Insider is a mix of sources and references used in compiling Stikky Stock Charts, trivia, things we left out, and known errors

It also provides some insight to anyone interested in the thinking behind the Stikky method.

If you have a specific question you would like to see addressed here, email : insider1 (at) stikky.com.

Some of the academic papers quoted here are highly mathematical. References to them are in the standard academic shorthand: "Int J Fin Econ 1997 October:2(4):267-280" means "The International Journal of Finance & Economics, October 1997, Volume 2, Issue number 4, pages 267-280."

Charts. Our aim was to create charts that look similar to those you would see online and print out to draw trendlines on. Charts were researched and created using the JavaCharts tool at www.prophet.net. They are 72dpi large screen shots, re-sized in Photoshop bringing them to a resolution of 300dpi. A small number of the charts, particularly ones early in the book, had one bar adjusted to make the trend easier to identify for beginners.

How to read this book: Many people think the stock market is too complex. Testing of an early draft of this book revealed that some readers automatically assumed they would not understand it and so, of course, didn't. The topic appears much more intimidating than it is. This paragraph attempts to allay their fears.

2: A distant relative has left you $100,000. We spent a long time looking for an opening that immediately engages lay readers before we hit on the idea of giving them $100,000. It seems to work.

6: Already in the chart. This is a fascinating point that is used to argue both for and against technical analysis. If all information is in the chart, reading the chart is as powerful as fundamental analysis. On the other hand, if all information is in the chart, isn't chartist information already in it too? The prevailing assumption on Wall Street is the 'efficient market hypothesis', namely that markets discount everything so that, if there were profits to be made from pattern-spotting, enough people would use them to eliminate the inefficiency. But the 'Grossman-Stiglitz paradox' points out that total informational efficiency is in fact impossible. It runs like this: if all information is already in the price, traders will have no incentive to study the fundamentals or pay someone else to, so information would not be uncovered in the first place. Where does this leave us? Perhaps that it's the interpretation of available information by market participants that is already in the price. Those interpretations evolve, even in the light of no new information, and, as more people 'catch on', trends are created. Genuinely new information (a company discovers the cure for cancer) could also arise, of course.

7: Recent research. Note the claim that is made here: that the chart can sometimes predict future prices. This is not necessarily evidence that technical analysis is profitable (since acting on predictions itself costs money-eg, commission). For evidence supporting this claim for foreign exchange markets see B LeBaron, 1996, 'Technical trading rule profitability and foreign exchange intervention' National Bureau of Economic Research Working Paper Series March 1996, Working Paper 5505, available for a fee at www.nber.org/papers. For stock markets: W Brock et al, 1992, 'Simple technical trading rules and the stochastic properties of stock returns' Journal of Finance 1992:47:1731-64. There is some evidence, though, that these rules were not superior to a buy-and-hold strategy during the 1990s bull market (TC Mills, 1997, 'Technical analysis and the London Stock Exchange: testing trading rules using the FT30' Int J Fin Econ 1997 October:2(4):319-331) and some evidence that they do not work at all (R Curcio, C Goodhart, et al, 1997, 'Do technical trading rules generate profits? Conclusions from the intra-day foreign exchange market' Int J Fin Econ 1997 October:2(4):267-280).

The key paper, though, was Andrew Lo's 2000 study, the first attempt to objectively identify patterns so that a computer, searching through stock market data, could identify them unaided. This was a crucial step since it allowed statistical analysis across a large number of pattern events to be performed. Much more work is needed--for instance, Lo examined only one-day returns for a triggered pattern. See A W Lo, H Mamaysky, J Wang, 2000, 'Foundations of technical analysis: computational algorithms, statistical inference, and empirical implementation' Journal of Finance August 2000:LV:4, which can be downloaded from Andrew Lo's homepage at http://web.mit.edu/alo/www/.

8: Professional traders make many of their decisions. During research, we interviewed a number of professional traders and one of our team was himself previously a bond trader. Evidence for the widespread use of technical analysis by professionals also comes from a recent survey of foreign exchange professionals that found 44 out of 45 chief dealers use technical analysis. See L Menkhoff, 1997, 'Examining the use of technical currency analysis' Int J Fin Econ 1997 October:2(4):307-318.

13: A series of days. We only use daily charts in this book. Why? Our view is that longer-term charts (weekly, monthly) were useful when daily charts covering multiple years were time-consuming to produce. Now a $1000 PC can produce them in seconds, so why use anything else? Shorter-term charts (15-minute) are primarily used for intraday trading, which is not our audience. Having only one type of chart to deal with also makes things easier for the reader.

