Global X Research TeamFor this post, I interviewed Justin Sibears, Managing Director and Portfolio Manager at Newfound Research. Justin shared his thoughts on the topic of smart beta, how to select a strategy for a portfolio, and the movement towards multi-factor.
Global X offers a suite of four multi-factor smart beta ETFs tracking Scientific Beta indexes, which is the indexing arm of the EDHEC Risk-Institute.
Some say smart beta is a gimmick and you cannot improve on a cap-weighted index. Are there inherent flaws in cap-weighted indexes or areas for improvement that can be addressed through smart beta strategies?
I wouldn’t say that there are flaws with cap-weighted indices. In fact, most investors could do far worse than parking their money in a low cost, cap-weighted index fund.
With that being said, we think that it is important to understand that by holding a cap-weighted index fund, investors are implicitly endorsing the market’s valuation of each security. If the market decides that the tech sector should be valued by clicks or page views instead of more traditional fundamentals, the cap-weighted investor goes along for the ride. The dot-com bubble exemplifies the pain that this type of approach can cause.
This might be an extreme example of mispricing. However, we know that historically some combination of behavioral biases, investor preferences, and structural inefficiencies have led to systematic over and undervaluation in certain groups of securities.
We believe that properly constructed smart beta or factor-based strategies have two potential advantages over cap-weighted strategies:
- Risk Management – Because these strategies are not cap-weighted they may be less likely to succumb to valuation bubbles in certain parts of the market. We believe this is especially true when multiple factor-based strategies are combined within the equity sleeve. For example, odds are low that a momentum strategy, a value strategy, and a low volatility strategy would all be concentrated in the same overvalued part of the market.
- Returns – We believe that smart beta strategies designed to take advantage of the behavioral biases, investor preferences, and structural inefficiencies that we alluded to before have the potential to outperform on a risk-adjusted basis.
The importance of risk-adjusted outperformance, or alpha, will vary by market environment. During strong bull markets, the equity markets may offer sufficient return potential for many investors to achieve their financial goals. In these markets where the return “pie” is large, we think it makes a lot of sense to use low cost, passive solutions so that you capture as much of the “pie” as possible.
In markets where equity returns are muted, the “pie” may not be big enough in the first place. In these markets, it becomes more important to look for ways to expand the “pie” with alpha-seeking strategies.
At Newfound, we believe that U.S. equities are overvalued relative to historical norms, increasing the likelihood that we are entering a regime where alpha1 will be quite valuable.
There are hundreds of smart beta ETFs on the market now. Where do you start when you are looking for an ETF for your model? What are your key considerations?
In our ETF selection process, we consider three factors: index construction, cost, and liquidity. The relative importance of these three factors will vary depending on the asset class and how we intend to use the ETF within a particular strategy.
When it comes to smart beta, especially in the large-cap equity space, liquidity is not too much of a concern because we can always work with market makers or tap into the creation/redemption process to achieve strong execution. Cost is always key, but with continued price compression we regularly find that index construction is the dominant factor in explaining relative performance differences between smart beta products.
Unfortunately, the only way to really get comfortable with the index construction process is to dig into the methodology document.
When it comes to smart beta or factor-based strategies, we typically ask questions like:
- Does the security selection jive with academic research and practitioner experience? We generally only consider smart beta strategies that provide exposure to certain factors (e.g. momentum, value, low volatility, size, quality). For consideration, a factor must meet four criteria to be included on this list:
- It must have a documented history of offering an attractive risk-premium over the market.
- There must be credible academic theories for why the premium exists.
- We must have conviction that the conditions necessary for the premium to exist will continue to prevail in the future.
- The premium must be able to be harvested in practice after accounting for real world frictions like transaction costs and fees.
- Once the securities are selected, how are allocations to each security determined? The best approach will vary based on portfolio use, but we always look to make sure stock-specific risk is appropriately diversified.
