Night Watch Fund Posts Modest Q3 Gains Amid Market Volatility

Night Watch Investment Partners LP returned 0.84% net of fees during the third quarter of 2026, with strong July performance offset by September weakness as inflation concerns drove risk-off sentiment. The fund maintains a strategy of roughly 120% gross long exposure and 95-100% net positioning, with refiners held as a geopolitical hedge. Marex was the top performer during the quarter while vocational training company Universal Technical Institute was a significant detractor.
Night Watch Investment Partners operates with a portfolio structure designed to reduce dependency on broad market movements. The fund maintains long positions totaling approximately 120% of assets while offsetting roughly 20% through short positions in market indexes, resulting in net exposure between 95–100%. This approach aims to generate returns independent of overall market direction, a challenging objective given current trends toward increased correlation among equities.
The fund's strategy includes holding refiner stocks as insurance against geopolitical escalation in regions including Ukraine and the Middle East. Year-to-date performance through October 2026 shows significant volatility, with strong gains in January and April partially erased by declines in March and September. Since inception, the fund has achieved approximately 18.45% annualized returns, reflecting both successful stock selection and periods of market-driven headwinds.
This fund's performance narrative may interest investors evaluating alternative investment strategies during periods of elevated market uncertainty. The fund's emphasis on counter-cyclical holdings and geopolitical hedging strategies could appeal to those concerned about portfolio concentration risk. However, the modest quarterly returns despite active management raise questions about whether such strategies consistently outweigh their costs, potentially affecting allocation decisions by institutional and retail investors seeking diversification beyond traditional index-based approaches.