In my previous article, I explained that the Human Factor (H-Factor) is an actuarially based portfolio tool, developed by New York-based asset managers New Age Alpha, aimed at mitigating the risk of human behaviour in stock picking. I explained the idea behind the H-Factor tool and its application within the investment space. Today I will be showing the H-Factor’s effectiveness as an equity fund.
The H-Factor does not seek to generate returns by applying traditional methods such as the common smart beta and factor exposure funds we have all come to know. Instead, the H-Factor quantifies and avoids the risk of human biases in stock picking – a risk that is often unpriced and undiversifiable.
To show how avoiding this unpriced risk contributes to outperformance, we have constructed a variable for the H-Factor Outperformance (HFO) which represents stocks with the lowest H-Factor scores drawn from the S&P 500 and have shown this portfolios outperformance relative to the S&P 500. In other words, we have created a hypothetical portfolio that systematically avoids stocks with the most Human Factor risk and shown its performance relative to the universe we selected the stocks from (S&P 500). The results are shown below (data is through June 30, 2020).
*Data Source: Bloomberg, New Age Alpha. The test period represents the entirety of the New Age Alpha database. These hypothetical results are not annualized. Please note that these hypothetical results do not reflect fees, expenses and capital market inputs and limitations. Any actual returns will be reduced by advisory fees and other expenses. Past performance is not indicative of future returns. See the Disclosures Section on New Age Alpha’s website for more information.
The above chart clearly shows how the H-Factor portfolio has consistently outperformed the S&P 500 over the period – the cumulative outperformance of 103.7% is quite astonishing.
This outperformance is all good and well, however, how do we know that it is not merely an artefact of taking on some underlying riskiness?
To show this, we decided to compare the H-Factor with the traditional Fama French factors – namely: the market, size, value, and momentum factors. The market factor incorporates the general market risk and return characteristics over some risk-free investments. The size factor states that small-cap companies tend to outperform relative to large-cap companies and thus overweight small-cap stocks.
The value factor states that companies with high book-to-market value ratios (value/cheap companies) tend to outperform companies with low book-to-market value ratios (expensive companies relative to fundamentals) and thus overweights the value stocks. The momentum factor states that stocks that have outperformed in the past tend to exhibit strong returns in the future and thus overweights stocks exhibiting this price momentum. Due to the empirical evidence supporting these factors many actively managed funds tend to target or overweight one of these factors to beat the market/benchmark.
The Fama French factor returns are undoubtedly positive over a very long history of the stock market, however, during shorter periods the factors have been shown to underperform. For instance, in 2015, the momentum factor was up 21.5% while the value factor was down 8.7%. The following year, momentum was down 17.6% while value was up 8.4%. This suggests that, at a minimum, factor performance is inconsistent.
To show the H-Factor’s consistency of returns relative to the Fama French factors we again took the H-Factor outperformance (relative to the S&P 500) to find the correlations of this variable with the performance of each of the Fama factors. What we would want to see is positive correlations when the factor returns are positive – implying that the H-Factor weighted the equity fund in favour of the specific factors that outperformed.
Likewise, we would also want to see negative correlations when the factor returns are negative – implying that the H-Factor did not include or underweighted the factors that did not perform and thus avoided the underperformance. The results are in the below table (data is through June 30, 2020):
*Factor data obtained from Kenneth French Data Library, 11/23/2020. Return data obtained from Refinitiv, HFO data obtained from New Age Alpha. The test period represents the entirety of the New Age Alpha database. Please note that these hypothetical results do not reflect fees, expenses and capital market inputs or limitations. Please carefully review disclosures on New Age Alphas website to understand the limitations of simulations.
In the above table, positive correlations with positive factor returns and negative correlations with negative factor returns are highlighted in yellow. If the factor is correlated in the wrong direction (i.e., positive correlation with a negative factor return) the cell is not highlighted.
It is apparent that HFO tends to be positively correlated with factor returns when the factor performs well, and negatively correlated when the factor performs poorly. This gives a visual depiction of how frequently HFO may perform well (highlighted blocks), whereas the factors themselves have performance that is more sporadic.
We believe the reason behind this is because the Fama French factors have little or no relationship to company fundamentals, nor do they capture the risk of investors misinterpreting vague or ambiguous information – they are merely based on empirical evidence. The H-Factor, on the other hand, uses both company fundamentals and stock price (which are both unambiguous pieces of information) to avoid this risk and stocks likely to underperform.
From the above analysis, we can conclude that the Human Factor Tool can undoubtedly produce outperformance whilst at the same time limiting exposure to any one of the traditional Fama French factors at any given time. This combined with the ability to avoid stocks whose prices are not in line with fundamentals, not only positions you to outperform but also to benefit from factor diversification and risk reduction.