Golden Toque Capital
Carrington Bradley, Patrick Kane, Peter Rudnick, Dennis Wu
Assignment 01 – Screening Based Stock Selection, Industry Comparisons
Professor Campbell R. Harvey
Table of Contents
1.Objective
2.Market Selection
3.Industry Selection
4.Time Period
5.Factor Selection
Factor 1: One month price momentum
Factor 2: Three month price momentum
Factor 3: Dividend yield
Factor 4: Rate of reinvestment
Factor 5: Consensus forecast earnings estimate revision ratio
Factor 6: Change in consensus FY1 estimates
Factor 7: Operating cash flow / Sales
Factor 8: Interest coverage (EBIT / Annual interest expense)
6.Findings
Industry: Oil and gas extraction (SIC code = 13xx)
Industry: Food and tobacco products (SIC code = 20xx + 21xx)
Industry: Electronic & other electric equipment (SIC code = 36xx)
Industry: Communication (SIC code = 48xx)
Industry: Depository institutions (SIC code = 60xx)
US Common Stocks
7.Analysis
8.Bi-variate Screens
Industry: Oil and gas extraction (SIC code = 13xx)
Industry: Depository institutions (SIC code = 60xx)
9.Optimizing Weights
Industry: Oil and gas extraction (SIC code = 13xx)
Industry: Depository institutions (SIC code = 60xx)
10.Closing Thoughts
11.Appendix A
12.Appendix B
Mutual Fund Evaluation
Asian Markets Evaluation
1.Objective
The main goal of our project is to explore and extend the findings from the “Stock Selection in Emerging Markets: Portfolio Strategies for Malaysia, Mexico and South Africa” (Harvey, et al.) into another area of interest. This idea was prompted by the recent availability of FactSet as a learning tool and the potential practical applications resulting. After a number of iterations, we found great interest in exploring applications segmented by industries within the US market, and the potential to optimize on these constraints.
2.Market Selection
In selecting a different area to explore based on discussions, we considered the mutual fund industry and other Asian countries beyond Malaysia. On the basis of data availability during initial trial, we opted for the more data-rich US market.An elaboration on why we thought these would be good projects is included in Appendix B, and perhaps worthy of another look given access to more information.
The refinement of our market selection was also very much influenced by the screening factors we chose, which will be explained in detail below. Specifically, we had found that the eight factors we chose produced very little tradable spreads in a US common stock universe within the historic in-sample time frame of 1990.01 – 2000.12.
In our discussion as to why these would vary in substance to the three countries considered in the paper, we hypothesized that perhaps the less developed nature of these countries were more dependent on some core industries, and that perhaps the US market does achieve noticeable spreads if we consider individual industries instead of the market as a whole, which is significantly more diversified. In addition, we saw value in exploring whether certain screening factors were industry specific, and how that would relate to the market as a whole.
3.Industry Selection
To standardize our selection of industries, we decided to base our study on Standard Industrial Classification (SIC) codes.This was reinforced by the fact that FactSet provides for the creation of universes by these codes. We evaluated each code based on a subjective assessment of the industry relevance to our study, as well as a dataset minimum of 100 companies. The minimum was set to allow us the ability to look for meaningful patterns within each industry without being adversely skewed by a disproportionate number of “N/A” listings for our factors.
The industries selected are listed below:
Code / Industry / Dataset size (Companies)13xx / Oil and gas extraction / 199
20xx + 21xx / Food kindred and tobacco products / 128
36xx / Electronic & other electric equipment / 506
48xx / Communication / 245
60xx / Depository institutions / 766
We recognize that both the companies listed within the SIC codes and the companies themselves can change over time.However, as new companies emerge and others falter, we believe this was a reasonable universe since the overall selection was still a representation of the particular industry.Some companies may change their main business over time as well by entering new markets or exiting their primary market.We felt that this impact would be minor overall.
4.Time Period
The aim was to balance the time period selected with the time it would take FactSet to run the factor screens.We decided on an 11 year period (1990.01– 2000.12) as our in-sample run.The desire was to include a small measure of the recent recession to balance the exceptional run in the US market in the 1990s. This left 2001.01–2003.12 for our out-of-sample comparisons.Although this is a relatively short out-of-sample period, we believe that the results will allow us to gain insight into whether these factors are still relevant for our chosen industries.In hindsight, although the 11 year period provided a great data screen, it proved cumbersome to run. Most runs took between 20 to 45 minutes to complete.
