2 Comments

  1. fed rate cut | Best data info December 30, 2007 @ 7:31 am

    […] An Exercise In Market Timing Using Dow Jones […]

  2. Raju March 29, 2010 @ 7:08 pm

    Awesome post. I feel for your your hardwork that you have done. But you did it handsomely & poosted it very nicely.

Timing The October High And November Bottom Of Dow Jones

Finance, Thoughts Comments (2)

Over the past two months (actually beginning October 1st) I conducted a documented experiment on this site. Although Reuters and other media republished some of the articles, my impression is that this exercise in market timing went completely unnoticed.

Since I do not, neither do I want to, provide financial advice, I will not publish any more articles related to market timing. What I will do here is dispel the myth that the markets are unpredictable.

First a clarification. I do not think that randomness has a place in the universe. I firmly believe that randomness is just a general name for the limit of our computational power. Whatever lies beyond our ability to compute complex results arising from simple rules we call random. In the quest for knowledge, whenever I read a reference to randomness, I read it as as confession of the limits of the knowledge of the writer.

This subject is not an easy one to tackle. We spent our Thanksgiving dinner with family and friends arguing about the limits of knowledge! At some point, somebody (I think it was my 15 year old son) created a mathematical response to Popper, to the effect that if you know 99.99%, you can, for any practical purposes, consider that you have complete knowledge of the thing (which I think explains mathematically why teenagers are always so certain about everything). My counter-argument was that it was 99.99% repeating, to which somebody else added that that infinitely moving 1 was the unknowable, where randomness occurs. With that, we moved to the certainty of the turkey, which was juicy without argument.

That said, I think I should write a longer article about the stupidity of black swans, but I will need time to write it so my head does not get chopped off.

And now for the results of the exercise

This chart shows the weekly Dow Jones and the dates of the relevant posts marked A, B, C, D and E.

image

By the way, for this article I use charts from www.stockcharts.com, the site of John J. Murphy, whose book Technical Analysis of the Financial Markets is a must read for anybody who tries to understand the markets.

Overall, I was able to predict

  1. An increase in volatility
  2. The beginning of the down move within 1%
  3. Intermediate turning points
  4. The end of the down move within 3%
  5. The timing of the Fed and Government interventions

The statistical and mathematical methods I use sometimes find two possible outcomes with the equal weight. In those cases, I will add charts and discuss a little further about the possible outcomes.

In the first article in October 1st I provided a statistical path for the following months in the Dow Jones

The next month should be interesting to watch, as October is an infamous month in the stock market and most market analyses seem to focus on some “mysterious” forces nobody can understand. The scenarios I see with more probability are:

  1. Double top in October, nasty correction, Fed bailout, break out in November
  2. Double top in October, nasty correction, more bad news on Black Wednesday, Fed lets the correction run its course and bails out in January
  3. Realization by the market that there may be a recession going on, bear market that can’t be stopped by the Fed
  4. Break out in October happy sailing

As I mentioned before, I think the Fed will do everything in its power to provide liquidity to the markets.

Expect Increased Volatility

So starting on that day, it was possible to predict both market direction, and an overall timing for the move. The direction was perfect, the timing is off due to a personal bias towards shorter moves. The “nasty correction” did not come in October but in November. October did have a correction, but, compared to the one in November, it is hard to call it nasty. In any event, in October 1st it was possible to predict

  1. What the collective mind of the market would do in October (double top)
  2. What the Fed would do to solve it (rate cut)
  3. The outcome of a breakout in October as the least probable

The next big call comes in October 12th with the article Bad News For the Dow (Short Term Bear Fest).

In that article I said

Yesterday was a key reversal day, a technical signal traders pay a lot of attention to. A Key Reversal Day is when the market makes a new high, reverses and closes below the previous low. If today’s close is down, expect a down Monday as well.

The Wall Street sponsored media should start next week pandering bad news and creating a climate of fear using the available examples of very bad October markets. Expect myriads of articles about 1929 and 1987. This climate should keep going to pressure the Fed to cut rates during the next meeting. The Fed should oblige preparing the stage for runaway inflation.

Some details were off, like the down close for the day. But within 1% it was possible to predict a short term down move that went from a high of 14118 to a low of 13407. In any event, the calls were not for intraday trades but rather weekly positions.

Calling for a down move

The reasons for the change of direction were correct as well, the Fed intervened to stop the fall in October 22nd and 23rd, and finally cut rates in October 31st.

