# A Gravitational Attraction? from Victor Niederhoffer

October 21, 2010 |

One wonders if by considering the distances and weight of one market from another one would create a gravitational attraction possibly related to square of distance. Would this be even better way to explain recent market moves than twitter? So many markets are up that they pull stocks with it. Every day the crude and the gold and the grains and the metals exert their gravitational attraction on stocks and it's hard for stocks to go down when gravity of everything else is pulling them up?

## Ken Drees comments:

Attraction theory may also pull monies from undervalued sectors-like nat gas for example– keeping these sectors starving for investment.

## Anatoly Veltman writes:

My take is the former recent relationship has been more a product of U.S. dollar's daily devaluation. Thus the commodity part of it was only a further derivative.

## Phil McDonnell writes:

Imagine we are on an island with only two things to trade stocks and gold. Naturally we use sea shells for money. At any given time there is only so much money M. So the total price of stock and gold is proportional to that. In fact we can visualize the possible prices as a circle with radius M and the X and Y axis are the prices of gold and stock respectively. The locus of possible points they can lie on is given by:

M^2 = G^2 + s^2

where I have changed the x and y to g for gold and s for stock.

Since M^2 is a constant at any given time we can just call it c and then we have.

s^2 = c - g^2

showing the relationship. This is all very pretty theory but does it stand up empirically?

The coincident correlation base on daily percent changes between gld and spy for the last 105 days was about 1%, so not much linear going on. but when we look at the relationship between spy^2 and gld^2 we get a 42% correlation consistent with the formula above. When we rewrite the formula for m and not m^2 we get:

m = ( g^2 + s^2 ) ^ .5

which is just the distance formula from high school.

Thought question: What happens when the Fed adds Q to M during QE 2?

## Sushil Kedia writes:

The House Money effect works the same way. There is more valuable collateral, there is a larger amount of mental wealth, there is a larger appetite for risk. Akin to the rabbit coming out of an empty hat, money grows in the minds of the market players, when things are moving up.

As one large down move comes in a widely betted asset it gravitationally sucks away the value of the collateral utilized for playing other assets. Like the invisible forces of gravity the various contracts naturally move by in varying proportions broadly in similar directions, mostly together.

I would be inclined to recognize the effect of the varying amount of bets inside different pits and the varying spread of those bets across hands of differing strengths. With that in place any static relationships in assets or contracts is less than likely to be existent for any periods of prediction worthy time horizons. The ever changing cycles are likely originating from this varying nature of the spread of the bets. The vector sum total of all current and past and future bets may indeed by hypothesized as zero. Yet the similar sum at the present moment is not zero. Every changing tick hurts or rewards different sets of people simultaneously.

So, without so much as trying to invoke my limited numeracy skills before the mighty minds, I lay a case, that the pursuit of discovering constant relationships in the markets is the innate desire of men to find a constant while knowing fully well that the meal for a lifetime indeed is the knowledge of ever changing cycles.

## Ralph Vince comments:

I lay a case, that the pursuit of discovering constant relationships in the markets is the innate desire of men to find a constant while knowing fully well that the meal for a lifetime indeed is the knowledge of ever changing cycles.

What could be more true than that statement?

We build models of the market– some, with ever-increasing complexity.

Take the stochastic differential equation for price changes in continuous time, where the second term is the Weiner process:

S0 = u S1 dt + dX

Involved math for many of us– but, as a model for how prices change, …it too is pathetically lacking. Our models are not reality, just little peepholes on it's behavior at times.

## Sushil Kedia replies:

To add, one early school beginner's physics question:

If gravity works the same way on a feather as well as on a stone, then why does the stone drop sooner to the ground?

Well, the air that provides so much of rest to the feather that it takes longer to come down.

Likewise, the "air" inside the markets that is the varying size of bets of any individual participant as well as the varying size of the total bets present in a market bring by the gravitational pulls to still carry wide and varying variances.

`SELECT * FROM wp_comments WHERE comment_post_ID = '5395' AND comment_approved = '1' ORDER BY comment_date`