5.Factor Selection
When picking screening factors we considered three criteria. First, we chose factors that had plenty of data available. Second, we chose at least two momentum, fundamental, and expectation factors. Third, we chose factors that performed well when used in predictive regression models. Note that in our analysis, we refer back to this as reference and generally refer to our factors by their designation number. Our eight main variables are:
Factor 1: One month price momentum
Data source: COMPUSTAT
Formula: G_PRICE_1MCHG
Reasoning: We chose one month momentum because we wanted to see if the contrarian nature of this factor held true over a range of industries. Additionally, momentum factors have performed well in past predictive regressions.
Factor 2: Three month price momentum
Data source: COMPUSTAT
Formula: G_PRICE_3MCGH
Reasoning: The three month momentum variable provides a nice comparison for the one month momentum results. We suspected that the one month momentum factor would lead to a superior long/short spread since that data set contained more information than the three month data. Additionally, momentum factors have performed well in past predictive regressions.
Factor 3: Dividend yield
Data source: COMPUSTAT
Formula: G_DIV_YLD(NR 0 L45D)
Reasoning: Strong dividend yields and dividend growth are consistently solid indicators of healthy companies.
Factor 4: Rate of reinvestment
Data source: COMPUSTAT
Formula: G_REINV_RATE(NR 0 L45D)
Reasoning: We believe this variable will capture the upside performance of the large, non-dividend paying firms. This variable also describes the general health of a firm. Firms with surplus cash and long term growth prospects tend to have greater reinvestment.
Factor 5: Consensus forecast earnings estimate revision ratio
Data source: COMPUSTAT
Formula: (IH_UP_FY1R(0)-IH_DOWN_FY1R(0))/IC_NEST_FY1R(0)
Reasoning: We are interested in factors that incorporated a change in values. We guess that firms with a greater number of upward (downward) revisions will over (under) perform.
Factor 6: Change in consensus FY1 estimates
Data source: COMPUSTAT
Formula: G_IH_MEAN_FYIR_3MCHG(0)
Reasoning: Again, we are interested in factors that incorporated a change in values. We guess that firms with greater per cent changes in earning estimates will benefit (suffer) from the effect of good (bad) news on the market
Factor 7: Operating cash flow/ Sales
Data source: COMPUSTAT
Formula: G_CASH_FLOW(NR 0 L45D)/G_SALES(NR 0 L45D)
Reasoning: An interesting valuation ratio. We guess that companies with higher quality of sales will have superior returns.
Factor 8: Interest coverage (EBIT/ Annual interest expense)
Data source: COMPUSTAT
Formula: G_INT_COV(NR 0 L45D)
Reasoning: An interesting valuation ratio. We looked for a variable that captured the idea of leverage. We guess that a firm with a lower coverage ratio will have lower returns since a larger portion of its capital is going towards interest commitments.
6.Findings
Upon evaluating our FactSet results (see MS Excel file A1_GTC.xls), we determined to focus on a few key factors and industries.Specific discussions are below.Note that some of the findings were difficult to rationalize.
