Recession Risk 34/100 — May 17, 2026

Recession Risk Score: 34/100 — MODERATE (-13 vs 30 days ago)

Today’s Recession Risk Score is 34/100 (MODERATE), down 13 points from 30 days ago (47 → 34). The signal mix continues to argue for continued expansion rather than an imminent contraction: labor market “real-time” stress markers remain contained and financial conditions remain broadly easy. The score has mean-reverted lower over the past month as the yield-curve backdrop normalized and credit stayed calm, even while a handful of early-cycle “canaries” (temp help, freight, sentiment) keep tail risk elevated. Net: soft-landing baseline, with non-trivial downside asymmetry if labor deterioration spreads beyond cyclicals.

Score Trend — Last 30 Days

The last 30-day window (2026-04-17 → 2026-05-17) shows a clear downshift in risk: Start 47 → End 34 (Δ -13), with a min of 33 and max of 47 across 24 samples. The distribution matters: an average score of 41 with repeated dips into the mid-30s suggests the model is increasingly comfortable that “recession in the next ~90 days” is not the base case.

The shape of the series looks choppy but trending down—a classic “two steps forward, one step back” pattern rather than a straight-line improvement. In the last 10 readings, we saw brief spikes back to 44 (May 9, May 11, May 12) followed by a step-down to 34 (May 13) and then stabilization at 33–34 (May 15–17). That profile is consistent with a regime where macro risk is receding, but headline-sensitive catalysts (energy/geopolitics, inflation prints, Fed speak) can still re-price sentiment quickly without confirming stress in jobs/credit.

Key Drivers

Here are the most important drivers behind today’s 34/100 reading, with the concrete datapoints that matter:

1. Sahm Rule remains decisively SAFE

   • Sahm Rule: 0.13 vs 0.50 trigger (SAFE). This is the single most important “real-time recession start” trigger in the dashboard, and it is not close to firing. With unemployment at 4.3% (WATCH), the gap to the Sahm threshold remains wide—consistent with cooling, not cracking.

2. Yield curve is no longer restrictive (2s10s positive)

   • 2s10s: +0.50 (SAFE). The curve being meaningfully positive removes a major historical recession precursor and aligns with easier forward financial conditions.
   • 2s30s: 1.02 (WATCH)—steepening is not automatically bullish if it’s driven by “cut expectations,” but with credit tight, this currently reads more like normalization than panic.

3. High-frequency labor stress remains contained

   • Initial jobless claims: 211K (SAFE) for the week ending May 9, 2026, still near cycle lows and consistent with limited layoffs. Recent reporting confirms the weekly increase but emphasizes claims remain historically low. (apnews.com)
   • Payroll growth is slower (April +115K) and unemployment has ticked up (4.3%), but so far the “fast” layoff data does not validate a recessionary labor unwind.

4. Credit conditions are not pricing recession risk

   • HY OAS: ~276 bps (SAFE)—tight spreads are inconsistent with near-term recession probability rising. A recent read from market data trackers puts high yield OAS around ~2.75% in early May. (ycharts.com)
   • Chicago Fed NFCI: -0.52 (SAFE)—loose conditions, and (per your own NFCI page) a trend of easier conditions vs mid-April. (recessionpulse.com)

5. The caution cluster: temp help + freight + sentiment

   • Temporary Help Services: 2,485K (DANGER)—a classic leading indicator; weakness here often foreshadows broader payroll softness.
   • Freight Transportation Index: 1.5 (DANGER)—goods-side weakness continues to flag a manufacturing/transport slowdown.
   • UMich sentiment: 53.3 (DANGER) in your tracker, with external reporting highlighting an even weaker prelim May reading of 48.2. (bloomberg.com)
     This trio matters because it’s the channel through which “softening” can become “self-reinforcing” (pessimism → pullback in discretionary spending → layoffs → more pessimism).

Category Breakdown

Using your CATEGORY BREAKDOWN counts:

• Primary Indicators: 4 safe / 4 watch / 1 danger
  Broadly balanced—primary recession markers are not screaming recession, but the watch-list is long enough to keep risk in MODERATE.

• Secondary Indicators: 2 safe / 0 watch / 1 danger
  Secondary signals lean okay, with one important weak spot (notably consumer sentiment).

• Housing & Construction: 1 safe / 0 watch / 1 danger
  Housing is mixed: starts are fine, but permits are soft (warning), consistent with “below-trend growth” rather than contraction.

• Business Activity: 2 safe / 1 watch / 0 danger
  Business activity is holding up; this category is not corroborating a recession call yet.

