Earlier than usual end of the year profit-taking. Sector rotation. Stretched valuations. All seem to have played a role in the retreat in momentum stocks that has driven U.S. stocks lower in the last 10 days.
And now we’re getting experts on program trading pointing at the role of computers in the slump and especially systematic trading programs that rely on weighting what the strategy calls “factors” to guide buys and sells.
Andrew Lapthorne of Societe Generale told Bloomberg today that he thinks systemic computer trading has actually played a bigger role in the drop than trends such as sector rotation. In this kind of trading, computers look at factors such as high price-to-earings ratios, or momentum, or leverage to see what combination of factors is performing best in the market. The computers then buy and sell to bring portfolios into line with those optimal factors. Lapthorne told Bloomberg that one sign of the influence of such factor trading is that the selling in the last week plus hasn’t been confined to sectors (so it’s not sector rotation and its not explained by macro changes) and instead bears the hallmark of selling momentum-factor and buying value-factor. Such a shift is visible in Bloomberg indexes that track momentum and value stocks. Through Friday of last week, the Bloomberg U.S. pure momentum portfolio had its worst week since April 2016. And in that same period value had its biggest rebound since March 2009. From this point of view, the dip in stocks has been caused by systemic computer trading strategies selling momentum–across all sectors–and buying value–across all sectors.
This isn’t the first time that the markets have seen such a shift out of one factor and into another. In fact, U.S. markets saw a similar factor-led move in June with technology stocks with high momentum factors selling off heavily.
That history is somewhat reassuring for investors in the current market. It took high-momentum factor technology stocks only about a month to recover their losses.