Smart Beta Theory and Practice

Transcription

1 Smart Beta Theory and Practice Burton G Malkiel Chemical Bank of Chairman s Professor of Economics, emeritus Princeton University November 5, 2014 RCI Rogers Investment Innovations Conference

2 What is Smart Beta? There is no universally accepted definition of Smart Beta strategies. What most people who use the term have in mind is that it may be possible to achieve greater than market returns using a variety of relatively passive investment strategies that involve no more risk than would be assumed by investing in a low cost total stock market index fund, which by definition, has a Beta of one. 2

3 Tilts to Portfolio to Beat the Market Value vs. Growth Small vs. Large Strong vs. Weak Low Volatility vs. High High Quality vs. Low 3

4 The Intellectual Cast for Factor Tilting 1. Value stocks are to be favored over growth stocks. Stocks with low price earnings multiples have produced higher returns than stocks wit high multiples. Data beginning in 1980 Is it an inefficiency or is it risk? 4

5 The Intellectual Case for Factor Tilting 2. The Small Firm Effect 5

6 The Intellectual Case for Factor Tilting 3. Momentum Increases in stock prices are slightly more likely to be followed by further increases than by price declines. For longer holding periods, mean reversion appears to be present. Explanations a) Behavioral b) Slow adjustment to news 6

7 The Intellectual Case for Factor Tilting 4. Low Volatility Can Produce High Returns Relationship of return to Beta is relatively flat. Therefore can leverage up a low volatility portfolio to produce excess returns. Or buy low volatility stocks and short high volatility ones. 7

8 The Intellectual Case for Factor Tilting 5. Quality Stable Profitability Low Leverage 6. Can Use Multiple Tilts Dimensional Fund Advisors RAFI Fundamental Indexes TM Equal Weighted Indexes 8

9 Appraisal of Smart Beta All Smart Beta strategies represent active management rather than indexing. Capitalization weighted portfolios are the market. To the extent that Smart Beta funds do generate excess returns, it is most likely because they are assuming greater risks. When Smart Beta portfolios are assessed with multi factor risk models (such as the Fama French three factor model or extensions of it), the typical finding is that alpha is zero. Advantage is that active tilts can be accommodated at fees well below those of active managers 9

10 Challenges to Implementation of Smart Beta Smart Beta funds require periodic rebalancing. (Higher transactions costs and taxes) All Smart Beta portfolios have undergone periods of underperformance Periods of excess performance are often followed by periods of disappointing returns. Mutual funds (and ETFs) designed to capture momentum and low beta effects have not demonstrated superior performance over their short history. Whether Smart Beta strategies will perform well in the future depends crucially on the market valuations existing at the time the strategy is implemented. 10

11 The Records of Smart Beta Funds and ETFs The actual records of Smart Beta portfolios run with real money do not always replicate the results suggested by academic studies. Value and Growth Mutual Funds ETFs Small and Large 11

12 Value Funds/Growth Funds Source: Bogle Research Institute The chart shows the return of value mutual funds divided by the return of Growth mutual funds. A number above (below) 1 indicates that value is outperforming (underperforming) Growth. 12

13 Reversion to Mean Growth Index vs Value Index ETFs The chart shows the return of value mutual funds (Value ETF) divided by the return of Growth mutual funds (Growth ETF). A number above (below) 1 indicates that value is outperforming (underperforming) Growth. 13

14 Russell 2000 (Small Cap) Index Versus Russell 1000 (Large Cap) Index (30 year) IWM (Russell 2000) ETF Versus IWB (Russell 1000) ETF (10 year) The chart shows the returns of the Russell 2000 index (ETF) divided by the returns of the Russell 1000 index (ETF). A number above (below) 1 indicates that the Russell 2000 is outperforming (underperforming) the Russell

15 Records for Blended Strategies Fundamental Indexing TM Equal Weighting DFA 15

16 Fundamental Index (PRF) and Equal Weight (EWRI) Versus Russell 1000 (IWB) The chart shows the return of the Smart Beta strategy divided by the return of the benchmark index. A number above (below) 1 indicates that the Smart Beta fund is outperforming (underperforming) the benchmark index. 16

17 DFA Large Cap Value (DFLVX) and DFA Small Cap Value (DFSVX) Versus Benchmarks (10 Year) The chart shows the return of the Smart Beta strategy divided by the return of the benchmark index. A number above (below) 1 indicates that the Smart Beta fund is outperforming (underperforming) the benchmark index. 17

18 Low Volatility and Momentum Neither strategy has outperformed over a brief history 18

19 Low Volatility (SPLV) and Momentum (AMOMX) Versus Russell 1000 (IWB) The chart shows the return of the Smart Beta strategy divided by the return of the benchmark index. A number above (below) 1 indicates that the Smart Beta fund is outperforming (underperforming) the benchmark index. 19

20 Conclusions Smart Beta strategies rely on a type of active management. They do not try to select stocks but provide factor tilts. (Expenses lower than traditional active managers. In general the records of Smart Beta funds and ETFs have been mixed. Some Smart Beta funds and ETFs have produced excess returns but some have not. My interpretation is that any excess returns should be considered as compensation for risk. Smart Beta portfolios have been the object of considerable marketing hype. Whether Smart Beta strategies will perform well in the future depends crucially on the market valuations existing at the time the strategy is implemented. 20

21 Questions 21