• Consumer Credit Stress: 0 safe / 3 watch / 1 danger
  This is the macro underbelly: delinquencies, debt service, and savings cushion are all uncomfortable—not an immediate trigger, but a shock amplifier.

• Market Signals: 7 safe / 2 watch / 5 danger
  The market is simultaneously calm on volatility/levels and dangerous on valuation/defensive ratios, creating a “priced for perfection” setup.

• Liquidity: 0 safe / 1 watch / 2 danger
  Liquidity is flashing caution—especially with ON RRP near depletion. This matters because liquidity regimes can flip quickly under stress.

• Real-Time / High-Frequency: 0 safe / 1 watch / 1 danger
  High-frequency data is split—claims are fine, but freight/goods and other fast signals are weak.

Biggest Movers

Top 5 by absolute 7-day % change (and what the move implies for recession risk):

1. ON RRP Facility ($647M): -84.7% (7D)
   Confirmatory (worsening tail risk): shrinking RRP can reflect cash shifting into the system, but “near-zero” also removes a buffer and can interact with funding volatility in a stress event.

2. GDP Growth (QoQ Annualized) (2.0%): -50.0% (7D)
   Confirmatory (worsening): a sharp downtick in growth-rate measures supports “below trend” momentum, even if still positive.

3. NY Fed Recession Probability (3.3%): -30.8% (7D)
   Contradictory (improving): model-implied recession odds falling supports the decline in the overall score.

4. VIX (17.3): +29.0% (7D)
   Confirmatory (worsening): volatility rising from a low base hints at fragility, though 17-ish is still not “stress.”

5. Personal Savings Rate (3.6%): +25.0% (7D)
   Contradictory (improving, but context matters): higher savings can rebuild cushion, but at 3.6% it’s still a thin buffer—one that can reverse quickly if inflation/energy costs bite.

90-Day Indicator Trends

Your 90-day history (as provided) shows a key pattern: financial conditions and “hard stress” indicators are stable-to-better, while select cyclicals and sentiment remain fragile.

Financial conditions & market stress: easy, but not “cheap”

• Chicago Fed NFCI stayed consistently loose, hovering roughly -0.57 to -0.51 from late Feb through mid-March—today -0.52 (SAFE) is still “easy money” in composite terms.
• HY spreads were ~294–312 bps in late Feb/early March (watch-ish in your labels), widening briefly to ~319 bps mid-March before easing. Today’s ~276 bps implies tightening of spreads vs those late-winter levels, a material improvement in recession pricing.
• VIX rose meaningfully in early March (peaking near ~29.5 in your series) and then cooled into the high teens. Today’s 17.3 is low, though the +29% 7D jump warns that complacency can break quickly.

Labor: stable, with leading-edge yellow flags

• Initial claims were remarkably stable around 206–213K from late Feb through mid-March, and remain 211K now—no trend break yet.
• Unemployment rate in your historical snippet moved from 4.3% into 4.4% in early March (then holding). That’s consistent with “cooling,” not recession.
• Sahm Rule moved down from 0.30 (WATCH) to 0.27 (SAFE) in early March in your history, and is 0.13 today—still safely below trigger.

Growth, housing, and goods: mixed-to-soft

• GDPNow (external) was 3.5% on May 1 for Q2 (Atlanta Fed). (atlantafed.org) Your internal GDPNow reading is lower (1.8% today), implying growth is positive but potentially below trend.
• Freight index fell sharply in early March in your history (from 1.3 to -0.5) and remains a danger signal—this is one of the clearest “real economy” weak spots.
• Building permits dropped from 1448K to 1376K in mid-March in your history (watch → warning), consistent with softer forward housing momentum even while starts are okay.

Consumer: the tail-risk channel

• Consumer sentiment in your history sits at 56.4 through late Feb/mid-March; today your tracker shows 53.3 (DANGER), while external reporting underscores a much weaker preliminary May read of 48.2. (bloomberg.com)
  Whether this pessimism bleeds into actual spending is the key swing factor for the next 30–90 days.

Stock Screener Signals

Today’s screener flags cluster into two themes: (1) high-yielding “value dividend” names and (2) a small pocket of “oversold growth.” In a recession-imminent regime, you’d typically expect a heavier skew toward defensives with balance-sheet quality and low beta; here, the list looks more like income/valuation hunting with selective mean-reversion trades.

The “value dividend” basket—ARCC, AIG, BBY, FNF, HMC, T, LTM, BCE—suggests the market is positioning around carry and valuation support rather than pricing widespread default stress. That is directionally consistent with tight HY spreads and loose NFCI: investors are still willing to own credit-like equities and dividend vehicles. The big caveat: the reported yields in the screener output (e.g., ARCC “1002%”) are almost certainly data artifacts; interpret the signal category (value/dividend) rather than the literal yield.

