Quote from Martinghoul:
Lower volatility is better because the assumptions required for such a conclusion are much "less strong" than the assumptions needed to conclude that there's some sort of evolution that eventually leads to a stronger outcome. This is, in fact, confirmed by basic statistics, and not just in the world of finance. Even in zoology (Earth is a vastly superior ecosystem than the market and the economy), survival is rare (overwhelming majority of all species that ever lived are extinct). This surely implies that the "survival through evolution" outcome isn't very likely at all, everything else being equal.
For firms, intuition would suggest that higher volatility causes the probability of survival to fall. There are some academic papers that provide empirical evidence that this is, in fact, the case. For example, see here: http://www.anderson.ucla.edu/documents/areas/fac/finance/1987-1.pdf
Fair enough, and of course there is going to be a body of evidence as to why academia has chosen the lower volatility route.
At the margins, though, I would argue things look more intriguing, and that conclusions generalized across a data set may not serve as well on a case-by-case basis or special case basis.
Obviously an investment strategy where one chose the more volatile investment ceteris paribus would not necessarily work well.
But in terms of looking at individual business models, or making highly individualized choices - such as, say, a career path, or where intrinsic volatility or placidity can be sourced to specific characteristics of the investment - the generalized logic can be turned on its head.
Take, for example, the career path of becoming a trader vs that of becoming a dentist. One is going to be far more volatile than the other, with a far greater risk of failure. One could argue that, for the general population, dentristry is obviously a saner and safer choice.
But on the individual level? Not necessarily, and not so much. Some of us would die of boredom as dentists, and are equipped with skills and temperaments that greatly improve our chances of success with the more volatile endeavor. And then too, when it comes to volatile vs smooth careers, one has to take self-sufficiency and portability of skills into account. Maybe the trader, through hard training and experience, has learned to support himself in all types of market environments, thus becoming more self reliant, whereas the dentist is exposed to potential competition glut and finds his practice failing, forcing him to deal with volatility he never trained for.
The same applying to individual investments, again on a case by case basis. General Electric in the Welch era was the king of smoothed earnings and artificial placidity - until it all went over a cliff. Cioffi and Tannin of the Bear Stearns credit funds went 40+ months in a row with steady 1% type returns before blowing up.
In contrast to that, you can find individual investments where earnings drivers appear to be quite volatile, but the business itself is quite robust - a survivor.
And, too, there is the question of generating outsized returns. Venture capitalists manage deliberately embraced volatility through diversity of bets. Or as Soros has said, the type of returns Quantum pursued in its glory days would not have been possible without significant volatility at times. Which then circles around to another question: Is the less volatile manager truly "safer?" Or is more volatility coupled with higher returns and more effective risk management in terms of cutting off mortality tail risk better?
All in all point being, there is certainly an academic case as to why "low volatility is better," in a very broad sense of generalizing across the data set. This is a given.
But the question gets interesting at the margins imho, for those individuals who recognize they are not like all the others, do not have to think like the masses or make decisions like the masses, etcetera.
(This also relates to a pet peeve of mine with academic studies. In so many instances, studying a generalized population of hundreds or thousands etc. is completely inappropriate, when what is attempting to be measured has a discretion-based selectivity component. The worst are studies that attempt to "disprove" technical analysis, by, say, looking at 5,000 head and shoulders top formations as defined by the study creator. Such is the height of uselessness.)