The Big Interview: Bernie Nelson at Style Analytics (part two)

Scott Longley

a microphone on a stand

In the second part of our interview with Bernie Nelson, president for North America at Style Analytics, he starts by looking at how the issue of factor persistence and why the changing business landscape means that strict adherence to academic rules might be detrimental. We start with the vexed issue of backtests.

ETF Stream: Is there the potential that by focusing on backtests, the academic research might be ignoring aspects of the real world, such as innovation and changing business practices?

That is always a risk although I think backtesting is a necessary and inevitable part of investment research. Ultimately there is always a tension in investing between ‘this time it’s different’ and its implicit converse. With factors the question is whether risk-based and behavioural arguments will ensure factor persistence or whether something is systematically changing in the world. Nobody knows for sure and only time will tell. Of course, respect should be paid to evidence-based findings on factors, based on long-term studies, but the practical world of investing cannot wait multiple human lifetimes until there is textbook statistical significance. Careers are made and broken in much shorter time periods and clients’ patience is tested well within those periods also.

Earlier we talked about book-to-price and the debate about its potentially changing relevance. For example, that the US has moved from a manufacturing to more of a service-based economy over the decades. Those same arguments can be made here.

a man wearing glasses

ETF Stream: You have spoken previously of the old Capital Asset Pricing Model and how everyone used to swear by it but it is now obsolete. Can you explain more about this and why it is a good example of why factor investing shouldn’t get stuck with age-old factor definitions?

Many investment professionals do pay careful attention to the latest factor-based thinking from academic finance researchers and often those ideas will inform how investment managers and investors might consider their investment portfolios. However they don’t necessarily restrict themselves to the prescribed factors from specific studies or even meta studies. Even if the authors have won Nobel Prizes.

An interesting way to think about the flaw of having to strictly adhere to recent academic factor models is to go back in time to earlier models. The first and most well-known factor-related model of the past 50 or so years was the Capital Asset Pricing Model (CAPM) (Sharpe, 1964; Lintner, 1965). CAPM essentially said that investors are only compensated for bearing market risk and therefore the most efficient strategy is for an investor to hold a broadly diversified portfolio. This idea essentially kick-started index funds. However, during the long reign of CAPM how many equity investors were buying stocks or assessing portfolios only using estimated stock betas? Was anyone? Certainly not the famous investors of that period including Ben Graham’s early disciples such as Warren Buffett or Walter Schloss. Consider also that the first edition of Graham and Dodd’s Security Analysis book was published in 1934, around 30 years before the Capital Asset Pricing Model. This reminds us that fundamental approaches to equity investing have been evolving over a long time and also that many academic papers on factor investing were motivated by the approaches of fundamental investors, not the other way around.

It’s also interesting that even seemingly stalwart factors such as size and value, as popularized in the three-factor model by Fama and French (1993), are still being debated over 25 years later. For example, the four factor (q-factor) model of Hou, Xhu, and Zhang (2012, 2015) included market, investment, profitability, and size factors – but the value factor was considered as being redundant. The size effect has also been called into question (See ‘Fact, Fiction, and the Size Effect’ Alquist, Israel, and Moskowitz; Journal of Portfolio Management Fall 2018). So the debate about which factors are relevant is rarely if ever settled, even between researchers in academic finance.

Today, even with the evolution to four-, five- or even six-factor academic models, the CFA curriculum for equity analysis does not just suggest that those factors are “all there is to know” about equity investing. Neither have the world’s most prestigious business schools abandoned their courses on financial statement analysis or equity valuation and sent their students home. Some might argue that fundamental equity analysts are wasting their time due to the growing appeal of index funds or factor-based enhanced index funds. Yet despite an increase of assets into index funds, the majority of equity funds are still currently actively managed, whether discretionary or non-discretionary. If we are going to compare funds objectively then we need to use a factor framework that recognizes different investment philosophies in addition to the latest academic factor models.

ETF Stream: Is this a classic schism in operation? Why are there disagreements in these areas? Presumably such arguments have very little relation to what investors experience in terms of return?

This is not such a classic schism as it may appear. I think it is a natural tension that exists in the field of investment where the very tough challenge of the fund manager who is trying to gain an edge to make real money overrides the elegance of any seemingly parsimonious academic theory. Establishing a factor framework to assess portfolios across many funds from differing investment approaches also needs to acknowledge that information must be able to be communicated to intermediaries and clients and to be easily understood. Making things understandable and easy to communicate is also part of the successful design of a portfolio analytics factor framework.

Ultimately I do think that these arguments do have a real bearing and importance in terms of what investors experience in terms of return. As I mentioned earlier, different factor choices can have a material impact on investment returns.

We are familiar with the phrase follow the money. When academics, practitioners, or vendors present a factor framework or model for assessing portfolios a more apt phrase is ‘follow the methodology’. In other words, are the narrow factor definitions and rigid model approach driven by interests other than providing a fair portfolio evaluation or comparison? For example, is the provider selling funds or selling indices as well as analytics? An over-zealous approach may be symptomatic of some other self-serving motivation. Style Analytics doesn’t sell funds or indices and we remain independent and objective when we analyse tens of thousands of funds worldwide, every month.

Ultimately a fair factor framework for fund comparison should serve the investor and allow the most objective and independent insights into the structure and intention of investment portfolios based on the underlying holdings. These insights need to be transparent, and easy to communicate and understand, in order to maintain confidence and trust when making investment decisions.

This is the second of a two part interview with Bernie Nelson. For part one, click here.

Featured in this article


No ETFs to show.