by Peter J. de Marigny
Introducing “The Red Heifer” in Capital Markets
A “Red Heifer” is a rare event condition, however, it is an event that differs from a “Black Swan” in that it is not a return observation within a sample of chained return measurements. A “Red Heifer” event is one that creates its own discrete data series. A “Red Heifer” is an event that is more than an outlier on a control chart, it is an event condition that creates a new control chart. For municipal bonds or real estate, it could be the change of a tax regime.
Banks/financials, pharmaceuticals, energy, telecom and technology face regulatory and other potential red heifer events. Technology could offer a transformative change creating a red heifer in multiple sectors. When a red heifer event condition does occur, it may not be reflective immediately in a series of return observations, but the red heifer event condition criteria become the model dialectic. In fact, the red heifer is the trigger event condition that determines the path of a new discrete future return series.
Conversely, in the article, “Death of the Black Swan” (http://bit.ly/DITMoj2t) it was proffered that Black Swans exist only in chaining return observations that may be unrelated and that pricing is unrelated to the true risk of an asset. Therefore, the presence of a Black Swan event is merely an apparition of unrelated data points and unperceived risk.
Back in the days of Western Electric, the telephone equipment monolithic monopoly, a man who far preceded Harry Markowitz named Edwards Deming, created statistical process control. Applying SPC to investment portfolios is the essence of Markowitz’ 1952 seminal work creating modern portfolio theory as we know it today. The application of SPC engineering methodology to financial portfolios begs the question of whether observations of the pricing of financial assets is equivalent to observations of product measurements.
This is tantamount to comparing pari-mutuel determined odds determined by bettors to odds of pure chance like dice. When data observations are determined by means other than pure chance the chained data series is specious. If the data series is corrupted by factors other than pure chance, such as discrete risk factors, the resulting statistical inferences setting probabilities are also specious. If an event with p<.3% occurs it would be defined as a Black Swan, but that assumes the observations are from the same chained series.
Black Swans assume a known distribution using the number of (perhaps unrelated) observations as degrees of freedom, that in turn establishes correlation significance at a lower measurement. The efficient frontier (and CML) is then drawn with an assumption of consistent correlation. Correlation itself is used by many hedge funds (along with the Sharpe Ratio) as a risk measurement which it is not. Correlation does not reflect scale and it is a linear measure, it changes often due to risk factors and is not an appropriate metric for portfolios susceptible to higher moment risks like many hedge funds. The Sharpe Ratio is easily “gamed” using derivatives and leverage, but it is reflected in the slope of the CML tangential to the market portfolio on the efficient frontier.
Red heifer event conditions create a new discrete data series unobservable in past return data series. Black swans are defined precisely by past return data. A red heifer may be observed by statistical inference only after return observations are chained because of its presence. The presence of a red heifer may remove a black swan due to the removal of otherwise chained data observations. In that case, the black swan may disappear as a normal observation within a new data series … created by a red heifer. *.*
Peter J. de Marigny, www.DITMo.net