Daily Speculations The Web Site of Victor Niederhoffer and Laurel Kenner

Home
 

Dept. of
Ever-Changing Cycles

Write to us at: (address is not clickable)

 

11/17/2005
ASK the CHAIR!

A Mr. S. asks why we predict a convergence of Sharpe ratio to minus infinity on delta neutral funds. Here is one answer from a fellow traveler:

I am surely no Mr. S., but it strikes me there are two things in life: where you are going and how you are getting there. Discussion of so-called "market neutral" strategies has often suffered from a confusion between means and ends:

  1. There is the" low correlation" end with respect to some targeted benchmark (' we make money regardless of the S&P's movement, etc'); or
  2. There is market neutral means (we make what the S&P or some other index does, but we do by means of a hedge or spread or offset technique), targeting a Sharpe or information ratio, etc.

The first type tends to be a sleight of hand or bald-faced deception based on a three card Monte set-up; the second is a pure illusion which by its very structure is generally doomed. In case one, for example, a famous refugee from LTCM, currently sends out a monthly letters for his new partnership in which the correlations with all the listed classes of alternative and standard benchmarks are shown to be next to zero, complete with a "risk budget" expressed as a volatility function of the notional value of the ten year treasury. He is known to be a fixed income convergence specialist as well as an options guru. But nowhere in his letter is a correlation depicting performance with ANY spread whatsoever - corporates to treasuries, TED, Mortgage backs to treasuries, High Yields, etc - even noted. Reading his letter one could be led to believe that his fund expresses the unitary, uncorrelated particle in the universe. And yet, it was only 3 months after AAA mortgages went from 40 bp over the curve to 250 over (in the summer of '98), that LTCM blew up. There is no such thing as "uncorrelated." There is always a correlation, and if you do not know what it is, then when you find out you will be dead.

But it is the second sense of "market neutral" about which the great Mr. S. inquires. And in this sense the question is even more curious.

Twice the transaction costs and dealer spreads for each position would be tough enough to beat, but the idea that one can correctly and continually forecast the simultaneous pathways of two sides in the same position - over many positions through time - and still give up the asymmetrical return possibility defies imagination. Being long a spread is not to decrease volatility: it is to transform it and express it in a different direction.

For example, when marking to market convertible arb positions, the bond prices are, like loans from your bank, good when you don't need them. As soon as you do, the relative illiquidity of one side of the "hedge" reveals itself with a vengeance, and the famous widening out of the conversion premium - say upon announcement of a credit problem - promptly disappears; the rate of decline of the bond price suddenly accelerates rather than slows. In this respect, while the convert specialist knows he is short a call through the farther out conversion strike of a bond and attendant call protection, he is also short a put to those senior in the capital structure to him. This hidden cost is quietly amortized by the yield spread on the bond/stock 'hedge'. But it ain't free money. There is some insurance being written. There is simply no free lunch, and the appearance of such a freebie at month end markings of over the counter securities does not prove the fact. Even in program trading, the "cleanest" spread trade, there is always a legged-in entry - a first side taken with the other laid off. This risk is paid for over time in little pieces and occasionally rebalances viciously.

The most remarkable thing about his question is that the vehicle through which one expresses a long bet in equities is always moving: the index itself has the huge survivor bias; the companies themselves grow or die; prodigious creative, competitive forces drive new monopoly rents to the marginal cost of production - thus infusing the economy with "saved" cash flows and high standards of living as long as there are no restrictions on exchange, etc - and are relentless, geometric, and evolutionary in scale and scope. The statistics on the table bear this out in terms of drift for all time periods, subject to volatility, etc. Taking the other side of that trade in the long term seems, on its face, to be akin to defying gravity.

The world view of one who would seek to eliminate risk is one of a flat-earther; there is no elimination. There is risk identification, estimation, and pricing , and variation in all. The rest is just big vig and the leaf of a fig.

11/16/04
Russell Sears adds:

Models fail to adequately adjust for their own effect on the market. One aspect of this has been said before, that good hunting grounds will bring in crowds of hunters. I would add that besides the models failing to adjust for diminishing returns, tougher competition, they fail to realize how their growing presence brings its own inefficiencies to the market. Where their assumptions fail, be it infinite liquidity, especially on the short side, or correlations being cause and effect, they bring in opportunities for other hunters. They forget, that they to are being hunted. Just because they are sheltered and their shelters are well camouflaged does not mean they are invisible, especially once their scent is all over the forest.

Further, while the long side has many advantages the short side has one. It takes years to develop a talent, but a moment for it to vanish. It always is wise to look for that player whose sport has lost the twinkle in his eye, and he is about to hang it up and cut him from your team. It's usually wise to replace those old sprinters with fresher hungry leg before his hamstrings give out. It is always the coach's job to destroy so he can rebuild. But the first and most important job is to develop and nurture the talent.

11/16/04
George Zachar offers:

As a fixed income futures/options guy, nearly everything I do can be modeled as either an insurance providing function, or a liquidity providing function. I find it downright funny that I spent much of the past few years at my desk, solo, in my jeans, providing what amounted to reinsurance for those who wrote portfolio coverage to Fannie and Freddie.

Nonetheless, the knowledge of my role in the great market ecosystem allowed me to both fix my risk parameters and sleep at night. My business concern going forward is that Fannie and Freddie [and by extension the mortgage market] no longer "need my services". The big GSEs are contracting their books, and mortgage borrowers are increasingly taking out loans with less optionality for their creditors. Much of the collapse in debt vol. is neatly explained by this. That cycle has changed, and it's time to move on.