As the world greeted the new year - 2009, I listened to New York State Insurance Department Superintendent - Eric Dinallo's Bloomberg interview with the ever-zippy Tom Keene. In the interview, Superintendent Dinallo said something that made me freeze and ruminate for a while: "You can't ask VaR (Value-at-Risk) first-order questions like 'Are we in a credit bubble?'. That is not a question you can ask VaR formulae."
What specifically caught my attention was the part where he said "Are we in a credit bubble?" I kept asking myself 'How can we establish whether we are in a bubble, or not?'. At that time, my knowledge on asset bubbles was chiefly influenced by the writings of Professor Robert J. Shiller, author of Irrational Exuberance. Hence, I naturally arrived at the conclusion that predicting, or knowing whether a speculative bubble is being inflated in your midst is only easy with hindsight - an assertion Shiller makes on countless occasions.
However, new insights I've gained since then, are challenging that conclusion.
A speculative bubble morphs over time like human beings, from an 'embryonic stage'; to infancy; to a juvenile stage; to adolescence; to maturity - 'the adulthood stage'; to the stage of decline - the equivalent of 'aging' in human beings; and, eventually death - when the bubble explodes, or bursts. Ascertainment of the existence of an asset price bubble, is difficult when a bubble is still at an embryonic stage, but it becomes easier with each progressive stage of maturity the bubble enters. Hence, I believe that I have found a way to establish if an asset price bubble that has been inflated is about to burst - i.e. if the bubble is making a transition from its 'aging phase' to its 'death phase', specifically in stock markets, and that can be done by examining the behavior of growth stocks and value stocks over time.
In their paper series on The Anatomy of Value and Growth Stock Returns, Eugene Fama and Kenneth R. French did a study of value stocks and growth stocks, between 1927-2006 - a period of 79 years, and found that value stocks outperform growth stocks over two-thirds of the time (which translates to ~ > 52.667 years, or 66.667% of the time) during that period, which in turn implies that growth stocks outperform value stocks one third of the time (which translates to ~ less than 26.333 years, or 33.333% of the time). Clifford Asness, the Principal of AQR, sums this observation as: "Cheap beats expensive more than it should.''
My belief is that stock price bubbles can be understood better, when people carefully scrutinize the periods when highly volatile growth stocks outperform highly volatile value stocks, because I believe that stock price bubbles lie within those time periods.
Hence, I suspect that when a portfolio of highly volatile growth stocks consistently outperforms a portfolio of highly volatile value stocks; a stock price bubble may be in your midst. To ascertain if the bubble is about to burst one would have to ascertain the differences between the rates of change of value both portfolios - with respect to time, and when that difference is at its greatest, it is a sign that the stock price bubble is ab0ut to burst.
Mathematically the last section preceding paragraph can be expressed as:
The bubble is about to burst when:
Where:Max means the maximum
dAVg/dt is the rate of change of the value of a portfolio of growth stocks with respect to time
dAVv/dt is the rate of change of the value of a portfolio of value stocks with respect to time
Let's test that hypothesis to establish how right/wrong I am.
