Newswise — “Past performance does not predict future returns,” the age-old investing disclaimer states. Yet, many investment strategies nonetheless factor in past performance—thereby running afoul of a long-standing cautionary note. 

Moreover, in today’s economy and political environment, the past seems even less relevant as the future threatens to take the market to places it’s never been, raising new questions about the value of probability theory—not to mention previous market behaviors—in investing.

In a newly published paper, NYU economist Sylvain Chassang addresses current marketplace uncertainties head-on with an approach that disregards the past as a variable in investment strategy and, instead, champions game theory. 

“Asset allocation strategies based on game theory rather than probability theory can provide a helpful tool when market behavior changes and past performance no longer predicts future returns,” says Chassang, a professor in NYU’s Department of Economics. 

His analysis, which appears in the Journal of Risk, focuses on the following question: How should one invest when the past does not predict the future?

The question is a complex one as it applies to old assets whose behavior might change, such as bonds in a rising-rate environment, and to new assets with little track record, such as crypto-currencies.

However, a game-theory approach does not consider precedent; rather it seeks to explain the behaviors of competitive parties who are assumed to act in a rational or self-serving manner.

In Chassang’s analysis, it’s investors vs. the market. 

“By viewing the market as an adversary eager to cause both losses and missed opportunities, it is possible to compute portfolios delivering the best possible performance guarantees against arbitrary market patterns,” he explains. “It cannot deliver the moon, but it can prevent bad surprises.”

Chassang’s work stems from Efficient Portfolio Theory, which was devised by Nobel Laureate Harry Markowitz. The latter allows the decision maker to express preferences with regard to risk and reward based on probable investment returns.

Chassang, on the other hand, adopts a “prior-free” approach that eliminates consideration of such probabilities and aims to maximize returns in a dynamic market. In the paper, he applies the method to different historic periods—stretching as far back as 1927—which calls for shifting asset allocations following small up and down movements in the market.

“These adjustments guarantee limited drawdowns in case a bull or bear market should emerge,” he writes.

“The portfolio allocation strategies derived from the prior-free approach offer a robust alternative to traditional portfolio allocation approaches,” adds Chassang. “They are particularly well-suited to deal with assets with little track record, or assets whose behavior seems likely to change. Importantly, these methods can be applied to manage the risk of investment strategies, including carry and contrarian strategies, which seek to avoid ‘hot’ stocks while grabbing those seen as out-of-favor.” 

The paper, “Mostly Prior Free Asset Allocation,” may be downloaded here: http://bit.ly/2EGcl9Q

DOI: 10.21314/JOR.2018.396

# # #

Journal Link: Journal of Risk