Jul

20

1. It is remarkable to note that for the S&P futures, there has been a complete symmetry in the number of big up and big down opens over the last 6 years. Using 1/2 % as the cut off, one finds that 340 have been big downs, and 345 have been big ups.          

Breaking it down by day of week, one finds that there is no day where the number of big ups diverges by more than 5 from the the number of big downs. Even more remarkable is that no day diverges from any other day by more than 15 in the total number of big ups or big downs. For example, on  Monday, there are 64 big downs, and 65 big ups. But of Friday, 70 big ups and 74 big downs.          

It used to be article of faith from efficient market types that news was generated randomly in time, like the number of horse kicks (v. Bortkiewicz 1898 ), i.e. a poisson process and that Mondays should have more bigs than the other days.

                 Big ups followed by big downs                                                

Mon 65 big ups, 64 big downs                                                     

Tue 69          73                                                              

Wed 68          61                                                            

Thur  73         68                                                            

Fri 70          74                                                                      

Even someone as prone to finding regularities in randomness as the signer, can't find any departures from randomness here. 

2. The concept of a bear market, i.e. a decline of 10% or 20% as an indicator of further bearishness is a hallmark of the charlatan. I think one recently saw millions of headlines to that effect of a bear market in commodities  (right before the 50% rise) but there has been a remarkable decline in Nikkei relative to the S&P, similar to other remarkable divergences notes in these humble speculums. [Note: Nikkei now 8670 vs 10185 on March 27, i.e. down about 15% vs S&P down only 3%].

3. There are many ephemeral things that move the market. One tends to shrug the shoulders and haul out some Shakespeare or Aesop on these occasions. "What fools these mortals be"  or "the ant and the grasshopper who played music for the day". But whether they are ephemeral or not in the market, they can have a lasting effect, and there is signaling to consider also.                                

To me, the  most ephemeral things in the market are the reaction to alcoa earnings, and the Philadelphia fed, and the various manufacturing surveys. All of these are random numbers based on a small sample representing 1/1000 th of the economy for one month even if they were not subject to so many sampling and seasonal errors.                                

Of course, 99% of the announcements are like this. However, they can have lasting effects as the meaningless monthly employment numbers show even though ofter an increase or decrease in the raw employment number of 900,000 or more can be converted to a 50,000 gain or loss through random and self interested adjustments.    

One is reminded of times I've tried to make a statement at a lecture before asking a question and in the middle a hundred agrarians from the audience shout out at me in disgust "what's your question".                      

In any case, what are the most ephemeral announcements and worthless things that move the market in your ken? And which ephemeral things tend to be reversed the most?                        

Okay. Are then any ephemera that have lasting effects?

Anatoly Veltman writes:

I can think of one that's puzzling: a Nikkei jump over night can create a follow the leader impression, which will trigger a wave of optimistic equity rises around the world. Is it ever considered that an oftentime .5% over night gain in the Nikkei is, in fact, unchanged in the neutral currency terms (by virtue of simply the Yen devalued .5% vis-a-vis the currency denominator of any other equity market)?


Comments

Name

Email

Website

Speak your mind

Archives

Resources & Links

Search