Does the Fama-French three-factor model adequately capture all information available in describing stock returns? According to Marlena Lee, PhD, VP of Dimensional Fund Advisors, the three-factor model is lacking one or two important components. Lee visited the Toronto offices of the CFA Society Toronto on the afternoon of June 19, 2014, to speak to over twenty financial experts about the evolution of asset pricing.
Lee was a funny and forthcoming lecturer. After her flight from the States up to Toronto, she said the suspicious Canada Border Services officer asked: “This CFA Society… what does ‘CFA’ stand for?” She momentarily blanked: Did the C stand for “certified”? For her trip back, she was prepared: C stands for “chartered.”
Asset pricing is a trickier matter—after all, direct questioning does not apply. “Prices reflect the expectations of market participants,” Lee emphasized. She drew on the case of the Denver Broncos vs Jacksonville Jaguars in the 2013 Superbowl, which had the largest spread in Superbowl history (-28). “The favourite should beat the spread about half the time” and the price is set by the marginal bettor.
Equity trading distills an enormous amount of information into security prices. With an average of 42 million trades world-wide every day, “it’s hard to find managers who can identify mispricing on a consistent basis,” she said. Sixty percent of funds underperform the benchmark in the first year, in part due to fees. After five years, 75 percent of funds underperform.
Lee described the “dimensions of expected returns” as laid out by the Fama-French model. The dimensions are market (equity premium), company size (small cap premium), and relative price (value premium, comparing value and growth companies). However, there appears to be more to asset pricing than these—something that is another indicator of future cash flows.
Lee delved into how such an indicator is identified. It must be a “persistent measure,” i.e., it can’t include one-off or exceptional items, it must include the major costs of doing business, and it must be “comparable across various sectors,” she said.
To the conventional three dimensions (or factors) she added “expected profitability,” defined as future profits divided by book equity, discounted to present value. Stocks with higher expected future profits have better returns.
“Do profits today contain information about profits in the future?” was the question posed by her firm, in order to use current profits as a proxy for future profits. “Profits are quite ‘sticky’,” Lee said, and they tend to move up and down together as companies go through the business cycle.
A colourful slide compared dimensions of expected stock returns for four geographic regions (Canada, US, other developed nations, and emerging markets) broken down by size, relative price, and profitability factors.
“Profitability is the other side of growth,” according to Robert Novy-Marx. Lee noted that Novy-Marx categorized profitability arising from average cash flow at low prices or “spectacular” cash flow at middling prices. He found the profitability premium correlated negatively with the value premium.
Profitability does indeed appear to be a significant “missing piece” to the asset pricing model. ª