Under the assumption that returns are normally distributed, a one-day move of greater than 7 percent in the Dow Jones Industrial Average should occur once every 300,000 years. Benoit Mandelbrot and Richard L. Hudson, however, report that 48 such moves occurred during the 20th century. This empirical evidence refutes the notion that stock market prices follow a random walk, as many proponents of the efficient market hypothesis (EMH) believe.
In Dynamic Asset Allocation: Modern Portfolio Theory Updated for the Smart Investor, financial journalist James Picerno contends that such thinking is out of date. He shows how research in the past few decades has refined the concepts introduced in seminal papers on modern portfolio theory (MPT) by Harry Markowitz, James Tobin, William Sharpe, Paul Samuelson, and Eugene Fama. The random walk hypothesis and the EMH are not equivalent, Picerno persuasively argues, nor does disproving the former undermine the usefulness of the latter.
Indeed, as early as 1965, Fama acknowledged the unlikelihood that the random walk hypothesis precisely describes the behavior of stock market prices. For his part, Mandelbrot concedes that the proponents of “postmodern” finance have not yet devised a better model of prices that is both accurate and practical for individual investors. Absent a reliable means of exploiting discrepancies between the random walk and actual price behavior, MPT’s prescription of holding the market portfolio (i.e., the entire investable universe) emerges as a not unreasonable solution.
The catch, Picerno points out, is that the market portfolio is optimal only for the average investor. Few, if any, individuals fit that hypothetical description. For example, it makes sense for investors who derive most of their livelihood from employment to avoid some of the recession risk embedded in the market portfolio. Once they have determined their personal optimal asset allocations, investors may choose to rebalance their holdings from time to time in order to maintain constant percentages of stocks, bonds, real estate, and so forth.
Picerno explains, however, that in contrast to early interpretations of MPT, subsequent research has shown that optimal asset allocation is not static. For one thing, the risk–reward trade-off varies over the course of the business cycle in a partially predictable and thus, in principle, exploitable way. Furthermore, he contends that investors should adjust their asset allocations over the course of their lives.
Picerno does not merely address these theoretical issues; he also offers guidance on customizing a portfolio for an investor’s individual needs and changing circumstances. Readers, however, will find no ready-made methodology they can apply by simply filling in the boxes. In Picerno’s view, not even the basic task of determining an investor’s risk preference is reducible to a formula.
Many readers will appreciate the author’s updating of MPT without resort to abstruse mathematics. Moreover, in dealing with theories developed mainly in academic settings, he is mindful of real-world constraints. For instance, he cites several papers on the tax and transaction cost impediments to frequent rebalancing.
Although Picerno is slightly prone to repetitiousness, no one could accuse him of lacking thoroughness. His 16-page bibliography attests to his meticulous tracing of the main lines of research that have challenged and ultimately enriched MPT. One lesser-known strand that Picerno helpfully highlights is the empirical work that links economic indicators with risk premiums.
The book’s comprehensive review of the literature pays off with numerous nuggets of investment knowledge. These include qualified endorsements of indexing from such unexpected authorities as behavioral finance proponent Richard Thaler and value investor par excellence Warren Buffett. As both a source of serendipitous enlightenment and a superb synthesis of recent advances in financial economics, Dynamic Asset Allocation is emphatically a book worth owning.
—M.S.F.