Industry: Oil and gas extraction (SIC code = 13xx)
Factor 1 / 1 / 2 / 3 / 4 / 5 / L - S / BenchmarkFractile returns / 1.84% / 8.50% / 10.37% / 19.93% / 31.34% / -29.51% / 16.30%
Std Dev / 31.97% / 28.83% / 31.34% / 31.27% / 37.41% / 24.07% / 13.68%
Avg Relative Performance / 3.00 / 2.50 / 2.75 / 2.00 / 1.63
Factor 2 / 1 / 2 / 3 / 4 / 5 / L - S / Benchmark
Fractile returns / 3.32% / 7.74% / 14.82% / 17.58% / 21.98% / -18.67% / 16.30%
Std Dev / 35.53% / 31.90% / 29.12% / 33.55% / 37.27% / 26.45% / 13.68%
Avg Relative Performance / 3.06 / 2.56 / 2.19 / 2.25 / 1.81
Factor 3 / 1 / 2 / 3 / 4 / 5 / L - S / Benchmark
Fractile returns / 12.86% / 15.60% / 18.35% / 33.33% / 11.44% / 1.42% / 16.30%
Std Dev / 25.51% / 33.32% / 38.75% / 63.94% / 35.72% / 16.11% / 13.68%
Avg Relative Performance / 2.38 / 2.44 / 2.19 / 2.44 / 2.44
Factor 4 / 1 / 2 / 3 / 4 / 5 / L - S / Benchmark
Fractile returns / 22.03% / 18.76% / 13.23% / 4.92% / 1.82% / 20.21% / 16.30%
Std Dev / 34.81% / 28.46% / 26.94% / 30.86% / 37.79% / 26.52% / 13.68%
Avg Relative Performance / 1.63 / 2.06 / 2.31 / 2.94 / 2.94
Factor 5 / 1 / 2 / 3 / 4 / 5 / L - S / Benchmark
Fractile returns / 13.73% / 18.24% / 16.04% / 9.26% / 11.76% / 1.97% / 16.30%
Std Dev / 30.89% / 28.08% / 31.31% / 30.55% / 30.40% / 21.51% / 13.68%
Avg Relative Performance / 2.38 / 2.19 / 2.44 / 2.44 / 2.44
Factor 6 / 1 / 2 / 3 / 4 / 5 / L - S / Benchmark
Fractile returns / 13.73% / 18.24% / 16.04% / 9.26% / 11.76% / 1.97% / 16.30%
Std Dev / 30.89% / 28.08% / 31.31% / 30.55% / 30.40% / 21.51% / 13.68%
Avg Relative Performance / 2.38 / 2.19 / 2.44 / 2.44 / 2.44
It appears that one month momentum is an attractive factor for a long/short strategy.The results are flipped however, as the strategy should change to one of going long on the bottom fractile and shorting the top fractile.One thought is that during the mid to late 1990’s, energy stocks enjoyed (at least on the upside) treatment very similar to technology stocks.Specifically, companies such as Enron and Dynegy seemed to take the world by storm with their impressive financial performance and unconventional styles.As a result, their stocks were more volatile than comparable stocks (the table also shows relatively high standard deviations of the returns) and that periods of relative calm (low momentum) were followed by bursts.What is surprising however is that this factor looks good for energy yet is not so for the communications or technology SIC codes.Perhaps the additional volatility of the crude oil market creates an additional opportunity within this sector.
Reinvestment rate also showed some promise in this sector.Typically, we assume that a high reinvestment rate is a favorable sign of future growth.As noted above, we feel that this may have reflected a sentiment that high fliers had tremendous potential in this sector compared to the relatively stodgy blue-chippers such as Mobil and Exxon that paid out dividends regularly.This factor also showed some life with electronics for perhaps the same reason, but telecom was not as strong.
When run using out of sample data, the momentum component lost luster while the reinvestment element produces very impressive results. The out of sample data are few, but this finding does give generate some concern over the usefulness of our first factor for this sector.
Industry: Food and tobacco products (SIC code = 20xx + 21xx)
Factor 1 / 1 / 2 / 3 / 4 / 5 / L - S / BenchmarkFractile returns / 18.02% / 17.06% / 18.77% / 18.80% / 16.35% / 1.67% / 16.30%
Std Dev / 20.69% / 13.72% / 14.01% / 18.80% / 24.68% / 20.67% / 13.68%
Avg Relative Performance / 2.44 / 2.63 / 2.00 / 2.06 / 2.75
Factor 2 / 1 / 2 / 3 / 4 / 5 / L - S / Benchmark
Fractile returns / 3.69% / 17.86% / 15.76% / 21.80% / 28.91% / -25.22% / 16.30%
Std Dev / 23.38% / 16.87% / 18.54% / 18.92% / 25.09% / 23.08% / 13.68%
Avg Relative Performance / 3.38 / 2.38 / 2.50 / 2.13 / 1.50
Factor 3 / 1 / 2 / 3 / 4 / 5 / L - S / Benchmark
Fractile returns / 18.93% / 15.23% / 9.59% / 12.29% / 13.22% / 5.71% / 16.30%
Std Dev / 20.96% / 19.11% / 19.86% / 30.19% / 25.34% / 21.27% / 13.68%
Avg Relative Performance / 1.88 / 2.19 / 2.63 / 2.63 / 2.56
Three month momentum appears to play a factor in this particular sector. Again, the long position on the bottom fractile offset by shorting the top fractile creates the desired return. This is probably the most surprising finding given the fact that many of these companies are relatively low-growth CPG concerns. We believe that there might be two items in play here however. The first is the drubbing that tobacco firms took in the 1990’s in the form of their public relations nightmare, class action and state’s lawsuits and the introduction of smaller firms unburdened by large settlements. This may have led investors to look for battered firms (low 3 month returns) that later generated large returns. Likewise, firms that were flying high for a period may have enjoyed a little “irrational exuberance,” making them prime candidates for a price drop. The second is that this SIC code has the smallest number of firms that we evaluated. Therefore, large swings by one or two large cap firms may have influenced our results more than anticipated.
