A New Approach to Measuring and Managing Investment Risk

Transcription

1

2 A New Approach to Measuring and Managing Investment Risk James Chong, Ph.D. *David T. Fractor, Ph.D. *G. Michael Phillips, Ph.D. June 19, 2010 (*presenting)

3 Part 1: The State of the Economy

4

5

6

7

8

9 S&P 500, NYSE Index, NASDAQ Index

10 The next few slides compare the S&P 500 index to several key economic factors.

11 S&P 500 vs. Gold

12 S&P 500 vs. U.S. Ag. Exports

13 S&P 500 vs. FTSE 100

14 S&P 500 vs. Tokyo Stock Exch. Index

15 S&P 500 vs. Housing Starts

16 S&P 500 vs. Monetary Base

17 S&P 500 vs. M2 Money Supply

18 S&P 500 vs. Corporate Cash Flow

19 S&P 500 vs. Auto Sales

20 S&P 500 vs. New Orders for Durable Goods

21 S&P 500 vs. Energy Price Index

22 S&P 500 vs. Unemployment Rate

23

24 The Center for Computationally Advanced Statistical Techniques has developed a new risk measure for stocks, funds, indexes, and portfolios based on their responses to 18 key economic variables, called Eta MacroRisk Factors. The next slide identifies those which have values that are outside the range that would have been expected given the past year s values.

25 The current MacroRisk Profile

26 Traditional measures of investment risk assume that economic turbulence is like the weather and you can t do anything about this systematic risk. Part 2 of our presentation will discuss a new approach to measuring and management investment risk that allows for better measurement and control of the economy s impact on investments.

27 Part 2: A New Approach to Measuring and Managing Investment Risk

28 First, let s consider the conventional wisdom Use Betas to measure volatility Select low Beta stocks (Beta < 1) Select high Beta stocks for aggressive portfolio Create your portfolio Equally weight? MPT, maximize Sharpe ratio? 7/12 Portfolio? Just pick the best stocks or funds?

29 Beta The ratio of an investment s risk premium to the overall market s risk premium is a constant called Beta. The slope of a line drawn against the plot of an investment s return over time against an index s return is an estimate of Beta. Beta is often assumed to be symmetric, the same value whether the market is going up or down.

30 The CAPM and ß are everywhere A review of capital asset pricing models, Don U.A. Galagedera, Managerial Finance v33 n pp

31 Investopedia.com says: 31

32 (more, Investopedia.com) 32

33 The Motley Fool ( 33

34 Investment Advice: For relative safety, pick Beta < 1. Suppose that Jo Investor has her money in an IRA and does her rebalancing right before tax filing each year. Suppose in April, 2008, she put her investments in stocks with Beta less than 1. Would that have mattered in the 2008 crash? The next slide compares the S&P 500 index and an equally weighted portfolio of NYSE stocks with Betas less than 1 in April, Performance statistics are computed for the following year, 4/2008 5/

35 Portfolio of NYSE Beta < 1 and SPX, 4/15/08 5/1/09 35

36 And, against the NYSE as a whole: 36

37 And, against the other NYSE stocks with Beta > 1: 37

38 Did different Betas matter? High Beta Stocks lost more value than Low Beta Stocks But, the safe stocks only did 100 bps better than the risky stocks Safe stocks (ß < 1) lost 30.17% Risky stocks (ß > 1) lost 31.18%

39 Another problem: Reported values of Beta are often reported inconsistently by popular sources Jennings and Phillips did an analysis of popular sources of estimated Beta used by investors, professors preparing lectures, and students performing class projects, and found a remarkable disagreement on the estimated values for those stocks listed in the Dow Jones Industrial Averages. Fun with Beta! (Or, applying Beta is a risky proposition) William P. Jennings and G. Michael Phillips Presentation to FMA Round Table

40 Fun with Beta! (Or, applying Beta is a risky proposition), William P. Jennings and G. Michael Phillips Presentation to FMA Round Table

41 Here are some updated results from June 15, 2010:

42 The Dow 30 are among the most liquid, most watched, securities. Even so, these commercially provided Beta estimates showed great variety. There did not appear to be a common bias across stocks; no one source always provided large estimates, no one source always provided small estimates. It is not obvious how the different sources are estimating their Beta statistics but the underlying assumptions matter. 42

43 Betas are NOT symmetric across market conditions. Just because a stock goes up with the market doesn t mean it will go down at the same pace. Up-market and Down-market Betas allow for separate measurements of how an investment tracks with the overall market.

