Value Averaging Beats Dollar-Cost Averaging on Paper. Here's Why Almost Nobody Actually Uses It
7 min read
Loading...
7 min read
You already know dollar-cost averaging: invest the same fixed amount every month, no matter what the price is doing. There is a less-known cousin of it that beat dollar-cost averaging in study after study going back decades. It is called value averaging, and almost nobody uses it, not even the robo-advisors that automate everything else about investing. Here is the value averaging vs dollar-cost averaging comparison, the real numbers behind it, and the catch that explains why it never caught on.
Dollar-cost averaging (DCA) fixes your contribution: the same rupiah or dollar amount every period, whatever the market does.
Value averaging (VA) fixes your target portfolio value instead. You decide how much you want your portfolio worth by the end of each period, and you contribute whatever is needed to hit that number, exactly:
This month's contribution = target value − actual value
That one change has a big effect. If the market fell last month, your actual value is below target, so you contribute more than usual. If the market rallied hard, your actual value may already be above target, so you contribute less than usual, or you sell the difference. VA mechanically buys more when prices are down and less (or sells) when prices are up, without you ever having to guess where the market is headed.
Here is what that looks like over five months, using a target that grows by a fixed amount each month and a made-up sequence of monthly returns, to show the mechanics rather than a real market result:
| Month | Target value | Return that month | Value before contribution | Contribution |
|---|---|---|---|---|
| 1 | $300 | n/a | $0 | +$300 |
| 2 | $600 | +5% | $315 | +$285 |
| 3 | $900 | −12% | $528 | +$372 |
| 4 | $1,200 | +20% | $1,080 | +$120 |
| 5 | $1,500 | +35% | $1,620 | −$120 (sell) |
Look at months 3 and 5. After the market drop in month 3, value averaging automatically calls for a bigger contribution than usual, $372 instead of the $300 a DCA plan would have put in. After the sharp rally in month 5, it calls for pulling $120 out, something a DCA plan never asks you to do. That is the whole mechanism in miniature: it is a built-in "buy more when cheap, trim when expensive" rule, with no market timing decision required from you.
This is not just a theoretical curiosity. Michael Edleson, at the time chief economist of NASDAQ and a finance professor at Harvard Business School, first published the value averaging method in the AAII Journal in August 1988, then expanded it into the 1993 book Value Averaging: The Safe and Easy Strategy for Higher Investment Returns.
A later, independently run test by AAII itself made the comparison concrete. In a Computerized Investing analysis by John Markese and John Bajkowski, first published in AAII's July/August 2001 issue and still hosted on AAII's site today, value averaging was tested against dollar-cost averaging using the Invesco QQQ Trust, the ETF that tracks the Nasdaq-100. Value averaging produced a lower average cost per share, $38.71 versus $41.40 for DCA, and a higher realized return: a 36.1% internal rate of return versus 33.6% for DCA over the same period.
Here is the finding that explains why professional researchers stopped recommending value averaging even after seeing it win.
In 2013, Simon Hayley of Bayes Business School (then Cass Business School) at City University London published Value Averaging and How Dynamic Strategies Bias the IRR and Modified IRR. His finding: value averaging's internal rate of return (IRR) advantage over dollar-cost averaging is, in large part, a statistical artifact of the IRR calculation itself, not extra real wealth in your account.
The reason is subtle but important. IRR weights early cashflows more heavily than late ones. Value averaging systematically invests less right after a strong run of returns and more right after a weak one, which shifts more of your money onto the periods that already performed well and less onto the periods that performed poorly, purely by the mechanics of the formula. That reshuffling inflates the measured IRR without necessarily leaving you with a bigger ending balance. Hayley's conclusion was that value averaging carries "substantial dynamic inefficiency," meaning the strategy's apparent edge shrinks or disappears once you measure the actual wealth created rather than the IRR alone.
Even setting the measurement debate aside, value averaging has three practical problems that dollar-cost averaging simply does not:
None of this means value averaging is a trap. If you have the discipline to track a target value path and top it up (or trim it) by hand each month, the mechanism genuinely tilts your buying toward cheaper prices. But for most people, the honest trade-off is: a small, uncertain, and debated edge, against a real ongoing chore and the risk of an unaffordable contribution landing at the worst possible time.
Whichever approach you actually use, fixed monthly DCA, occasional lump sums, or a hand-run value averaging plan, what matters most is knowing your real, money-weighted return afterward, not which strategy you assumed would win going in. Check the IRR figure on your NetWort dashboard: it is calculated from the exact size and timing of every contribution you actually made, so it reflects what really happened to your money rather than a theoretical comparison between strategies.