CFA Series - Revisiting the Fama And French Model

Revisiting the Fama And French Model

Serious investors whom have a decent foundation in quantitative finance will have came across the Fama and French 3 Factor model in his readings. Huge amounts of investment research and literature and quantitative asset pricing models are based on this. Even hedge funds or robo-advisers whom reject the efficient market hypothesis / CAPM have their key assumptions / origins based on the writings of Fama and French.

With the large number of market commentary commenting about the death of the value premium (including a professor I greatly respect), I am exceptionally lucky to come across a recent commentary by French (March 2020) whom originated the model. His views will be more factual compared to the marketing speak of robo-advisers / investment trainers whom like to tout recent market volatility as a permanent change of things, or based their materials on outdated data and analysis.

For an active investor whom utilises hybrid active stock picking and ETF selection, French's quantitatively rigorous views will be valuable to assess the changes that may be noteworthy. One point to note that his research is based on financial and stock price data and not qualitative factors such as structural changes in the economy. As such, his views are valuable but may not fully represent market reality.


Key Takeaways based on my understanding

1)    Based on Mutual fund data, which is audited and most reliable,
Generally, there is no reason you should choose Active managed funds as most funds only deliver market results after fees and expenses.

It is very hard to reliably identify fund managers that can beat the market consistently. Luck and skill is only distinguishable over a long period of time (beyond 10 years). There are managers that can beat the market but they are incredibly rare and few, difficult to identify, with no guarantee whether it will continue to work.

There are no fund managers that can reliably predict market changing events.

Most people overreact to short term news and allocate capital to fund managers which has good stellar recent performance, which the fund manager may underperform subsequently.

2)    If you are a rational investor
Market corrections should not change your preferences for stocks regardless of market boom or crashes.

Just because you are poorer after market shocks doesn’t mean portfolio construction rules have changed. Volatility should have already been priced in before the fact and not after the market corrects.

If everyone knows the market is going to crash, and you rotate to safe havens, you do not have any special insight, and is simply mirroring herd behavior. The knowledge is already priced in.

The market correction DOES NOT teach you anything new about the markets. It merely educates the investor about his true risk appetite.

3)    Actual returns for investors depends on expected and UNEXPECTED returns.
Unexpected returns can often completely dominate the expected returns based on French personal experience.

This explains why an unassuming unprofitable company selling books on the Internet (Amazon) and a small and boring search engine company (Google) turn out to be multibaggers. This also explains why great companies with great expectations priced in may turn out to deliver subpar returns.

The equity premium is earned based on the client’s risk appetite. If you are not behavioral suitable for volatility in stocks, you are not likely to achieve out-sized returns.

4) Problems with the asset management industry

Measurement
5 years is too short to observe any possible change in risk premiums or identify any fund manager whom is truly skilled.

False Precision
There are no fund managers that can predict market changing events. If that is the case, why are you paying extra fees for something you don’t get?

False extrapolation based on < research using  limited sample size>
Having witnessed too many repeated use of the excuse <This time it is different> across the years , French believers research based on small sample sizes is an convenient excuse to justify whatever behavior that is convenient, rather than a tool for good decision making.

5)    Shortcomings of the Fama-French Model
There are many ways to measure value. French have no proof that Book value is the best method to spot the winners.

But book is the most stable metric to measure value compared to noisy signals such as sales, expected future cash flows, intangible assets, brands, earnings. The choice of book value as a stable metric is an attempt to reduce portfolio turnover.  

Low portfolio turnover is a key metric to achieve good returns as it reduces transaction cost. A high turnover Fama-French strategy will erode any returns it is designed to capture.

6)    Discussion about momentum
Momentum => In the short term (months), a strong performer will continue to outperform
Mean Reversion Strategy => Stocks that do well in one period will do poorly in the next

Momentum delivers so much returns that it cannot be ignored.
But French believes a momentum-based strategy has high portfolio turnover that will reduce returns over time. Momentum is also very hard to identify.

French personally believes Momentum should only be used when the strategy already involves active trading and should be disregarded for passive strategies.

7)    Is home country bias a bad thing?
A home Country Bias depends on the allocation of the portfolio to the local stocks.

Home country bias is useful to hedge against forex risk. There are possible tax advantages in investing in local stocks through tax advantaged accounts.

BUT if most of the portfolio is allocated to local stocks because the investor is biased, it is a dumb thing to do. The global economy may diverge significantly from the home country economy.

8)    Share Buybacks VS Dividends
From a Finance professor view
Share buybacks are equivalent to dividends as a capital return tool, if dividend taxes and transaction costs are ignored.

Share buybacks are great if there is no better way to utilise cash. It is not inferior to dividends as it will push up the stock price.

9)    No clear reason that ESG companies should deliver superior returns.
If there are no incentives provided by the government to companies supporting ESG policies, there should not be superior returns delivered. French have not studied the issue extensively or cross checked with others to make strong conclusions.

If ESG is used as moral suasion to reduce resource consumption (operational expense) or boost sales / revenue, it may then deliver superior returns. However, there should be a real cost of implementation and it may not be reliable.

If investors overweight their portfolio towards ESG friendly companies, they are burdening the company with extra covenants and they should expect a drag to share prices. Even if there is short term boost to share prices, the future expected returns will be lower.