20: Touch at least three separate bars. Most manuals on technical analysis require only two touches for a trendline. But basic geometry says that you can always draw a line through two points, so any two bars would define a trend. Clearly this is not a strict enough requirement, hence our three-point definition.

22: It's worth a little more practice. We found that readers had far more difficulty drawing trendlines than we expected, so we doubled the number of pages devoted to the topic. This is a good example of a skill that experts consider trivial but is very challenging to beginners.

24: Did you draw one of these? Examples of incorrect answers were taken from real responses of test readers.

52: The release of Windows. Windows 3.0. It soon became apparent that this was the first version of Windows that could challenge the Apple Macintosh in terms of usability.

61: Hint. Hints like this one (and the one on page 63) were added on pages in Sequence One that more than half of the test readers got wrong. There is a careful balance to be struck between making the questions too easy and making them reflect the challenges of real charts. Early in the book, we are worried about readers becoming demoralized, and so the balance tips in favor of hints; later in the book the main concern is making sure readers have experience in situations similar to those they will encounter in charts they analyze themselves, so the balance tips the other way. At the end of the book, though, we want to make sure we don't leave people feeling down, so the challenges get a little easier again. If the difficulty level were plotted from the beginning to the end of the book, it would form an inverted-U.

68: Traders use them and nothing else. Our research turned up a notable difference between traders and market commentators who use technical analysis. Traders tend to use trendlines, few patterns, and maybe one or two indicators such as MACD. Commentators often use the full range of patterns and indicators. This may reflect the fact that traders have less time for analysis.

79: Self-fulfilling prophecy. The book conveys the two main explanations for the persistence of trends: the self-fulfilling prophecy (which is logical, but difficult to prove) and the asset flow model, essentially the idea that the market is a bandwagon with some investors getting on board later than others after a change in sentiment.

87: What would you actually do? Our research showed that fear of speaking to brokers presents a barrier for some people to enter the market, even when they have formed an investment plan. This may be one reason--besides price--for the popularity of internet trading.

99: Support and resistance are stronger. A study of the effectiveness of support and resistance levels provided to customers by trading firms found that they were quite successful in predicting intraday trend interruptions and the predictions held good for five days after their release. A disproportionate number of these levels were round numbers. See C Osler, 2000, 'Support for resistance: technical analysis and intraday exchange rates' Federal Reserve Bank of New York Economic Policy Review July 2000:53-66, available at www.newyorkfed.org/research/epr/00v06n2/0007osle.pdf.

102: You may want to take a closer look. It's unusual to find a 'trick question' in a Stikky book. We used one here to emphasize the need to challenge other people's trendlines.

124: Charles Dow. Charles H Dow wrote what came to be called Dow Theory in a series of Wall Street Journal editorials between 1900 and 1902. William P Hamilton developed the theory further and Robert Rhea published Dow Theory in 1932 based on the work of both Dow and Hamilton.

124: Einstein did the same thing. Albert Einstein published the special theory of relativity in a paper in 1905 and followed it with the general theory of relativity in 1915. The first experimental evidence for relativity came during a total eclipse in 1919.

125: Dow's theory began to attract cranks. A favorite example from one of the best-selling books on technical analysis: a section entitled "The Importance of the Number Three" explains how some chart features appear, spookily, in threes.

125: Fibonacci. Of course, if enough people belief it works, it will. This is one advantage technical analysts have over astrologers.

126: Professor Andrew Lo. See the note to page 7, above.

131: A clue is in the volume chart. Actually Andrew Lo did not find evidence of predictive power in volume, though other researchers have and it is very widely used by analysts.

136: Megaphone. Also called, less memorably, a 'broadening formation'.

139: Price to move up. Some analysts would forecast an upward move only after the chart breaks out above the upper line--a 'confirming' breakout. Similarly, they would wait for a breakout on the downside to confirm a downward move following a descending triangle.

146: Amazon.com. The original draft included the comment, "was AMZN ever worth $110?" If 'worth' means 'what someone will pay and someone else will sell for' then obviously AMZN was worth $110 in late 1999. Researchers have shown that artificial markets run by robot traders also produce unsustainable prices using just a few simple trading rules. This is a long way from traditional financial theory. See the notes on page 191, below.

157: Stop loss. The example given here is a stop loss set $1 below a $10 stock--ie, the stop loss would be triggered after a 10% move. A problem with tighter stops, 1% for example, is that the exiting trade may be triggered too easily, by normal market volatility, effectively ensuring you make a loss. In highly volatile markets, the same thing can happen even to wider stop losses.