Industry: Electronic & other electric equipment (SIC code = 36xx)
Factor 1 / 1 / 2 / 3 / 4 / 5 / L - S / BenchmarkFractile returns / 32.46% / 21.03% / 33.84% / 20.04% / 37.40% / -4.94% / 16.30%
Std Dev / 39.85% / 28.39% / 33.82% / 35.33% / 36.79% / 28.92% / 13.68%
Avg Relative Performance / 2.50 / 2.63 / 1.88 / 2.75 / 2.13
Factor 2 / 1 / 2 / 3 / 4 / 5 / L - S / Benchmark
Fractile returns / 35.48% / 31.55% / 20.85% / 29.69% / 28.87% / 6.61% / 16.30%
Std Dev / 38.02% / 32.63% / 31.93% / 32.32% / 38.80% / 31.05% / 13.68%
Avg Relative Performance / 2.13 / 2.19 / 2.63 / 2.44 / 2.50
Factor 3 / 1 / 2 / 3 / 4 / 5 / L - S / Benchmark
Fractile returns / 26.77% / 23.81% / 30.45% / 29.79% / 33.02% / -6.26% / 16.30%
Std Dev / 27.06% / 39.75% / 36.01% / 40.71% / 36.98% / 19.76% / 13.68%
Avg Relative Performance / 2.13 / 2.75 / 2.50 / 2.38 / 2.13
Factor 4 / 1 / 2 / 3 / 4 / 5 / L - S / Benchmark
Fractile returns / 37.00% / 22.42% / 21.72% / 31.52% / 16.78% / 20.21% / 16.30%
Std Dev / 34.97% / 26.69% / 31.03% / 38.08% / 42.56% / 31.01% / 13.68%
Avg Relative Performance / 1.69 / 2.56 / 2.56 / 2.19 / 2.88
Factor 5 / 1 / 2 / 3 / 4 / 5 / L - S / Benchmark
Fractile returns / 32.20% / 31.40% / 13.00% / 19.31% / 26.02% / 6.17% / 16.30%
Std Dev / 35.78% / 32.76% / 31.69% / 34.14% / 28.07% / 24.87% / 13.68%
Avg Relative Performance / 1.81 / 2.06 / 2.81 / 2.75 / 2.44
Factor 6 / 1 / 2 / 3 / 4 / 5 / L - S / Benchmark
Fractile returns / 31.18% / 31.55% / 21.25% / 24.58% / 31.01% / 0.17% / 16.30%
Std Dev / 39.64% / 32.97% / 25.25% / 27.19% / 35.77% / 25.06% / 13.68%
Avg Relative Performance / 2.50 / 1.94 / 2.38 / 2.50 / 2.56
The only factor that gave a hint of hope was the reinvestment factor. As noted in our energy discussion, we believe that this factor is a proxy for growth as companies with high growth potential prefer to reinvest earnings rather than pay them out as dividends. As IT grew at enormous clips during the 90’s it is reasonable to believe that the electronics firms pursued this strategy and reaped the rewards in returns. Therefore, this factor seems to be explained fairly well within industry. What will be interesting to learn is whether this factor continues to be important during the 2000’s.
Industry: Communication (SIC code = 48xx)
Factor 1 / 1 / 2 / 3 / 4 / 5 / L - S / BenchmarkFractile returns / 16.30% / 9.95% / 10.62% / 18.29% / 31.63% / -15.33% / 16.30%
Std Dev / 25.93% / 20.03% / 24.36% / 26.25% / 47.16% / 39.31% / 13.68%
Avg Relative Performance / 2.38 / 2.63 / 2.63 / 2.25 / 2.00
Factor 2 / 1 / 2 / 3 / 4 / 5 / L - S / Benchmark
Fractile returns / 12.67% / 18.91% / 10.50% / 24.59% / 41.72% / -29.06% / 16.30%
Std Dev / 27.78% / 19.72% / 24.91% / 28.27% / 50.83% / 42.93% / 13.68%
Avg Relative Performance / 3.06 / 2.19 / 2.75 / 2.31 / 1.56
Factor 3 / 1 / 2 / 3 / 4 / 5 / L - S / Benchmark
Fractile returns / 14.79% / 18.41% / 10.22% / 20.14% / 24.78% / -9.99% / 16.30%
Std Dev / 16.67% / 23.02% / 31.88% / 36.74% / 36.22% / 28.31% / 13.68%
Avg Relative Performance / 2.13 / 2.50 / 2.88 / 2.25 / 2.13
Factor 4 / 1 / 2 / 3 / 4 / 5 / L - S / Benchmark
Fractile returns / 19.39% / 13.45% / 14.99% / 22.54% / 14.36% / 5.04% / 16.30%
Std Dev / 21.14% / 20.25% / 27.96% / 36.16% / 44.04% / 34.47% / 13.68%
Avg Relative Performance / 1.88 / 2.31 / 2.56 / 2.38 / 2.75
Factor 5 / 1 / 2 / 3 / 4 / 5 / L - S / Benchmark
Fractile returns / 19.45% / 16.76% / 14.62% / 17.68% / 15.92% / 3.53% / 16.30%
Std Dev / 24.33% / 20.85% / 18.87% / 25.54% / 26.30% / 26.71% / 13.68%
Avg Relative Performance / 2.31 / 2.38 / 2.50 / 2.31 / 2.38
Factor 6 / 1 / 2 / 3 / 4 / 5 / L - S / Benchmark
Fractile returns / 19.45% / 16.76% / 14.62% / 17.68% / 15.92% / 3.53% / 16.30%
Std Dev / 24.33% / 20.85% / 18.87% / 25.54% / 26.30% / 26.71% / 13.68%
Avg Relative Performance / 2.31 / 2.38 / 2.50 / 2.31 / 2.38
As with food & tobacco, the telecom industry did respond to the 3 month momentum factor. This is not entirely surprising given the discussion noted above in energy. What inspires greater discussion however is why the 3 month momentum piece is effective for food & tobacco and telecom, while energy showed life in both the 1 and 3 month momentum swings (although to a lesser extent in 3 month). An examination of the levels of volatility within each sector may give us some insight. The energy sector exhibits much higher volatility in both 1 and 3 month momentum measurements than do food & tobacco and telecom. Perhaps this extra volatility makes quicker strategies (1 month momentum) effective, with effects lasting as long as 3 months. However, our next industry casts that theory into some doubt.
