Factor Investing

Quant Strategies: Will Returns Seen in Research Be Achieved?

Hype and Results:

The following paragraphs, as it relates to quant strategies performance was initially published by the CAIA Association’s AllAboutAlpha blog (link) and further added to by Factor Research.

Before we get into CAIA’s research, let’s take a look at the plethora of these algorithmic strategies that are offered by the primary player among factor investing quant asset managers, AQR.

When marketing a quant strategy with a backtested history, the constant reaction from potential investors is that they have never observed a poor performing backtest. This is valid. There is no reason for presenting a strategy with a poor backtest as investors have zero enthusiasm for losing money. They additionally will not pursue theory but reality, aka actual strategy returns.

Most quantitative managers have a scientific education and are probably not going to advocate a discretionary approach. Given that most active fund managers have failed to beat their benchmarks over the short-, medium-and long haul. Backtesting ought to be respected critically, even though it’s dependant on the type of backtesting. The more unpredictable and creative a methodology is, the warier a potential investor ought to be.

The returns of plain-vanilla algo methodologies, as found in factor investing research, ought to be attainable. We will differentiate between accessible factor returns and examine if investors had the option to capture the gains.

The Dark Side of Backtesting:

A research analysis broke down the backtested and live excess returns of 215 these strategies supported by fifteen banks between 2005 and 2015. The study covered strategies from equities, fixed income, currencies, commodities, and multi-assets. The examination shows a noteworthy contrast between the in-sample and out-of-sample returns. No technique was found to perform reliably. All strategies create inherently lower returns once live. The fact is that the strategies are either the aftereffect of data mining or are affected adversely by trading and market liquidity costs that were not anticipated.

The difference between backtested and live returns can easily be seen in equity strategies, which may manage a large number of stocks versus just a few commodities or currencies. The authors presume that all backtested returns require a discount, which ought to correspond to the complexity of the approach.

Looking at Factor Investing Return Data:

The measure of factor investing writing keeps on developing, and a significant piece of that depends on the freely accessible factor returns that come from the Fama French dataset. The information isn’t utilized for just featuring positive returns coming from factors like value or momentum when advertising performance. It also is used for clarifying the returns of active managers using factor analysis. Notwithstanding, returns can be challenged. Various reasons explain the distinction in performance.

Data source: The universe of stocks and information quality rely upon the supplier of price and necessary information, which may be different for CRSP, Bloomberg, Factset, or S&P Capital IQ.

Market cap limitations: Fama-French data and fund managers have no market cap confinements. At the same time, FactorResearch rejects all stocks with a market cap underneath $1 billion.

Value definition: Fama-French characterizes Value utilizing price-to-book. At the same time, FactorResearch uses a blend of price-to-book and price-to-earnings multiple.

Trading costs: Fama-French expects zero exchange costs while FactorResearch incorporates ten basis points per transaction. Portfolio development: Rebalancing recurrence, dollar versus beta-neutrality, time lags for fundamentals, and portfolio size is further purposes of differentiation.

It is, to some degree, astonishing that the patterns in performance were not increasingly heterogeneous. The strength of factor performance is alluring from a backtesting standpoint.

As a subsequent contextual analysis for contrasting backtesting information, the Momentum factor was examined. The Fama-French and fund manager, AQR, performance show a practically indistinguishable difference, as opposed to lower returns from FactorResearch.

The two explanations behind the lower return of the Momentum factor, as determined by FactorResearch, are likely the market cap levels and trading costs. Fama-French and AQR do not consider these two. The Momentum factor has high turnover and trading costs that have an altogether negative effect on returns.

Conclusion:

Given the vast differences between the backtested and live execution of quant strategies, investors ought to be wary of alluring returns from factor investing. Luckily, asset managers have developed long-short strategies that give long-short factor analysis and present investors with hypothetical and realized gains.

We can likewise think about the acknowledged returns of smart beta ETFs to their hypothetical returns, which feature that smart beta ETFs had lower returns than the theoretical benchmark, yet can be clarified by the absence of the management expenses for the hypothetical portfolios. Smart beta ETFs that are producing performance like the hypothetical portfolios are in-line with expectations.

Fortunately, returns found in factor investing literature can be captured by investors through liquid alternative and smart beta mutual funds. The bad news is that factor investing has not produced attractive returns in recent years.

Print Friendly, PDF & Email