Most of the return from a diversified portfolio is derived from stock market exposure. If an investor can increase or decrease exposure to the stock market at the right time, he or she can avoid much downside loss and gain much upside return. There is more reason to believe that asset classes are inefficiently priced relative to each other than that individual assets are mispriced within asset classes. If that is true, then tactical asset allocation, if done well, can be extremely beneficial.
Strategic vs Tactical Asset Allocation
The objective of strategic asset allocation is to determine long-term “normal” benchmark allocations to various asset classes. At the very highest level, these would typically be stocks, bonds, and cash. At the next level down, stocks and bonds might be divided into major categories, such as U.S. and international. Another layer down might include different groups of stocks such as small cap and large cap, growth and value, and developed and emerging international. Allocations to bond groups might include Treasury and Agency, investment grade and high yield corporate, municipal, TIPs, and developed and emerging market international. Strategic allocations are based upon long-term considerations, particularly long-term expected returns and risks.
Tactical asset allocation (TAA) involves short- to intermediate-term forecasts of expected return among asset classes. The underlying assumption is that markets are not “macro efficient.” (“Micro efficiency,” a separate question, has to do with whether individual securities within an asset class are mispriced relative to each other.) That is, at times, the prices of some asset classes and groups may not appropriately reflect their risks. They may be “mispriced.” Some may be overvalued, and others undervalued. Tactical asset allocation attempts to take advantage of this mispricing.
Market Timing vs Tactical Asset Allocation
At its most extreme, tactical asset allocation borders on “market timing.” Traditionally, market timing involves completely selling out of certain asset classes, especially stocks, from time to time. This “all in or all out” approach may involve high transaction costs (although using futures rather than cash markets may alleviate that). It may also involve paying higher taxes if the moves are frequent enough to result in short-term rather than long-term gains.
Market timing has a very negative reputation. Perhaps the origin of this low opinion can be traced to a 1975 article by William F. Sharpe entitled “The Likely Gains from Market Timing.” (Sharpe, a titan in modern finance, was one of the originators of the Capital Asset Pricing Model, and won a Nobel prize in 1990.) Using U.S. annual stock market returns 1929-1972, Sharpe found that because of the high transaction costs involved (he assumed a 2% one-way cost), the skill required to add value net of transaction costs would present a very high hurdle. His analysis assumed that a market timer would make once a year 100% in or out decisions regarding stock market exposure. He found that a market timer with no forecasting skill (a 50% probability of being right) would underperform the buy-and-hold stock investor by about 4% per year. Just to break even with the buy-and-hold investor, he found that the market timer would have to be right about 75% of the time. Very few forecasters would claim that level of prescience.
On the other hand, a more recent study of the “live” performance of 30 professional market timers (Chance and Hemler, 2001) using daily data over the period 1986-1994 found “significant” forecasting ability even net of transaction costs. However, when tested on a monthly basis rather than daily, the significance disappeared. Apparently, the signals that the forecasters used had a quick decay, making nimbleness in execution very important. This is not a surprise. High-frequency trading tends to have the strongest performance among investment styles and processes.
If markets are informationally efficient, then current asset prices reflect all known information. At its most extreme, this would be true for both individual securities and for asset classes. However, there is more reason to believe that asset classes are not informationally efficient—the markets are macro inefficient. The rationale for this assertion has to do with the “limits to arbitrage”—the fact that not all assets are subject to effective arbitrage that will drive their prices toward intrinsic value. Arbitrage can be applied fairly easily to individual stocks that are mispriced. For example, overpriced stocks can be sold short, and other related risks, such as market, sector, and industry, can be neutralized with hedging techniques. That sort of low-risk arbitrage is not possible with asset classes.
There is evidence that the stock markets are not always appropriately priced. Prices sometimes fall dramatically with no apparent change in the underlying fundamentals. For example, on October 19, 1987, the S&P 500 dropped 20.47% with no news that would explain it. Those who study behavioral finance theorize that markets are subject to crowd psychology, and that euphoric pricing bubbles can form that ultimately pop. Selling pressure can feed on itself and spiral into a panic in some cases, as it did on that day.
Robert Shiller was one of the first academics to point out that stock indexes are far too volatile relative to changes in their underlying fundamentals. His 1981 article “Do Stock Prices Move Too Much to be Justified by Subsequent Changes in Dividends?” was a thunderbolt against the backdrop of absolute faith in market efficiency among academics at the time. His book, Irrational Exuberance (2000), which was published at the height of the dot-com bubble, argued that stocks had become grossly overvalued. A second edition in 2005 included cautionary remarks on a growing housing bubble. His latest edition (2015) includes warnings on the risks of long-term bonds. (Shiller won a Nobel prize in 2013.)
The Importance of Managing Stock Market Risk
Managing stock market risk is vitally important for investors. That single risk exposure determines a majority of the return from a diversified portfolio. However, stock market returns are exceedingly difficult to forecast. The combination of difficulty and importance leads most investors to seek some accepted and easy solution, such as the sacrosanct 60/40 strategic stock/bond asset mix. This head-in-the-sand approach is often accompanied by blind faith in the magic of periodic rebalancing back to the inviolable strategic asset mix.
Part of the problem is the fact that most investors, whether amateur or professional, do not believe that they have enough information or skill to actively manage their asset mix. However, I have argued elsewhere that the simple fundamentals of current bond yields and stock dividend yields provide a rich and potentially valuable dataset with which an investor can tactically manage the stock/bond asset mix around the strategic benchmark allocations. Applied modestly, it can function as a substitute methodology for calendar-based or percentage-based rebalancing.
In addition, there are other tactical asset allocation factors that have historically proven helpful in over- and under-weighting asset classes relative to their strategic allocations. In the case of stocks, these TAA factors generally fall into the following categories:
I will explore the potential benefits of some of these factors in future articles.
Investors must of necessity manage their asset mix. They will do it somehow, one way or another. They will do it unconsciously or consciously, based on rules-of-thumb or thorough analysis, with neglect or attention. Given its importance, I believe that asset allocation, both strategic and tactical, should be the focus of every investor.