Skip to content

Some Internal Market Measures, by Victor Niederhoffer

2006-12-09 · Victor Niederhoffer
Original source

There are many approaches to measuring the health of the stock market:

  • relation to other markets such as interest rates, foreign exchange or energy
  • the diffusion of extremes
  • the distribution of the ensemble, i.e., the cross section of price changes
  • the health of the economy as measured by one of 500 indicators like employment or consumer sentiment
  • sales of this or that key retailer
  • movements of earnings estimates
  • the steadiness and velocity of recent moves
  • the sentiment of advisers
  • flows into and out of funds
  • distributions of returns in the past historical record
  • performance of new issues
  • performance of bellwethers or key sectors
  • issuance of debt versus security
  • relation to economic indicators
  • the yield curve
  • opportunity costs
  • yields from dividends
  • seasonality
  • cyclicality
  • volatility
  • strength of the close
  • returns on various balance sheet items, including total capital
  • technical factors, including moving averages, oscillators, relative strength, candlesticks, stochastics, parabolics, Hursts, Bollingers, Chaikins,Williamses, McClellans, point and figures, channels, envelope , pivots –and, yes, Fibonaccis)
  • flow of acquisition activity
  • tax policy
  • key world stock markets
  • options activity, especially put-and-call ratios
  • the activities of those who are survivors — such as insiders, big trading firms and specialists — and the activities of those who tend to lose more than they have to, including as the friend who’s always wrong, readers of the venerable columnist at the leading financial weekly, followers of the prudents, followers of fixed systems, and the recommendations of those who tend to be right or wrong

Readers can feel free to suggest a few others from the endless list.

I have long thought an alternate approach might be apt. The basic idea is to see how individual stocks respond to day-to-day things that are part of the fray.

Start with earnings reports. How do companies perform on the day they report earnings, taking into account how the earnings compare to expectations, and how the expectations are changing? Then expand to specialized things that are always happening. When big company like Pfizer loses a product and the stock drops 15% at the opening, as it did on Dec. 4, how quickly is the drop retraced over the next one to four days? What is the average premium on acquisitions, and how frequent are they? What is the daily volume of reported insider buys and sells, and how did the stocks perform in the days before the purchases and disclosures? What is the relation of value to growth on a particular day, and of big versus small (this is the easiest question because a glance at a few indexes provides the answer.) How many stocks close at various high or low parts of their ranges, relative to the stock market as a whole? How did the Value Line Group 1 champs do on the days preceding their earnings reports? How are the big moves in big option stocks distributed?

It would be laborious to construct such a series, and it would require skill to use it for estimates. The beauty is that because of the cost and extent of the input required, one might come up with something that actually has not already been discounted by the hundreds of thousands of excellent minds always looking for an edge.