Factor investing — sometmes referred to as “smart beta” investing — is one of the fiercest battles now being fought among ETF providers. But for investors and advisors alike, it can be a confusing battle, because of widespread disagreement about which strategy is superior.

First, it’s important to understand that diverse factors have been analzyed and debated in academic ciricles. Examples of factor exposures include value, dividend yield, momentum, low volatility, and size.

Let’s examine a few popular factor strategies being used today within the ETF marketplace and compare their recent historical performance.

iShares MSCI USA Momentum Factor (MTUM)

Momentum stocks are typically high growth companies with rising share prices. These are the kinds of stocks that MTUM owns and the strategy is to hold companies with better relative performance compared to out-of-favor companies with lagging share prices.

In the short term, recent performance may persist, but stocks that have outperformed for several years have the tendency to become expensive and offer lower future returns. MTUM owns 125 large and mid-cap stocks with positive upward momentum and the $1 billion fund charges 0.15% annually.

RevenueShares Large Cap Fund (RWL)

As its name implies, RWL is focused strictly on stocks with superior corporate revenue. The fund owns the same 500 stocks within the S&P 500, but rather than weighting stocks by market size it ranks and weights companies according to the companies with the best top line revenue growth.

The fund rebalances its holdings quarterly based upon updated revenue data. RWL’s top three holdings are Wal-Mart, Exxon Mobil and Apple.

The fund has $339 million in assets and charges 0.49% annually.

PowerShares FTSE RAFI US 1000 ETF (PRF)

PRF is best described as a multi-factor ETF that weights its holdings according to fundamental measures of size, including book value, cash flow, sales and dividends, rather than market capitalization. The fund offers broad exposure to U.S. large- and mid-cap stocks.

PRF’s approach tends to give smaller weightings to more-expensive stocks and larger weightings to cheaper stocks than they would receive if they were weighted by simply market capitalization or size. As a result, this fund has a value tilt and its performance will be influenced by how value stocks are performing.

The fund has $4.2 billion under management and charges 0.39% annually.

WisdomTree Total Earnings Fund (EXT)

This ETF follows a fundamentally weighted index that measures the performance of earnings-generating companies within the broad U.S. stock market. To be selected, companies must have generated positive cumulative earnings over their most recent four fiscal quarters prior to the index measurement date.

WisdomTree Investments uses “Core Earnings,” computed by Standard & Poors, as the weighting metric. Core Earnings is a standardized calculation of earnings developed by Standard & Poors designed to include expenses, incomes and activities that reflect the actual profitability of an enterprise’s ongoing operations.

The index is earnings-weighted in December to reflect the proportionate share of the aggregate earnings each component company has generated. Companies with greater earnings generally have larger weights in the index.

EXT has $64 million under management and charges 0.28% annually.

Which Strategies Do Advisors Favor?

A 2015 study by FTSE Russell found that almost 70% of the advisors polled have adopted factor strategies inside their portfolio management.

“Factor-based and alternatively weighted indexes have transformed the investment landscape,” says Rolf Agather, managing director of North America Research for FTSE Russell “It is clear that retail advisors are embracing investment products based on these indexes as a way of incorporating new ideas into their clients’ portfolios.”

The most popular factor or alternative beta strategies used by advisors, according to the poll, were dividend, high quality, equal weight and fundamental weightings.

Among the 307 survey respondents were wirehouse brokers (29%), regional broker dealers (23%), RIAs (23%) and independent broker-dealers (21%). Also, 81% of polled participants manage more than $50 million and 52% manage more than $100 million.

Which Factor Wins?

A quick gander at recent performance tells us which factors have performed the best. With a handsome gain of 92.07%, the RevenueShares Large Cap Fund (RWL) is the clear winner among factor ETFs over the past five years. It edged out the SPDR S&P 500 ETF (SPY), which recorded a gain of 89.21%.

The worst performing factor was PowerShares FTSE RAFI US 1000 ETF (PRF) which had a five-year gain of just 82.58%. Although value stocks easily outperformed growth stocks over the past five years, PRF’s underperformance runs counterintuitive to what a stock market investor might expect. Despite having a bias toward value stocks, PRF still underperformed and its blended approach toward weighting multiple factors didn’t produce alpha compared to single factor ETFs.

Summary

Looking ahead, which factors will offer the best future performance? The answer is it won’t necessarily be the factors with the best historical performance today, because there’s obviously no guarantee that past observed risk-adjusted returns will continue going forward.

“It’s important to note that not all factor exposures are expected to earn a return premium over the long term,” said Scott N. Pappas, CFA, and Joel M. Dickson, Ph.D., at Vanguard in a white paper titled “Factor-Based Investing”: “Factor exposures can be compensated or uncompensated, a critical distinction in any factor-based investing framework.”

Deciding which factors to use within a client portfolio will largely depend on the client’s investment goals, the current market cycle and the advisor’s philosophical approach toward equity investing.

In the end, regardless of the specific factor strategies chosen by an advisor, a key driver to capturing the potential long-term premium of a factor is for clients to hang tight even during difficult periods of lackluster performance.