Saturday, April 23, 2016

How I Work: A Hunger Games for Traders

A trader, and reader of Private Trader, recently sent in an email asking how I work, not so much the rules or options strategy that I follow (posted here), but the nuts and bolts of my daily routine: What am looking for practically in making my decisions? What tools am I using to get to the decision point?

(A note: All times in this essay are New York time, except where noted; I live in Portland, Oregon, so I'm actually doing the work three hours earlier local time.)

My daily routine is a series of screens that potential trades must pass through before reaching the point where I'll consider doing a full analysis. Think of it as a series of challenges such as those served up to the contestants in the Hunger Games, and of the trader/analyst as the Gamemaker Seneca Crane.

Many stocks compete; few survive.

Step by step, in my role as gamemaker, here are the challenges that I send against the 500 or so stocks that I analyze each day.

Let the games begin, O Stocks, and may the odds be ever in your favor.

Actually, I'm running two games simultaneously, one based on trading signals, and the other based on earnings announcements.


A must read for traders:

Option Volatility and Pricing
by Sheldon Natenberg


1. For the trading signals game, my first step is to define the contestants among whom I shall search for the few Katniss Everdeens that I shall eventually consider for trades with a full analysis. I'm looking for liquid stocks trading on the U.S. markets and having options.

For this work, I turn to the Stock Screener on Zacks Investment Research (zacks.com), to which I subscribe. The query is as follows: Average volume is 1 million shares a day or greater, market capitalization is $10 billion or greater, which means that all are in the upper reaches of the Large Cap category.

I run the screen each week around 3 p.m. on Friday, and it will generally serve up around 450 or so stocks. To that I add a fixed list of liquid exchange-traded funds that will bring the total near to 500. The actual number of stocks fluctuates as market cap and average volume fluctuates for symbols on the margin.

I post the precise number of stocks analyzed at the start of the daily Prospects post and also as an item in my Week Ahead discussion of upcoming events, which is generally posted each Sunday

For the earnings announcements game, I use the list of earnings announcements published by Yahoo! at https://biz.yahoo.com/research/earncal/today.html, copying and pasting the information into a Google Sheets spreadsheet. I do this analysis once a week, generally on a  Thursday. Sometimes, during peak earnings season, the number of potential earnings announcements for a week exceed 1,000 stock symbols.

For either game, I call the potential trades my trading "universe".

2. For the trading signals game, the next step of screening relies on tech, in the form of a program called ptdata that I wrote, using the perl programming language. The program screens the symbols in my universe for breakouts beyond the 20-day price channel, using a variation on the Turtle Trading method.


Way of the Turtle
by Curtis Faith


The program runs each day from about 8:30 p.m. and compares the closing price for the day of the analysis to see if it is higher or lower than the highest or lowest closing price of the past 20 days. Those that meet the test in either direction are considered to have passed the screen and are written to a comma-delimited spreadsheet file that I can import into Google Sheets.

For symbol that give a trading signal, the program also calculates historical odds of a trading signal being successful, going back over the past 12 months of trading signals, as a way of flagging symbols that are prone to whipsaws, the absolute bane of any technical trader.

The number of symbols in each report varies wildly, depending upon what the markets do, running anywhere from none to 30 or 40, rarely more.

In truth, there's nothing magic about price channels and Turtle Trading. Any signaling system will do, as long as it gives unambiguous signals that can be calculated programmatically. Moving average breakouts would be another possible signal to use; a price close above the moving average is a bull signal, and below, a bear signal.

An example of a method that won't work well is Bollinger Bands, which are much nuanced in its signaling and usually requires the human excellence in pattern recognition to work.

The program uses a database of daily prices for all U.S. stocks -- open, high, low, close, volume -- which I update daily using EODData LLC (eoddata.com) as my source.

The ptdata program runs on a small Ubuntu Linux server built by an Israeli company: A Utilite Pro from CompuLab, based in Haifa, running an ARM Cortex-A9 chip. That means very low power consumptions, so I can keep it up and running 24-7 for very little cost.

Servers, of course, all have names. Mine is named Buffy, in honor of Joss Whedon's iconic character, Buffy the Vampire Slayer

For the earnings announcement game, I take the Google Sheets spreadsheet and perform a series of manual screens.

First, I eliminate the foreign announcements (Yahoo! is quite comprehensive) by screening out all stock symbols that have a period in them.

Next I screen for price ($30 or greater) and volume (Usually 1 million shares a day or more; sometimes higher if the number of prospects is greater than I can reasonably handle.)

I then normalize Yahoo!'s text tables describing the time of day of an earnings announcement into three categories: am, pm, mm (for the middle of the trading day) and xm (for time-of-day unknown).

