| ||||||||||||
| Optionetics.com Platinum provides users with quick access to a security correlation tool which can help traders in a few ways. This article focuses on assessing the extent of diversification in a portfolio and if it makes sense to look at a broad index exchange traded fund [ETF] to hedge portions of the portfolio. This tool is available in the Stock Rankers section. Correlations Correlation is a statistical measure, in this instance supplying information about movement between two pairs of securities. It provides both the magnitude and direction of movement, for securities which include individual issues (i.e. a stock) or baskets of securities taken together (i.e. an index or portfolio). It’s important to note that correlation data provides expected movements and is not causal. When considering stock/ETF movement, returns are used to take advantage of a normal distribution that generally emerges from this data. Correlation values range from -1 to +1, with +1 indicating a perfect positive relationship and -1 indicating a perfect negative relationship. In other words, if a pair of securities has a +1 correlation, a 5% move upward in one means the trader would expect a 5% upward move in the other. When extreme movements in the overall market occur, you may find stocks rising or falling together. While this can occur on both the upside and the downside, there are bearish periods that can result in strong positive correlations across the board. For instance correlations that include market data from last fall will skew pairs towards a +1 correlation due to the movement occurring over that period. Proxy ETFs The three main broad index ETFs associated with the US stock market include:
While correlations across all three of these are strongly positive using daily and weekly data over a variety of timeframes, the recent bear market has slightly weakened QQQQs relationship with DIA and SPY. This is due to index construction which excludes financial stocks; names that were particularly volatile over the period. It’s helpful to consider such market conditions when assessing your results. Applying the Correlation Analyzer Using the List Tools menu, create two new lists which include the following securities:
If you’re uncertain as to whether you include a specific stock in the hedge, add it to the Portfolio list. ProfitSource users can transfer a folder of portfolio names into Platinum to create their list. Go to the Correlation Analyzer from the Stock Rankers menu and select your Portfolio List and ETF Proxy list for analysis. To obtain correlations for the last year’s worth of trading, select 250 for the Correlation Days Length. Allow the analyzer to use the default Daily Rate of Return data. Select Run and review the table of results which provides the correlation for each name in the portfolio list with each of the three ETF proxies. You can also cut and paste the data to a spreadsheet for analysis. Values of +0.80 or higher indicate a strong positive, correlation while those with a value of -0.80 or lower indicate a strong negative, correlation. If the correlation values exceed +0.90 or are less than -0.90, the respective relationship is consider very strong. As a slight aside, when the analyzer was run before the close of trading on 11/3/09, the last 250 days of trading resulted in a strong positive correlation between both DIA & SPY (+0.98) and QQQQ & SPY (+0.94). Diversification When the value between a pair of securities is close to zero, they are considered not correlated. There is no relationship between the movement of one and that of the other in statistical terms. In portfolio diversification, investors seek to combine non-correlated securities or negatively correlated securities so that declines in one can be off-set by advances in the other. In a two security portfolio of non-correlated securities, the effect of losses in one security may or may not be dampened by the movement in the other. The main goal of the correlation analysis is to identify securities that are positively correlated to the ETF proxies. This will allow the trader to hedge a group of stocks using options for a liquid ETF, rather than having to hedge all of the individual names separately. This assumes that the trader does not have a delta neutral focus since the hedge will be approximate rather than more precise. Using the +0.80, strong positive level as a cut-off, consider which securities can be included in the ETF proxy hedge. The individual can decide if they wish to use more or less stringent criteria given their preference for a less or more precise hedge (lower correlations will likely lead to less precise hedge). A sample portfolio is used in this portfolio to test two different hedge approaches long with a “no hedge” approach. The stock list or sample portfolio I use is based on names that have recently appeared in LiveTrade case studies and represent a random group. The names and techniques provided here are not recommendations. Remember that portfolio selection and decision-making is a very personal thing and if you feel uncomfortable with your current holdings, take some time when the markets are closed to assess this separately. Create active management plan that will get you to the appropriate portfolio as a starting point. Portfolio Correlations Figure 1 provides the correlations for each stock/ETF in the sample portfolio versus each of the three ETF proxies using daily returns for the last 250 days. The first column (List 1 Stocks) provides the portfolio stocks. Note that SPY is also a security in the sample portfolio.
Figure 1: Sample Portfolio with ETF Proxy Correlations (250 days of Daily Returns) To gain a sense of the diversification that already exists in the portfolio, Figure 2 displays a test of the correlation of daily returns (250 days) for each security versus the other securities in the portfolio. Using the +0.80 threshold to identify a strong positive correlation, you can see that there is only one security pair that meets this criteria (EBAY-SPY @ +0.795).
Figure 2: Sample Portfolio with Paired Correlations (250 days of Daily Returns) Given only one security pair in ten with a strong positive correlation, there is some inherent diversification built into the portfolio which should result in losses realized by one security to be dampened by smaller losses in another. Hedge Approaches Five different portfolios will be tracked on a weekly basis for the next four weeks. These include three that are hedged and two that are not hedged. They include:
Since the focus is a comparison of different hedge approaches the portfolio values will vary to start. Results will be compared from changes in each portfolio. Jan 2010 options are available for each of the portfolio securities, so this month is used for the hedge. Portfolios 2 & 3 use a delta neutral hedge when the individual security’s put is used and the security’s correlation to SPY when SPY is used. No adjustments are made to the hedge during the holding period (11/3 – 11/27). An example of using the security’s correlation to SPY, consider Portfolio 2. The only security that will be hedged with SPY is EBAY, with a strong positive correlation of +0.80. As a result, the SPY put requires -180 deltas to hedge both SPY and EBAY (100 + 80). Portfolio 3 will use each security’s correlation to add to the SPY hedge; these values can readily be reviewed using the last row in Figure 2 (=393 deltas).
Figure 3: Hedged Portfolios The goal of using SPY puts to hedge as much as the portfolio as possible is to reduce costs from slippage and commission. If all it accomplishes is reducing costs, while providing in inadequate hedge, it’s not worth the savings. A midterm update will be posted to my discussion boards on November 17th. Clare White, CMT Questions for Clare? Please visit the discussion board on the homepage of Optionetics.com. For more information on learning how to make money with options, go to the Optionetics.com full site! We empower investors through knowledge.
|  | |||||||||||