The following call was in October 22nd, with the article Dow Jones 400 to 600 more points of pain. The Dow not only did not move down 400 to 600 points then, but it moved up. I published the article at 10 AM, and 12 of that day, Fed Officials came out to sustain the fall announcing that there was going to be a rate cut in the next Fed meeting.

Short term you could say that the call was wrong. And this is were the proponents of the randomness of the markets will claim victory. My argument is that I was not trying to predict short term moves, and I was looking into a longer term picture (in terms of weeks), so, yes, I did not predict the intra week movements, but that was because I was not looking into them. If I had wanted to predict intra week movements I could have done so. As a matter of fact, my dear friend David Elliott does exactly that, and for free.

So, if you were in a weekly short position (a trade position that bets the Dow is going to go down and does not care for intra-week volatility), you could have stayed on the position, because this is what happened immediately after the Fed announcement.

image

Pretty neat right? From the low of the day of the call to the bottom there was exactly a drop of 683 points. If you are not convinced that randomness of the markets is just an admission of ignorance by the analysts, keep on reading.

More importantly, midway through the fall, on November 9th I wrote Dow Jones at 13,099 should see a rest. The bottom that week was 12975, around 1% lower than my call. If you sold that day because of what you read, for instance, in the Boston Herald (Latest Dow dive strikes fear on US economy), Bloomberg (US Stock-Index Futures Drop; Qualcomm Falls on Forecast Cut) or CNNMoney (Stocks battered by mortgage mess), you would had sold 200 points from the bottom, after having endured 400 points of pain.

The climate, indeed, matched my description of the situation.

The Wall Street sponsored media should start next week pandering bad news and creating a climate of fear using the available examples of very bad October markets. Expect myriads of articles about 1929 and 1987. This climate should keep going to pressure the Fed to cut rates during the next meeting.

Starting on November 15, I started warning about a “surprise move” by the Fed, or some other way to inject liquidity into the markets. I did not expect much to happen during Thanksgiving week, and nothing much happened.

Finally, on November 23rd I said that I thought the Dow Jones was Ready for Next Leg Down. Of course, there was the need again to qualify the statements, because we were indeed so close to a bottom that the statistical probabilities pointed to two equally weighted outcomes. For the Dow to go up you will need an external force of some kind so I said:

Barring a surprise move from the Federal Reserve next week, the next meaningful movement for the Dow should be below 12,5000 (about 4% down from here). The Federal Reserve will intervene to protect America’s economy in times of war.

On November 28, we had the first “surprise move” with some Fed official suggesting a rate cut during the next meeting, that day I said that it was not all clear yet, but had to qualify it with the following:

Hopefully, the Short Term Bear Fest that I announced in October 12th was starting, is coming to an end.

Finally, in November 30, I said that the Dow Jones had produced a long term buy signal, after which normally there is a 2 to 8 day pullback.

image

We had exactly 2 days of pullback that I am sure were full of news to justify the move. The reality is that after the 4-day rally the “pioneers” of the move stopped pouring money into the markets, and waited to see if the move was “for real”. When the sellers did not show interest in further selling, the buyers entered the markets again.

Furthermore, on December 6 we had the “other way to inject liquidity into the market” with the Administration announcement of measures to hold interest rates frozen for variable rate mortgages. I think that this cemented the conditions for an inflationary run that will leave our heads spinning.

Whether my longer term prediction of the Dow doubling in price within two years comes to pass (because as the amount of unknowledgeable variables increases, the possibility of predicting outcomes diminishes) or not, this exercise should suffice to prove that predicting market moves is not only possible, but also relatively easy. What’s more, with an adequate model of social interaction and social psychology, it is even possible to predict to some extent the reaction of the media, the public and government officials.

The failures in the predictions should be taken as failures of the model, and the model should be adjusted consequently. But I think that eventually it will be possible to develop a theoretical framework that allows for the mathematical prediction of future events that currently are considered impossible to predict and belonging to the magical real of randomness. I still think that absolute knowledge about the future is hard to come by due to our lack of computational power, this brings about very interesting issues related to Moore’s law. But that, together with the stupidity of the black swans belongs to another post.

Franklin @ December 9, 2007

Leave a comment

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>

Entries (RSS)
Comments (RSS)

82 queries.
0.866 seconds.

Politics blogs Politics blogs
Links to Site
%d bloggers like this: