ML K5/H60 · 0.30 Hedge

The identical ML K5/H60 book with a standing short of 0.30 SPY units per dollar of book value. A long/short structure is often assumed to be market-neutral; this one is not — the raw book carries beta 0.614 because the long leg is systematically higher-beta than the short leg. This series tracks the configuration that would actually be traded rather than inferring it from the unhedged one.

• Overlayshort 0.30 × book value in SPY shares, resized at each monthly cohort rebalance, not chased daily
• Effect on riskbeta 0.614 → 0.314; average loss on the 10 worst SPY days −1.60% → −0.59%; worst single day −4.49% → −3.02%
• Costbacktest Sharpe 1.99 → 1.86, return +35.4% → +29.1%/yr. Full neutrality (0.61 units) would cost 0.46 Sharpe for little further vol reduction, so 0.30 is the efficient point on the curve
• At $10k$5k long + $5k short stocks + ~$3k short SPY = $13k gross, ~$6.5k Reg T margin
• Honest caveatthe hedge is a linear overlay on the same paper book — it inherits every caveat of ML K5/H60, including that the backtest headline should be haircut roughly in half. Paper only, no orders.