BofA: With Rise Of Quants Comes Meltdown Risk

“Everyone is becoming a quant,” Savita Subramanian, the Head of US Equity and Quantitative Strategy at Bank of America Merrill Lynch, wrote in a 247 page “Quantitative Primer 2017.”  As quants begin to take on an increasingly dominant position in the investing landscape, what are the ramifications?

BofA: With rise of quants, risk of global financial meltdown is increasing

Available job postings for data scientists and quantitative analysts significantly outnumber those for fundamental analysts by a factor of eight, Subramanian wrote in the report. But is the quant space getting “crowded?”

BofA’s quantitatively orientated clients use three times more factor components today than they did 20 years ago, when the nascent industry was little known or respected. But market cycles change and often over-correct, which is what could be happening today.

“One of today’s greatest market inefficiencies may stem from the scarcity of capital devoted toward long-term, fundamental investing,” the report stated, pointing to “a seismic shift” in investing resources toward systematic strategies, many of which rely on shorter-term time horizons.

With quants increasingly looking to real-time data feeds, big data and machine learning, the landscape has become more competitive, as market opportunity can be more fleeting. But with this migration to quantitative strategies, there is a concern the BofA quant notes.

“The risk of an August 2007-like meltdown may be increasing,” Subramanian wrote, pointing to a herd of quants pointing in the same direction.

How can investors combat this problem? “Being different from one’s peers is of paramount importance,” he wrote, outlining in his report exactly where the herd was headed and how to take a different path.

 

Earnings vs risk-based value metrics

When looking at the diverse market panoply of quantitative equity investing strategies over time, different factors move in and out of vogue, but valuation at some level has typically been involved to various degrees.

Value factors are not designed for market timing nor are they precise at modeling short-term three month valuations, for instance. They should be expected to operate over a longer period of time. Valuation factors also carry different weight and meaning in a formula relative to the industry.

Price earnings ratio has been a value factor that has different relevance depending on sector and time frame consideration. One method to account for single-period bias is to estimate the underlying earnings power based on the historical trend, adjusting for cyclicality, as Subramanian explains:

We estimate normalized earnings based on a linear log normal regression and our analysis shows that this measure of market valuation explains over 80% of the variability of equity market returns over the next 10 years. In the late-1990s, equity valuations were near peak levels, and we subsequently saw negative returns over the following decade. In contrast, valuations in the wake of the financial crisis reached extreme levels far below those seen during the 1980s and 1990s, and were similarly followed by strong equity market returns. Over the last two years, valuations have returned to levels in line with the history, and the S&P 500’s current normalized PE ratio of 20x suggests annualized 10-year price returns of +7%, which would represent more than a doubling of the market’s current levels.

There is typically no one valuation metric that handles all complaints. While normalized PE ratios account for the single-period bias well, it is a backward looking metric that does not account for changes in the risk free rate of return, the report noted.

This begins to move valuation methods into the realm of risk / reward models, which is where the Equity Risk Premium model enters the discussion. This look at valuation considers the potential risk associated with a market and puts a value on compensating investors based on the risk taken.

“When investor fear is high, and the market perceives equities to be very risky, the equity risk premium is high to compensate for higher perceived risk,” the report outlined, pointing to a similarity to volatility as a measure of risk.  “When the risk premium is rising, this typically coincides with higher quality investments outperforming, and when the risk premium is falling, this typically coincides with lower quality investments outperforming.”