- Is the index construction “pure” enough to deliver the factor exposure we are looking for? We see many smart beta ETFs that hold so many securities that they start to behave exactly like their cap-weighted counterparts. If you want cap-weighted exposure, just buy a cap-weighted ETF.
- Does the process appropriately address real world issues that come into play when moving from index to actual implementation? This would include topics like liquidity and turnover.
Many of the older smart beta ETFs are single factor strategies, but now there have been a series of multi-factor ETF launches. What are the potential advantages of a multi-factor strategy over single factor?
Alpha isn’t free. While a single factor strategy may outperform over the long run, it will inevitably go through periods of relative underperformance. In fact, it is precisely these bouts of underperformance that allow the alpha to exist in the first place.
If underperformance never happened, the factor premium would be considered risk free. Investors would crowd into the strategy and drive up valuations until the alpha compressed to zero. Sustained periods of lagging relative performance causes weak hands to “fold”, passing outperformance to those with the fortitude to “hold”.
Luckily, the relative out and underperformance of different factors has tended to be rather uncorrelated. A multi-factor approach capitalizes on this opportunity of diversification. For “folders”, this may sufficiently reduce the depth and duration of underperformance, potentially turning them into “holders.” For “holders”, the diversification can make the ride a bit more enjoyable.
What separates EDHEC’s approach to multi-factor that appeals to you?
We believe successful investment strategies have three characteristics: simplicity, consistency, and thoughtfulness. In our opinion, EDHEC’s approach checks all three of these boxes.
- Simplicity: EDHEC uses a single metric for each factor in the portfolio. Book-to-market (for Value), one-year return excluding the most recent month (for Momentum), market-cap (for Size), and trailing volatility (for low volatility). Our research shows that simple investment processes are more robust amid changing market conditions. Using more than one metric per factor isn’t necessarily bad, but it does increase complexity. More complexity increases the risk that the strategy was data-mined to look good in a backtest.
- Consistency: The beauty of a passively managed, index-based ETF is that it ensures process consistency. The ETF is managed according to the pre-defined rules in the index methodology. We do not worry that EDHEC may abandon or materially change the process if it goes through a bout of short-term underperformance.
- Thoughtfulness: The level of thoughtfulness that went into building the EDHEC indices is what initially drew us to the approach. EDHEC’s approach is faithful to the academic literature.
It recognizes the inherent risks in any single portfolio weighting scheme, and opts instead for the elegant solution of combining five independent approaches to diversify those risks. In our view, this significantly reduces the negative impact that historical estimates for returns, standard deviations, or correlations can have on the portfolio.
We especially like the way that EDHEC combines the four individual factors into a single portfolio. Each individual factor portfolio is constructed independently. This is a bit of a nuanced point, but an important one. A multi-factor portfolio can be constructed one of two ways. For simplicity, let’s just consider two factors: value and momentum. One method would be to find stocks that offer good value and high momentum. The other would be to find stocks with good value or high momentum. EDHEC opts for the “or” approach, which we prefer. The “or” approach has more academic and empirical support and we believe it is better able to capture the diversification benefits of combining multiple factors together.
How does Newfound combine smart beta or factor-based ETFs with tactical asset management?
As a firm, we were founded on the idea that investors care deeply about capital preservation. Therefore, all of our strategies prioritize risk management. What risk management means depends on the asset class or strategy. When it comes to equities, we believe that risk management means avoiding large drawdowns.
Broadly speaking, we think large equity drawdowns can be classified into one of two camps. Some are caused by crashes in a subset of the universe and so they are more likely to be able to be diversified away. The dot-com bubble would fit into this camp. Others are so broad-based that no amount of stock-level diversification is enough. The global financial crisis of 2007 to 2009 fits in this category.
We combine a multi-factor equity portfolio with the ability to tactically move to cash or short-term Treasuries. With this dual approach to risk management, we look to lean on the ETF level diversification as much as we can, saving our tactical ammunition for more extreme sell-offs.