Definition: The attempt to anticipate of predict the future movement and direction of a market, historically through the use of technically based indicators or fundamental economic data
I believe many traders, especially 'professorial' types believe the markets are always "efficient", meaning, it's relatively impossible to time any market. They believe that at any given point in time, the market is digesting all available information...and filters that information through buyers and sellers to find the "best price". There's startling evidence to the contrary though, especially in light of the evidence from traders who use market timing strategies successfully to accumulate better than market average profits year after year after year.
Jack Schwager in his book, "The new market Wizards" wrote this in regards to the 'efficient market theory':
"After interviewing some several dozen profitable successful traders, I have come to the conclusion that the markets are not random".
There are 3 market timing techniques that both relatively new traders as well as seasoned veteran market traders can employ to see an immediate increase in success when employing market timing strategies
Annual market timing cycles
When looking at market charts, especially daily charts, it's relatively easy to discover days that are "pivot days"...days where the market direction changes and the trend reverses anywhere between several days or longer. What most traders don't know, and what many trading companies don't teach, is that these days, year after year after year, produce reversals out into the next year. That is to say, if you scan back across the previous one or two years worth of pivots, and project those dates forward - you'll see that the past pivot dates of the prior 365, 730 days turn out to be the pivots for the upcoming year.
In the above chart, you'll see what happens at the days from the prior year's movements and pivots. Note the percentage of times that the market turns on those dates the following year. This powerful technique is very rarely taught to beginning traders because of its relative ease of study and forward implementation. Many traders are overwhelmed by the rapid inflow of information from live tv feeds, as well as complex assortments of technical indicators - as opposed to discovering years later that it was simply "time" for the stock or index to go up, or reverse.
Many veteran traders ascribe these daily bar reversals to large institutional fund balancing...times of year when heavily traded portfolios involving these instruments are sold off to pay for dividends and subsequently liquidated. Whatever the theory behind it, these powerful daily market timing pivot strategies are relatively simple to study, and powerful - statistically speaking- when those dates are projected into the next year's market.
Intraday market timing strategies
Intraday performance of the major equity indices has an observable and cyclic/repetitive/rhythmic structure. Traders that have a few years of experience will tell you that they are waiting for a "10:00 reversal" or are looking to buy "weakness on the 2:15 dip". Are there time periods - statistically speaking, where the market is most likely to reverse - go up or down? A few years ago, Tradestation did some intraday studies with regards to times of market throughout the day and found some very interesting trends. While a certain degree of price movement will always be random, we'll see that there are biases - "self fulfilling prophecies" that can provide a statistically relevant edge with regards to market timing strategies intraday.
In their paper, "Mapping the Intraday Price Movement in the S&P Index" , technicians constructed a relatively fine resolution (60 minute increments) study of positive (bullish) and negative (bearish) behavior. When we first look at this study, you'll note that there is a significantly large number of positive periods in the ten o'clock hour, as well as in the four o'clock hour. The majority of the returns from the ten o'clock session come from the pre-market session.
As we go back to 1987, 21 out of the 25 events had average gains that were positive for the four o'clock interval, or approximately 84%...far from a 'random' distribution. Moreover as we look closer, it appears that the market 'rests' in the eleven o'clock hour after it's initial morning run towards positive gains. Even closer examination shows that if stocks close higher on average into the three o'clock time period the probability of their moving even more upwards into the next hour is almost seventy percent. Again....far from average.
Incorporating data mining to uncover leading intraday turning times
Finally, there's an exponential surge in the equities and futures, as well as forex markets of algorithmic trading. Computers that are programmed to trade for the large funds and investment institutions. This phenomenon has breached the half way point, with over 60% of US and UK funds being traded electronically and programatically by these algorithms.
The idea of the markets being predictable is not a new one - and was first demonstrated as such by WD Gann back in the 1920's and 1930's. Gann theorized that the markets were projecting future movements in time, via the averages of the past movements of the instrument. Specifically, " by studying the past records of the averages of stocks you will determine for yourself that history does repeat, and that by knowing the past you can tell the future". The modern day equivalent of that technique involves data mining - a technique championed and funded as of this article by C.I.A. and leading hedge funds through a software company called "Palantir" whose "Prisim" software studies the market's complex timing patterns and feeds traders and algorithms the best times to trade as per the results of the data mining.
BackToTheFutureTrading.com produced a retail-trader friendly Ninjatrader indicator version (in our opinion) of this data mining software that studies the past records of the averages of any instrument between 1 minute and 240 minutes to discover the times the market has been moving up, or down - consistently - for how long and in what direction. Similar to the Tradestation study above, they've discovered that these sorted times discovered by the data mining software are accurate about 70% of the time out into the future of the market the next day. This allows traders to sit and wait at pre-determined times and wait for moves to form and take off, as opposed to the conventional technique of applying multiple lagging technical indicators to a chart and using that lagging information to base an idea for what the market will do next.
In this chart, you can see what the data mining software discovered could potentially happen the next day. In the bar chart on the right, you'll see what actually happened at those times the very next day, some 24+ hours later. Analysts theorize that the ability to study the markets this way and project turning times forward into the future (a data mining market timing strategy) only increases in efficiency as more algorithms are programmed to trade electronically.
To verify this is happening, traders can (using an excel sheet) use the same technique described above for daily bars go back and see what times over the past 2 weeks the market is acutally turning and trending at. After a few hours of study, any intraday trader will discover that there are certain times, on certain days, where the market has a 70%+ chance of going up or going down....defying academics who claim the market cannot be forecast.
One trader interviewed by Schwager in his book said this to those academics, "That's why their professors, and I'm making money trading".
However you want to address it, the market you trade is not random. We've uncovered these, and even more sophisticated techniques for uncovering "what the market will do next". The more we study these techniques, the more we're convinced that Schwager's traders were right...and deservedly profitable.