Prediction Wizard

Additional details to the  Prediction Data  Puzzle.

For market participants,  playing  a bit longer timeframe (Investors, Swing traders)   the Weekly predictions might play a more important role in the decision making process, than the Daily predictions.

Some of the characteristics of the Weekly Predictions:

-          As for Daily predictions it is true, that the probable movement of the market will be in the predicted direction,  at least at the beginning of the next week. The movement in the predicted direction might  start the first day of the week or later but the highest the probability is   for the first 1 – 2 days.

-          We might  use Hourly, Daily or Weekly charts to estimate  the potential extent of a move in the predicted direction, along with classical technical analysis measures, similar to  estimates of  potential moves for  predicted Daily directions.

-          The weekly predictor filters out more noise from the markets and for this reason  a bit more reliable and have  a bit higher probability of  being successful. Win probability varies for the different indices, but  usually 5 – 15% higher than the win probability of the Daily predictions.

-          It is interesting that both for the Daily predictions and for the Weekly predictions we have Internal divergences  for example between the NASDAQ  and the S&P500 predictions about 16 – 17% of the time, so out of  100 Daily predictions the NASDAQ prediction is different then the S&P500 prediction 17%  or 17 times and similarly for 100 weekly predictions we got that  NASDAQ weekly predictions point to different direction than the S&P500 weekly predictions 16% of the time or 16  weeks. (This statistics is calculated from the April.  2006 – Dec. 2010 period for the weekly predictions and Jan. 2007 – Dec 2010 period for the Daily predictions.)

-          It is also True for the Weekly predictions, that the  successful predictions  or actual Win probability is better for Longs  in a Bull market  or prolonged Uptrend and Win probability is better for shorts in a Bear market period or prologed Downtrend.

To help you better  evaluate the performance and behavior or the weekly predictors,  we created  new sets of weekly prediction charts, based on the Daily historical data. Those charts can be downloaded from the Downloads section of  our WEB site, and  printed out for further analysis.

As for the Daily predictions we  researched the most  probable reasons, a specific weekly prediction might Fail. We found, that the three most  often  occurring reason is the following:

-          Bearish predictions in a strong Uptrend or after a Bullish trend reversal / Bullish predictions in a strong Downtrend or after a Bearish trend reversal.

-          Bullish predictions at highly overbought conditions / Bearish predictions  at highly oversold conditions (Both occurring in trending and range market periods.)

-          Bearish prediction in a strong Uptrend at a time of a pullback below the SMA20 / Bullish prediction in a strong Downtrend after a move above SMA20. (This can be considered  a special case of the first reason, which needs further attention to  avoid considerable trading losses.) A lot of trend continuation players  use these spots on charts for market entry.

To get some long-term  view about the performance of the  weekly predictions we need to download the weekly prediction history for the index  of our interest and analyse the available data.

For  a quick  overview I created a table about the weekly predictoin win  probability for the NASDAQ index  for the past few years.

The table  below  show the prediction win probability separately for the Bullish and for the Bearish predictions:

It is  very interesting in many respect.

We know that long-term prediction win probability is usually between 74 – 88% for the NASDAQ index.

But in this table we see some outlyers in both directions.  For example the Win probability for Bullish predictions during 2009 for the NASDAQ was the extremely good  96%.  24 / 25 of our Bullish Weekly predictions were correct.  That year  was  clearly above our  historical range.  The most probable reason is, that the FED and the government  entered the market, and their objective was to push up the indices as much as possible as quickly as possible, withouth thinking much about the costs of  the market manipulation long – term.

The 2010 year was by far the weakest year for the Bearish weekly predicttions.  Only 63% of the predictions  came out as expected.  Our data suggest,  that  market manipulation was prevailing throughout most of the year, but  this time the objective was not to push up the indices as much as possible,  but only not to allow them to be pulled down by the natural market forces. This game definitelly cost much less for the manipulators, than  it cost  in 2009 do complete their mission.

In this environment  to  reap the true benefits of the weekly Bearish predictions became extremely difficult,  and  action definitelllly required  caution  and other confirmations (The most obvious is the price confirmation.)  to  initiate  market participation  as a  responce  for  the Bearish  predictions.

We have reason to belive, that  playing the Bearish predictions will continue to be challenging  untill  the market manipulation forces present in the market.

The  so-called QE2, initiated by the FED  in Nov. 2010 is only part of the big picture.

You can  follow the  day-to-day actions of the  QE2-related  FED actions on the following WEB link, which details the  schedule of the operations:

http://www.newyorkfed.org/markets/tot_operation_schedule.html

Similar tendecies  could be noticed  after analyzing the Daily Bearish predictions.

During the previous  20 – 25 years we did not  have  such overwhelming  market manipulation, that will impact  all kind of  market analysis results.  Obviously  the  many different market cycles  will also be  impacted. When Bearish forces not allowed to  take their turn for long period of time,  their  impact  could be much more  violent  at a later time, when the  FED and the government run out of ammunition and   get forced to leave the  battlefiled. That could happen  when the dollar loses the global reserve currency status, the latest, if that is about to happen.