Industry: Depository institutions(SIC code = 60xx)
Factor 1 / 1 / 2 / 3 / 4 / 5 / L - S / BenchmarkFractile returns / 5.09% / 21.22% / 19.55% / 29.70% / 39.03% / -33.94% / 16.30%
Std Dev / 20.40% / 20.27% / 21.42% / 23.68% / 22.48% / 12.97% / 13.68%
Avg Relative Performance / 3.63 / 2.56 / 2.63 / 1.94 / 1.13
Factor 2 / 1 / 2 / 3 / 4 / 5 / L - S / Benchmark
Fractile returns / 6.55% / 18.08% / 19.50% / 21.38% / 35.98% / -29.42% / 16.30%
Std Dev / 21.17% / 19.60% / 20.59% / 22.73% / 26.07% / 17.29% / 13.68%
Avg Relative Performance / 3.38 / 2.44 / 2.44 / 2.25 / 1.38
Factor 3 / 1 / 2 / 3 / 4 / 5 / L - S / Benchmark
Fractile returns / 32.95% / 22.54% / 16.49% / 13.45% / 17.90% / 15.05% / 16.30%
Std Dev / 23.20% / 23.69% / 22.44% / 20.91% / 23.22% / 13.28% / 13.68%
Avg Relative Performance / 1.25 / 2.13 / 2.69 / 3.06 / 2.75
Factor 4 / 1 / 2 / 3 / 4 / 5 / L - S / Benchmark
Fractile returns / 22.95% / 18.25% / 19.15% / 15.67% / 19.16% / 3.79% / 16.30%
Std Dev / 23.28% / 22.03% / 21.24% / 21.09% / 23.23% / 15.01% / 13.68%
Avg Relative Performance / 1.94 / 2.69 / 2.31 / 2.81 / 2.13
Factor 5 / 1 / 2 / 3 / 4 / 5 / L - S / Benchmark
Fractile returns / 27.44% / 20.05% / 22.96% / 19.52% / 13.65% / 13.79% / 16.30%
Std Dev / 21.80% / 20.04% / 19.75% / 23.07% / 25.00% / 12.34% / 13.68%
Avg Relative Performance / 1.88 / 2.19 / 2.00 / 2.56 / 3.25
Factor 6 / 1 / 2 / 3 / 4 / 5 / L - S / Benchmark
Fractile returns / 24.00% / 29.16% / 17.79% / 16.73% / 18.76% / 5.24% / 16.30%
Std Dev / 22.86% / 22.65% / 22.57% / 23.00% / 25.85% / 13.62% / 13.68%
Avg Relative Performance / 2.13 / 1.81 / 2.44 / 2.94 / 2.56
The banking sector showed a promising stretch for both 1 and 3 month momentum factors.Again, going long on the bottom fractile and shorting the top brings great returns.It is not entirely clear why this happened for the staid banking industry.We postulate that constant consolidation may cause some pretty generous spreads, while large banking fees in the mid to late 1990s may have also contributed to momentum swings. The consensus forecast revision ratio also showed some promise, but the returns were very similar for long/short hedge to the S&P 500.
Upon running out of sample data, the 3 month momentum factor also diminishes in importance. We wonder whether this is a reflection of markets reacting quicker to information (becoming more efficient in some ways) as a 3 month reaction period is simply too slow, even for the less volatile banking sector.
US Common Stocks
For a comparison reference, we ran these same factors for a universe of all US based common stocks.While the dividend yield, forecast revision ratio and change in consensus FY estimates showed good returns, they did not beat the S&P during that period of time.This brings us to an important part of our evaluation.
7.Analysis
SIC codes / Factor 1 / Factor 2 / Factor 3 / Factor 4 / Factor 5 / Factor 6 / Factor 7 / Factor 813xx / good / good
36xx
48xx
60xx / good / good
20xx + 21xx / good
On the surface, it would appear that these factors are not effective in setting up a long/short hedge fund upon examining the output for US common stocks.However, when we dig a little deeper into sectors, it appears that some of these factors may be important for certain types of businesses.This leads us to believe that there are still promising searches within sectors, geographies or other assets that may offer solid returns.
We also have some misgivings about these findings however.To get those high returns within a sector, the investor often has to take on much more risk.We believe that this a reflection of the much smaller pool of companies within each SIC code.While this might reduce the overall level of transactions costs, the volatility may not justify the larger returns.Thus, we will continue our pursuit of higher returns with reduced volatility…
8.Bi-variate Screens
Of the two industries where we found multiple interesting factors–Oil and gas extraction (SIC code 13xx) and Depository institutions (SIC code 60xx) – we were intrigued by how the factors would interplay. By using the FactSet bi-variate functionality, we were able to generate sorting for each industry, and group them independently.
Industry: Oil and gas extraction (SIC code = 13xx)
While there is no interaction between the two factors based on the FactSet bi-variate independence postulate, it appears that by letting the one month price momentum factor subdivide the original bucket formation on reinvestment rate, we lose coherence originally found in the single variate sorting. This is a case where the aggregate of each bucket per one factor provides more information in a predictive sense than in finer granularization. This is true vice versa as well. So while we may still be able to trade off a spread by taking the bucket-to-bucket corners, this will more than likely prove inconsistent across time.