April 1997
Letters to the Editor
OLD STOCKI have some old stock from the early 1900s and I was wondering how to go about researching these six different stocks. Do you have any suggestions?
ELIZABETH BUMGARDNERYou could try searching the Internet for information. Speaking for myself, if a stock is not actively traded, I have little interest in it. Companies are like people: They grow old and die.
via E-mailI do, however, think old printed stock certificates make great mementos or wallpaper. There is a market for old stock certificates, just as for stamps. You can find such a market at the Web site https://www.robinsoft.com/. -- Publisher
Editor,
NOVICE TRADER'S NOTEBOOK
I am a pleased new subscriber to your magazine, having been driven from a competing magazine after suffering through years of transition from an informative, news- and analysis-laden tome to its current existence as a promotion piece for commodity funds, trend-following systems and the assorted detrius of self-serving promoters.I enjoy the Novice Trader's Notebook feature. I have noticed that the online version of this section at your Web site only permits me to access certain specific patterns, despite listing many others. How can I retrieve this information?
In addition, the listing does not mention key reversals or "outside days" (that is, at a new trend extreme, prices establish a higher high and a lower low than the session/period immediately prior, and they close in the opposite direction of the trend). Is this intentional, or have you listed this pattern under another topic?
SETH DIAMONDThe topics that you cannot access have not yet been completed but will be soon. We'll continue to add more entries and more topics, such as the pattern you request, but my time constraints with our magazine deadlines require me to focus on completing the currently outlined set of topics first. I'll look into additional topics later on this year. -- Editor
via E-mail
Editor,
OPTIONS ON FUTURES
I presently trade (sell) stock index options. I have been looking into options on futures. It appears that margin requirements for the Chicago Mercantile Exchange (CME) are about one-fourth to one-third what they are for the Chicago Board of Trade (CBOT). I'm still trying to find out what differences exist between the two exchanges. My risk could be significantly reduced with futures options, all else being equal, as I can sell more options further away from the index price (that is, the S&P 500), and make the same amount of money as I am making now on the Standard & Poor's Options Exchange (SPX).Does your magazine rank stock and futures brokerages? If so, please send me the list. I'm looking for good service at a reasonable price (neither the most expensive nor the superdiscounted services).
Moreover, I'm looking for a low-cost, basic options graphing package with the capability of graphing the strike price versus the option price, option decay versus time-to-expiration, and so on.
Any info you send will be greatly appreciated!
ROY W. BRANDT
via E-mailSorry, we don't rank stock and futures brokerage services. As for options graphing software, check the Options Analysis Software and Options Trading Systems sections of our Readers' Choice Awards in the 1997 STOCKS & COMMODITIES Bonus Issue. You will also find a complete listing of software at our Web site. See the Software Comparison Table at https://www.traders.com. -- Editor
Editor,
MONEY MANAGEMENT ARTICLES
How about a few articles on money management in your magazine? Specifically, I have two questions. First, what is fixed-ratio trading, and why, according to some promoters of it, is it better than fixed fractional trading?Second, what criteria do you use when trying to allocate resources to various systems? For example, I trade three daytrading systems on the S&P 500, one of which is active and the other two special situation systems that trade infrequently. In addition, I trade two other methods (for the intermediate term) on a basket of markets. All these systems can be described in terms of common characteristics such as net profits, profit per trade, profit factor, and so on. How do I combine these to get the smoothest equity line?
GEORGE FAMYThank you for your article topic suggestions. Readers, we would welcome article submissions that would answer these questions! -- Editor
via E-mail
Editor,
GLOSSARY OF TERMS
I am a long-time reader of STOCKS & COMMODITIES and wish to say you have a first-rate magazine.I have a comment/question for you about two terms in your Traders' Glossary: Under accumulation/distribution, a reference is made to the Chaikin indicator, while under the Chaikin indicator entry, a reference is made to accumulation/distribution. Can you please define accumulation/distribution?
JEFF L. ZEILER
via E-mailSee the sidebar titled "Chaikin indicators" in this month's interview, "On Rational Group Structure: John Bollinger and Group Analysis." -- Editor
Editor,
TELLING IT LIKE IT IS
I usually read your magazine cover to cover when I receive it each month. But when the January 1997 issue came, I noticed on the cover a reference to an interview inside with Mark Douglas, who is the author of the book, The Disciplined Trader. I immediately jumped right to the interview. I also suggested to all my customers that they do the same. There isn't a single commodities trader out there (rookie or veteran) who couldn't get some excellent, profitable information from this well-spoken author, who simply tells it the way it is.When I first started as an S&P floor trader, a fellow trader suggested I read Douglas's book. I decided to pick it up, and the wealth of information was worth 10 times the price. The Disciplined Trader should be in every trader's library and should be reread at least a few times each year.
Way to go, STOCKS & COMMODITIES, for an excellent interview. You've no doubt given your readers some superb information that they can use to improve their trading. Keep up the good work!