Meanwhile, “oversold growth” flags like CHTR (RSI 28) and TLK (RSI 30) imply selective risk appetite for rebound setups even as macro is noisy. In practice, this combination—income/value preference + pockets of oversold growth—is consistent with a moderate-risk, late-cycle expansion mindset: investors want to be paid to wait, but they aren’t abandoning cyclical risk entirely.

Latest Economic Developments

In the past few days, the hard data reinforces the model’s central message: the labor market is cooling at the margin but not deteriorating in a way that typically precedes recession within ~90 days.

• Initial jobless claims rose to 211,000 for the week ending May 9, 2026, up from the prior week’s revised level, but remain historically low—multiple reports framed it as “up, but still low.” (apnews.com)
• Consumer sentiment has been reported as extremely weak: Bloomberg reported preliminary May sentiment at 48.2 vs 49.8 in April—a level consistent with acute pessimism. (bloomberg.com)
• Atlanta Fed GDPNow (as of May 1, 2026) estimated Q2 real GDP growth at 3.5% (SAAR), down slightly from the prior day—still a pro-growth signal, though your internal tracker is more cautious. (atlantafed.org)
• Inflation tone remains a swing variable. Recent coverage of the April 2026 CPI emphasized a hotter print (reported 0.6% m/m; 3.8% y/y) amid energy-price pressure—important because sticky inflation can prevent the Fed from easing quickly if growth wobbles. (kiplinger.com)

Near-Term Outlook (Next 30 Days)

Base case for the next month: risk score stabilizes in the low-to-mid 30s unless labor data weakens materially or credit spreads widen.

What could move the score higher (worse) quickly:

• Claims trend break: a sustained move toward ~250K+ (your stated threshold logic) would likely flip high-frequency labor from safe to watch/danger.
• Credit repricing: HY OAS widening from ~276 bps toward 350–400 bps would be a meaningful “recession risk is getting priced” regime shift.
• Consumption confirmation: if weak sentiment translates into weaker retail/control-group spending, recession tail risk rises fast given low savings cushion.

What could move the score lower (better):

• Continued stable claims around ~210–220K.
• Sentiment rebound (even modestly) alongside steady real income.
• Further easing in financial conditions without a valuation blow-off (hard to achieve, but possible if earnings breadth improves).

Calendar catalysts to watch over the next 30 days:

• The next weekly jobless claims prints (each Thursday) and any notable revisions.
• Final University of Michigan sentiment (late May) as a check on whether prelim pessimism was overstated.
• Fed communication (speeches/minutes) for evidence that policy reaction function is shifting given inflation/energy dynamics.

Long-Term Outlook (3-6 Months)

Over the next 3–6 months, the macro picture is best described as “expansion with fragile internals.” The stabilizers are real: a positive 2s10s curve, tight credit spreads, and low claims all argue against an imminent recession. But the destabilizers are also real: temp help contraction, freight weakness, crisis-level sentiment, and thin household buffers (low savings, rising delinquency stress).

The 90-day pattern in your indicators suggests the economy is not in a classic pre-recession cascade yet (where claims rise, spreads gap wider, and financial conditions tighten in unison). Instead, the setup resembles periods where the economy can keep expanding while risk builds under the surface—until a catalyst (energy shock, policy mistake, credit event, geopolitics) forces a rapid repricing. In those regimes, recessions don’t arrive because the economy is already broken; they arrive because the system is less resilient to shocks.

Historically, when credit stays tight and claims stay low, recessions usually require either:

• A sudden financial tightening (spreads + liquidity), or
• A sudden labor inflection (claims + unemployment) that becomes self-reinforcing.

Right now, your dashboard says neither has happened—yet.

What to Watch

Concrete thresholds/events that would most likely move the needle:

• Initial jobless claims
  • Watch for a 4-week trend pushing and holding >235K, with ~250K+ as a clear escalation level.
• Sahm Rule
  • Anything moving materially toward 0.50 would change the regime; today’s 0.13 is comfortably safe.
• High yield spreads
  • A move from ~276 bps to >350 bps would be an early warning; >450 bps would be a serious stress signal.
• Financial conditions (NFCI)
  • A reversal from -0.52 toward 0.0 would indicate tightening consistent with recession risk rising.
• Temp help & manufacturing employment
  • If temp help keeps falling and manufacturing employment rolls over decisively, layoffs can broaden beyond cyclicals.
• Consumption reality check
  • If sentiment stays depressed and real spending slows, the probability of a fast deterioration rises.

Sources

No data available for this window.