by improved long-term earnings-growth prospects or continued compression in credit spreads. Our base case assumes consensus on interest rates, falling corporate credit spreads and 10% EPS growth from 2H22-1H23. These inputs suggest a 22.6x trailing P/E for the S&P 500 in the year ahead, about 1.7 turns below the index’s current 24.3x valuation.

Sharply improved economic growth in 2022, which would push rates higher, could be easily digested by the market if EPS growth follows suit. Our bull case assumes a 75-bp two-year rate and a wider-than-expected yield curve to arrive at 23x. A bearish case — featuring no EPS growth in 2H22-1H23 — could pressure that valuation down to 19.3x.

S&P 500 gains may be a bit below those implied by the aggregated bottom-up consensus, according to our baseline scenario, but likewise, stocks could surpass the 5,000 mark if economic progress surpasses our relatively conservative assumptions. Our base case of 4,628 for the index trails the aggregated sell-side target of 4,979, which is bolstered by expectations of 19.1% EPS growth year-over-year vs. our model’s implied 13.8% over the next year. However, our regression’s bullish EPS case — for 24.3% — suggests upside to 5,155 in the year ahead — a gain of 18.4% if a rise in rates is accompanied by improving fundamental expectations.

If the economy sputters as the Fed becomes more hawkish, stocks could fall back closer to our bear case of 3,600.

The equity-risk premium (ERP) side of our model has typically tracked close to reality, but during the 2020 Covid-19 recession and recovery, our estimates suggested stocks could become even more expensive than they actually were. Over the past 20 years, the ERP model has undershot observed risk premiums — defined as the difference between the S&P 500’s earnings yield and the yield on the two-year Treasury note — by an average 13 bps, implying the model slightly overvalued stocks vs. reality. However, since the end of 2019, that underestimation averaged 236 bps, as a stronger and quicker economic recovery than expected, coupled with rock-bottom two-year yields, kept the ERP lofty.

Since 2001, our risk-premium model has a mean absolute error of 1.15% and an r^2 of 0.62.

The EPS part of our model has closely aligned with actual results since 2001. Over the past 20 years, our EPS growth model has averaged a minimal miss of actual trailing 12-month earnings growth. However, since the end of 2019, the model has overshot by an average 386 bps as economic consensus suggested a slightly stronger EPS recovery than actually materialized.