LARRY LEVIN
Chicago, IL
Editor,
HISTORICAL VOLATILITY DEBATE
Historical volatility in option theory is defined as the annualized standard deviation of the natural logarithm of the daily rate of return on a security. We calculate each daily rate of return as today's price divided by yesterday's price. Then we take the natural logarithms of these numbers and compute the standard deviation in the usual way. If the standard deviation is taken over a period of less than one year, it must be annualized (by definition) to arrive at the historical volatility.If a normally distributed (Gaussian) random variable has a standard deviation over a time period t, then the standard deviation over a different time T is found by:
STD(T) = STD(t) * SQRT(T/t)Equis International's Web site (www.equis.com) has a page displaying MetaStock custom formulas. There, historical volatility is defined as follows:
10 Day Historical Volatility Std(Log(C/Ref(.C,-1)),10) * Sqrt(365) * 100 100 Day Historical Volatility Std(Log(C/Ref(C,-1)),100) * Sqrt(365) * 100It is my understanding that to annualize a standard deviation taken over any time period, we must multiply by the square root (Sqrt) of the number of such periods in a year. Thus, to annualize the 10-day standard deviation, we must multiply by Sqrt(365/10) and not by Sqrt(365), as Equis's equation shows. Similarly, to annualize the 100-day standard deviation, we must multiply by Sqrt(365/100), not by Sqrt(365), as Equis's equation shows. Multiplication by Sqrt(365) would only be appropriate to annualize a one-day volatility.
Equis's logic says no matter what time period you use to compute the standard deviation, you multiply the answer by Sqrt(365) to get the historical volatility. Extension of this fallacious reasoning would mean that if you computed the standard deviation over a one-year period (365 days), you would still multiply by Sqrt(365). This is absurd. A standard deviation taken over a one-year period is already annualized by its inherent nature. Using my procedures, you would multiply the one-year standard deviation by Sqrt(365/365), or unity. Using Equis's method, the answer would be almost 20 times as large.
I hate to question the logic of such a large software company as Equis, but the folks at Equis insist they are correct. [See Equis's response below.-Ed.] Can you render a decision as to which of us needs enlightenment?
JOHN R. ZIMMERMAN
via E-mailBill Forman, Equis International's customer support manager, replies:
We gladly accept your questions. It keeps us on our toes and can help other traders.
When we reread Sheldon Natenberg's writings on historical volatility and recalculated his formulas, we found that he does not take the standard deviation lookback period into account when he annualizes. The square root taken in his example is for a weekly time period, and therefore the square root of 365/7 is taken, even though the standard deviation lookback periods are 10. When this is extrapolated for a daily time period, the square root taken would be 365/1. The same standard deviation time periods are used. The divisor is the time period, not the lookback period.
We have discussed this with another published options trader and author who uses historical volatility extensively. His daily historical volatility calculation is identical to the daily historical volatility calculation we show on our Web site -- which we did in fact change after you brought the original error to our attention
[std(log(C/ref(c,-1)),10)*sqrt(365)].Again, please feel free to bring anything concerning our products and services to our attention. Open discussion helps us all.
Editor's reply:
Based on my sources, neither of you is correct. Page 463 of Options as a Strategic Investment by Larry McMillan explains it well. But to recap, when you are using standard deviation, you are calculating the square root of the squared deviations from the mean, summed over n periods; then you divide by n or n-1. In other words, the n in the denominator results in a daily basis figure, so it doesn't matter if you use a 10-day or a 100-day lookback - the result is a daily average.
Thus, to annualize the volatility, you multiply the result by the square root of the number of trading days in the year. McMillan uses 260 trading days, but one of my Contributing Editors, Dennis Meyers (who has a Ph.D. in applied mathematics, owns a seat on the OEX, is a former floor trader and was once a quantitative analyst for a primary dealer), points out that there are actually 252 trading days in a year, and therefore the one-day average volatility should be multiplied by the square root of 252, not 260.
Editor,
FUTURES LIQUIDITY
I've been reading the Futures Liquidity feature for years now in STOCKS & COMMODITIES, and the explanation never changes. Unfortunately, it is not very clear! Can you give me a simpler explanation than that given by the general introduction at the top of the page and by the equation at the bottom of the chart? This section appears to offer very useful information, but I cannot make heads or tails on how you come up with the numbers.GARYFutures Liquidity is presented as a tool to help readers select and use liquid contracts when testing and trading systems. Although more contracts are available to trade than are listed in this table, the ones listed in Futures Liquidity are some of the most liquid.
via E-mailA mistake that some traders make is not properly assessing the value, volatility and liquidity in the contracts they are trading. In deciding which contract to trade, some traders only look at the margin requirements or the contract value. But using our Futures Liquidity table, traders can get an idea of how the contracts compare when taking into consideration volume, open interest and the contract's maximum price excursion (volatility).
For example, if you simply compared, say, Lind-Waldock's February 1997 margins for the March 97 Standard & Poor's 500 contract ($15,120) with the March 97 US Treasury bond contract ($2,700), you may conclude that trading 5.6 bond contracts would equal one S&P contract -- which is not exactly true. Meanwhile, if you simply compared the contract value of the two contracts, you might conclude that trading four bond contracts ($100,000/contract) would equal one S&P contract ($800 x 500 = $400,000/contract). Again, this is not exactly true.
A more meaningful calculation might take into consideration the maximum adverse excursion of the contract. For instance, if you multiply the tick dollar value (that is, $31.25 for bonds, $25 for the S&P) by the three-year maximum price excursion, you'll find that it takes 13 T-bond contracts to equal one S&P 500 contract. This calculation helps you to identify the relative size of each contract. This is especially important when you're trying to diversify your trading system to minimize market risk. If your trading system works with several markets, you may choose to trade multiple markets that are unrelated. That way, if you encounter a limit day in one market, your exposure may not be as great, since it's unlikely that unrelated markets would also experience a limit day.
The relative contract liquidity, the determining factor by which our Futures Liquidity table is sorted, multiplies three variables: the number of contracts to trade, the total open interest and the volume factor. -- Associate Publisher
Any reader interested in submitting an article to STOCKS & COMMODITIES magazine may send for our Author Guidelines, which describe categories of topics accepted, style and submission requirements, copyright information and remuneration. Call 206 938-0570 or view the Guidelines at our Web site.
EDITORIAL SUBMISSIONS