161: You may be tempted. This is known to psychologists as the 'endowment effect'--people place an extra value on what they already own. Recent research showed that experienced traders are less susceptible to the effect. See J A List, 2003, 'Does Market Experience Eliminate Market Anomalies?' Quarterly Journal of Economics 2003:118(1):41-71 available for a fee at http://mitpress.mit.edu/journal-home.tcl?issn=00335533.

174: Neckline slopes. Some analysts would be uncomfortable with a neckline sloping as much as this, preferring one that is nearly horizontal or gently upsloping.

176: Double top. Some analysts would prefer less difference in the levels of the two tops than there are in this chart.

178: Eight patterns. With two trendlines, the first sloping up, sloping down, or being horizontal, and the second sloping up more than the first, parallel to the first, sloping up less than the first, horizontal, sloping down less than the first, or sloping down more than the first, there are 13 possible patterns. They are: ascending/descending wedge, broadening ascending/descending wedge, upsloping/downsloping channel, broadening formation (megaphone), right-angled ascending/descending broadening formation, ascending/descending triangle, symmetrical triangle, and rectangle. Though the technical analyis literature covers all of these, the book focuses on those that (a) appear to work, according to the research, and (b) real traders use.

191: Bubbles. The classic text on bubbles, including tulip mania, is Charles MacKay's Extraordinary Popular Delusions and the Madness of Crowds. The classical model of efficient markets has a hard time explaining bubbles since it asserts that prices are always at a fair value. But a simple model of two types of investors--value-oriented and trend-oriented--can reproduce bubbles in the laboratory (and also patterns including triangles and head and shoulders tops). For instance, if corporate earnings rise over time, value-based investors create a rising trend in stock prices. Trend investors (whose principal attribute is that they know a good trend when they see one) then get on board and accelerate the rally. But when there are no more trend investors left to join in (or they run out of funds, or corporate earnings falter) the rally stalls. Everyone then decides to get out at the same time, resulting in a crash. The final stages of the rally are so strong that it continues over the rational objections of value-based investors (as was the case for Amazon, see the chart on page 145). See G Caginalp, D Balenovich, 1996, 'Trend-based asset flow in technical analysis and securities marketing' Psychology & Marketing July 1996:13(4):407-444, available free at www.pitt.edu/~caginalp.

191: Computer simulations. Researchers at the Santa Fe Institute have built a simulated market to allow closer study: the Santa Fe Artificial Stock Market. See W B Arthur, J H Holland, B LeBaron, R G Palmer, and P Tayler, 1997, 'Asset pricing under endogenous expectations in an artificial stock market' in W B Arthur, D Lane, and S N Durlauf, editors, The Economy as an Evolving Complex System II, 1997, Addison-Wesley. Also, K Steiglitz, D Shapiro, 1998, 'Simulating the Madness of Crowds: Price Bubbles in an Auction-Mediated Robot Market' Computational Economics, 1998:12:35-59, available free at www.cs.princeton.edu/~ken.

213: Mexico Fund. The analysis presented here is also based on G Caginalp, D Balenovich, 1996, 'Trend-based asset flow in technical analysis and securities marketing' Psychology & Marketing July 1996:13(4):407-444.

223: Bubble of 2000. The similarity between the chart in the background image on this page and that of the 1929 crash on page 225 is striking.

224: Crash of 1987. A survey carried out immediately after the crash found that about a third of respondents said they were influenced by prices dropping through a long-term trend line. See RJ Shiller, 1987, 'Investor behavior in the October 1987 stock market crash: survey evidence' National Bureau of Economic Research Working Paper Series November 1987, Working Paper 2446, available for a fee at www.nber.org/papers. The background photograph here is the floor of the New York Stock Exchange around 1987. The photograph on the next page is the NYSE floor from the 1920s.

230: Gaps. Typical descriptions of how to trade gaps are good examples of the circular thinking that has given technical analysis a bad name. For instance: a 'measuring gap' is interpreted as meaning the market will continue in the same direction whereas an 'exhaustion gap' means it will reverse direction. How do you tell one from the other? A measuring gap occurs in the middle of a move and an exhaustion gap at the end!

230: Moving average. Evidence for the success of moving average rules can be found in JWC Kwan et al, 2000, 'Forecasting and trading strategies based on a price trend model' Journal of Forecasting 2000:19:485-498. The paper is interesting because it derives a trading rule from the data that turns out to be similar to a popular moving average trading rule (rather than the less reliable method of starting with the rule and finding data to fit).