Bond Short Squeeze Ignites Cross-Asset Momentum Strategies, CTA Weekly Losses Hit Ninth-Percentile Low
nashnova research
A wave of bond short-covering has dragged the cross-asset momentum factor down over 4% in five days, pushing CTA strategies to a ninth-percentile weekly loss — Nomura's McElligott calls it a crowded-position unwind, not a regime change.
Why did "hot" inflation data push bond prices up, not down?
Nomura cross-asset strategist Charlie McElligott offers a counterintuitive explanation: above-consensus CPI prints gave bond shorts a clean window to lock in profits — better to exit on friendly data than risk a weak print forcing a scramble.
This means → last week's string of hot inflation readings coincided with the 10-year Treasury yield falling to a three-week low — not because bulls were buying, but because shorts were voluntarily leaving.
In plain terms = the shorts weren't beaten; they chose to cash out while ahead. Too many ran for the exit at once, and the stampede became a squeeze.
How violent was the bond reversal?
Nomura's trend model now flags virtually every major bond contract as "reversing." The long end took the biggest hit: the 30-year Treasury future returned 3.25% in one week; the 20-year returned 2.77%.
Yet these same contracts still show negative tracking Sharpe ratios — a measure of return relative to volatility — across every horizon from three months to five years. This means → years of accumulated short-side profits were abruptly interrupted — not a small pullback, but a trend-level reversal.
The same reversal signal has spread to G10 FX and short-term rate futures — EUR/USD, GBP/USD, SOFR, and euro-area overnight rate futures all show similar position unwinds.
Inside equities, what's rising and what's falling?
Stocks are showing "reverse dispersion": the equal-weight index outperformed the S&P 500 by 2.2 standard deviations over five days, and the short-term reversal factor gained 1.8 standard deviations — put simply, last week's losers are this week's winners, and vice versa.
The high-momentum-minus-low-momentum factor dropped 2.5 standard deviations, roughly 6.8% in one week. The U.S. equity momentum factor fell 5.87% weekly and 7.08% month-to-date, nearly erasing its 7.29% year-to-date gain.
In plain terms = McElligott describes it as "everything outside AI" beating the former AI winners — the market's most crowded momentum trades are unwinding all at once.
How much did CTAs lose, and where does that rank historically?
Nomura's CTA replication portfolio lost 2.0% last week — a ninth-percentile outcome, meaning only 9% of weekly returns in history were worse. Société Générale's trend index fell 1.7% over the same period.
Attribution breakdown: since October 1, bond positions alone account for roughly 1.5 percentage points of drawdown; equities, short-term rate futures, and G10 FX added smaller drags. The trailing one-month return has turned negative, though year-to-date gains still stand at 12.7%.
This reflects the core vulnerability of trend-following strategies — CTA logic is to ride momentum, which means they earn from trends and absorb the full shock when trends snap.
Is this a temporary shake-out or the end of the trend?
McElligott's explicit call: this is not a macro regime change. It is a "crowded reversal of high-Sharpe trend trades" — a crowded, long-profitable, low-volatility set of positions whose profit-taking snowballed into a broader deleveraging impulse.
He flags one tail risk: a growth-scare scenario. Leveraged-fund net positioning has fallen to one-year, three-year, and five-year lows — deleveraging pressure has not fully cleared.
This means → the key verification point is whether momentum factors can rebuild positive Sharpe after the deleveraging ends. If they can, this was a crowded-position flush. If they cannot, the trend may genuinely be turning.
市场有风险，内容仅供研究参考，不构成投资建议。