The Balance of Risks (August 2026)
Equities have pushed to new all-time highs while liquidity stays restrictive and the Federal Reserve holds with a hawkish tilt, and the memo argues that combination reads as a late-cycle expansion rather than one that is ending. The base case into the fourth quarter is a midterm correction that resolves into recovery rather than a recessionary top, because the recession cluster has not formed: year-over-year payrolls remain positive near 0.2 percent, unemployment is contained at 4.1 percent, initial claims sit near 209,000, and the macro recession composite reads 0.024, near the floor of its range. That same correction window is where Bitcoin has historically carved its cycle low and where gold has often bottomed in midterm years, which keeps Bitcoin on bottom-watch near $63,000 rather than at a confirmed low. The principal risk is energy re-accelerating inflation and forcing the Fed to hike into a labor market whose hiring momentum has already stalled, and the view is carried as a distribution rather than a single path: roughly 65 percent late-cycle base case, 20 percent deeper policy-error correction, and 15 percent recessionary transition.
The S&P 500 has pushed to new all-time highs near 7,786, up about 13.7 percent year to date, while liquidity remains restrictive and the Federal Reserve holds its target range at 3.50 to 3.75 percent with three members dissenting in favor of a hike. That combination, restrictive policy alongside loose financial conditions, is the signature of a mature expansion rather than an ending one. Risk has spent two years migrating up the quality ladder: altcoins bled to Bitcoin, speculative growth bled to durable cash flow, smaller caps bled to larger caps, and capital has concentrated into the largest indices. Those indices are the last rung, so new highs at the top of that ladder are what the late stage looks like rather than evidence the transmission failed.
2026 is a midterm election year, historically the weakest of the four-year cycle, and its weakness concentrates in the back half. Across roughly two dozen midterms back to 1930, September is the weakest month and the seasonal low usually falls in the September-to-October window before a fourth-quarter recovery. The three most recent midterms ran later and deeper, drawing down about 7 percent in 2014, roughly 20 percent in 2018, and on the order of a quarter in 2022. 2026 has tracked the earlier, stronger part of that path, front-loading its gains into April and May, and it enters the difficult stretch expensive: the Shiller CAPE sits at 41.2, second-richest in a century behind only 2000, with an excess CAPE yield near 0.97 percent. Rich valuation does not time a correction, but it leaves little cushion beneath the seasonal setup.
The labor market is what separates a correction from a recession, and it has not made the turn. Unemployment has drifted down to 4.1 percent, initial claims sit near 209,000, and job openings per unemployed worker have normalized to about 1.0 rather than collapsing. The caveat is that hiring momentum has all but stalled: year-over-year payroll growth is roughly 0.2 percent, the July establishment survey printed a decline of about 23,000 alongside 103,000 of downward revisions, and part of the drop in unemployment reflects some 264,000 people leaving the labor force. That is a low-hiring, low-firing market, consistent with late-cycle rather than contraction. The live risk is inflation. Headline CPI eased to 3.4 percent on an oil decline that has already partly reversed, crude has recovered into the low-$80s from near $70 in early July, and the GDP-weighted global policy rate has stopped falling as other central banks resume hiking. A Fed forced to tighten into stalling payrolls is the policy error that has historically shortened late cycles.
The cross-asset read follows from that. The dollar carries a bullish lean on the presidential-term analog, energy leadership is intact and has topped after the index in both prior cycle tops on record, and gold's spike to a late-January high near $5,560, fade to a late-June low near $3,970 and recovery toward $4,370 reads as a midterm reset rather than a top. Bitcoin remains bottom-watch rather than a confirmed low, chopping near $63,000 at about four times its last cycle bottom, with composite on-chain risk at 0.181 and dominance at 67 percent, and its cycle low has historically formed during the equity correction rather than ahead of it. The framework carries the view as a distribution: roughly 65 percent late-cycle base case, 20 percent deeper policy-error correction, and 15 percent recessionary transition. A material broadening of breadth, liquidity risk falling out of its tight band, Bitcoin reclaiming its longer-term weekly moving averages, and the dollar breaking to new lows would together argue the expected drawdown simply does not arrive.