Abstract
of Dissertation
The research effort described in this dissertation is conducted
to investigate the performance of an “automatic” investment
technique known as dollar value averaging relative to the performance of
dollar cost averaging and asset allocation.
Dollar value averaging has not been widely published and few
practical implementation issues have been discussed in the literature.
Therefore, this dissertation addresses the following questions:
•
When is dollar value averaging superior to dollar cost averaging
and asset allocation?
•
How significant is this superior performance?
•
How can the small investor best implement dollar value averaging
using real world investments?
A
desired by-product of this effort is the creation of a step-by-step plan
to use dollar value averaging on real investments.
This plan includes guidelines to maximize dollar value averaging
performance and minimize the impacts of real world constraints or
limitations such as commissions, taxes, and minimum investment amounts.
The research effort is conducted in three phases.
The first phase or method is a literature search.
The literature search included book and periodical searches using
on-line, multi-library index systems as well as printed indexes over the
last ten years. A thorough
search was also conducted over the Internet using the “alta vista”
spider indexing system. The
second phase of research is a mathematical derivation of the equations
defining the performance criteria used in this effort for dollar value
averaging and dollar cost averaging.
These derivations are performed to prove mathematically one of
the hypotheses of the proposal (i.e., dollar value averaging and dollar
cost averaging are equivalent in a perfectly flat market) and to predict
the performance of pure dollar value averaging and dollar cost averaging
relative to long-term trends versus short-term volatility of the market.
Mathematical derivations for asset allocation were not conducted.
The third phase of research consists of empirical simulations of
dollar value averaging, dollar cost averaging, and asset allocation over
periods up to thirty years. The
simulations are facilitated through the creation of a large spreadsheet
that simultaneously calculates each technique’s actual results and
differential results between the techniques for a given market.
Published literature for value averaging is very limited, and
except for the book, Value Averaging:
The Safe and Easy Strategy for Higher Investment Returns, by
Michael E. Edleson (the founder of value averaging), none of it goes
very deep. Most of the
literature simply defines value averaging and states that it is better
than dollar cost averaging. Only
Edleson goes into the circumstances of when value averaging is better
and provides some detail on how to invest using value averaging.
The general conclusions regarding base rates and volatility are
listed below:
•
Dollar value averaging is best in up markets, but it swings more
violently (and inversely) with volatility.
•
Increasing volatility magnitude and frequency have slightly
positive lasting effect on all three techniques.
•
In flat markets, asset allocation performs better than dollar
cost averaging which performs better than dollar value averaging, but
the margin between the techniques is small.
•
In down markets, dollar value averaging is the worst, but the
situation quickly reverses on an uptrend.
•
Dollar value averaging performs best, from a return on investment
perspective, in bursty and uptrending markets.
•
Dollar cost averaging has no tax consequences (ignoring dividends
and capital gains and losses for trades made within the investment
vehicle [e.g., a mutual fund]).
•
Dollar value averaging suffers heavily when taxed, but it is
still the best technique when it is best in the absence of taxes.
Real world constraints that had significant effects on dollar
value averaging returns on investment are summarized below:
•
Generally, no-load, or no-expense, investing is best.
However, it is better to have one-time commissions, either upon
investment or redemption or both, than to have on-going expenses (e.g.,
storage fees for metal accumulation accounts) based on the value of the
account.
•
The impact of taxes range from no effect on dollar cost averaging
to dramatic negative impacts on dollar value averaging.
•
The use of growth factors results in dramatically higher account values
with little cost in return on investment.
•
More frequent adjustment tends to be better unless a
predominantly flat market occurs.
•
Reinvesting dividends tremendously boost returns.
•
Generally, insufficient money to invest in down markets is good
with little degradation in return on investment on upswings and high
resistance to poor returns on investment on down swings.
Real world constraints investigated that resulted in little or no
significance to the return on investment for dollar value averaging are
presented below:
•
Initial investment amount
•
Minimum additional investment or redemption amounts
•
Threshold of adjustment
•
Maximum number of exchanges per year
•
Size of monthly investments
•
Timeliness (NOTE: This
test has biases as discussed in the body of the dissertation).
Overall, the results of this research are consistent with
publications by others. Most notably, this research confirms Mr.
Edlesons’ claims that “no one strategy is strictly better than all
others,” and the investment vehicle and performance is more important
than the mechanical rules of a strategy.
The research presented in this paper shows that of the three
techniques (dollar value averaging, dollar cost averaging, and asset
allocation), the best technique from a return on investment perspective
depends on the trends of the market and none of the techniques will
counteract the direction of the trend.
The comprehensive
dissertation is well over 300 pages.
If you want access to the entire text, Buy It or Get It Free using the
buttons below.
Download your own complete copy today .... or Click "Next
Page" to keep reading for now.
|