Daily Speculations The Web Site of Victor Niederhoffer and Laurel Kenner

Home
 

The Chairman
Victor Niederhoffer

 

 

About Victor Niederhoffer

2006 All content on site protected by copyright

 
Write to us at:(address is not clickable)

27-Apr-2006
A Good Market, by Victor Niederhoffer

The question arises: What makes a good market?

As a preliminary step in this direction, I counted the number of times that various representative non-stock markets are up in conjunction with an S&P 500 rise of 40 full points, in the last 20 days, a mark which it set some 17% of all days in the last 6 1/2 years.

All independent variables are measured over the same contemporaneous 20-day period that the stock market was up 40 points, i.e., the query is what were the concomitants of a great good market. (Thus in part we are making the same mistake that Sornette and so many others make by concentrating on the likelihoods of an extreme up, not taking into consideration that the same variables that enhance the likelihood of an extreme up might enhance the likelihood of an extreme down.)

The answers were as follows:

Percentage of times that Markets Moved in Conjunction With The S&P 500 up.

   Market               % of co-movements
   bonds up                  53%
   bunds up                  58%
   Gold down                 58%
   Eurodllr down             62%
   Yen down                  60%
   Crude oil up              59%
   Natural gas down          55%

Now assuming perfect knowledge that a move of more than 40 full points has occurred in the S&P 500 over the past 20 days, what is the best quantitative forecast given perfect knowledge of the independent variables.

Here's the best perfect knowledge descriptive equation for the S&P 500: S&P 500 = 45 points + 3 times the move in bunds less 1 times the move in us bonds less 230 times the move in the euro less 1.5 the move in yen per dollar( a fall in yen per dollar means the yen is up and the dollar down). The key variable that describes and accompanies a big move in the S&P 500 is a decline in the dollar and a rise in bunds.

That equation has a r2 of 0.10 which corresponds to a correlation between prediction and actual of 0.3, quite helpful assuming you had perfect knowledge of all the independents and you already knew that the S &P 500 was up over 40 points over the last 20 days. None of the other independents such as gold or crude are significant.

Now, consider an unconditional description of the S&P 500 move over the next 20 days without any knowledge that it went up or down. That descriptive equation is S&P 500 = 3 points + 2 times the move in bunds less 2 times the move in bonds less 264 times the move in euro less 1.5 times the move in yen per dollar less 1 times the move in crude .

That equation has an r2 of just 0.04, thus the squared error in forecasting the S&P 500 over the next 20 days is reduced by just 4% even if you had perfect knowledge of all the other key independent variables during that same 20- day contemporaneous period. (Of course this excludes markets that are part of the US stock market like the Dax or the Nikkei.)

I have seen various other approaches to the philosophy of what makes for a good market including the amount of good news, the state of the p/e, the number of terrorists detained, the amount of positive news about the economy, the number of years into the Presidential cycle.

I prefer the approach based on market co-movements as it's less wishy-washy, but then again I know almost nothing about philosophic questions like this and would appreciate guidance from my friends and readers.

Abe Dunkelheit replies:

I would assume that people as a group tend to be more happy if they have steady and predictable returns over a long time period. Therefore, I'd define a Good Market as 'low' volatility around a 'normal' and 'long-lasting' drift rate.

In analogy to the life of a human being, 'long-lasting' means a long life; 'low' volatility means no exposure to stress or physical disease; a 'normal' drift rate means the right mix of emotional and mental stimulation (not too much nor too little) to have a fulfilled life.

Sushil Kedia comments:

A capital market has as its core function the allocation of capital. So, a good capital market would be one that refrains from repeatedly or regularly getting either into states of manic euphoric bubble conditions or even the severe depressive conditions where it saves the economy and its people from not allocating capital to fly by night fanciful operations or does not hesitate ever in facilitating the formation and dissemination of new capital.Through expansion of investment wisdom, sharing of knowledge and embracing openness to change societies, possibly, are endeavoring in sustaining the primary goals of their capital markets.

A good market is one which would reward the enterprising astute risk taker and one which does not bring about passive investing as the best option. Such a market would also continue to place premium on credibility, prevalence of trust and, ideally, be self-regulated well enough to keep external regulators, the public and the media in harmony with itself.

A long-lasting attribute akin to the human life is for a good market to show its resilience to overcome challenges and difficulties that would be part and parcel of its existence. A good market would foster institutions of disseminating education to its participants, instill proper risk controls for executing fairness of dealing and sustain liquidity to ensure longevity and survive in the face of adversity to come out as a stronger institution. It would be a vehicle for helping achieve and sustain the cherished societal goals of fostering freedom of choice, fair exchange and productive economic contribution.

 

 

 

More writings by Victor Niederhoffer