44 The following chart shows estimated overall Betas, up-market Betas, and down-market Betas for the Dow 30 components. These are estimated using the S&P 500 as the market index and using 1 year of daily returns. Calculations were performed using EconomicInvestor s Beta+ Tool.

45

46 Investment Hint: For safety, select investments with ß < 1. Note: ß < 0 often display larger overall risk and so we avoid them. If up-market and down-market days have an equal chance of happening, then favor investments with up-market ß greater than down-market ß.

47 (On 6/15/2010, 17 of the Dow 30 have higher Up than Down Betas)

48 Another related statistic is the investment s Alpha, which is the average return expected in excess of money market rates regardless of the overall market return. When estimated using annual data, this can be a proxy for stock momentum. Like Betas, Alphas are not symmetric. If a stock behaves very differently in up-market and down-market conditions, it is even possible for the overall Alpha to be positive while both Directional Alphas are negative! The following table compares Alphas for the Dow 30.

49

50 Investment Hint: For safety, consider selecting investments with > 0. (Why invest in assets with built-in loss?) In a perfect world, investments would have (up) > 0 and (down) > 0 as well.

51 (On 6/15/2010, 6 of the Dow 30 have positive Alpha measures in both down and up market conditions. 23 had positive overall Alphas but a negative directional Alpha.)

52 Suppose that you created a portfolio from the S&P 500 of those stocks with Alpha greater than zero. Then, suppose you created another one with stocks having Alpha greater than zero and Beta less than one. Total Returns (No rebalancing) 4/15/08 6/15/10 S&P: -18.2% Positive Alpha: % Beta < 1: -8.68% Alpha > 0, Beta < 1: -6.13%

53 Investment Hint: Portfolios of investments with both positive Alpha and with Betas less than 1 often provide superior returns to portfolios constructed using just one of the measures.

54 Now, consider a Modern Portfolio Theory Approach The Sharpe Ratio is the average excess return per unit of risk (measured by the standard deviation of returns). Consider the portfolio selected from S&P 500 stocks that maximizes the Sharpe Ratio in 4/15/2008. (Portfolio weights were obtained using the MacroRisk.com professional optimizer.)

55 Compare MPT Optimized Portfolio (Maximum Sharpe Ratio) to Equally Weighted >0, ß<1 Portfolio Total Returns (No rebalancing) 4/15/08 6/15/10 S&P: -18.2% Maximum Sharpe: % Alpha > 0, Beta < 1: -6.13%

56 The and ß relate asset returns to a single benchmark. However, there are many economic factors that influence investment performance.

57 The Eta Profile is a graphical tool that helps investors understand how 18 key MacroRisk Factors impact investments These key factors are the same for all U.S. stocks, indexes, mutual funds, ETFs, ADRs, etc. These factors apply to many exchange rates. The impact of each factor can be compared across investments, across factors, and across time. The impact of each factor can be combined at the portfolio level.

58 18 Eta MacroRisk Factors statistically explain a substantial part of most stock, fund, and index prices. The Red Bars indicate factors which are outside the normal range and therefore are having extra impact on some assets.

59 The MacroRisk Status indicates the current standardized value of each of the 18 factors. This describes the overall economy at a point in time. The Eta Profile is unique for each asset or portfolio. It shows the relative impact of a change in each MacroRisk factor on the asset price. The profile here is for Microsoft (MSFT). For example, MSFT stock prices historically have responded positively to Unemployment and Energy Prices, and negatively to the CPI. (All 18 Eta responses are shown on the chart.)

60 Now a Diversification Experiment consider Microsoft vs. EcoLab

61 Microsoft vs. EcoLab Microsoft (MSFT) is a technology company specializing in applications software. MSFT has about 93,000 employees and about $59.5 billion in revenues. EcoLab (ECL) is a consumer goods firm specializing in cleaning products and janitorial supply. ECL has about 26,000 employees and about $5.98 billion in revenues.

62 These two assets have similar Eta Profiles (though MSFT has more economic exposure than ECL).

63 MSFT and ECL, with similar Eta Profiles, have similar price patterns.

64 Here s another one: MSFT vs. Lorillard (LO), the cigarette manufacturing company

65 In contrast, consider Microsoft vs. Tyson Foods (TSN)

66 MSFT and TSN, with generally opposite Eta Profiles, tend to have different price patterns.

67 Here s another one: MSFT vs. Compass Minerals (CMP)

68 Investment Tip: When Eta Profiles are similar, stocks tend to move up and down together; when they are opposite shaped, they tend to move up and down oppositely. When diversification is the goal, search for investments with opposite Eta profiles to smooth out the resulting portfolio profile. When an investment is identified that should do well in the current economy, purchasing other investments with similar Eta profiles can diversify firm specific effects while maintaining the same economic exposure.

69 The CMRI (Composite MacroRisk Index) is computed by adding the absolute values of Eta Measures which are statistically important (t statistic >= 1). The CMRI value changes from asset to asset and may change over time. Because of economic diversification, the average of CMRI statistics for a portfolio is usually not the portfolio s CMRI.

70 A recipe for reducing economic risk: On this particular date, combine 4 parts of the white portfolio Eta Profile with 1 part of the red MSFT Eta Profile to get the blue Portfolio (20% MSFT, 80% other) Profile. (These weights were determined using the MacroRisk.com professional optimizer.)

71 This compares MSFT with a portfolio designed to have a low CMRI, containing 20% MSFT and 80% other stocks chosen to offset the MSFT economic exposure. These data are as of 6/15/2008.

72 The resulting portfolio still had Microsoft s resiliency but a far smoother path as economic risk is removed. MSFT is blue, the Low CMRI portfolio (with 20% MSFT) is red and the S&P Index is grey.

73 The CMRI forecasts future volatility. Theoretically the values range from 0 to infinity, but for the most part CMRIs range from 0 to The following report from EconomicInvestor.com shows some current CMRI values (as well as a few other statistics) for eight sample stocks.

74 The Alpha, Beta, and CMRI filters can be combined to create statistically safer portfolios. (Nothing can prevent financial craziness, but on average you can improve the odds by reducing the likely negative outcomes.) In the following example, an equally weighted portfolio of NYSE traded stocks was created on 4/15/2008. The stocks had positive Alpha, Beta less than 1, and CMRI less than 300. (They also all had an economic climate rating of at least 4. This statistic, shown on the previous table, measures how favorable the current economic conditions are for a given investment and can help in forecasting performance. It is based on a combination of the MacroRisk Status and the Eta Profile of the investment being analyzed. This provides very different information than other stars ratings.)

75 Equally weighted portfolio using basic risk screens and NYSE traded stocks

76 Now, using same criteria, compare equally weighted and minimum CMRI portfolios (using 4/15/08 data)

77

78 Investment Tips (Summary): Consider both Alpha and Beta. Consider both up-market and down-market values. Don t assume that two stocks in different industries provide diversification; examine their Eta Profiles. Use CMRI as a measure of economic risk. When expected holding periods are longer than a year, try to obtain the lowest CMRI portfolio from your buy-list. This all applies to ETFs and Mutual Funds too. For shorter term holdings, use the Alpha, Beta, CMRI, and Economic Climate ratings (demonstrated in the appendix).

79 Some current NYSE screens

80 Some current NASDAQ screens

81 Some current ETF screens

82 (Appendix: An experiment using economic climate rating >= 4 as a filter on NYSE stocks, and a minimum CMRI portfolio based on those same stocks, both portfolios constructed using information as of 6/11/2009.)

83 Want to see how the Economy Matters to you? Want access to patented Investment Tools for a Changing Economy? Use this exclusive offer to get a FREE month-long Premium Subscription to ($129 value)! To get your free account: 1. Go to 2. Fill in your information and above offer code 3. Select your subscription level To run an overview report: 1. Scroll to online research 2. Click Overview Report (under Eta Analysis) 3. Select your security and click get report