The final step is to assign an action date: The day of the announcement when it is scheduled for after the closing bell, and the day before the announcement when it is happening before the opening bell.

3. Now the real work begins, a series of manual screens. I do them by hand because I have no easy (read inexpensive) way to bring the data into a my server.

For the trading signal game, I eliminate all signals that are within six weeks of earnings announcements, or that were given in the first trading session during an announcement. This step eliminates almost all symbols when peak earnings season looms.


Technical Analysis Explained
by Martin Pring

I sort the signal out by bull and bear, and within each by those having high odds of success and low odds of success. Those with high odds as candidate for vertical option spreads, which are directional. Those with low odds, having a tendency to whipsaw, are candidates for non-directional trades, such a iron condors.

I report those who survived the screening in the daily Prospects post, which is published genrally after 10:30 p.m.

For both games, and for my own information, I determine implied volatility (IV) relative to its history, sometimes where it stands within the most recent range, or swing, and if that proves onerous due to the number of signals, where it stands relative to the 12-month range, which I can get directly from the ThinkOrSwim platform run by my brokerage.

The relative level of implied volatility is expressed as a percentile of the range. For the recent range, I require the IV be in the 60th percentile for greater; for the 12-month range, in the 50th percentile or greater.

I select the expiration month for the options that I use in the trade: Those expiring between 30 days and 42 days (four to six weeks) of the trading date, whichever options series expires soonest.

I also note the open interest, a measure of liquidity, on options filling the the key positions in building a trade. I want open interest to be 100 contracts or greater on call and put options at strikes having an 89% or lower chance of expiring out of the money for maximum profit.

I get the IV, open interest and chance information from the ThinkOrSwim platform. I don't use either one for screening at this point, but only to get a preliminary idea of how viable the trades might be.

At this point I gather some data to help engage my pattern recognition sense, the excellent toolkit that resides in the brains of every human and that is, in my opinion, the most useful arrow the private trader's quiver.

I take a look at the stock chart to identify historical trends, and,  identify what brokerages and their analysts assessments are. I have have no hard and fast rules for using this data, but it just gives me a sense of whether I like the trade or not. What does my gut tell me?

In addition to looking at trends, I'll often use Elliott Wave analysis as a means of better understanding where a stock stands in its cycle.


Elliott Wave Principle
by Robert Prechter

4. Implied volatility and open interest can be rapidly changing. The final screening for both games happens on the day of the trade.

I do a final calculation of implied volatility percentile and of open interest. I reject without a full analysis prospects that fail either test, reporting that decision in my daily Agenda post, which generally is published on Private Trader about 10:30 a.m.

Those fortunate survivors of the Hunger Games go to full analysis.

I won't go into the full analysis process here, which is entirely transparent. Each is headlined with the symbol and "Analysis" (as in "AAPL Analysis", and includes a decision on whether to trade or not.

I also do a summary of the day's trades, and rejections, in the daily Outcomes post, which generally is published between 3:15 p.m.and 4 p.m.

5. And after that, about 1 p.m. my local time, the games are over and the day is mine. I walk three miles, ending up at my favorite coffee house where I hang with friends and discuss every topic under the sun.

But there's still prep to do for the next day. Ptdata, running on Server Buffy, sends me an email around 5:30 p.m. my local time saying that the initial screening is done, and I swing into step 3, preparing the Prospects report.

In truth, the markets are a never ending cycle, and so must be the days in a private trader's life.

-- Tim Bovee, Portland, Oregon, May 23, 2016


The Bible of Options Strategies
by Guy Cohen


References

Tradecraft: Playing the odds to build winning stock market trades from options, a description of how I trade, can be read here.


Elliott wave analysis tracks patterns in price movements. StockCharts has a good explainer. The principal practioner of Elliott wave analysis is Robert Prechter at Elliott Wave International. His book, Elliott Wave Principle, is a must-read for people interested in this form of analysis, as is his most recent publication, Visual Guide to Elliott Wave Trading

I can be reached via comments on Private Trader posts or by email at datnillc@gmail.com.

Alerts


Two social media feeds provide notification whenever something new is posted.

Disclaimer
Tim Bovee, Private Trader tracks the analysis and trades of a private trader for his own accounts. Nothing in this blog constitutes a recommendation to buy or sell stocks, options or any other financial instrument. The only purpose of this blog is to provide education and entertainment.
No trader is ever 100 percent successful in his or her trades. Trading in the stock and option markets is risky and uncertain. Each trader must make trading decisions for his or her own account, and take responsibility for the consequences.
License

Creative Commons License

All content on Tim Bovee, Private Trader by Timothy K. Bovee is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.

Based on a work at www.timbovee.com.

No comments:

Post a Comment