Value averaging

From Wikipedia, the free encyclopedia
Jump to: navigation, search

Value averaging, also known as dollar value averaging (DVA), is a technique of adding to an investment portfolio to provide greater return than similar methods such as dollar cost averaging and random investment. It was developed by former Harvard University professor Michael E. Edleson. Value averaging is a formula-based investment technique where a mathematical formula is used to guide the investment of money into a portfolio over time. With the method, investors contribute to their portfolios in such a way that the portfolio balance increases by a set amount, regardless of market fluctuations. As a result, in periods of market declines, the investor contributes more, while in periods of market climbs, the investor contributes less. In contrast to dollar cost averaging which mandates that a fixed amount of money be invested at each period, the value averaging investor may actually be required to withdraw from the portfolio in some periods.

Value averaging incorporates one crucial piece of information that is missing in dollar cost averaging – the expected rate of return of your investment. The investor must provide this information for the value averaging formula. Having this data allows the value averaging formula to identify periods of investment over-performance and under-performance versus expectations. After the investment has over-performed, the investor will be required to buy less or sell (selling high). After the investment has under-performed, the investor will be required to buy more (buying low). Some research suggests that the method results in higher returns at a similar risk, especially for high market variability and long time horizons. Some research suggests otherwise.


American financial theorist and money manager William J. Bernstein has stated that value averaging is superior to lump sum investing and dollar cost averaging for deploying a large sum into a portfolio. In this case, Professor Edleson recommends a VA period of three years. He suggests an infusion or withdrawal of capital every three or six months. For example, if one were to win or be bequeathed one million dollars, roughly 8.33 percent, with the exact amount being set by the formula, could be invested every quarter. It is important to note that the quarterly or semiannual amount can vary greatly, even resulting in a withdrawal, as mentioned above. Opponents argue that this misses the opportunity of already being fully invested when a large market upswing occurs. This argument against value averaging and dollar cost averaging and in favor of lump sum investing ignores the suggestion of Ben Stein and Phil DeMuth that it is more important to avoid a large market downswing, which is theoretically equally possible, since market movements are essentially unpredictable. Participating early on in a large market downswing has been shown to be devastating to the success of long term retirement, for example.

Author Timothy J. McManaman further outlines the benefits of Value Averaging when applied to the popular 401(k) tax qualified investment vehicle. As stated in McManaman's book, Building a 401(k) Fortune, Value Averaging a 401(k) is a precise method of making periodic internal transfers between Equity and Money Market funds within a 401(k) to take advantage of market fluctuations. This is accomplished by initiating minor movements out of Equity funds when the overall market trends higher and back into Equity funds when the market moves lower. It is essentially buying fund shares at a lower base price and selling them a higher base price within a tax qualified 401(k) on a monthly or quarterly interval. As outlined by Michael E. Edleson and Paul S. Marshall, Value Averaging can provide for an increased rate of return when compared to dollar cost averaging and other investment techniques.


Any return advantage that value averaging provides derives from mean reversion in the market, which implies that there is some market force that causes a market disturbance in one direction to be balanced out on average by an opposite market disturbance so that the market reverts to the mean. (This not the same concept as regression to the mean as used by statisticians where retesting a non-random sample of a population tends to produce results that are closer to the mean than the original test.) The existence of mean reversion in financial markets is controversial and very difficult to detect, and is a subject of active research. If it exists, then it is almost certainly very slight.

Value averaging makes use of a side cash account to hold the funds that are generated by value-averaging sales and that provide the funds for value-averaging purchases. Consider the following example. In its simplest form, the value averaging concept can be applied to a portfolio with no net inflow (or outflow) or lump sum to invest. In this case, the value path is the projected value of the portfolio alone based on its expected return, i.e., one does not need to make any adjustments to the path for net new investments. If at time t, the current value of the portfolio is above the value path, securities are sold to bring the value back down to the value path and deposit the proceeds in the side cash account. If the current portfolio value is below the value path, securities are bought using the cash account to try to bring the portfolio back up to the value path. The concept is to sell high and buy low.

However, for this approach to work, there must be a mean reversion process in the market. If there is not, then the increase in the value of a portfolio today tells nothing about what it will do tomorrow. If one had an expectation of a return of X% today, one should have an expectation of X% tomorrow. The same is true if the value dropped. Without mean reversion, the fluctuations in portfolio value can only be attributed to random noise in the expected value of the return sequence. In general, the value of the portfolio will increase because the expected value of the returns is positive but the value curve is generally not monotonically increasing and there is no relationship between departures from a monotonic curve. Independence between today's return and tomorrow' return is similar to the situation that if you have flipped a fair coin ten times and the results have all been heads, you still have a 50/50 chance of heads or tails on the eleventh flip.

Without mean reversion, the expected value of the value averaged portfolio described above at some point in the future will be less than if one did nothing, because any money that is in the side cash account due to random fluctuations in the portfolio value generally has a lower expected return than the portfolio. In other words, any money that randomly ends up in the cash account is effectively earning less than it could if you had done nothing. This also applies to the more general case of value averaging with net inflows.

Independent of the issue of mean reversion, the cash side account required for value averaging will always cause some amount of reduced return on the overall portfolio since the money in the cash account, on average, will be earning less than if it was in the main portfolio. Any benefits that value averaging provides in terms of market timing need to overcome this factor.

Because value averaging sometimes calls for the sale of assets even during an overall accumulation phase, there can potentially be additional transaction costs and restrictions. For example, some mutual funds have frequent trader policies. Some funds forbid additional investment in the fund within N months of a redemption from the fund. Some funds charge an additional fee for a redemption if there has been an investment in the last N months.



External links[edit]